UNIVERSIDAD COMPLUTENSE DE MADRID FACULTAD DE CIENCIAS QUÍMICAS Departamento de Ingeniería Química y de Materiales TESIS DOCTORAL Pretratamientos en una etapa, multietapa y proceso "one- pot" de madera de pino y eucalipto, basados en el empleo de líquidos iónicos apróticos, próticos y bioderivados MEMORIA PARA OPTAR AL GRADO DE DOCTOR PRESENTADA POR Victoria de los Ángeles Rigual Hernández Directoras Mercedes Oliet Palá Mª Virginia Alonso Rubio Madrid Ed. electrónica 2019 © Victoria de los Ángeles Rigual Hernández, 2019 UNIVERSIDAD COMPLUTENSE DE MADRID Facultad de Ciencias Químicas Departamento de Ingeniería Química y de Materiales Tesis Doctoral Pretratamientos en una etapa, multietapa y proceso “one-pot” de madera de pino y eucalipto, basados en el empleo de líquidos iónicos apróticos, próticos y bioderivados MEMORIA que para optar al Título de Doctor por la Universidad Complutense de Madrid en el Programa de Doctorado de Ingeniería Química presenta Victoria de los Ángeles Rigual Hernández Tutora: Mercedes Oliet Palá Directoras: Mercedes Oliet Palá y Mª Virginia Alonso Rubio Madrid, 2019 Portada: Imágenes de microscopía electrónica de barrido adquiridas con el microscpio electrónico Jeol JSM 6400 de muestras de madera de pino pretratadas con autohidrólisis a 150 °C y posteriormente con el líquido iónico acetato de 1-etil-3-metilimidazolio a 50 °C en horno microondas. Sobre la imagen de fondo hay superpuestas imágenes de la estructura química de algunos de los líquidos iónicos utilizados en este trabajo. Dª Mercedes Oliet Palá y Dª Mª Virginia Alonso Rubio, Catedrática y Profesora Titular del Departamento de Ingeniería Química y de Materiales de la Facultad de Ciencias Químicas de la Universidad Complutense de Madrid, CERTIFICAN: que el presente trabajo de investigación titulado “Pretratamientos en una etapa, multietapa y proceso “one-pot” de madera de pino y eucalipto, basados en el empleo de líquidos iónicos apróticos, próticos y bioderivados” constituye la memoria que presenta Victoria de los Ángeles Rigual Hernández para optar al grado de doctor, que ha sido realizada mayoritariamente en los laboratorios del Departamento de Ingeniería Química y de Materiales de la Universidad Complutense de Madrid bajo su dirección. Y para que conste a los efectos oportunos, firman el presente certificado en Madrid, febrero de 2019. Mercedes Oliet Palá Mª Virginia Alonso Rubio Una imagen vale más que mil palabras. (Refrán popular) A Pixel is worth a thousands words. En estas líneas quiero mostrar mi agradecimiento a todas aquellas personas que, de una u otra manera, han hecho posible la realización de este trabajo. En primer lugar quiero agradecer a mis directoras Mercedes Oliet Palá y Mª Virginia Alonso Rubio la confianza que mostraron en mí para realizar la tesis doctoral con ellas, su saber estar y la libertad que me han dejado para probar nuevas ideas a lo largo de estos años. También quiero agradecerles su ayuda cuando ha sido necesaria, especialmente en esta última etapa de la tesis doctoral y su paciencia. Del mismo modo, me gustaría agradecer a Francisco Rodríguez Somolinos la oportunidad que me ha brindado de formar parte de su grupo de investigación durante estos años. Su apoyo, confianza y cercanía, así como a Julián García González, miembro de mi grupo de Investigación. Este trabajo no habría sido posible sin la financiación del Ministerio de Economía y Competitividad a través del proyecto CTQ-2013-42006-R, el contrato predoctoral BES-2014- 067788 y la ayuda a la movilidad EEBB-I-17-12049. También debo agradecer la colaboración de la empresa DuPont, especialmente del Dr. Sharief Barends, quien nos donó los complejos enzimáticos para realizar nuestros ensayos, así como a la Dra. Sandrine Bouquillon de la Universidad de Reims, quien me proporcionó los líquidos iónicos basados en la prolina. Gracias también a toda la gente que conocí en The Joint BioEnergy Institute (JBEI). Primeramente gracias a la Dra. Seema Singh por aceptarme para realizar la estancia, sus palabras amables y su seguimiento durante mi estancia allí, así como meses después. Gracias al Dr. Blake A. Simmons por sus ideas, sus puntos de vista, sus comentarios y su opinión sobre mi trabajo en JBEI. Gracias a John Gladden y a Aindrila Mukhopadhyay por su ayuda con las levaduras. Por último, gracias a todos los postdocs por el enorme trabajo y apoyo que me dieron: Maren, Kwang, James, Alberto y especialmente Gaby. No puedo olvidar tampoco a los amigos que hice durante mi estancia en Berkeley: Nadine, Seb, Raphael, Arthur y Pablo Nunca hubiera podido comenzar mi tesis doctoral de no ser por Rubén Miranda, quien me mostró el trabajo de investigación y, junto con Juan Carlos, pensó en mí para realizar la tesis cuando surgió la oportunidad: muchas gracias por todos vuestros consejos. En la UCM tenemos la suerte de tener unos Centros de Apoyo a la Investigación que me han permitido realizar análisis que, sin ellos, nunca hubiera podido hacer. Gracias a Pedro por su amabilidad y a Lola por su predisposición para poner a punto un nuevo método para hacer 2D- RMN de lignina. También han pasado por el laboratorio TFMs con los que aparte de aprender, he pasado grandes momentos: gracias a Álvaro, a Miguel, a Marta, a Patri, a Carmen y en especial a mi ya amiga Loreto. Por último, agradecer al mejor técnico que se puede tener, Jesús, por sacarnos del apuro a diario, despertarnos por las mañanas y sobre todo por ser tan buena gente. Durante estos años he compartido cafés, laboratorio y comidas con un grupo espectacular. Gracias a Pepelu, Jhiret, Antonio, Miguel, Noelia, Noemí, Marcos, Cynthia y Javi por las risas. También quiero agradecer a Pablo su apoyo y ayuda en los momentos de bloqueo, a Patricio por su genialidad fractal (sin la cual no existiría el apartado 3.5 de la tesis) y a Gonzalo por sus chistes y momentos compartidos. Mención especial tienen mis compañeros de laboratorio: Sandra por su conocimiento y dosis de realidad y Antonio por su buen rollo: el alumno superará al maestro en menos que canta un gallo. Previamente a ellos, éramos Tamara y yo: juntas conseguimos sacar grandes ideas y disfrutar de la biorrefinería. Gracias por todo lo que me has enseñado y por escucharme. Esta tesis tampoco habría sido posible sin toda la gente de Corre con Nosotros: Luis, Juan Carlos, Paco, Vane, Eva, Laura, Almu, Alberto, David, Isma, Héctor, Marta, Jeffrey, y unos cuantos más. Gracias a Amaya y Pablo por engancharme a este gran deporte y hacer que cada lunes y miércoles esté deseando ponerme las zapatillas. Aquí también tienen culpa Gonzalo (menudo final de geolodía y primera experiencia de trail) y Edu, que no había quien le siguiera. También quiero agradecer la paciencia que han tenido mis amigos de siempre. Gracias a las patas y patos por ser mi gente, por las quedadas por Guada y Madrid y por el buen rollo, las risas y ya gran colección de momentos juntos. A Mateo y a Isra por sus ideas locas, sus ganas de hacer cosas y por ser mis amigos. A David, Ana y Luci por su sencillez y alegría continua, a Ana por su gran iniciativa, a Fon por todas las risas y a María por comprenderme en la distancia. No puedo olvidarme de mi gran descubrimiento en esta tesis: Mari y Cris. Nunca pensé que durante estos años podría sacar momentos tan divertidos: hacemos un gran equipo y espero teneros siempre, sois unas grandes. Gracias a mi familia por hacerme sentir tan afortunada: Gracias a mi padre por preocuparse pero dejarme seguir mi camino, a mi madre por cuidarme siempre y a mi hermana porque…ES LA MEJOR. Por último agradecer a Óscar su apoyo, paciencia y esfuerzo infinito todo este tiempo: ahora ya sí subiremos montañas…juntos. Through these lines I want to thank all the people who have made possible this work. First of all, I want to thank my supervisors Mercedes Oliet Palá and Mª Virginia Alonso Rubio their reliance, their patience and the freedom they gave me to test new ideas through these years. I also want to thank them their help, always I have needed it, especially in this last stage of my PhD. In the same way, I would like to thank Francisco Rodríguez Somolinos the opportunity he gave me to be part of his research group during all these years, as well as his support, reliance and good character, and also Julian García González, member of this research group. This work would have been impossible without the funding of Ministerio de Economía y Competitividad through the project CTQ-2013-42006-R, the contract BES-2014-067788 and the mobility grant EEBB-I-17-12049. But this work would have been impossible too, without the collaboration of the company Dupont, especially Dr. Sharief Barends who donated us the enzymes complexes to run our experiments and Dr. Sandrine Bouquillon of the University of Reims, who gave me some samples of her prolinate-based ionic liquids. Thank you to all the people I knew in The Joint BioEnergy Institute. First of all thanks to Dr. Seema Singh for accepting me as invited PhD student and her nice words and follow-up during my stay there and after; thanks to Dr. Blake Simmons for his ideas, points of view, and valuable comments about my research; thanks to John Gladden and Aindrila Mukhopadhyay for their help with the yeasts; and thanks to all the postdocs I knew there, for the huge work and support they gave me: Maren, Kwang, James, Alberto and specially Gaby. I cannot forget my months in Berkeley without thanking my friends: Nadine, Seb, Raphael, Arthur and Pablo. I wouldn’t have started my PhD if Ruben Miranda, had not shown me the research work, and Juan Carlos had not thought about me when the opportunity arose. Thank you very much to both of you for all your advices. We are very lucky at the Complutense University of Madrid because, through the Centros de Apoyo a la Investigación, I have been able to develop different analytical techniques that I would have never done without this service. Thanks to Pedro for his kindness and thanks to Lola for setting up a new method to run 2D-NMR for lignin samples. Thank you also to all the Master thesis students who have been in my lab: Álvaro, Miguel, Marta, Patri, Carmen and especially my friend Loreto. Finally, I would like to thank to the best technician we can have, Jesus, for his last-minute help, waking up every morning and for being such a good guy. During all these years I have shared coffees, lab and lunches with great people. Thanks to Pepelu, Jhiret, Antonio, Miguel, Noelia, Noemí, Cynthia, Marcos and Javi for all the laughs. I also want to thank Pablo his support when I was blocked, Patricio for his fractal brilliance and Gonzalo for his jokes and good moments spent together. Special words are for my lab mates: thanks to Sandra for her knowledge and reality-pills and Antonio for his good vibes: the student will beat the teacher. Before them, we were only Tamara and me: we fulfilled great ideas together, and enjoyed with biorefinery: thanks for everything you taught me and for listening me. This work would have also been impossible without my running group of Corre con nosotros: Luis, Juan Carlos, Paco, Vane, Eva, Laura, Almu, Alberto, David, Isma, Hector Marta, Jeffrey, etc. Thanks to Amaya and Pablo for calling my attention with this great sport that makes me desire to footwear my sport shoes every Monday and Wednesday. Gonzalo and Edu are also part of this habit. I also want to thank my life-long friends. Thanks to patas y patos for being my people, the good feelings, laughs and huge collection of great moments. Thanks to Mateo and Isra for their crazy ideas and being my friends. Thanks to David, Ana and Luci for their simplicity and happiness, thanks to Ana for her initiative, thanks to Fon for all the good moments and thanks to Maria for understanding me in the distance. I cannot forget my best discovery in this PhD: Mari and Cris. I never thought that during these years I would spend so funny times: we make a great team and I hope to have you forever, you are the best. Thanks to my family for making me feel so lucky: thanks to my father for worrying about me but let me follow my own way, my mother for taking care of me and my sister because…SHE IS THE BEST. Finally I want to thank Óscar all his support, patience and unlimited effort during all this time: now we will climb mountains…together. ÍNDICE Índice i ÍNDICE RESUMEN / ABSTRACT ................................................................................................................ 1 Capítulo 1: LÍQUIDOS IÓNICOS EN EL PROCESADO DE BIOMASA ................................... 11 1.1. El concepto de biorrefinería ................................................................................................. 13 1.1.1. Tipos de biorrefinería .................................................................................................... 15 1.1.2. Biorrefinería de materiales lignocelulósicos ................................................................. 15 1.2. La madera como materia prima en biorrefinería .................................................................. 16 1.2.1. Estructura y composición de la madera ......................................................................... 17 1.2.2. Tipos de maderas ........................................................................................................... 20 1.2.3. Aprovechamiento de la madera en biorrefinería ........................................................... 22 1.3. Líquidos iónicos en el pretratamiento de biomasa lignocelulósica ...................................... 29 1.3.1. Clasificación de los líquidos iónicos ............................................................................. 29 1.3.2. Propiedades de un líquido iónico para ser empleado como disolvente en el pretratamiento de madera ........................................................................................................ 30 1.4. Interés y relevancia del pretratamiento de maderas con líquidos iónicos ............................ 37 Capítulo 2: OBJETIVOS Y PLANTEAMIENTO DE LA TESIS DOCTORAL ........................... 41 Capítulo 3: PRETRATAMIENTOS EN UNA ETAPA .................................................................. 49 3.1. Autohidrólisis ....................................................................................................................... 51 ii 3.1.1. Estructura y procedimiento experimental ...................................................................... 52 3.1.2. Caracterización de las fracciones obtenidas .................................................................. 54 3.1.3. Resultados experimentales. Interpretación y discusión ................................................. 54 3.2. Pretratamiento con líquidos iónicos con calefacción convencional ..................................... 58 3.2.1. Estructura y procedimiento experimental ...................................................................... 58 3.2.2. Caracterización de las fracciones obtenidas .................................................................. 61 3.2.3. Resultados experimentales. Interpretación y discusión ................................................. 62 3.3. Pretratamiento con líquidos iónicos con calefacción microondas ........................................ 75 3.3.1. Estructura y procedimiento experimental ...................................................................... 75 3.3.2. Caracterización de las fracciones obtenidas .................................................................. 77 3.3.3. Resultados experimentales. Interpretación y discusión ................................................. 78 3.4. Recuperación y reutilización del líquido iónico ................................................................... 82 3.4.1. Estructura y procedimiento experimental ...................................................................... 82 3.4.2. Caracterización de las fracciones obtenidas .................................................................. 84 3.4.3. Resultados experimentales. Interpretación y discusión ................................................. 84 3.5. Estudio de la accesibilidad del sólido a través de técnicas microscópicas ........................... 92 3.5.1. Cálculo de la distribución de tamaños de partícula, dimensión fractal y lacunaridad a través de microscopía electrónica de barrido .......................................................................... 93 3.5.2. Cálculo de la relación holocelulosa/lignina superficial a partir de microscopía de fluorescencia confocal ............................................................................................................. 95 3.5.3. Resultados experimentales. Interpretación y discusión ................................................. 96 3.5.4. Validación del estudio de la accesibilidad del sólido a través de técnicas microscópicas ................................................................................................................................................. 98 3.6. Evaluación de los pretratamientos en una etapa ................................................................. 100 Capítulo 4: PRETRATAMIENTOS MULTIETAPA: AUTOHIDRÓLISIS + LÍQUIDOS IÓNICOS ....................................................................................................................................... 101 4.1. Estructura y procedimiento experimental ........................................................................... 103 4.2. Caracterización de las fracciones obtenidas ....................................................................... 106 4.3. Resultados experimentales. Interpretación y discusión ...................................................... 106 4.4. Estudio de la accesibilidad del sólido a través de técnicas microscópicas ......................... 112 4.5. Evaluación de los pretratamientos multietapa .................................................................... 115 Capítulo 5: PROCESO “ONE-POT”: PRETRATAMIENTO + HIDRÓLISIS ENZIMÁTICA .. 117 5.1. Estructura y procedimiento experimental ........................................................................... 119 Índice iii 5.2. Caracterización de las fracciones obtenidas ....................................................................... 123 5.3. Resultados experimentales. Interpretación y discusión ...................................................... 124 5.3.1. Ensayos previos: selección del sistema madera-LI-levadura ...................................... 124 5.3.2. Prueba de concepto: Proceso “one-pot” seguido de fermentación con Rhodosporidium toruloides............................................................................................................................... 127 5.4. Evaluación del proceso “one-pot” y evaluación conjunta de todos los procesos estudiados ................................................................................................................................................... 134 5.4.1. Evaluación del proceso “one-pot” ............................................................................... 134 5.4.2. Evaluación de los procesos en una etapa, multietapa y “one-pot” .............................. 134 Capítulo 6: CONCLUSIONES / CONCLUSIONS....................................................................... 137 Capítulo 7: NOMENCLATURA .................................................................................................. 147 Capítulo 8: BIBLIOGRAFÍA ........................................................................................................ 157 ANEXO I: LISTADO DE LÍQUIDOS IÓNICOS Y ESTRUCTURA QUÍMICA ...................... 173 ANEXO II: PUBLICACIONES .................................................................................................... 179 Publicación I: “Evaluation of hardwood and softwood fractionation using autohydrolysis and ionic liquid microwave pretreatment” Publicación II: “Protic, aprotic and choline-derived ionic liquids: towards enhancing the accessibility of hardwood and softwood” Publicación III: “Recovery and reuse of 1‑allyl-3-methylimidazolium chloride in the fractionation of Pinus radiata wood” Publicación IV: “Application of microscopy techniques for a better understanding of biomass pretreatment” Publicación V: “Combining autohydrolysis and ionic liquid microwave treatment to enhance enzymatic hydrolysis of Eucalyptus globulus wood” Publicación VI: “Autohydrolysis and microwave ionic liquid pretreatment of Pinus radiata: imaging visualization and analysis to understand enzymatic digestibility” Publicación VII: “One-pot conversion of woody biomass feedstocks using protic ionic liquids” Índice v ÍNDICE DE FIGURAS Figura 1.1: Modelo de fijación de CO2 y reemplazo de combustibles fósiles ................................. 14 Figura 1.2: Esquema de biorrefinería de biomasa lignocelulósica .................................................. 16 Figura 1.3: Estructura y organización de las células en la madera .................................................. 17 Figura 1.4: Estructura de la celulosa ............................................................................................... 18 Figura 1.5: Estructuras de la hemicelulosa ...................................................................................... 18 Figura 1.6: Ejemplo de estructura de la lignina en madera de chopo.............................................. 19 Figura 1.7: Esquema de etapas de aprovechamiento lignocelulósico ............................................. 22 Figura 1.8: Pretratamiento de biomasa lignocelulósica ................................................................... 23 Figura 1.9: Acción de las enzimas celulolíticas .............................................................................. 28 Figura 1.10: Líquidos iónicos apróticos .......................................................................................... 29 Figura 1.11: Líquidos iónicos próticos ............................................................................................ 30 Figura 1.12: Líquidos iónicos bioderivados .................................................................................... 30 Figura 1.13: Características que deben cumplir los LIs para ser utilizados en el pretratamiento de biomasa ........................................................................................................................................... 31 Figura 1.14: Publicaciones por año introduciendo la palabra clave “líquido iónico” y “biomasa” 39 Figura 2.1: Planteamiento de la Tesis doctoral ............................................................................... 46 Figura 3.1: Planteamiento experimental del proceso de autohidrólisis ........................................... 53 vi Figura 3.2: Composición y rendimiento del sólido pretratado en la autohidrólisis ......................... 56 Figura 3.3: Digestibilidad en glucano y rendimiento en glucosa de los sólidos autohidrolizados .. 57 Figura 3.4: Planteamiento experimental del proceso de pretratamiento con LIs con calefacción convencional .................................................................................................................................... 59 Figura 3.5: Composición y rendimiento de madera pretratada con LIs con calefacción convencional .................................................................................................................................... 65 Figura 3.6: Digestibilidad en glucano y rendimiento en glucosa de eucalipto pretratado con LI y calefacción convencional ................................................................................................................ 67 Figura 3.7: Digestibilidad en glucano y rendimiento en glucosa de pino pretratado con LI y calefacción convencional ................................................................................................................ 68 Figura 3.8: Espectros 2D-NMR de la región aromática .................................................................. 69 Figura 3.9: Espectros 2D-NMR de la cadena lateral ....................................................................... 71 Figura 3.10: Espectros FTIR-ATR de los LIs empleados ............................................................... 72 Figura 3.11: Derivadas de los termogramas de los LIs utilizados ................................................... 74 Figura 3.12: Ejemplo de termograma del [Mim][Cl] ...................................................................... 75 Figura 3.13: Planteamiento experimental del proceso de pretratamiento con líquidos iónicos y calefacción convencional ................................................................................................................ 76 Figura 3.14: Composición y rendimiento del sólido obtenido tras el pretratamiento con LIs con calefacción microondas ................................................................................................................... 79 Figura 3.15: Lignina acumulada en [Emim][OAc] ......................................................................... 80 Figura 3.16: Digestibilidad en glucano y rendimiento en glucosa de eucalipto pretratado con LI y calefacción con microondas ............................................................................................................ 81 Figura 3.17: Planteamiento experimental del proceso de recuperación y reutilización del LI [Amim][Cl] ...................................................................................................................................... 83 Figura 3.18: Rendimiento de sólidos recuperados con [Amim][Cl] recuperado en cada ciclo para madera de eucalipto ......................................................................................................................... 85 Figura 3.19: Composición del SRL obtenido con [Amim][Cl] recuperado en cada ciclo para madera de eucalipto ......................................................................................................................... 86 Figura 3.20: Espectros de FTIR del SRC recuperado obtenido con [Amim][Cl] en cada ciclo para madera de eucalipto ......................................................................................................................... 87 Figura 3.21: Cristalinidad del SRC recuperado en cada ciclo para madera de eucalipto ................ 87 Figura 3.22: Recuperación de LI, DMSO y metanol en cada lavado para la madera de eucalipto . 88 Figura 3.23: Espectros FTIR-ATR del [Amim][Cl] recuperado en cada ciclo para madera de eucalipto .......................................................................................................................................... 89 Índice vii Figura 3.24: Fotografías del [Amim][Cl] fresco y recuperado en cada ciclo para madera de eucalipto .......................................................................................................................................... 90 Figura 3.25: Lignina acumulada en el [Amim][Cl] recuperado en cada ciclo para madera de eucalipto .......................................................................................................................................... 90 Figura 3.26: Fotografías de a) [Amim][Cl] recuperado + agua y b) [Amim][Cl] recuperado + H2SO4 .............................................................................................................................................. 91 Figura 3.27: Digestibilidad en glucano de SRC y SRL pretratado con LIR ................................... 92 Figura 3.28: Ejemplo de conteo de cajas en imagen binarizada de madera de pino pretratada ...... 94 Figura 3.29: Procesado de imágenes para la determinación de la lignina superficial ..................... 95 Figura 3.30: Matriz de correlaciones del análisis de imagen digital de pino y eucalipto autohidrolizados y pretratados con [Emim][OAc] con calefacción por microondas ...................... 99 Figura 4.1: Esquema del proceso de combinación de pretratamientos AH + LI ........................... 104 Figura 4.2: Composición y recuperación de madera pretratada mediante el proceso multietapa . 107 Figura 4.3: Lignina acumulada en el LI recuperado del proceso multietapa ................................ 110 Figura 4.4: Digestibilidad y rendimiento de madera pretratada mediante el proceso multietapa . 111 Figura 4.5: Matriz de correlaciones del pretratamiento de pino multietapa, resultado del análisis de imagen digital ................................................................................................................................ 114 Figura 4.6: Balance de materia por 100 g de eucalipto sin tratar combinando los pretratamientos de autohidrólisis y líquido iónico ....................................................................................................... 116 Figura 5.1: Planteamiento experimental del proceso “one-pot” .................................................... 120 Figura 5.2: Reacciones de síntesis de los LIs próticos .................................................................. 122 Figura 5.3: Digestibilidad enzimática de pino y de eucalipto tras el proceso “one-pot” .............. 125 Figura 5.4: Crecimiento de levaduras en presencia de LI al 5 % en peso ..................................... 126 Figura 5.5: Caracterización del sólido obtenido tras el “one-pot” con eucalipto y [OHEtAm][OAc] ....................................................................................................................................................... 128 Figura 5.6: Morfología de los sólido obtenidos con eucalipto y [OHEtAm][OAc] ...................... 130 Figura 5.7: Crecimiento de levaduras en hidrolizados obtenidos tras el “one-pot” ...................... 131 Figura 5.8: Crecimiento de levaduras en hidrolizados obtenidos tras el “one-pot” en tubos de fermentación .................................................................................................................................. 131 Figura 5.9: Efecto del ajuste de pH en el consumo de azucares por R. toruloides en presencia de [OHEtAm][OAc] ........................................................................................................................... 132 Figura 5.10: Cromatogramas del consumo de monómeros de lignina por R. toruloides en hidrolizados en presencia de [OHEtAm][OAc] ............................................................................ 133 Figura 5.11: Resumen de las variables estudiadas en las muestras más representativas ............... 135 Índice ix ÍNDICE DE TABLAS Tabla 1.1: Ejemplos de definiciones de biorrefinerías ................................................................... 14 Tabla 1.2: Composición de maderas duras y blandas ..................................................................... 20 Tabla 1.3: Efectos principales de los procesos ................................................................................ 25 Tabla 1.4: Trabajos realizados sobre pretratamientos multietapa ................................................... 26 Tabla 1.5: Trabajos realizados sobre reciclado de LI en el procesado de biomasa ......................... 33 Tabla 1.6: Trabajos realizados sobre el estudio de procesos “one-pot” .......................................... 35 Tabla 1.7: Proyectos de investigación financiados por la UE para apoyar la industria de la biorrefinería a partir de biomasa lignocelulósica ............................................................................ 38 Tabla 3.1: Condiciones de operación en los ensayos de autohidrólisis ........................................... 52 Tabla 3.2: Azúcares, oligosacáridos y productos de degradación obtenidos en la fase líquida ...... 55 Tabla 3.3: Suministrador y pureza de los líquidos iónicos utilizados ............................................. 60 Tabla 3.4: Resumen de los resultados del “screening” de LIs ........................................................ 63 Tabla 3.5: Perfil de temperaturas utilizado en el horno microondas ............................................... 77 Tabla 3.6: Recuperación de [Emim][OAc] en cada lavado ............................................................. 79 Tabla 3.7: Temperatura de descomposición, de pérdida de 25 % y 50 % de masa y cenizas a 400 °C para madera de eucalipto ..................................................................................................... 91 Tabla 3.8: Imágenes al microscopio de muestras pretratadas por autohidrólisis ............................ 96 x Tabla 3.9: Imágenes al microscopio de muestras pretratadas con LIs y calefacción convencional 98 Tabla 4.1: Condiciones de operación en ensayos de autohidrólisis + LI ...................................... 105 Tabla 4.2: Recuperación de [Emim][OAc] en cada lavado en pretratamiento multietapa ............ 109 Tabla 4.3: Imágenes al microscopio de muestras de pino y eucalipto pretratadas con procesos multietapa ...................................................................................................................................... 113 Tabla 5.1: Suministrador y pureza de los líquidos iónicos utilizados ........................................... 121 Tabla 5.2: Condiciones de ajuste de pH en el proceso “one-pot” ................................................. 125 Tabla 5.3: Productos derivados de la pirólisis de la lignina y tiempos de retención encontrados en las muestras de eucalipto sin tratar, y tras el “one- pot” con [OHEtAm][OAc] ............................ 129 RESUMEN / ABSTRACT Resumen / Abstract 3 RESUMEN Título: Pretratamientos en una etapa, multietapa y proceso “one-pot” de madera de pino y eucalipto, basados en el empleo de líquidos iónicos apróticos, próticos y bioderivados. Introducción La biomasa lignocelulósica y en concreto la madera, es una estructura tridimensional, no uniforme, que necesita emplear pretratamientos para romper los enlaces entre los compuestos que la constituyen. Los procesos de pretratamiento de biomasa lignocelulósica pueden suponer hasta un 40 % de los costes de procesamiento totales. Los líquidos iónicos son sales formadas por un catión, generalmente orgánico y un anión, cuyo punto de fusión es inferior a 100 °C. Su capacidad de solvatación, su naturaleza no volátil, su no inflamabilidad, su alta conductividad iónica, su reciclabilidad y su facilidad de diseño y síntesis atendiendo al uso que se les quiera dar, los convierte en una alternativa a los disolventes convencionales. En los últimos años se ha comprobado que algunos líquidos iónicos son capaces de romper la red de la biomasa lignocelulósica, reduciendo la cristalinidad de la celulosa, incrementando la accesibilidad y favoreciendo la posterior sacarificación enzimática. Sin embargo, existen todavía muchos estudios que deben llevarse a cabo en lo relativo a la recuperación del líquido iónico, la reducción de costes del proceso, la eliminación de etapas de lavado, la toxicidad y la actualización 4 continua en base a las nuevas familias de líquidos iónicos que van surgiendo (líquidos iónicos bioderivados) o la aplicación de otras ya existentes (líquidos iónicos próticos). Objetivos El objetivo principal de la tesis doctoral consiste en estudiar distintos pretratamientos basados en la utilización de líquidos iónicos, con o sin autohidrólisis previa, para tratar madera de pino y eucalipto y seleccionar el proceso más efectivo de cara a la posterior etapa de hidrólisis enzimática. La principal novedad de la tesis radica en el empleo de diferentes pretratamientos en una etapa, multietapa y procesos “one-pot” sobre madera dura y blanda, lo que no se ha estudiado de manera conjunta en ningún trabajo anteriormente. Aparte de los líquidos iónicos apróticos, muy utilizados en bibliografía, se van a utilizar líquidos iónicos surgidos recientemente que sean menos tóxicos o de menor coste que los convencionales. El trabajo se verá enriquecido al poder comparar una madera dura frente a una madera blanda debido a las diferencias en cuanto a composición, estructura y comportamiento. El procedimiento experimental y los resultados derivados de la investigación se estructuran en tres bloques: pretratamientos en una etapa, pretratamientos multietapa y procesos “one-pot”. Resultados y conclusiones En los pretratamientos en una etapa se estudia la autohidrólisis y los pretratamientos con líquidos iónicos por separado. Los líquidos iónicos empleados se dividen en próticos, apróticos y bioderivados. En los pretratamientos con líquidos iónicos se emplea calefacción convencional y calefacción con microondas, ya que los líquidos iónicos son buenos absorbedores de la radiación por microondas y, por tanto, los tiempos de pretratamiento se pueden ver reducidos. Otro de los aspectos clave del empleo de los líquidos iónicos en el pretratamiento de biomasa es el estudio de la capacidad de reutilización. En este caso, se evalúa la capacidad de reciclado del líquido iónico cloruro de 1-alil-3-metilimidazolio, el cual se reutiliza hasta cinco veces analizando los cambios estructurales y de propiedades que experimenta a lo largo de los sucesivos ciclos. Por último, se propone aplicar técnicas de análisis de imagen digital basadas en el cálculo de la dimensión fractal, la lacunaridad y las relaciones de holocelulosa/lignina superficial para el estudio de la accesibilidad enzimática del sólido, parámetros que estarían relacionados con la digestibilidad del sólido pretratado. En el segundo bloque se plantea la combinación de pretratamientos de autohidrólisis seguido del pretratamiento con líquidos iónicos, con el objetivo de extraer hemicelulosas y a su vez aumentar la digestibilidad enzimática del material. Los resultados evidencian las diferencias existentes entre ambas maderas frente a estos pretratamientos. Mientras con la madera de pino la etapa de autohidrólisis limita la eficacia de pretratamiento con el líquido iónico posterior, la madera de eucalipto permite extraer hemicelulosas en la etapa de autohidrólisis y aumentar la digestibilidad del Resumen / Abstract 5 sólido pretratado tras la etapa de pretratamiento con líquido iónico, permitiendo así el fraccionamiento de la biomasa. En el tercer bloque se evalúa un proceso en el que pretratamiento e hidrólisis enzimática se llevan a cabo en el mismo equipo, sin etapa de separación sólido-líquido intermedia y, por tanto, sin incluir la etapa de lavado del sólido. A este proceso se le conoce como “one-pot”. Los resultados de este bloque han demostrado que los líquidos iónicos próticos probados no son capaces de ofrecer una mejora considerable de la digestibilidad en la madera de pino. Sin embargo, en la madera de eucalipto se puede plantear un proceso “one-pot”, utilizando el líquido iónico prótico acetato de 2- hidroxietilamonio. Este proceso, además de no necesitar la etapa de lavado, evita incluir una etapa de ajuste de pH normalmente requerida para operar en condiciones óptimas para las enzimas. De cara a una posterior etapa de fermentación, los hidrolizados obtenidos se han utilizado como sustrato de crecimiento del microorganismo Rhodosporidium toruloides, para evaluar su biocompatibilidad con el mencionado líquido iónico. Los procesos de autohidrólisis tanto de madera de eucalipto como de pino son capaces de extraer en torno al 70 % de las hemicelulosas y, en el caso del eucalipto, consiguen digestibilidades del sólido de hasta el 72 % aunque con bajos rendimientos finales a glucosa (38 %). En la madera de pino, se ha conseguido una digestibilidad del 84 % utilizando el líquido iónico aprótico acetato de 1-etil-3-metilimidazolio. Este valor es superior al conseguido con madera de eucalipto (80 %) y fue el más alto de todos los obtenidos en esta tesis doctoral. Los pretratamientos con líquidos iónicos próticos y bioderivados dieron mejores resultados con madera de eucalipto que utilizando madera de pino. Los pretratamientos multietapa con madera de eucalipto permitieron extraer hemicelulosas y al mismo tiempo obtener digestibilidades del sólido del 84 %. Además, el proceso “one-pot” con el líquido iónico prótico acetato de hidroxietilamonio resultó efectivo, obteniendo un rendimiento de glucosa final del 75 %. Resumen / Abstract 7 ABSTRACT Title: Single step, multi-step pretreatment and one-pot process of pine and eucalyptus wood, based on the use of aprotic, protic and biobased ionic liquids. Introduction Lignocellulosic biomass is a recalcitrant non-uniform three-dimensional material that requires pretreatment processes to deconstruct the linkages and disrupt the structure into their main compounds. Pretreatment technologies can constitute up to 40 % of the total processing costs of lignocellulosic biomass conversion. Ionic liquids are salts formed by a cation and an anion whose melting point is below 100 °C. High solvation capacity, negligible volatility, non-flammability, high conductivity, recyclability, and switchability, are some of the ionic liquids advantages against other organic solvents. Due to all these properties, ILs are alternative and promising solvents in a wide range of applications, including biomass pretreatment. During the last years, the ability of certain ionic liquids to disrupt the lignocellulosic network has been confirmed, reducing cellulose crystallinity, increasing cellulose accessibility and favoring the subsequent saccharification step. However, further research is required to ensure ionic liquid recovery, reduce costs, removing washing-steps, promote the use of non-hazardous ILs and continue 8 studying application of emerging families of ionic liquids (e.g. biobased ionic liquids) and existing ones (e.g. protic ionic liquids). Aim and results The aim of this work is to study ionic liquid based pretreatments with or without a prior autohydrolysis step, to pretreat pine and eucalyptus wood, and select the most effective pretreatment for the subsequent enzymatic hydrolysis step. The main novelty of this work is to develop three different strategies, i.e. one-step pretreatment, multi-step pretreatment or one-pot processes for a hardwood and a softwood material. These approaches have never been studied together before in literature for the same type of biomass. Apart from studying well-known aprotic ionic liquids, novel less toxic, low cost ionic liquids will be tested in this thesis. The work will be enriched by comparing hardwood vs softwood results, due to their differences in terms of structure, composition and behavior. The experimental procedure and results derived from this research are divided into three main chapters, i.e. one-step pretreatments, multi-step pretreatments and one-pot processes. Results and conclusions Firstly, autohydrolysis and ionic liquid one-step pretreatments are studied separately. The ionic liquids employed for this work can be divided into three groups: protic, aprotic and biobased ionic liquids. Furthermore, conventional and microwave heating are employed, as ionic liquids are good microwave absorbers and times employed in the pretreatment can be reduced by using microwave heating. Another key factor in biomass pretreatment with ionic liquids is their recovery and reuse capacity. In this work, an evaluation of the recovery and reuse capacity of the ionic liquid 1-allyl-3-metylimidazolium chloride is demonstrated, reusing the ionic liquid up to five cycles and analyzing its properties after the recovery and reuse. Finally, digital image analysis based on fractal dimension, lacunarity and surface holocellulose/lignin ratios calculations are proposed for the study of the enzymatic accessibility of the pretreated solid. These parameters are correlated in this work to the enzymatic digestibility of the pretreated solid. Secondly, a two-step process combining autohydrolysis + ionic liquid pretreatments is studied, aiming at hemicellulose extraction as well as enzymatic digestibility increase. Results evidence opposite behaviors in the hardwood and softwood species. Thus, autohydrolysis of pine wood limits the efficiency of the subsequent ionic liquid pretreatment. In contrast, autohydrolysis pretreatment of eucalyptus wood enhances hemicellulose extraction, while favoring the subsequent pretreatment with ionic liquids, and enhancing the saccharification. Finally, a pretreatment and saccharification process developed in the same vessel, without any solid-liquid separations and without including a washing step is studied. This alternative configuration is the so-called “one-pot” process. Results highlight that the tested protic ionic liquids Resumen / Abstract 9 are not capable of increasing pine digestibility. However, eucalyptus wood digestibility can be successfully enhanced using the “one-pot” process with the protic ionic liquid 2- hydroxyethylammonium acetate. This process prevents the need for including a washing step with water, also avoiding the pH adjustment step that is crucial for other types of ionic liquids. In the implementation of a subsequent fermentation step, hydrolysates were employed as substrate to study the growth of Rhodosporidium toruloides in the presence of this ionic liquid. Autohydrolysis of eucalyptus and pine wood led to the extraction of approximately 70 % of the total hemicelluloses present in wood. In eucalyptus wood a maximum digestibility of 72 % was achieved, although glucose yields were reduced to 38 %. Using the aprotic ionic liquid 1-ethyl-3-methylimidazolium acetate, and conventional heating, pine wood yielded the highest digestibility of the studied in this work (84 %). The yield obtained for pine is even higher than the one obtained for eucalyptus wood. Eucalyptus wood yielded higher digestibility than pine, using protic and biobased ionic liquids. Multi-step pretreatments enhanced hemicellulose extraction and synergistically improved digestibility up to 84 %. Furthermore the employment of the ionic liquid 2-hydroxyethylammonium acetate in a one-pot process yielded a final digestibility of 75 %. Líquidos iónicos en el procesado de biomasa Capítulo 1: LÍQUIDOS IÓNICOS EN EL PROCESADO DE BIOMASA Líquidos iónicos en el procesado de biomasa 13 1. LÍQUIDOS IÓNICOS EN EL PROCESADO DE BIOMASA La disminución de las reservas de combustibles fósiles, el aumento de sus precios, el crecimiento de la población y la creciente inquietud por la emisión de gases efecto invernadero y el cambio climático, están llevando a la búsqueda de combustibles alternativos (Erythropel et al., 2018). Para el año 2020, la Unión Europea se ha planteado aumentar en un 20 % la contribución de la energía de fuentes renovables, aumentando un 20 % la eficiencia energética (IDAE, 2011). Para el año 2050, el Consejo Europeo ha fijado reducir las emisiones de CO2 de un 80 a un 95 % (Hassan et al., 2019). De cara a cumplir con estos objetivos, la biorrefinería surge como una alternativa para ayudar a solucionar estos retos. 1.1. El concepto de biorrefinería La aplicación de la biorrefinería al cumplimiento medioambiental y energético, permite tanto fijar CO2 a través de la biomasa, como sustituir parcialmente los combustibles fósiles (Figura 1.1). En este contexto, la biomasa se define, de acuerdo a la Directiva Europea 2009/28/EC, como la fracción de productos biodegradables y residuos de origen biológico procedentes de la industria agrícola (incluyendo sustancias vegetales y animales), forestal e industrias relacionadas incluyendo la acuicultura, así como la fracción biodegradable de residuos municipales e industriales (De Corato et al., 2018). Capítulo 1 14 Figura 1.1: Modelo de fijación de CO2 y reemplazo de combustibles fósiles Existen numerosas definiciones para el concepto de biorrefinería, aunque todas incluyen una serie de factores comunes, como son el empleo de biomasa, la integración de procesos y las analogías con las refinerías de petróleo convencionales. En la Tabla 1.1 se recogen algunas de estas definiciones. Tabla 1.1: Ejemplos de definiciones de biorrefinerías (Bauer et al., 2017) Definición Fuente Planta de procesado en donde la biomasa se utiliza como materia prima y se convierte en una amplia gama de productos valorizables. Este concepto está basado en la refinería petroquímica Departamento de Energía (1997) Procesado sostenible de biomasa en un rango de productos de base biológica (ingredientes alimentarios, productos químicos, materiales) y bioenergía (biocombustibles, calor y energía). Agencia Internacional de la Energía (IAE) (Cherubini et al., 2007) Sistemas en los que la biomasa puede ser utilizada de manera completa, mediante la conversión a través de múltiples procesos en un número de productos valorizables (Menrad et al., 2009) Instalación multiproducto que integra procesos de conversión de biomasa y equipos para producir combustibles y productos químicos a partir de biomasa basada en madera (biorrefinería forestal) (Hämäläinen et al., 2011) Instalación que integra procesos de conversión de biomasa y equipos para producir combustible, energía y productos químicos. Laboratorio Nacional de Energías Renovables (NREL, 2015) Líquidos iónicos en el procesado de biomasa 15 1.1.1. Tipos de biorrefinería El concepto de biorrefinería, relacionado con procesos de valorización de biomasa, va a variar dependiendo de la materia prima empleada, los procesos utilizados, los productos intermedios producidos (plataformas) y los productos finales (Kamm et al., 2007; Cherubini et al., 2009). La clasificación por materia prima utilizada, permite distinguir las biorrefinerías según si se emplean cultivos energéticos (hierbas, pastos, cultivos lignocelulósicos oleaginosos o alcoholígenos) o residuos (de origen agrícola, forestal o urbano). Otro criterio para clasificar las materias primas las divide en primarias (procedentes de cosecha forestal o agraria), secundarias (procedentes de los residuos de la industria forestal) y terciarias (procedentes de los residuos generados por el consumo humano) (Cherubini et al., 2007; Cherubini et al., 2009; Cherubini, 2010). La clasificación a partir de los procesos de conversión, organiza las biorrefinerías en función de si se emplean procesos mecánicos/físicos, bioquímicos o termoquímicos. La clasificación por plataformas organiza las biorrefinerías en función de los productos intermedios que se obtienen. La conversión de estos productos intermedios, plataforma, a productos de mercado, se lleva a cabo posteriormente a través de diferentes procesos de conversión. Atendiendo a este criterio de plataforma, los tipos de biorrefinería más importantes son: de biogás, de gas de síntesis, de hidrógeno, de azúcares C6, de azúcares C5, de lignina, de pirólisis líquida, de aceites, de zumos, de electricidad y de calor (Cherubini, 2010). La clasificación por productos finales distingue dos tipos de biorrefinería: de productos en sí mismos (materiales, piensos, alimentos y productos químicos) y de energía (bioetanol, biodiesel o combustibles avanzados) (Cherubini, 2010). 1.1.2. Biorrefinería de materiales lignocelulósicos La biorrefinería de base lignocelulósica engloba los procesos de refino de biomasa lignocelulósica (madera, paja, etc.) a productos intermedios (celulosa, hemicelulosa y lignina), para procesarlos en un amplio abanico de productos y bioenergía (Cherubini et al., 2009). En la Figura 1.2 se muestran las materias primas, plataformas, productos y procesos de conversión de una biorrefinería lignocelulósica. En este tipo de biorrefinerías los materiales lignocelulósicos se tratan con diferentes disolventes para liberar celulosa, hemicelulosa y lignina. La celulosa y la hemicelulosa se convierten en azúcares C6 y C5 mediante hidrólisis enzimática. Estos azúcares se utilizan generalmente para la producción de biocombustibles por fermentación (etanol, butanol, etc.) o productos químicos de alto valor añadido (Cherubini, 2010). Capítulo 1 16 Figura 1.2: Esquema de biorrefinería de biomasa lignocelulósica Por otro lado, la lignina se aprovecha de manera térmica para la obtención de calor y energía. Además, esta lignina se puede procesar para producir compuestos químicos de alto valor añadido como los compuestos fenólicos o formar parte de los “composites” (de Jong & Jungmeier, 2015). 1.2. La madera como materia prima en biorrefinería El empleo de biomasa lignocelulósica como materia prima es más barato que otras alternativas como la biomasa agrícola, tiene una amplia disponibilidad y se puede generar en grandes cantidades. La biomasa lignocelulósica se puede clasificar en: residuos agrícolas (bagazo de caña de azúcar, paja de trigo), maderas duras (eucalipto, álamo), maderas blandas (pino, abeto), residuos de celulosa (papel de periódico, papel de oficina) y biomasa herbácea (heno de alfalfa, hierba) (Arenas- Cárdenas et al., 2016). La biomasa lignocelulósica es una fuente de energía renovable, ya que posee ciclos de generación cortos y accesibilidad a escala mundial (Shill et al., 2011a; van Spronsen et al., 2011; Van Dyk & Pletschke, 2012; Arenas-Cárdenas et al., 2016; Stand et al., 2018). De manera particular, la madera (dura y blanda) y el sector forestal posee la ventaja de no competir con la producción de alimentos y de ofrecer cultivos densamente energéticos (Octave & Thomas, 2009; Cherubini, 2010; Peleteiro et al., 2015; Hassan et al., 2019). Además, el sector forestal puede encontrar, en el aprovechamiento integral de la biomasa lignocelulósica, una alternativa más diversificada a su industria tradicional, mejorando la huella de carbono (Dessbesell et al., 2016). Líquidos iónicos en el procesado de biomasa 17 De las diferentes biomasas lignocelulósicas existentes, este trabajo se va a centrar en la madera, describiendo su estructura, composición y finalmente enfocándose en las materias primas de este trabajo: pino y eucalipto. 1.2.1. Estructura y composición de la madera La estructura de la madera y su organización se esquematiza en la Figura 1.3. Las células de la madera están compuestas, principalmente, por lumen y pared celular. El lumen es una conducción hueca utilizada para el transporte de agua. La pared celular se subdivide en tres regiones: lámina media, pared primaria y pared secundaria, estando esta última subdivida a su vez en tres subcapas (Schwarze, 2007). Figura 1.3: Estructura y organización de las células en la madera (Nishimura et al., 2018) De manera muy general, la pared celular de la madera está compuesta principalmente por celulosa, hemicelulosa y lignina. La celulosa se encuentra en la parte interna y la hemicelulosa y la lignina se encuentran entrecruzadas en la parte externa (Doherty et al., 2011; Nishimura et al., 2018). 1.2.1.1. Celulosa La celulosa es un homopolisacárido compuesto por una cadena lineal de unidades de D- glucosa, unidas por enlaces β-1,4-glucósidicos (Speight, 2011; Arenas-Cárdenas et al., 2016). Debido a su estructura polisacárida, existe una gran cantidad de grupos hidroxilo a lo largo de la cadena principal. El grado de polimerización varía desde cientos hasta miles de unidades (Sun et al., 2016c). En la Figura 1.4 se muestra la estructura de la celulosa. Capítulo 1 18 Figura 1.4: Estructura de la celulosa (Lin & Dufresne, 2014) La celulosa se agrega en haces largos y finos, denominados fibrillas elementales. La agregación se organiza en regiones con estructuras de haces muy ordenadas (regiones cristalinas) o estructuras más irregulares (regiones amorfas) (Arenas-Cárdenas et al., 2016; Sun et al., 2016b). Las fibrillas elementales se agregan a su vez para formar microfibrillas, formando estructuras difíciles de disolver en agua, alcoholes, éteres o acetona. Además, esta organización confiere a la celulosa resistencia a la hidrólisis química y enzimática (Arenas-Cárdenas et al., 2016; Elgharbawy et al., 2016). 1.2.1.2. Hemicelulosa La hemicelulosa es un heteropolisacárido formado por estructuras complejas que contiene hexosas (glucosa, manosa, galactosa), pentosas (xilosa, arabinosa) y ácidos urónicos (ácido glucurónico, ácido 4-O-metil-glucurónico y residuos de ácido galacturónico) (Speight, 2011; Arenas-Cárdenas et al., 2016). En la Figura 1.5 se muestran las estructuras constituyentes de las hemicelulosas. a) b) c) d) e) f) g) h) Figura 1.5: Estructuras de la hemicelulosa a) D-glucosa, b) D-manosa, c) galactosa, d) D-xilosa, e) D-arabinosa, f) ácido D-glucurónico, g) ácido 4-O-metil-glucurónico, h) ácido D-galacturónico (van Osch et al., 2017) Las hemicelulosas tienen una estructura aleatoria, amorfa y ramificada, con poca resistencia a la hidrólisis. Son más fáciles de degradar con productos químicos y/o calor que la celulosa y se hidrolizan fácilmente mediante ácidos a sus componentes monoméricos. Su función es mantener la Líquidos iónicos en el procesado de biomasa 19 flexibilidad, para evitar que las fibras de celulosa de la pared celular se peguen entre sí. Su estructura generalmente se ramifica en los extremos y posee cadenas laterales. Su grado de polimerización, de en torno a 200 unidades, también es considerablemente menor que el de la celulosa (Sun et al., 2016b). La composición de la hemicelulosa depende del tipo de madera. En la madera dura, las hemicelulosas se encuentran predominantemente como xilano (pentosanos), mientras que, en la madera blanda, es más común el glucomanano (hexosas: glucosa y manosa) (Akhtar et al., 2015). Los componentes hemicelulósicos se asocian con de puentes hidrógeno a la celulosa y con enlaces covalentes a la lignina (Cherubini, 2010; Akhtar et al., 2015; Arenas-Cárdenas et al., 2016) 1.2.1.3. Lignina La lignina es un heteropolímero aromático, insoluble en agua, altamente complejo, que proporciona impermeabilidad, rigidez estructural y resistencia al ataque biológico y físico (Cherubini, 2010; Prado et al., 2014; Arenas-Cárdenas et al., 2016; Dutta et al., 2017). En el caso de las maderas este compuesto es de especial importancia, ya que se encuentra en gran proporción, lo que les aporta esa resistencia y dureza extra comparada con otro tipo de biomasas. La lignina está formada por unidades de fenilpropano unidas mediante enlaces éter que forman complejos con la hemicelulosa para encapsular la celulosa, haciéndola resistente a la hidrólisis química y enzimática. En comparación con los demás componentes de la biomasa (polisacáridos), la lignina posee enlaces covalentes fuertes, lo que la hace más resistente a la disolución (Prado et al., 2014; Akhtar et al., 2016; Elgharbawy et al., 2016a; Arenas-Cárdenas et al., 2017). En la Figura 1.6 se muestra un ejemplo de la estructura de la lignina de una madera dura. Figura 1.6: Ejemplo de estructura de la lignina en madera de chopo (Vanholme et al., 2010) Las unidades de fenilpropano pueden proceder de tres tipos de alcoholes: sinapílico, coniferílico y p-cumarílico (Arenas-Cárdenas et al., 2016; Dutta et al., 2018). A partir de estos alcoholes, la lignina se forma por la eliminación de un átomo de hidrógeno fenólico, que conduce a la formación de radicales fenoxilo, que se convierten en unidades guayacilo (G), siringilo (S) y Capítulo 1 20 p-hidroxifenilo (H) (Parthasarathi et al., 2011; Prado et al., 2014). En las maderas en particular se presentan principalmente anillos guayacilo y siringilo con muy baja proporción de unidades p- hidroxifenilo (en maderas duras) o sólo guayacilo (en maderas blandas) (Evtuguin et al., 2001; Crestini et al., 2017). La unión de las unidades de la lignina se realiza mediante enlaces covalentes de tipo C-C o C-O-C. Los enlaces más importantes en la lignina y más fáciles de romper son los β-O-4 (fenilpropano β-aril éter). En menor proporción, se pueden encontrar en la lignina otros enlaces éter (C-O-C) como los α-O-4 (fenilpropano α-aril éter) y 4-O-5 (diaril éter). Finalmente, los enlaces C- C son β-5 (fenilcumaran), 5-5 (bifenilo y dibenzodioxocin), β-1 (1,2-diaril propano) y β-β (resinol), más resistentes y difíciles de romper (Chakar & Ragauskas, 2004). 1.2.1.4. Extractos y cenizas Los extractos son componentes minoritarios en la madera, que suponen entre un 1 y un 10 % de la composición total, de naturaleza muy variada. Entre los extractos más comunes se incluyen quinonas, grasas, aceites, terpenos, ceras y compuestos fenólicos. Son uno de los compuestos más fáciles de separar del material lignocelulósico, utilizando disolventes orgánicos como etanol, acetona o diclorometano, o agua (Sjostrom, 1993). Las cenizas son sustancias remanentes en la madera tras haber sido incinerada y suponen entre un 0,2 y un 0,5 % de la composición total. Entre las cenizas se encuentra el potasio, el calcio, el magnesio o el sodio (Hofer, 2009). 1.2.2. Tipos de maderas En la Tabla 1.2 se muestra el contenido en celulosa, hemicelulosa y lignina presente en distintas maderas. Tabla 1.2: Composición de maderas duras y blandas (Windeisen & Wegener, 2009) Especie Celulosa (%) Hemicelulosa (%) Lignina (%) Extractos (%) Maderas duras Acer rubrum 42,0 28,9 25,4 3,2 Fagus sylvatica 39,4 33,3 24,8 1,2 Acacia mollissima 42,9 33,6 20,8 1,8 Maderas blandas Picea abies 41,7 28,3 27,4 1,7 Larix sibirica 41,4 29,6 26,8 1,8 Abies balsamea 38,8 28,5 29,1 2,7 Líquidos iónicos en el procesado de biomasa 21 Las maderas se pueden clasificar en dos grandes grupos: maderas duras (o angiospermas) y maderas blandas (o coníferas) (Ek et al., 2009). La proporción de cada compuesto en la madera varía según la especie y en función de si son duras o blandas. En el presente trabajo, se van a emplear una madera dura, Eucalyptus globulus, y una madera blanda, Pinus radiata. A continuación, se van a describir algunas de sus características. 1.2.2.1. Madera dura: Eucalyptus globulus Las maderas duras tienen mayor diversidad de células que las maderas blandas y poseen hojas caducas (Sjostrom, 1993). Las hemicelulosas de la madera dura contienen, principalmente, 4- O-acetil-4-O-metilglucuronoxilano (10-35 %) (Barsett et al., 2005; Asmadi et al., 2017). Además poseen un contenido considerable de grupos O-acetil y del resto ácido 4-O-metilglucurónico mayor que el que se da en maderas blandas. La lignina de la madera dura está compuesta tanto por grupos guayacilo como por grupos siringilo (Asmadi et al., 2017). Poseen vasos conductores alargados por los que circula el agua o la savia del árbol, y poseen fibras que actúan como soporte (Sjostrom, 1993). Esto dota a las maderas duras de un aspecto fibroso y alargado. Además, poseen unas células híbridas denominadas traqueidas, que cumplen tanto la función de conducción del agua y la savia, como de soporte (Sjostrom, 1993). La madera de eucalipto es una de las más comunes en las regiones templadas, cubriendo una superficie en 2016 de 1,3 millones de hectáreas en Europa (un 80 % en la Península ibérica) (Cerasoli et al., 2016; Kandhola et al., 2017; Wang et al., 2017). La especie Eucalyptus globulus, también conocida en inglés como “Tasmanian blue gum” es un árbol de hoja ancha capaz de crecer hasta 70 metros de altura, siendo la especie más alta en Europa. Esta especie es originaria del sureste Australiano, introduciéndose en Europa (principalmente por el noroeste de España y Portugal) a medidados del siglo XIX para la obtención de pasta y papel a nivel industrial (Cerasoli et al., 2016). Actualmente, es una de las fuentes de materia prima principales para la fabricación de la pasta de papel en España (ASPAPEL, 2018). 1.2.2.2. Madera blanda: Pinus radiata La estructura de las maderas blandas está formada principalmente por traqueidas (en un 90- 95 %) y células radiales (en un 5-10 %). Las traqueidas aportan el sostén mecánico y permiten que el agua fluya a las distintas partes de la planta. Por su parte, las células radiales son mucho más pequeñas y componen el sistema horizontal del xilema y floema (Sjostrom, 1993). Las hemicelulosas de las maderas blandas se componen de 4-O-acetil-galactoglucomanano (10-15 %) y de arabino-4-O-metilglucuronoxilano (5-10 %) (Barsett et al., 2005; Asmadi et al., 2017). La madera de pino es la más abundante en el hemisferio norte (Wang et al., 2017). La especie Pinus radiata, también conocida como pino insigne o pino de Monterrey, es originaria de Capítulo 1 22 California, pero fue exportada a Australia, Nueva Zelanda, América del Sur y Sur de Europa. El crecimiento y la expansión de esta especie se ha dado sobre todo en la última mitad del siglo XX, pasando de 650.000 hectáreas a 4 millones de hectáreas nivel mundial en el 2011. En España, en el año 2006 la superficie de Pinus radiata era de 287.000 hectáreas (Mead, 2013). 1.2.3. Aprovechamiento de la madera en biorrefinería De cara al aprovechamiento de la madera en una biorrefinería, se plantean dos estrategias de procesado: el pretratamiento o el fraccionamiento. Aunque son dos conceptos diferentes, pueden combinarse y utilizar el mismo tipo de tecnologías, en función de si el fin es preparar el material de cara a una etapa posterior (pretratamiento) o separar uno o varios de los compuestos principales de la madera (fraccionamiento). El proceso convencional de aprovechamiento de la madera y de cualquier material lignocelulósico se lleva a cabo en cuatro etapas (Figura 1.7). En primer lugar, se pretrata el material para preparar el sólido para una etapa posterior de hidrólisis enzimática, en la que la celulosa se convierte a glucosa. Posteriormente, los hidrolizados obtenidos se llevan a una etapa de fermentación, para obtener productos como alcoholes (etanol, butanol,etc) u otras moléculas (biocombustibles avanzados) (Lee et al., 2009). Finalmente, se emplean tecnologías de separación, para la recuperación y aislamiento del producto principal de la fermentación. Figura 1.7: Esquema de etapas de aprovechamiento lignocelulósico El pretratamiento se integra en el proceso con el objetivo de facilitar la etapa de hidrólisis mediante la ruptura/eliminación de lignina, el aumento de la porosidad y la disminución de la cristalinidad de la celulosa. De manera particular, la madera presenta resistencia a las enzimas debido a sus características estructurales, físicas y químicas. Es por ello que en este tipo de materiales, la implementación de un pretratamiento eficaz que mejore la etapa de hidrólisis enzimática posterior es indispensable (Maitan-Alfenas et al., 2015; Sun et al., 2016). En la Figura 1.8 se esquematiza cómo la biomasa se pretrata rompiéndose la estructura. Líquidos iónicos en el procesado de biomasa 23 Figura 1.8: Pretratamiento de biomasa lignocelulósica (Kumar et al., 2009) Un pretratamiento eficaz debe evitar la reducción de biomasa, mejorar la hidrólisis enzimática, minimizar la degradación de los azúcares derivados tanto de celulosa como de hemicelulosa, prevenir la pérdida o degradación de carbohidratos, evitar la formación de inhibidores para posteriores procesos de hidrólisis y fermentación, permitir la recuperación de lignina y ser rentable económicamente (Zhao et al., 2009; Alvira et al., 2010; Asakawa et al., 2015; Arenas- Cárdenas et al., 2017; Bhutto et al., 2017). Con un fraccionamiento efectivo se puede lograr una utilización eficiente de la biomasa y desarrollar procesos económicos, robustos y de confianza para una biorrefinería basada en los tres componentes mayoritarios: celulosa, hemicelulosa y lignina (Peleteiro et al., 2015; Mussatto, 2016). El fraccionamiento de la madera supone un paso clave en la conversión de biomasa lignocelulósica en combustibles, productos químicos y materiales valiosos. 1.2.3.1. Procesos de aprovechamiento de la madera en biorrefinería En función del aprovechamiento que se pretende dar, existen una gran cantidad de procesos que pretratan y/o fraccionan el material lignocelulósico. Estos procesos se pueden clasificar en (Akhtar et al., 2015):  Procesos físicos. El objetivo es incrementar el área superficial de los materiales lignocelulósicos, reduciendo el tamaño de partícula y alterando su estructura. La gran desventaja de los pretratamientos físicos es su alta demanda de energía (Sun et al., 2016b). Dentro de estos pretratamientos se encuentran: los pretratamientos mecánicos (astillado, molido, fresado o extrusado) y los pretratamientos por irradiación (irradiación de rayos gamma y haces de electrones o pretratamiento de campo eléctrico pulsado) (Akhtar et al., 2015). Capítulo 1 24  Procesos químicos. Enfocados hacia el pretratamiento, buscan mejorar la digestibilidad de la celulosa mediante la eliminación de lignina y/o hemicelulosa, así como reducir la cristalinidad de la celulosa (Behera et al., 2014). Si se pretende el fraccionamiento, permiten aislar algunos componentes mayoritarios de la biomasa lignocelulósica. Algunos de los pretratamientos químicos más empleados son: el pretratamiento ácido, la hidrólisis alcalina, el proceso organosolv, la oxidación y los líquidos iónicos (LIs) (Baruah et al., 2018; Ghosh et al., 2018). En esta tesis doctoral se va a estudiar en profundidad el empleo de LIs para pretratar el material lignocelulósico.  Procesos físico-químicos. Combinan tanto procesos químicos como físicos para mejorar la accesibilidad de las enzimas a la celulosa. Dentro de este grupo se encuentran: la autohidrólisis, el pretratamiento con amoníaco, el pretratamiento con fluidos supercríticos y la explosión con vapor entre otras (Sun et al., 2016c). En esta tesis doctoral se va a utilizar el pretratamiento de autohidrólisis para la extracción de hemicelulosas de la madera.  Procesos biológicos. Son aquellos que emplean hongos, capaces de generar enzimas que degradan la hemicelulosa, la lignina o algunas sustancias derivadas de la lignina (compuestos fenólicos). Los hongos más empleados son los de podredumbre blanca, entre los que destaca Phanerochaete crysosporium, Ceriporia lacerata, Cyathus stercoletus, Ceriporiopsis subvermispora, Pycnoporus cinnabarinus, o Pleurotus ostreatus (Sun et al., 2016b). Uno de los grandes inconvenientes de estos procesos son los largos tiempos de pretratamiento (10-14 días) (Singh et al., 2016). El pretratamiento ácido, el alcalino, la explosión con vapor y la autohidrólisis producen una gran eliminación de hemicelulosas. Por otro lado, el pretratamiento alcalino, la deslignificación oxidativa, el organosolv, el pretratamiento con LIs y el biológico permiten una extracción eficaz de la lignina. Muchos de los pretratamientos disponibles permiten un aumento del área superficial, ya sea producido por el hinchamiento de las fibras o por la disminución del tamaño de partícula. Uno de los principales inconvenientes de algunos pretratamientos, como los tratamientos ácidos, la autohidrólisis o la explosión con vapor, es la formación de inhibidores debido a la degradación del material lignocelulósico. Esto hace necesario incorporar etapas de detoxificación en los procesos. En la Tabla 1.3 se muestran los principales efectos que producen algunos de los procesos más utilizados sobre la biomasa lignocelulósica. Se ha identificado como alto (A), medio (M) y bajo (B), el efecto que produce sobre cada una de las propiedades que aparecen en la Tabla los procesos más empleados sobre biomasa lignocelulósica. Líquidos iónicos en el procesado de biomasa 25 Tabla 1.3: Efectos principales de los procesos (Sun et al., 2016c) Eliminación lignina Eliminación hemicelulosa Aumento área superficial Reducción cristalinidad celulosa Aumento porosidad Generación inhibidores Molienda – – A A – B Alcalino A A A – A B Ácido M A A – M A Oxidativo A – – – A B Organosolv A B A – M B LIs M B A A A B Explosión con vapor B A A B A A Autohidrólisis M A M – M A AFEX M B A A A B SC-CO2 B B A – A B Biológico A M A – A B Alcalino – – A A – B A= efecto alto, M= efecto medio y B= efecto bajo Como se puede observar, no existe un proceso universal que permita maximizar todos los efectos, por lo que atendiendo al material empleado y al objetivo que se pretende alcanzar, se debe de buscar el tratamiento más adecuado. 1.2.3.2. Procesos multietapa para el aprovechamiento integral de la biomasa lignocelulósica La combinación de pretratamientos físicos y químicos permite potenciar las ventajas de cada uno de los pretratamientos individuales. Además, los procesos multietapa pueden mejorar la digestibilidad y maximizar la recuperación de lignina y hemicelulosa. La combinación de pretratamientos puede tener un enfoque multiobjetivo, de manera que se consigue fraccionar la madera a la par que se pretrata el material. Otras aplicaciones van orientadas a la obtención de un producto en concreto, como la celulosa nanocristalina, con altos rendimientos (Kassaye et al., 2016). Por último, la combinación de pretratamientos permite también solventar algunos de los problemas que alguno de los pretratamientos individuales pueden plantear a ciertos tipos de biomasas, como es el caso de la recondensación de las lignina en el pino, lo que se puede Capítulo 1 26 evitar al utilizar una combinación de pretratamientos alcalino seguido de ácido diluido (Das et al., 2019). En la Tabla 1.4 se muestran algunos de los trabajos más representativos realizados en los últimos años mediante combinación de pretratamientos, indicando la materia prima, la combinación de pretratamientos empleada y los principales resultados obtenidos. Tabla 1.4: Trabajos realizados sobre pretratamientos multietapa Materia prima Secuencia de pretratamiento Resultados Referencia Paja de arroz Amoníaco + LI Se recupera el 82 % de celulosa y se obtiene una digestibilidad del 97 % (Nguyen et al., 2010) Eucalyptus globulus Autohidrólisis + organosolv Se extrae una corriente rica en hemicelulosa, otra en lignina pura y un sólido rico en celulosa hidrolizable enzimáticamente (Romaní et al., 2011) Fibra de fruto de palma Ácido + alcalino Se obtiene un sólido con un 82 % de celulosa, un 30 % de lignina y menos de un 1 % hemicelulosa (Kim et al., 2012) Abedul Autohidrólisis + LI Se obtiene un sólido rico en celulosa con 13-15 % de lignina y una fracción de lignina pura (Hauru et al., 2013) Bagazo de caña de azúcar Ácido + LI Se obtiene una digestibilidad enzimática del sólido del 95,5 % (Jiang et al., 2013) Eucalyptus globulus Autohidrólisis + hongos Mejora el rendimiento de la hidrólisis 8,5 veces (Martín- Sampedro et al., 2015) Rastrojo de maíz Ácido + alcalino Se recupera un sólido que contiene un 81 % de glucano y un 6 % de lignina en la fase sólida (Lee et al., 2015) Angelim Vermelho Explosión por vapor + LI Se extrae un 47 % de celulosa en forma de nanocristales (Abushammala et al., 2016) Abedul AH + LI Se disuelve completamente la madera a tamaño de partícula de hasta 6 mm (Deb et al., 2016) Abeto LI + alcalino Aumenta la conversión a azúcares hasta un 50 % (Heggset et al., 2016) Celulosa microcristalina Ácido + LI Se consigue un 92 % de azúcares reductores totales (Kassaye et al., 2016) Paja de arroz Explosión por vapor + LI Se incrementa el rendimiento a glucosa hasta un 74 % (Liu et al., 2016) Líquidos iónicos en el procesado de biomasa 27 Tabla 1.4 (continuación): Trabajos realizados sobre pretratamientos multietapa Materia prima Secuencia de pretratamiento Resultados Referencia Deshecho de caña de azucar FeCl2 + alcalino Se obtiene un rendimiento del 79,6 % de azúcares reductores (Raghavi et al., 2016) Abeto, sauce y paja de soja LI + Na2SiO3 En el sauce, se obtiene un rendimiento de hidrólisis del 98,6 % (Sun et al., 2016d) Eucalyptus grandis Autohidrólisis + alcalino Se obtiene un sólido con un 58,8 % de glucano y un 28,3 % de lignina (Cebreiros et al., 2017) Residuos del café Ácido concentrado + acetona Elimina lignina y hemiceluolsa (Ravindran et al., 2017) Paja de arroz Presión ultra alta + microemulsiones de LI Se obtiene un rendimiento a azúcares reductores del 89,6 %, se extrae un 66 % de la lignina y se disminuye en un 33 % la cristalinidad del sólido (Gao et al., 2018) Tallo de girasol Alcalino + LI Se incrementa la digestibilidad en un 163 mg/g biomasa (Nargotra et al., 2018) Pino Álcali-ácido diluido Se rompen los enlaces β-O-4 de la lignina y se forman unidades G monolignol (Das et al., 2019) Rastrojo de maíz Ácido clorhídrico + oxidación húmeda alcalina Se recupera el 83 % de xilano, se extrae el 86 % de lignina y se consigue una digestibilidad del 71 % (An et al., 2019) En esta tesis doctoral se va a aplicar la combinación de pretratamientos de autohidrólisis y LIs, con el objetivo principal de poder separar la fracción hemicelulósica en una primera etapa para su valorización y mejorar la accesibilidad del material utilizando líquidos iónicos. 1.2.3.3. Hidrólisis enzimática (HE) La etapa de hidrólisis enzimática (HE) tiene como objetivo la conversión de la fracción de celulosa y hemicelulosa en sus azúcares monoméricos constituyentes. La celulosa se hidroliza con un conjunto de enzimas llamadas celulasas, que pueden ser endoglucanasas, exoglucanasas o celobiohidrolasas y β-glucosidasas (Balat, 2011). Todas estas enzimas actúan conjuntamente para convertir la celulosa en glucosa, según se muestra en la Figura 1.9. Capítulo 1 28 Figura 1.9: Acción de las enzimas celulolíticas (Balat, 2011) Las endoglucanasas escinden enlaces glucosídicos dentro de microfibrillas de celulosa, actuando preferentemente en regiones de celulosa amorfa accesibles. Estas enzimas fragmentan las cadenas de celulosa para generar extremos reactivos para las celobiohidrolasas, que corta las cadenas de celulosa en los extremos con la finalidad de liberar celobiosa o glucosa soluble. La celobiosa, a altas concentraciones, inhibe la actividad de las celobiohidrolasas. Por lo tanto, a menudo se requieren β-glucosidasas, que convierten la celobiosa en glucosa para reducir la inhibición del producto final cuando se acumula celobiosa (Waldron, 2010; Balat, 2011). La hidrólisis de la celulosa catalizada por enzimas está controlada por dos etapas: la adsorción de enzimas en partículas celulósicas y la formación de complejos enzima-sustrato (ES). Los factores que afectan a la HE se dividen en dos grupos: factores relacionados con las enzimas y factores relacionados con los sustratos. Los factores que afectan a la enzima en la realización de la HE pueden ser la inhibición por productos, el pH y la estabilidad térmica (Hendriks & Zeeman, 2009). Las temperaturas y pH óptimos de operación para trabajar con celulasas varían entre los 40 y los 50 °C y pHs de entre 4 y 5. Los factores relacionados con el sustrato incluyen la presencia de hemicelulosa y lignina, la cristalinidad de la celulosa, el grado de polimerización y las superficies internas y externas de la celulosa (Elgharbawy et al., 2016;Karimi & Taherzadeh, 2016). Las características que deben tener las celulasas para llevar a cabo la hidrólisis enzimática son: poseer elevada actividad específica (disminuye la cantidad de enzima requerida), poseer alta eficiencia catalítica frente a la cristalinidad de la celulosa (reduce los tiempos de la hidrólisis), tener alta termoestabilidad (permite operar a mayor temperatura y se produce un aumento de la velocidad de la reacción), ser resistentes a la inhibición del producto final (produce mezclas más concentradas de azúcares) y ser estables contra fuerzas de cizallamiento (útil cuando se necesita agitación para tener en suspensión la celulosa sólida en el reactor) (Shuddhodana et al., 2016). Líquidos iónicos en el procesado de biomasa 29 1.3. Líquidos iónicos en el pretratamiento de biomasa lignocelulósica Los LIs son compuestos formados por un catión generalmente orgánico y un anión orgánico o inorgánico, cuyo punto de fusión es inferior a 100 °C (Brandt et al., 2013). Estos compuestos tienen propiedades físico-químicas únicas, como la capacidad de solvatación, la naturaleza no volátil (baja presión de vapor), la no inflamabilidad, el amplio rango electroquímico, la alta conductividad iónica, la reciclabilidad o el amplio rango en estado líquido y la buena capacidad de disolución, entre otras (Olivier-Bourbigou et al., 2010). De este modo, los LIs se han aplicado en muchos campos como la electroquímica, la catálisis heterogénea, la bio-catálisis, los procesos de extracción, etcétera. Las oportunidades que surgen en torno a la aplicación de los LIs en el procesado de biomasa radica en que algunos de ellos son capaces de extraer la hemicelulosa y la lignina de la biomasa, reduciendo a la vez la cristalinidad de la celulosa (Li et al., 2010). La aplicabilidad de los LIs a cualquier tipo de biomasa es muy amplia, e incluye maderas tanto duras como blandas (Sun et al., 2009). Además, en los últimos años también se ha observado cómo se pueden utilizar los LIs en presencia de complejos enzimáticos como proceso alternativo a los procesos de pretratamiento seguidos de hidrólisis enzimática y fermentación que se vienen empleando con otros pretratamientos (Wahlström & Suurnäkki, 2015). 1.3.1. Clasificación de los líquidos iónicos Los LIs se pueden clasificar atendiendo a su naturaleza, el tipo de cationes/aniones o su hidrofobicidad/hidrofilicidad. De manera general, los LIs se clasifican en cuatro tipos: próticos, apróticos, inorgánicos o solvo-quelatos (Angell et al., 2012). En los últimos años han aparecido nuevas familias de LIs con características interesantes como la biocompatibilidad o la capacidad de ser destilados, tales como los “conmutables”, destilables, o bioderivados (Domínguez de María, 2014). De todos estos grupos, los más estudiados en el procesamiento de biomasa han sido los LIs apróticos y los bioderivados. Los líquidos iónicos apróticos engloban la mayoría de los LIs existentes. Suelen estar compuestos por cationes amonio, imidazolio, piridinio o pirrolidinio entre otros. En el caso de la aplicación de LIs a procesamiento de biomasa, destacan los LIs derivados del catión imidazolio, como el [Amim][Cl] y el [Emim][OAc] (Figura 1.10) (Angell et al., 2012). a) b) Figura 1.10: Líquidos iónicos apróticos: a) [Amim][Cl] y b) [Emim][OAc] Capítulo 1 30 Los líquidos iónicos próticos son aquellos que presentan un protón, ya sea en el catión o en el anión (o en ambos). Se sintetizan de forma sencilla a través de una reacción de transferencia de un protón entre un ácido y una base de Brønsted-Lowry (Rocha et al., 2017). En este tipo de LIs destaca por su sencillez de síntesis (Achinivu, 2014; Rocha et al., 2017). Algunos de los LIs empleados en el pretratamiento de biomasa son [Mim][Cl] o [OHEtAm][HCO2] que han sido probados en algunos materiales como el algodón en el caso de [OHEtAm][HCO2] o la madera de pino en el caso del [Mim][Cl](Figura 1.11) (Cox & Ekerdt, 2013; Haykir et al., 2013). a) b) Figura 1.11: Líquidos iónicos próticos: a) [Mim][Cl] y b) [OHEtAm][HCO2] Para superar los problemas económicos, tóxicos y ambientales que presentan los LIs basados en el catión imidazolio, ha surgido una nueva generación de LIs que contienen iones derivados de bases naturales (por ejemplo, colina) y ácidos (por ejemplo, aminoácidos y ácidos carboxílicos). Estos LIs contienen cationes de colina combinados con aniones basados en aminoácidos ([Ch][AA]) o aniones basados en ácidos carboxílicos ([Ch][CA]). Estos LIs son más baratos, como consecuencia de los menores costes del material de partida del catión, más biocompatibles y más biodegradables en comparación con los LIs basados en el catión imidazolio. Además, los materiales de partida utilizados para producirlos son renovables (Ninomiya et al., 2013; An et al., 2015; Asakawa et al., 2016). Algunos de los LIs empleados en el pretratamiento de biomasa son [Ch][Lys] o [Ch][OAc] utilizados sobre todo con biomasas herbáceas y más recientemente en el caso de [Ch][Lys] en maderas (Figura 1.12) (Sun et al., 2017; Dutta et al., 2018; Sundstrom et al., 2018). a) b) Figura 1.12: Líquidos iónicos bioderivados: a) [Ch][Lys] y b) [Ch][OAc] 1.3.2. Propiedades de un líquido iónico para ser empleado como disolvente en el pretratamiento de madera Existe un gran número de LIs que ha ido aumentando a lo largo de los años, pero la resistencia de la madera y la complejidad de los procesos de pretratamiento, hidrólisis y fermentación, hacen que sea baja la proporción de LIs que puede ser utilizada para este fin. Para que un líquido iónico pueda ser empleado en el pretratamiento de biomasa, debe cumplir las características mostradas en la Figura 1.13. Líquidos iónicos en el procesado de biomasa 31 Figura 1.13: Características que deben cumplir los LIs para ser utilizados en el pretratamiento de biomasa La Figura 1.13 organiza las características que debe poseer el LI en función de su importancia. Así, la principal característica de un LI para ser utilizado en el pretratamiento de la biomasa es que sea capaz de interaccionar con la celulosa, hemicelulosa y lignina. A continuación, se comenta cada una de estas características. 1.3.2.1. Parámetros de Kamlet-Taft: Interacción del LI con los componentes de la biomasa La principal característica que debe tener un LI para pretratar un material lignocelulósico está relacionada con su interacción con los componentes de la misma: celulosa, hemicelulosa y lignina, de manera que pueda disolverlos. Existen tres modelos que predicen la capacidad de solvatación de los LIs: el modelo de solubilidad de Hansen, el modelo COSMO-RS86 y el modelo de polaridad de Kamlet-Taft, siendo este último el más empleado (Sun & Cheng, 2002; Badgujar & Bhanage, 2015). El modelo de Kamlet-Taft implica tres parámetros: α (dador de puentes de hidrógeno o acidez de puentes de hidrógeno), β (aceptor de puentes de hidrógeno o basicidad de puentes de hidrógeno) y π* (polarizabilidad o dipolaridad). Estos 3 valores se calculan por espectroscopía de UV-VIS en presencia de tintes solvatocrómicos (El Seoud et al., 2007). Se ha observado que los LIs que tienen aniones como el acetato, el cloruro o el dimetilfosfato poseen valores de β ≥ 0,82, y son capaces de romper los puentes de hidrógeno de la celulosa, proporcionando una mayor disolución de la misma. Sin embargo, los LIs con valores más bajos de β no son capaces de disolver la celulosa (Badgujar & Bhanage, 2015). Por otro lado, valores altos de β-α dan lugar también a la disolución Coste Biocompatibilidad Recuperación del LI y propiedades físiscas Interacción del LI con los componentes de la biomasa: celulosa, hemicelulosa y lignina Capítulo 1 32 de la celulosa, y su posterior precipitación con la adición de un antidisolvente, como por ejemplo el agua (Doherty et al., 2010). A modo de ejemplo, dos de los LIs más estudiados para pretratar biomasa han sido los derivados del catión imidazolio, como el acetato de 1-etil-3-metilimidazolio ([Emim][OAc]) o el cloruro de 1-butil-3-metilimidazolio ([Bmim][Cl]). Los valores de α para el [Emim][OAc] y [Bmim][Cl] son 0,47 y 0,44, respectivamente, y los valores de β para el [Emim][OAc] y [Bmim][Cl] son 1,14 y 0,84 (Fang, 2014; Sun et al., 2014). Estos valores han dado lugar a que tanto el [Emim][OAc] como el [Bmim][Cl] sean dos de los LIs que más eficazmente pretratan cualquier tipo de material. 1.3.2.2. Recuperación del LI y propiedades físicas La baja o nula presión de vapor de los LIs les aporta una ventaja que, a priori, los convierte en potencialmente reutilizables en este proceso. Existen diferentes métodos para la recuperación de LIs como la destilación, la separación por membranas, la extracción o la adsorción. La destilación es el método más simple para la separación de compuestos del LI, debido a la baja presión de vapor y alta estabilidad térmica de los LIs, lo que permite separar fácilmente los compuestos más volátiles. Debido a sus altos costes energéticos, el proceso de destilación se utiliza como etapa final en gran parte de los estudios de recuperación y reciclado de LI (Mai et al., 2014). La separación por membranas tiene como principales ventajas su menor consumo energético y de disolventes, lo que la haría una tecnología aplicable a escala industrial de no ser por el elevado coste de las membranas y su vida útil. En comparación con la destilación, las membranas permiten separar compuestos poco volátiles, operación compleja mediante destilación, debido a la baja presión de vapor de los LIs. Los LIs, al estar compuestos por iones, pueden ser separados empleando membranas de nanofiltración (Mai et al., 2014). La extracción es un método utilizado frecuentemente para la recuperación y reutilización de LIs. Diversos disolventes, incluyendo el agua, son inmiscibles con algunos LIs. La formación de fases separadas, hace posible la extracción o separación de diferentes compuestos de los LIs (Mai et al., 2014). La adsorción permite utilizar adsorbentes para recuperar los LIs de soluciones acuosas. (Mai et al. 2014) han investigado la separación de LI de corrientes acuosas mediante sólidos como el carbón activado, la sílice o la alúmina. En ese estudio, sus autores han demostrado que estos sólidos son más eficientes en la separación de compuestos de bajo peso molecular y no polares, en comparación con especies iónicas y de gran tamaño como los LIs. En bibliografía se encuentran diferentes estudios relacionados con la reutilización de diferentes LIs para el pretratamiento de diferentes tipos de biomasa, como se muestra en la Tabla 1.5. Líquidos iónicos en el procesado de biomasa 33 Tabla 1.5: Trabajos realizados sobre reciclado de LI en el procesado de biomasa Materia prima LI Recuperación Nº ciclos Referencia Rastrojo de maíz [Emim][OAc] Eliminación de etanol y acetona (antidisolventes) por destilación 1 (Dibble et al., 2011) Rastrojo de maíz Miscanthus giganteus [Emim][OAc] Adición de sal de fosfato para formar un sistema trifásico S-L-L 3 (Shill et al., 2011) Celulosa de algodón [Emim][OAc] [Emim][MeOPO2] Destilación a vacío 5 (Auxenfans et al., 2012) Celulosa [Bmim][Cl] Destilación a vacío 5 (Lozano et al., 2012) Celulosa [Amim][Cl] Destilación a vacío 5 (Huang et al., 2013) Pino [Bmim][Cl] Electrodiálisis 1 (Trinh et al., 2013) Serrín de roble y abeto [Emim][OAc] Destilación a vacío 7 (Auxenfans et al., 2014) Celulosa y lignina [Emim][OAc] Destilación 1 (Castro et al., 2014) Paja de arroz [Emim][OAc] Destilación a vacío 10 (Weerachan chai & Lee, 2014) Paja de arroz [Ch][OAc], [Ch][Lys], [Ch]Gly], [Ch][Arg] y [Ch][Glc] Neutralización + destilación a vacío + metanol +filtración + destilación 8 (An et al., 2015) Bagazo [Ch][OAc] Destilación a vacío 5 (Ninomiya et al., 2015a) Poda de manzano [Bmim][MeSO4] Destilación a vacío 4 (Prado et al., 2015) Eucalyptus globulus [Amim]Cl] [Bmim][OAc] Destilación a vacío 4 (Xu et al., 2017) Bagazo [Ch][OAc] Liofilización 3 (Hamidah et al., 2018) Bagazo [Me(NH2)(CH2)2OH] [OAc] Destilación 0 (Pin et al., 2019) Capítulo 1 34 Muchos de los estudios se basan en el empleo del LI [Emim][OAc] recuperado mediante procesos de destilación a vacío y con biomasas herbáceas o componentes aislados presentes en la biomasa (celulosa microcristalina, hemicelulosa y lignina). Además, la evaluación de la capacidad de reutilización del LI se suele realizar en base a la digestibilidad enzimática del sólido pretratado. En este trabajo, este estudio se lleva a cabo para evaluar la capacidad de reutilización del LI [Amim][Cl] con madera de pino y eucalipto. La metodología que se va a seguir está enfocada en evaluar en primer lugar la capacidad del LI para disolver el material lignocelulósico y posteriormente su digestibilidad enzimática, como consecuencia de emplear LI reutilizado en el proceso en sucesivos ciclos. 1.3.2.3. Biocompatibilidad del LI: procesos “one-pot” Los LIs empleados deberían de ser biocompatibles con las etapas posteriores del proceso: hidrólisis enzimática y fermentación (Erythropel et al., 2018). La biocompatibilidad implica la tolerancia de los microorganismos o enzimas presentes en el proceso con el sistema a tratar, que en este caso contendría el LI. Uno de los mayores obstáculos para llevar a cabo la sacarificación de biomasa pretratada con LI radica en que las enzimas se desactivan incluso en presencia de bajas concentraciones de ciertos LIs. Por esta razón, la etapa de lavado del sólido pretratado con LIs resulta esencial para asegurar la completa eliminación del mismo. Hasta la fecha, se han barajado tres alternativas posibles para aumentar la biocompatibilidad de las enzimas con LIs: encontrar enzimas que sean tolerantes a los LIs, estabilizar y activar las enzimas en presencia de LIs o buscar nuevos LIs que toleren las enzimas (Chen et al., 2017). Todos estos procesos tienen como fin último la intensificación del proceso llevando a cabo las etapas de pretratamiento e hidrólisis enzimática en un mismo equipo. Este nuevo esquema de proceso se conoce como “one-pot” o hidrólisis enzimática in situ. Para llevar a cabo el proceso “one-pot”, es necesario incluir una etapa intermedia en la cual se ajusta el pH de la mezcla, para hacerlo adecuado para las enzimas (Hou et al., 2017). Este proceso se ha venido desarrollando en los últimos 10 años, siendo el primer trabajo encontrado el realizado por Kamiya et al. (2008), en donde se evaluó la inhibición de las enzimas celulasas en presencia de diferentes concentraciones de LI. En la Tabla 1.6 se muestran algunos de los trabajos más relevantes realizados. Líquidos iónicos en el procesado de biomasa 35 Tabla 1.6: Trabajos realizados sobre el estudio de procesos “one-pot” Materia prima LI Complejo enzimático Resumen Referencia Celulosa Avicel [Emim] [DEP] Celulasa de T. reesei Las enzimas no se inactivan en una disolución de LI:agua que contiene 25 % de LI (Kamiya et al., 2008) Celulosa Avicel Rastrojo de maíz [Emim] [OAc] Enzimas hidrotermófilas Las enzimas se inactivan en disoluciones con concentración de [Emim][OAc] > 15 % (Datta et al., 2010) Hierba [Emim] [OAc] Enzimas termoestables Con disoluciones que contienen 10 % de LI, se consigue una digestibilidad del 81 % (Shi et al., 2013) Carboxime- tilcelulosa [Dmim] [DMP] Enzimas inmovilizadas en quitosano La mayor actividad de la enzima se presenta a concentraciones > 25 % de LI (Su et al., 2013) Papel Wattman [Bmim][Cl] Celulasa halofílicas En disoluciones al 20 % de LI se consigue un 50 % de actividad de la enzima (Gunny et al., 2014) Chopo LIs derivados de imidazolio Celluclast 1.5 L No se consigue mantener la actividad de la enzima para desarrollar el proceso “one- pot” (Ungurean et al., 2014) Celulosa microcrista- lina [Emim] [OAc] [Dmim] [DMP] Celulasa de T. reesei En disoluciones con más de un 50 % de LI no se liberan azúcares. El [Emim][OAc] daña más a las enzimas que el [Dmim][DMP] (Wahlström, 2014) Paja de arroz [Emim] [DMP] Celulasa de A. fumigatus Se demuestra la sacarificación de paja de arroz en disoluciones con un 25 % de LI (Xu et al., 2014) Bagazo [Ch][OAc] Cellic CTec2® Se alcanza una digestibilidad del 80 % con un 10 % de [Ch][OAc] (Ninomiya et al., 2015b) Paja de arroz [Emim] [DMP] Celulasa encapsulada en alginato La celulasa mantiene el 76 % de su actividad en disoluciones del 40 % de LI (Xu et al., 2015a) Paja de arroz [Emim] [DMP] Celulasa de T. aureoviride Se alcanza una digestibilidad del 61 % en disoluciones que contienen un 15 % de LI (Xu et al., 2015b) Capítulo 1 36 Tabla 1.6 (continuación): Trabajos sobre el estudio de procesos “one-pot” Materia prima LI Complejo enzimático Resumen Referencia Hierba [Ch][Lys] Cóctel Cellic CTec2/HTec2 Se alcanza una digestibilidad del 87 % en disoluciones que contienen un 10 % de LI (Sun et al., 2016a) Rastrojo de maíz [Ch][Lys]: H2O Cóctel Cellic CTec2/HTec2 Se alcanza una digestibilidad del 75 % en disoluciones que contienen un 10 % de LI (Xu et al., 2016a) Paja de arroz [Bmim][Cl] Celulasa inmovilizada con polietilenglicol Se alcanza una digestibilidad del 85 % en disoluciones que contienen un 15 % de [Bmim][Cl] (Xu et al., 2016b) Celulosa [Emim] [MPO2] Celulasa de T. reesei Se alcanza una digestibilidad del 60 % en disoluciones que contienen un 10 % de LI (Auxenfans et al., 2017) Hierba [2-HEA] [OAc] Cellic CTec2 Se alcanza una digestibilidad del 55 % en disoluciones que contienen un 10 % de LI (Sun et al., 2017) Hierba, eucalipto y pino [Ch][Lys] Cóctel Cellic CTec2/HTec2 Digestibilidad de hierba del 93 %, de eucalipto del 50 % y de pino del 23 % (Dutta et al., 2018) Los procesos “one-pot” están todavía en una fase muy temprana de desarrollo, pero se ha conseguido avanzar mucho en los últimos años sobre todo utilizando biomasas herbáceas menos recalcitrantes que la madera. En esta tesis doctoral se intentará, llevar a cabo el proceso “one- pot” en maderas. A través del proceso integrado, se ha conseguido obtener altos rendimientos de azúcares fermentables y etanol, usando mezclas de enzimas disponibles comercialmente. Sin embargo, aún quedan grandes desafíos de ingeniería que deben abordarse antes de que se pueda aplicar el procesamiento de biomasa de alta carga de sólidos (≥ 10 %), utilizando el enfoque de proceso de una sola etapa. Estos desafíos se basan en:  La limitación de transferencia de materia que existe durante las operaciones de pretratamiento, sacarificación y en la unidad de fermentación debido a la restricción de agua.  La generación de productos inhibidores a altas cargas de sólido, que podría plantear problemas para el procesamiento posterior y los productos finales concentrados que podrían reducir la actividad enzimática general.  La reducción de la viabilidad de microorganismos, dado el aumento de la osmolaridad, como resultado de una alta concentración de sustratos de carbono, sales y productos finales. Líquidos iónicos en el procesado de biomasa 37 1.3.2.4. Coste de los LIs Uno de los mayores inconvenientes que se exponen normalmente acerca del empleo de los LIs es el relativo a su coste de síntesis. Muchos de los LIs suelen ser sintetizados en procesos con varias etapas de purificación, lo que generaría a nivel industrial un elevado coste final de producto. Sin embargo, la mayoría de los LIs actualmente se sintetizan a escala de laboratorio, por lo que los precios son muy superiores a los de compuestos sintetizados a escala industrial. Existen todavía pocos estudios acerca de la viabilidad del pretratamiento con LIs en biorrefinerías. Klein‐Marcuschamer et al. (2011), George et al. (2014) y Brandt et al. (2017) han llevado a cabo estudios preliminares acerca de cómo afecta la economía de un proceso de biorrefinería basado en el empleo de LIs. El coste de procesos de pretratamiento con LIs oscila entre 2,5 y 50 $/kg. Este valor está unido por un lado al coste de producción del LI en sí mismo, a la capacidad de recuperación y reutilización de LI y al desarrollo de tecnologías que permitan un uso efectivo del mismo. Además, la valorización de otros subproductos como la lignina al emplearla como fuente de energía en la propia biorrefinería puede reducir los costes del proceso (Klein‐Marcuschamer et al., 2011). De manera general, los LIPs son la alternativa más rentable, calculándose que pueden llegar a costar aproximadamente 0,50 $/kg, asumiendo que se recupera y reutiliza un 99 % del LI (Brandt et al., 2017). Pin et al. (2019) estimaron el precio de diferentes LIPs con aniones acetato e hidrogenosulfato, basándose en los datos de la plataforma Global TRade-Atlas durante los años 2015-2017 y sin tener en cuenta los pasos de purificación. Los resultados muestran precios de entre 0,98 y 2,79 $/kg, siendo el LI hidrogenosulfato de trietilamonio el más económico (0,98 $/kg) (Brandt et al., 2017; Pin et al., 2019). Al comparar los precios de los LIs con otros disolventes comunes, el precio de los LIs todavía tiene en los casos más favorables precios más elevados que el etanol (0,49 $/kg), el ácido clorhídrico (0,042 $/kg) o el ácido sulfúrico (0,27 $/kg). Sin embargo, otros factores como la no volatilidad y menor corrosión que supone el empleo de LIs hay que tenerlo en cuenta de cara a valorar el disolvente a escoger en cada proceso (Pin et al., 2019). 1.4. Interés y relevancia del pretratamiento de maderas con líquidos iónicos La biorrefinería de base forestal se basa frecuentemente en los procesos de obtención de pasta y papel (Ragauskas et al., 2006). A nivel global, la producción de madera alcanzó en 2016 los 1.874.000.000 m3/año. En particular, Europa es la región que más madera produce, un 32 % de la producción total mundial (FAO, 2016). España es el cuarto país de Europa con más bosques, detrás de Rusia, Suecia y Finlandia. Capítulo 1 38 Los bosques españoles están compuestos principalmente de árboles de eucalipto y pino con una superficie total en 2015 de 18,4 millones de hectáreas (ASPAPEL, 2018). Sin embargo, en los últimos años algunas de las instalaciones han tenido que cerrar debido a la enorme competencia surgida en Asia y Sur América, por lo que la diversificación de los productos sería la forma de permitir que esta industria pueda salir adelante (Stern et al., 2015). De este modo, a nivel europeo se han estado potenciando proyectos de investigación y desarrollo basados en la biorrefinería lignocelulósica y principalmente en la madera. En la Tabla 1.7 se muestran los proyectos que a fecha de Enero de 2019 se encuentran en vigor. Tabla 1.7: Proyectos de investigación financiados por la UE para apoyar la industria de la biorrefinería a partir de biomasa lignocelulósica (Hassan et al., 2019) Proyecto/página web Objetivo BIOFOREVER https://www.bioforever.org Demostrar la viabilidad comercial de la biorrefinería lignocelulósica (de biomasa forestal) para la industria química. BIOSKOH http://bioskoh.eu Demostrar la primera de una serie de biorrefinerías de segunda generación en Europa. EUCALIVA http://eucaliva.eu Crear una cadena de suministro a partir de la lignina utilizando residuo de eucalipto como materia prima. GRACE http://www.grace-bbi.eu Explorar el potencial de cereales industriales no alimenticios como fuente de biomasa. GREENSOLRES http://www.greensolres.eu Demostrar la viabilidad comercial de convertir la biomasa lignocelulósica en ácido levulínico. HYPERBIOCOAT http://www.hyperbiocoat.eu Desarrollar polímeros biodegradables a partir de subproductos del procesado alimentario. IFERMENTER Conversión de corrientes residuales de biomasa forestal a proteínas antimicrobianas por fermentación inteligente. LIBRE http://www.libre2020.eu Fabricación de composites a partir de fibras de carbono basadas en lignina. LIGNIOX http://www.ligniox.eu/ Tecnología de oxidación de lignina para producir dispersantes de lignina versátiles. LIGNOFLAG http://www.lignoflag-project.eu Producción de bioetanol a través de una cadena de suministro construida a partir de biomasa lignocelulósica. PEFERENCE Producción de ácido furano dicarboxílico, un intermedio de reacción que produce productos de alto valor añadido. SSUCHY https://www.ssuchy.eu/ Producción de “biocomposites” estructurales sostenibles a partir de fibras naturales híbridas y polímeros de base biológica. https://www.bioforever.org/ http://bioskoh.eu/ http://eucaliva.eu/ http://www.grace-bbi.eu/ http://www.greensolres.eu/ http://www.hyperbiocoat.eu/ http://www.libre2020.eu/ http://www.ligniox.eu/ http://www.lignoflag-project.eu/ https://www.ssuchy.eu/ Líquidos iónicos en el procesado de biomasa 39 Tabla 1.7 (continuación): Proyectos de investigación financiados por la UE para apoyar la industria de la biorrefinería a partir de biomasa lignocelulósica (Hassan et al., 2019) Proyecto/página web Objetivo SWEETWOODS Producción de lignina de alta pureza y compuestos químicos de plataforma asequibles a través de azúcares de base maderera. UNRAVEL Desarrollo de pretratamientos avanzados, separación y tecnologías de conversión para biomasa lignocelulósica compleja. US4GRENCHEM http://www.us4greenchem.eu/ Combinación de tratamientos con ultrasonidos y con enzimas de biomasa lignocelulósica para aplicaciones biotecnológicas basadas en azúcares. VALCHEM http://www.valchem.eu Obtención de productos químicos de valor añadido y lignina a partir de madera. WOODZYMES Empleo de enzimas extremófilas para la producción de productos químicos de base derivados de la madera: de las fábricas de papel a productos conglomerados y de revestimiento. ZELCOR http://www.zelcor.eu Valorización integrada de lignina mediante biorrefinerías lignocelulósicas con cero generación de residuos. Por otro lado, los LIs se empezaron a aplicar en procesos de pretratamiento de biomasa a partir del año 2002, cuando por primera vez se observó que el LI cloruro de 1-butil-3- metilimidazolio [Bmim][Cl] era capaz de disolver hasta un 10 % de celulosa de alto grado de polimerización (DP≈1000) (Swatloski et al., 2002). A partir de ese momento, se publicaron un gran número de trabajos en los que diferentes materiales se han pretratado con distintos LIs. Actualmente (a fecha de enero de 2019) se han encontrado 1875 artículos al buscar las palabras claves “biomasa” y “líquidos iónicos” (Figura 1.14). Figura 1.14: Publicaciones por año introduciendo la palabra clave “líquido iónico” y “biomasa” (Scopus, 2019) http://www.us4greenchem.eu/ http://www.valchem.eu/ http://www.zelcor.eu/ Capítulo 1 40 Esto demuestra que en los últimos años, la aplicación de los LIs en procesos de pretratamiento de biomasa está siendo muy estudiada, aunque el avance de esta tecnología todavía está en fase de desarrollo y mejora. A la vista de todo lo expuesto en este capítulo, la aplicación del concepto de biorrefinería a maderas supone una alternativa que permite intensificar el proceso de reducción de gases invernadero, sustituir combustibles fósiles y diversificar la industria de la pasta y el papel. Dentro de los pretratamientos disponibles, los líquidos iónicos se presentan como disolventes capaces de disolver todos los componentes de la madera, sin degradar el sólido. En un enfoque de fraccionamiento de la madera los procesos multietapa supondrían una ventaja que permite separar los componentes constituyentes de la madera. Por otro lado, en un enfoque de pretratamiento, los últimos estudios de biocompatibilidad que buscan la integración de procesos supondrían un ahorro de etapas y costes considerable. Líquidos iónicos en el procesado de biomasa 41 Capítulo 2: OBJETIVOS Y PLANTEAMIENTO DE LA TESIS DOCTORAL Objetivos y planteamiento de la tesis doctoral 43 2. OBJETIVOS Y PLANTEAMIENTO DE LA TESIS DOCTORAL La madera de pino y de eucalipto, así como sus residuos, constituyen materias primas de gran potencial para su aplicación en el sector biorrefinero. No obstante, la necesidad de un pretratamiento que maximice el rendimiento de las etapas posteriores es indispensable debido a su naturaleza recalcitrante. Esta tesis parte del interés surgido en torno a la aplicación de los líquidos iónicos en procesos de biorrefinería y más concretamente en procesos de pretratamiento, aplicados a biomasa lignocelulósica. La tesis se enmarca en el Grupo de Investigación de Desarrollo de Procesos y Productos de Bajo Impacto Ambiental de la Universidad Complutense de Madrid; concretamente en la Línea de Investigación centrada en el procesado de biomasa con líquidos iónicos y disolventes orgánicos, para su fraccionamiento y aprovechamiento. El objetivo principal de la tesis doctoral consiste en estudiar una serie de pretratamientos basados en la utilización de líquidos iónicos, con o sin autohidrólisis previa, para pretratar madera de pino y eucalipto con el objeto de poder seleccionar el proceso más efectivo de cara a la etapa posterior de hidrólisis enzimática. El interés de los pretratamientos de madera con líquidos iónicos se basa en la capacidad de los mismos para liberar determinados compuestos de la madera, aumentando así la accesibilidad del sólido para que las enzimas puedan sacarificarlo y alcanzar digestibilidades elevadas de celulosa a glucosa. La madera, a pesar de ser un material lignocelulósico de gran disponibilidad y poseer un Capítulo 2 44 elevado contenido en celulosa en comparación con otros materiales, posee una estructura compleja y un elevado contenido en lignina, lo que dificulta su pretratamiento. La novedad de esta tesis radica en el empleo de diferentes esquemas de pretratamiento en una etapa, multietapa y procesos “one-pot” sobre dos maderas distintas: pino y eucalipto, lo que no se ha estudiado de manera conjunta en ningún trabajo anteriormente. Además, se van a estudiar familias de LIs surgidas recientemente, como los líquidos iónicos derivados de la colina, menos tóxicos y basados en compuestos presentes en la naturaleza, así como los líquidos iónicos próticos cuyo proceso de síntesis tiene menor número de etapas, lo que podría traducirse en menores costes de producción. El uso de los líquidos iónicos puede resultar muy atractivo desde el punto de vista medioambiental siempre que se puedan recuperar y reutilizar, aspecto novedoso que también se ha contemplado en esta tesis. La investigación realizada conlleva un exhaustivo análisis y caracterización de todas las fracciones obtenidas, siendo la digestibilidad enzimática la respuesta principal para la evaluación de los pretratamientos. No obstante, se busca maximizar la separación de los demás componentes de la biomasa sin degradarlos: hemicelulosa y lignina. Para alcanzar el objetivo global se han planteado una serie de objetivos específicos que se enumeran a continuación: - Objetivo específico 1: estudiar el pretratamiento de madera de eucalipto y pino mediante procesos de autohidrólisis, analizando cada una de las corrientes sólidas y líquidas del proceso y evaluando la digestibilidad del sólido pretratado y el rendimiento a glucosa en ambas maderas. - Objetivo específico 2: identificar los líquidos iónicos más adecuados para pretratar madera de pino y eucalipto. - Objetivo específico 3: estudiar el pretratamiento de madera de eucalipto y pino mediante procesos con líquidos iónicos, analizando cada una de las corrientes sólidas y líquidas del proceso y evaluando la digestibilidad del sólido pretratado y el rendimiento a glucosa. Comparar los resultados obtenidos con ambas maderas. - Objetivo específico 4: estudiar el pretratamiento de madera de eucalipto y pino con líquidos iónicos utilizando calefacción por microondas para reducir el tiempo del proceso respecto a la calefacción convencional. - Objetivo específico 5: determinar la capacidad de recuperación y reutilización de un LI y los factores que limitan su eficacia al ser recuperado y reutilizado. - Objetivo específico 6: identificar los factores que influyen en la accesibilidad por vía enzimática de los sólidos pretratados. Objetivos y planteamiento de la tesis doctoral 45 - Objetivo específico 7: estudiar y comparar las maderas de eucalipto y pino pretratadas mediante procesos multietapa de autohidrólisis seguido de disolución con líquidos iónicos, para evaluar cómo afecta la combinación de pretratamientos a la digestibilidad del sólido pretratado y al rendimiento en glucosa. Determinar las condiciones más favorables para también obtener hemicelulosas y un sólido rico en lignina. - Objetivo específico 8: estudiar el proceso “one-pot” en madera de eucalipto y pino y evaluar las condiciones más favorables para obtener el mayor rendimiento en glucosa. Analizar cada una de las corrientes sólidas y líquidas del proceso y evaluar la biocompatibilidad del proceso con una posterior etapa de fermentación. Como se puede observar, muchos de los objetivos específicos tienen un punto común de comparación entre los resultados obtenidos con madera de pino y eucalipto. Debido a sus diferencias en cuanto a composición, estructura y comportamiento, el empleo de ambas maderas y su comparación permite ampliar el ámbito de aplicación de los resultados obtenidos a maderas duras y blandas. Para cumplir con los objetivos se divide la experimentación en tres bloques siguiendo el esquema de la Figura 2.1. A continuación se explica cada uno de los bloques planteados, que coinciden con los siguientes capítulos incluidos en esta tesis doctoral. Capítulo 2 46 Figura 2.1: Planteamiento de la Tesis doctoral F ig u ra 2 .1 : P la n tea m ien to d e la T esis d o cto ra l Objetivos y planteamiento de la tesis doctoral 47 - Pretratamientos en una etapa: Responde a los objetivos específicos 1-6. En el Capítulo 3 se estudia el proceso de autohidrólisis, pretratando las maderas de pino y de eucalipto e identificando las ventajas e inconvenientes de la autohidrólisis como pretratamiento para cada una de las maderas (apartado 3.1). Para evaluar el pretratamiento con líquidos iónicos, primero se llevó a cabo un “screening” para seleccionar los que ofrecían resultados más prometedores, y posteriormente se estudió el pretratamiento de ambas maderas con aquellos que ofrecieron mejores resultados (apartado 3.2). A continuación, las dos maderas se pretrataron con líquidos iónicos en horno microondas para evaluar el empleo de este sistema de calefacción como alternativa a la calefacción convencional (apartado 3.3). Debido a que los procesos con líquidos iónicos en muchos casos sólo son viables si se consigue reciclar y reutilizar el LI sucesivas veces, para el caso particular del [Amim][Cl] se estudió la recuperación y reutilización en 5 ciclos con ambas maderas (apartado 3.4). Para identificar los factores que influyen en la accesibilidad del sólido a las enzimas durante la sacarificación, se aplicaron técnicas microscópicas y análisis de imagen digital que permitieron relacionar parámetros morfológicos y de la superficie de los sólidos pretratados con su digestibilidad (apartado 3.5.). - Pretratamientos multietapa: Responde al objetivo específico 7. En el Capítulo 4 se estudian, para las dos maderas, combinaciones de pretratamientos de autohidrólisis seguidos de disolución con líquidos iónicos de manera secuencial. A partir de los resultados de caracterización y cuantificación de cada una de las fracciones obtenidas se seleccionó el proceso más favorable de los probados, que además maximiza la extracción de hemicelulosa, la producción de glucosa en la hidrólisis enzimática y la obtención de un sólido rico en lignina. - Proceso “one-pot”: Responde al objetivo específico 8. En el Capítulo 5 se lleva a cabo un proceso “one-pot”, pretratando madera con LI y realizando la hidrólisis enzimática en el mismo recipiente, en ausencia de la etapa de lavado. Uno de los puntos más importantes de este proceso es conocer la biocompatibilidad de LI con vista a la posterior etapa de fermentación. Así pues, se evaluó la toxicidad de diferentes líquidos iónicos con distintas levaduras fermentativas y se llevó a cabo una prueba de concepto de un proceso “one-pot” seguido de fermentación para la producción de bisaboleno. Capítulo 3: PRETRATAMIENTOS EN UNA ETAPA 3. Pretratamientos en una etapa 51 3. PRETRATAMIENTOS EN UNA ETAPA En este capítulo se recoge el estudio de los pretratamientos en una única etapa realizados en la tesis doctoral. Se han llevado a cabo dos pretratamientos: autohidrólisis y disolución con líquidos iónicos. En ambos pretratamientos se utilizan diferentes condiciones de temperatura y tiempo. En el pretratamiento de autohidrólisis se utiliza un reactor a presión. En el pretratamiento con líquidos iónicos se utiliza tanto calefacción convencional con líquidos iónicos próticos, apróticos y derivados de la colina, como calefacción por horno microondas empleando líquidos iónicos apróticos. Además, en el caso del líquido iónico cloruro de 1-alil-3-metilimidazolio, se estudia su recuperación y su reutilización. Finalmente, se evalúa la digestibilidad del sólido obtenido tras los distintos pretratamientos ensayados, con el fin de determinar su viabilidad para una posterior etapa de hidrólisis enzimática. 3.1. Autohidrólisis El proceso de autohidrólisis (AH) se estudia sobre madera de pino y de eucalipto bajo diferentes condiciones de temperatura y tiempo. El procedimiento experimental seguido, la caracterización y los resultados se recogen en la Publicación I y se resumen brevemente a continuación. Capítulo 3 52 3.1.1. Estructura y procedimiento experimental El planteamiento experimental que se ha llevado a cabo se muestra en la Figura 3.1. En este esquema se recogen las etapas de acondicionamiento, reacción, lavado e hidrólisis enzimática. De cada una de las etapas se obtienen corrientes sólidas (corrientes 1 y 3) y/o líquidas (corrientes 2, 4 y 5) para analizar y caracterizar. Las materias primas de partida utilizadas en este proceso, y a lo largo de toda la tesis doctoral, son madera de Eucalyptus globulus y Pinus radiata, suministradas por el Instituto Nacional de Investigación y Tecnología Agraria y Alimentaria (CIFOR-INIA). El proceso se lleva a cabo en un reactor a presión modelo 4567 de la marca Parr Instrument Company, Moline, IL. El equipo consta de un controlador de temperatura, presión y velocidad de agitación modelo 4848. La vasija donde se lleva a cabo la reacción es de acero inoxidable T316 con capacidad para 450 mL. Para enfriar la vasija se utiliza un criotermostato de la marca Huber, modelo MPC-K6, que usa aceite de silicona (Sil Oil, M20, 195/235,20) como fluido refrigerante. En el proceso se introducen 20 g de madera de pino/eucalipto en el reactor en forma de serrín a un tamaño de partícula (0,3-2 mm) y una relación líquido: sólido en peso de 8:1 y de 10:1 para el eucalipto y el pino, respectivamente (García-Domínguez et al., 2013; Romaní et al., 2014; Santos et al., 2018). Las condiciones de reacción y el factor de severidad (S0) utilizados se recogen en la Tabla 3.1. Para asegurar condiciones de baja y alta extracción de hemicelulosas, así como de degradación de las mismas, se han escogido diferentes condiciones de autohidrólisis para la madera de pino y de eucalipto, debidas a la variación de estructura, densidad y reactividad al pretratamiento de cada una de ellas. El perfil de temperaturas de cada ensayo se divide en tres etapas: en la primera se calienta el reactor desde 30 °C hasta la temperatura de operación en 50 minutos. A continuación se lleva a cabo una isoterma durante el tiempo establecido. Finalmente, se enfría la mezcla de reacción hasta los 40 °C. Tabla 3.1: Condiciones de operación en los ensayos de autohidrólisis Madera Temperatura (℃) Tiempo (min) S0 Muestra Eucalyptus globulus 150 30 2,95 AH150E 175 30 3,69 AH175E 200 30 4,42 AH200E Pinus radiata 150 30 2,95 AH150P 175 60 3,99 AH175P 200 90 4,90 AH200P Pretratamientos en una etapa 53 Figura 3.1: Planteamiento experimental del proceso de autohidrólisis F ig u ra 3 .1 : P la n te a m ie n to e x p er im en ta l d el p ro ce so d e a u to h id ró li si s. L a s fl ec h a s en d is co n ti n u o i n d ic a n l a s co rr ie n te s a a n a li za r , id en ti fi ca n d o c o n n ú m er o s la s té cn ic a s d e ca ra ct er iz a ci ó n a p li ca d a s a c a d a c o rr ie n te . Capítulo 3 54 Terminada la AH se filtra el contenido y el sólido autohidrolizado se lava y se caracteriza (corriente 3). Tanto la fracción líquida (corriente 2) como la fracción de lavado (corriente 4) se recogen y analizan para conocer su contenido en oligosacáridos, azúcares monoméricos (glucosa, xilosa, galactosa, arabinosa y manosa), ácido acético y productos de degradación (hidroximetilfurfural, furfural, ácido acético, ácido fórmico y ácido levulínico). La digestibilidad enzimática será el parámetro principal que se va a considerar para evaluar la eficacia del pretratamiento estudiado: autohidrólisis. El desarrollo de la hidrólisis enzimática se ha llevado a cabo de acuerdo a la norma NREL/TP-5100-63351, en la cual se trabaja con baja carga de sólidos con una dosis de enzima de 0,25 mL de cóctel/g glucano. El cóctel enzimático que se utiliza para llevar a cabo la hidrólisis enzimática se denomina Accellerase 1500®, suministrado por la empresa Dupont Industrial Biosciences. El contenido en proteína de este cóctel es de 70 mg de proteína/mL disolución. Este cóctel es el que se utiliza mayoritariamente a lo largo de toda la tesis doctoral (Publicaciones I, II, III, V y VI) para evaluar la digestibilidad enzimática de cada pretratamiento. 3.1.2. Caracterización de las fracciones obtenidas Tras el proceso de autohidrólisis de ambas maderas, se caracterizan tanto la fracción sólida como la fracción líquida. La fracción sólida se caracteriza en lo que a su rendimiento en el sólido pretratado se refiere (como cociente entre el sólido inicial y el sólido final obtenido), así como en cuanto a su contenido en carbohidratos y lignina de acuerdo a la norma NREL/TP-510-42618 adaptada a pequeñas cantidades (Sluiter, 2011; Ibáñez & Bauer, 2014). La determinación de azúcares y productos de degradación de la fracción líquida se lleva a cabo tras la autohidrólisis, siguiendo la norma estandarizada NREL/TP-510-42623 (Sluiter et al., 2006). La determinación de los monosacáridos presentes en la fase líquida se realiza de manera directa mediante cromatografía en fase líquida usando un HPLC Agilent 1260 Infinity. La descripción detallada de la caracterización realizada para ambas fracciones se recoge en la Publicación I. 3.1.3. Resultados experimentales. Interpretación y discusión En la Tabla 3.2 se recoge la cantidad de monosacáridos, oligosacáridos y productos de degradación obtenidos en la fase líquida tras el tratamiento de autohidrólisis, tanto de madera de eucalipto como de pino, frente a S0. Como se puede observar, las hemicelulosas principales extraídas del eucalipto fueron los xilo-oligosacáridos (12,90 g/100 g de madera), mientras que los manano- oligosacáridos fueron las principales hemicelulosas extraídas del pino (9,15 g/100 g de madera), ambas para un factor de severidad intermedio. En el caso del pino, también se extraen 2,61 g de gluco-oligosacáridos/100 g de madera a un S0= 3,99, procedentes de los galactoglucomananos. Estos Pretratamientos en una etapa 55 resultados confirman la limitación de este pretratamiento para extraer glucano estructural por lo que a S0= 3,69 y 3,99 la celulosa apenas se degrada en ninguna de las dos maderas (Yildiz & Oner, 2014). Tabla 3.2: Azúcares, oligosacáridos y productos de degradación obtenidos en la fase líquida Eucalyptus globulus Pinus radiata Factor de severidad (S0) 2,95 3,69 4,42 2,95 3,99 4,90 Azúcares monoméricos (g/100 g madera) Glucosa 0,05 0,35 1,29 0 0,08 1,61 Xilosa 0 1,51 0 0 1,55 0,35 Galactosa 0,06 0,61 0 0 0,62 0,62 Arabinosa 0 0,04 0 0,90 0,94 0 Manosa 0 0 0 0 0,77 1,22 Azúcares oligoméricos (g/100 g madera) Gluco-oligosacáridos 0,54 0,61 0,75 0,71 2,61 0,31 Xilo-oligosacáridos 1,39 12,90 0 0,90 1,91 0 Galacto- oligosacáridos 0,55 1,04 0 0,94 1,62 0,06 Arabino- oligosacáridos 0,04 0 0 0,54 0,36 0 Manano- oligosacáridos 0,10 0,52 0 2,10 9,15 0,26 Ácido acético (g/100 madera) 0,07 0,90 5,27 0,09 0,52 2,36 Productos de degradación (g/100 g madera) Ácido fórmico 0,02 0,22 1,38 0,04 0,14 1,05 Ácido levulínico 0 0 0,20 0 0 0,06 Hidroximetilfurfural 0 0,02 4,73 0 0,23 5,59 Furfural 0 0,18 2,88 0 0,42 2,57 En el caso del eucalipto, a condiciones suaves (S0 = 2,95) se recuperó el 7,5 % de todo el xilano en forma de xilo-oligosacáridos, y a condiciones intermedias (S0 = 3,69) el 69,7 %. A condiciones severas (S0 = 4,42), no se obtienen xilo-oligosacáridos, el ácido acético aumenta y también los productos de degradación (sobre todo hidroximetilfurfural). Por el contrario, en el caso del pino a condiciones suaves (S0 = 2,95) y a condiciones intermedias (S0 = 3,99), se recuperó el 15,8 % y el 68,8 % del manano en forma de manano- Capítulo 3 56 oligosacáridos, respectivamente. A condiciones severas (S0 = 4,42) sólo se obtuvo un 2% de todo el manano inicial, lo que supone menos de un 1 % de la biomasa inicial introducida. A condiciones severas, se producen mayoritariamente furanos y ácidos. Tanto el contenido en ácido acético como en furfural producido es superior en los licores de autohidrólisis de la madera de eucalipto que en la de pino, primeramente debido a la mayor relación sólido:líquido utilizada en la autohidrólisis de pino respecto a la de eucalipto, pero también debido al mayor contenido en grupos acetilo y en pentosas de la madera de eucalipto (Nitsos et al., 2016). Además, las elevadas cantidades de hidroximetilfurfural producidas en eucalipto y pino se pueden atribuir a la deshidratación de gluco-oligosacáridos (Nitsos et al., 2016). De manera general, se observa que tanto las condiciones intermedias como las condiciones severas dan lugar a la obtención de productos valorizables (Li et al., 2013a). En la Figura 3.2 se muestran los rendimientos del sólido recuperado en el proceso de autohidrólisis, así como la composición de este sólido. Figura 3.2: Composición y rendimiento del sólido pretratado en la autohidrólisis en a) eucalipto y b) pino La cantidad de sólido recuperado tras la autohidrólisis disminuye al aumentar la severidad de la AH. Al comparar la composición relativa de la madera pretratada con la madera sin tratar, se observa que las fracciones de hemicelulosa se eliminan de manera selectiva en este pretratamiento (Yildiz & Oner, 2014). La disminución del contenido en hemicelulosas ocasiona un aumento del Pretratamientos en una etapa 57 porcentaje en glucano y en lignina. Al comparar ambas maderas, también se comprueba que las variaciones de composición, al aumentar el factor de severidad son más pronunciadas al emplear madera dura (eucalipto) que blanda (pino), lo cual ya indica que la madera blanda es más recalcitrante que la dura (Nitsos et al., 2016). En la Figura 3.3 se muestra la digestibilidad de la madera pretratada por AH a las 72 h. En el caso del eucalipto, la menor digestibilidad se obtiene con madera de eucalipto sin tratar y aumenta gradualmente debido a la ruptura de la red lignocelulósica del material y la reducción del contenido en hemicelulosa (Alvira et al., 2010a; Silva-Fernandes et al., 2015). Sin embargo, a pesar de que la mayor digestibilidad se obtiene con las condiciones más severas, la degradación de glucano producida en el pretratamiento de autohidrólisis hace que el rendimiento final (50 %) sea mayor en condiciones intermedias (AH175E). Figura 3.3: Digestibilidad en glucano y rendimiento en glucosa de los sólidos autohidrolizados En el caso del pino, la tendencia es opuesta, dando lugar a una disminución de la digestibilidad cuando aumenta la severidad del tratamiento. Este resultado se atribuye al efecto inhibidor que produce la lignina de madera blanda debido a estructuras recondensadas de la misma (Nitsos et al., 2016). Además, en la mayoría de los trabajos de autohidrólisis la mejora de la digestibilidad por la etapa AH es baja, pero en pocos se observa una disminución de la digestibilidad con el aumento de la severidad como aquí ocurre (Pielhop et al., 2015; Carvalheiro et al., 2016; Nitsos et al., 2016). Cabe destacar que la etapa de hidrólisis enzimática en todas las muestras se llevó a cabo al mismo tamaño de partícula. Esto hace que en la determinación de la digestibilidad no se tenga en Capítulo 3 58 cuenta el efecto que produce la disminución del tamaño de partícula al aumentar S0 y por tanto en los resultados sólo influya la accesibilidad del material (MacAskill et al., 2018). 3.2. Pretratamiento con líquidos iónicos con calefacción convencional En este bloque se ha estudiado el proceso de pretratamiento de madera de pino y de eucalipto con un total de 14 líquidos iónicos diferentes. El procedimiento experimental, la caracterización y los resultados obtenidos se recogen en la Publicación II, y se resumen brevemente a continuación. 3.2.1. Estructura y procedimiento experimental El planteamiento experimental del pretratamiento con LIs utilizando calefacción convencional se muestra en la Figura 3.4, donde se incluyen las etapas de acondicionamiento, “screening”, pretratamiento con LIs, regeneración, lavado e hidrólisis enzimática. De cada una de las etapas se obtienen corrientes sólidas (corrientes 1, 2, 4 y 5) y/o líquidas (corrientes 3, 6 y 7) para analizar y caracterizar. La madera se acondiciona llevando a cabo una etapa de eliminación de extractos (con acetona/ agua) para impedir la formación de espumas, una etapa de molienda y una etapa de tamizado (Kostamo & Kukkonen, 2003; Casas et al., 2013). Como resultado se obtiene una madera libre de extractos con un tamaño de partícula <150 µm. De manera previa al pretratamiento, ha sido necesario llevar a cabo un “screening” de todos los LIs que potencialmente podrían ser válidos para pretratar las maderas. La selección de los LIs a probar en el “screening” se basó en los resultados que se habían aportado en bibliografía hasta la fecha, con diferentes biomasas, y se añadieron otros basándose en su estructura química (estructuras similares a otros LIs que ya se había comprobado que funcionaban) o a sus propiedades físicas (Fang, 2014). Se decidió buscar familias de líquidos iónicos que fueran polares, para interaccionar con la celulosa, con valores de β elevados o con valores de β-α elevados (Doherty et al., 2010). Los LIs utilizados en el “screening” se recogen en la Tabla 3.3. Se preparan mezclas de 1 g de LI con madera de eucalipto libre de extractos (<150 μm), madera de pino libre de extractos (<150 μm), celulosa o lignina con una relación en peso de 25:1, en un bloque termostático sin agitación modelo Labnet AccuBlock 1. Los LIs se manipulan en atmósfera inerte de N2 usando una cámara de guantes. El “screening” se inicia a T0= 30 °C, observando la disolución de la mezcla a t=1 h y t=24 h, tanto de manera visual como al microscopio (corriente 2). El microscopio empleado es un Axio Scope A.1 para confirmar la disolución total o parcial. Si la disolución de la mezcla no se ha producido, se incrementa la temperatura 20 °C dejando disolver durante 24 h más y observando de manera visual y al microscopio el resultado final. Pretratamientos en una etapa 59 Figura 3.4: Planteamiento experimental del proceso de pretratamiento con LIs con calefacción convencional F ig u ra 3 .4 : P la n te a m ie n to e x p er im en ta l d el p ro ce so d e p re tr a ta m ie n to c o n L Is c o n c a le fa cc ió n c o n v en ci o n a l. L a s fl ec h a s en d is co n ti n u o s o n l a s co rr ie n te s a a n a li za r , in d ic a n d o c o n n ú m er o s la s té cn ic a s d e ca ra ct er iz a ci ó n a p li ca d a s a c a d a c o rr ie n te Capítulo 3 60 Si alguna de las muestras se disuelve totalmente, se añade agua como antidisolvente y se observa si el sólido disuelto se vuelve a regenerar. Si por el contrario, una vez disuelto ya no se regenera, se considera que el sólido se ha degradado. Este proceso se realiza hasta una temperatura final de 150 °C, debido a que algunos de los LIs empleados degradan a esa temperatura. Tabla 3.3: Suministrador y pureza de los líquidos iónicos utilizados Líquido iónico Fórmula Procedencia Pureza (%) LIs apróticos Cloruro de 1-alil-3- metilimidazolio [Amim][Cl] Iolitec Gmbh. >98 Dimetilfosfato de 1, 3- dimetilimidazolio [Dmim][DMPO4] Iolitec Gmbh. >98 Dietilfosfato de 1-etil-3- metilimidazolio [Emim][DEPO4] Iolitec Gmbh. >98 Acetato de 1-etil-3- metilimidazolio [Emim][OAc] Iolitec Gmbh. >95 Bis(trifluorometilsulfonil)imida de 1-etil-3-metilimidazolio [Emim][Tf2N] Iolitec Gmbh. 99 LIs próticos Hidrogenosulfato de 1-etil-3- metilimidazolio [Emim][HSO4] Iolitec Gmbh. 99 Cloruro de 1-metilimidazolio [Mim][Cl] Iolitec Gmbh. 98 Formato de 2- hidroxietilamonio [OHEtAm][HCO2] Iolitec Gmbh. >97 LIs con catión basado en colina Acetato de colina [Ch][OAc] Iolitec Gmbh 98 Lisinato de colina [Ch][Lys] Iolitec Gmbh 95 LIs con anión basado en prolina R-prolinato de tetrabutilamonio [TBA][R-Pro] Instituto de Química Molecular de Reims - S-prolinato de tetrabutilamonio [TBA][S-Pro] Instituto de Química Molecular de Reims - Trans-4-hidroxi-L-prolinato de tetrabutilamonio [TBA][t-4OH-LPro] Instituto de Química Molecular de Reims - S-prolinato de tetrabutilfosfonio [TBP][S-Pro] Instituto de Química Molecular de Reims - Pretratamientos en una etapa 61 Una vez seleccionados los LIs más prometedores, se lleva a cabo el proceso de pretratamiento con LIs con calefacción convencional en un bloque termostático Barnstead Steam RS9000 con agitación orbital. En el pretratamiento se utilizan 0,4 g de madera en una relación LI:sólido de 25:1. El perfil de calentamiento consta de una sola etapa isoterma a 120 ℃ durante 6 horas con una agitación orbital de 200 r.p.m. Una vez terminada la etapa de disolución y sin separar el sólido que queda sin disolver, se añade agua destilada como antidisolvente a la mezcla en una relación en peso de agua:LI de 2,5:1. La no separación de la fracción sólida de la líquida permite, además de ahorrar una etapa de separación, evitar pérdidas de sólido y maximizar el rendimiento. La mezcla se lleva a un baño de agua con agitación a 40 °C durante 10 minutos. Transcurrido el tiempo de regeneración de la madera pretratada, se filtra a vacío separando la fracción sólida de la líquida. La fracción sólida se lava sucesivas veces para asegurar la eliminación del LI de la madera pretratada (corriente 4). La cantidad final de agua de lavado respecto al LI es de 17,5 a 1. El agua de la fase líquida se elimina en un rotavapor para recuperar el LI y caracterizarlo (corriente 6). Finalmente, la fracción sólida (corriente 7) se hidroliza enzimáticamente para determinar su digestibilidad. 3.2.2. Caracterización de las fracciones obtenidas Tras el proceso de pretratamiento con LI, se caracterizan tanto la fracción sólida como la fracción líquida. La fracción sólida de madera sin tratar (corriente 1) y pretratada (corriente 4) se caracteriza de acuerdo a la norma NREL/TP-510-42618, como ya se ha explicado en el apartado 3.1.1. El sólido obtenido tras la etapa de hidrólisis enzimática (corriente 5) se analiza por resonancia magnética nuclear 2D (Heikkinen et al., 2003). Para ello, se solubilizan parcialmente 15-25 mg de muestra en 500 µL de DMSO-d6 y se someten a ultrasonidos durante 210 minutos. Los análisis se realizan en un espectrómetro Bruker AVANCE III (Bruker BioSpin, Rheintetten, Alemania) operando a 700,17 MHz de frecuencia de protón y con una criosonda TXI de 5 mm. El espectro se analiza utilizando un programa de pulsos perfect_hsqc (Castañar et al., 2015). Las corrientes líquidas 3 y 6 están compuestas por LI y por LI y agua, respectivamente. La eliminación de agua se lleva a cabo en un rotavapor Büchi R-114, obteniéndose corrientes de LI concentrado que se pueden comparar con el LI fresco. La estabilidad térmica del LI se determina mediante un TGA/SDTA851e, de Mettler Toledo Instrument. Para llevar a cabo el análisis, se pesan 7-9 mg de LI en crisoles de PCA/zafiro. La rampa de calentamiento utilizada va de 30-105 °C (para eliminar la posible humedad de la muestra) manteniendo la temperatura a 105 °C durante 30 minutos. A continuación, se sigue calentando a una velocidad de 10 °C/min hasta una temperatura final de 800 °C. Los experimentos se realizan en una atmosfera de nitrógeno a un caudal de 20 mL/min. La espectroscopía FTIR-ATR del líquido iónico se lleva a cabo utilizando un equipo Jasco FTIR 4700 al que se incorpora el accesorio de reflectancia total atenuada para muestras Capítulo 3 62 líquidas Golden Gate TM diamond accessory Specac 10542. Los espectros se adquieren en el intervalo de 4000-400 cm-1 para posteriormente cortarlos, en cada caso, en el intervalo de bandas representativas, y normalizarlos. Para cada muestra, se realizan 64 barridos con una resolución de 4 cm-1. La corriente 7 obtenida tras la hidrólisis enzimática, realizada de acuerdo a la norma NREL/TP-5100-63351, se caracteriza tal y como se explica en el apartado 3.1. La descripción detallada de la caracterización realizada, tanto para la fracción sólida como la líquida, se muestra en la Publicación II. 3.2.3. Resultados experimentales. Interpretación y discusión A continuación se exponen los resultados obtenidos en el “screening” de los LIs, así como en el pretratamiento con LIs. 3.2.3.1. “Screening” de líquidos iónicos En la Tabla 3.4 se resumen las observaciones del “screening” con cada uno de los LIs. En el caso de los LIs [TBA][R-Pro], [TBA][S-Pro], [TBA][t-4OH-LPro] y [TBP][S-Pro], sólo se llevó a cabo el ensayo en madera de pino. En ninguno de los LIs ensayados con aniones derivados de la prolina se consiguió disolver por completo la madera de pino a 150 °C. Además, se comprobó que los LIs derivados de la prolina cambiaron de aspecto y de coloración al ser almacenados, por lo que se decidió descartar su uso en ensayos posteriores. Se ha podido comprobar que la lignina y la celulosa se disuelven con más facilidad que las maderas y que la madera de pino es más difícil de disolver que la de eucalipto. Además, a temperaturas superiores a 130 ° C, algunos de los LIs utilizados comienzan a degradar, adquiriendo un color oscuro, burbujeando o generando sobrepresiones en el vial. La degradación se produce, principalmente, en los LIs derivados del catión colina y en el LI [Emim][HSO4]. También se ha observado que la lignina álcali se puede disolver a temperatura ambiente utilizando el LI [OHEtAm][HCO2]. El “screening” realizado, en conjunto con la información bibliográfica de partida, muestra que el comportamiento de cada tipo de LI es diferente, y que responden de manera distinta a la madera de pino y a la de eucalipto. Los líquidos iónicos apróticos (LIAs) o convencionales ([Amim][Cl] o [Emim][OAc]) son capaces de disolver tanto celulosa como lignina, mientras que la disolución total de madera no se produce siempre a la máxima temperatura de ensayo (150 °C). Los líquidos iónicos próticos (LIPs) estudiados son capaces de disolver lignina, mientras que la disolución de celulosa resulta más difícil, y en algunos casos se produce la degradación antes de conseguir la disolución total, probablemente debido al efecto catalítico que provoca la presencia de iones H+ por la higroscopicidad del LI (Cox & Ekerdt, 2013). Los LIs derivados del catión colina parecen ser menos estables térmicamente que los LIs apróticos al oscurecerse las mezclas a 150 °C. Pretratamientos en una etapa 63 Sin embargo, son capaces de disolver la lignina, pero no la celulosa. Además, algunos LIs como el [Ch][Lys] producen mezclas que visualmente parecen tener una consistencia más viscosa con la madera. Tabla 3.4: Resumen de los resultados del “screening” de LIs LI LI-Celulosa LI-lignina LI-Eucalipto LI-Pino LIs apróticos [Amim][Cl] Disuelta a 90 °C y 24 h Disuelta a 90 °C y 24 h DCT a 150 °C y 24 h Disolución oscura DCT a 150 °C y 24 h Disolución oscura [Dmim][DMPO4] Disuelto a 130 °C y 1 h Disuelto a 70 °C y 24 h No disuelto a 150 °C y 24 h No disuelto a 150 °C y 24 h [Emim][DEPO4] Disuelto a 90 °C y 24 h Disuelto a 90 °C y 24 h DCT a 150 °C y 24 h DP a 150 °C y 24 h [Emim][OAc] Disuelto a 90 °C y 1 h Disolución marrón oscuro Disuelto a 90 °C y 1 h Disolución marrón oscuro Disuelto a 150 °C y 24 h Disolución marrón oscuro Disuelto a 150 °C y 24 h Disolución marrón oscuro [Emim][Tf2N] Disuelto a 70 °C y 24 h Disuelto a 70 °C y 24 h No disuelto a 150 °C y 24 h No disuelto a 150 °C y 24 h LIs próticos [Emim][HSO4] No se disuelve y se degrada A 70 °C y 1 h se disuelve y burbujea No se disuelve y se degrada No se disuelve y se degrada [Mim][Cl] Disuelto a 150 °C y 24 h. Líquido muy oscuro DCT A 150 °C y 24 h. Se ha oscurecido DP a 150 °C y 24 h. Se ha oscurecido DP a 150 °C y 24 h. Se ha oscurecido [OHEtAm][HCO2] Disuelto a 90 °C y 24 h Disuelto a Tamb DCT a 150 °C y 24 h DCT a 150 °C y 24 h LIs con catión basado en colina [Ch][OAc] DCT a 130 °C y 24 h Presión en vial a 150°C DCT a 130 °C y 24 h Presión en vial a 150°C DCT a 130 °C y 24 h Presión en vial a 150°C DCT a 130 °C y 24 h Presión en vial a 150°C [Ch][Lys] DCT a 150 °C y 24 h y presión en vial Disuelto a 130 °C y 24 h Gelifica a 110 °C y 24 h DCT a 150°C y presión en vial Gelifica a 110 °C y 24 h DCT a 150 °C y 24 h DCT= Disolución casi total; DP= Disolución parcial; Capítulo 3 64 En base a las observaciones experimentales de todos los LIs estudiados se descartaron 8 de ellos, y se decidió seguir trabajando con 2 LIAs convencionales, [Emim][OAc] y [Amim][Cl]; 2 LIPs [Mim][Cl] y [OHEtAm][HCO2]; y 2 LIs derivados de la colina [Ch][OAc] y [Ch][Lys] por las siguientes razones:  [Emim][OAc] es el LI más empleado en bibliografía en el pretratamiento de biomasa (Elgharbawy et al., 2016). Los resultados del “screening” muestran que este LI es capaz de disolver celulosa, lignina, madera de eucalipto y de pino a 150 °C.  [Amim][Cl] es también muy usado en bibliografía en el pretratamiento de biomasa. Los resultados del “screening” muestran que es capaz de disolver celulosa y lignina. Además, casi disuelve totalmente la madera de eucalipto y de pino a 150 °C. Por último, su estabilidad térmica (Tonset=273 °C) es bastante elevada, lo que permite un mayor intervalo de operatividad de temperaturas (Zhang et al., 2005; Hao et al., 2010).  [Mim][Cl] disuelve la celulosa y casi la totalidad de la lignina. Además, en disolución acuosa tiene un pH ácido que puede favorecer la ruptura de biomasa lignocelulósica. Por último, dentro de los LIPs presenta una elevada estabilidad térmica.  [OHEtAm][HCO2] es capaz de disolver lignina álcali a temperatura ambiente, lo cual se ha considerado una ventaja muy atractiva de cara a realizar pretratamientos a baja temperatura.  [Ch][OAc] y [Ch][Lys]: se observa que se produce disolución parcial de todos los componentes. Además, la tolerancia de estos LIs en celulasas y levaduras son mayores (Asakawa et al., 2016; Xu et al., 2016). 3.2.3.2. Pretratamiento con los LIs seleccionados En la Figura 3.5 se muestra la composición del sólido obtenido tras el pretratamiento (120 °C, 6 h, agitación orbital) con cada LI. En el caso de los LIPs, el rendimiento del sólido recuperado es similar entre la madera dura (eucalipto) y la madera blanda (pino). Al comparar los dos LIPs utilizados, el rendimiento con [OHEtAm][HCO2] es mucho mayor (83-85 %) que con [Mim][Cl] (40-42 %). Esto es debido a las diferencias existentes en los parámetros Kamlet-Taft β- α, en lo que a disolución de celulosa y su consecuente regeneración se refiere, con los valores de 0,43 y 0,31 para [OHEtAm][HCO2] y [Mim][Cl], respectivamente (Hauru et al., 2012). A pesar de que en el “screening”, el LI [OHEtAm][HCO2] parecía un candidato prometedor por su gran capacidad de disolución de la lignina, la falta de interacciones con toda la pared celular de la biomasa (celulosa, hemicelulosa y lignina) limita la ruptura de toda la red lignocelulósica. Utilizando el LIA [Emim][OAc] se obtienen rendimientos del sólido pretratado más bajos con la madera dura (rendimiento del sólido del 76 %) que con la blanda (rendimiento del sólido del Pretratamientos en una etapa 65 89 %). La madera de pino y de eucalipto es diferente en estructura, composición y también en el tipo de lignina, ya que la lignina de pino está formada principalmente unidades guayacilo y la de eucalipto por unidades siringilo y guayacilo. Figura 3.5: Composición y rendimiento de madera pretratada con LIs con calefacción convencional de a) eucalipto y b) pino Capítulo 3 66 Las diferencias existentes en el tipo de unidades aromáticas, podría implicar una diferente afinidad del [Emim][OAc] a cada tipo de estructura de lignina y justificar las diferencias de rendimiento entre madera dura y blanda observadas. El rendimiento del sólido con el LIA [Amim][Cl] es similar con eucalipto (86 %) que con pino (84 %). En este caso, el valor β [Emim][OAc] es mayor (1,14) que el del [Amim][Cl] (0,83) y el contenido en celulosa del eucalipto es mayor que en el pino (Sun et al., 2014). Los rendimientos de la recuperación de sólidos con los LIs derivados de la colina son bajos, debido a la elevada interacción que ambos LIs tienen con la celulosa: β=1,22 y 1,30 para el [Ch][OAc] y [Ch][Lys], respectivamente. El efecto producido en la composición de los sólidos pretratados con cada LIP utilizado es diferente. El LI [OHEtAm][HCO2] prácticamente no modifica la composición de las maderas. El porcentaje en lignina del pino tratado con [OHEtAm][HCO2] (34,6 %) es muy similar al no tratado (34,1 %). Se produce un incremento del contenido relativo de glucano, debido a la extracción selectiva de algunas hemicelulosas. En el caso del eucalipto, se extrae más hemicelulosa que en el pino y en consecuencia aumenta el porcentaje de lignina en el sólido pretratado. Por otra parte, el efecto catalítico del [Mim][Cl] causa la degradación de todos los componentes (Cox & Ekerdt, 2013). La extracción de hemicelulosas es casi completa en ambas maderas, y se extrae también lignina, más en la madera dura que en la blanda. En los LIAs se demuestra que el contenido relativo de lignina y hemicelulosa disminuye en las dos maderas tanto con [Emim][OAc] como con [Amim][Cl]. Estos resultados están en concordancia con lo observado en literatura por otros autores (Sun et al., 2013; Torr et al., 2016). Cabe destacar que la eliminación de hemicelulosas es mayor en la madera de eucalipto, en donde el contenido de xilano disminuye, que en madera de pino, donde el manano disminuye mucho menos, probablemente debido a la mayor facilidad de extraer pentosas como el xilano que hexosas como la manosa. En el caso de los LIs con catión colina, presentan diferencias entre el comportamiento que ofrece el LI [Ch][OAc] y [Ch][Lys]. Utilizando [Ch][OAc] se extrae un 11 y un 19 % de lignina; un 39 y un 69 % de hemicelulosa y un 5 y un 18 % de celulosa para pino y eucalipto, respectivamente. Otros autores ya demostraron el potencial de este tipo de LIs con otros materiales como bagazo de caña de azúcar o cáscara de arroz (Sun et al., 2014; Ninomiya et al., 2015a; Xue-Dan Hou, 2015). Por último, el LI [Ch][Lys] extrae un 21,6 % y un 40 % de lignina; un 37 y un 76 % de hemicelulosa y casi no extrae celulosa en pino y eucalipto, respectivamente. El efecto deslignificante del [Ch][Lys] queda demostrado y es probablemente debido a la basicidad de este LI, al contar con un aminoácido como anión. Este efecto deslignificante se ha observado previamente con Pretratamientos en una etapa 67 pretratamientos de mezclas de LI y agua (Dutta et al., 2018). Con este trabajo se demuestra la capacidad deslignificante que tiene el LI por sí solo. En la Figura 3.6 se muestra la digestibilidad a 72 h de la madera de eucalipto pretratada con LIs y calefacción convencional. Los resultados están expresados en g de glucano/100 g glucano introducido en la etapa de hidrólisis enzimática. Los resultados de digestibilidad, con cada LI, siguen el siguiente orden: Eucalipto sin tratar < [OHEtAm][HCO2] < [Mim][Cl] < [Amim][Cl] < [Ch][Lys] < [Ch][OAc] < [Emim][OAc] El pretratamiento con los LIPs, aunque produce un aumento significativo en la digestibilidad del sólido, este es menor que el de otro tipo de LIs. Los LIs que dan lugar a mayores digestibilidades son el [Emim][OAc] con un 80,4 % y el [Ch][OAc] con un 69 %. El [Emim][OAc] es un LI ampliamente utilizado en bibliografía, pero la alta digestibilidad obtenida tras el pretratamiento con el [Ch][OAc] lo convierte en una alternativa atractiva por ser más biocompatible y menos tóxico que los LIs convencionales derivados del catión imidazolio. Estos resultados también demuestran la efectividad del anión acetato, común en ambos LIs. Figura 3.6: Digestibilidad en glucano y rendimiento en glucosa de eucalipto pretratado con LI y calefacción convencional En el caso de la madera de pino, los resultados de digestibilidad y rendimiento de glucosa se muestran en la Figura 3.7. Los resultados de digestibilidad, con cada LI, siguen el siguiente orden: [Mim][Cl] < Pino sin tratar < [OHEtAm][HCO2] < [Ch][Lys] < [Ch][OAc] < [Amim][Cl < [Emim][OAc] Capítulo 3 68 Figura 3.7: Digestibilidad en glucano y rendimiento en glucosa de pino pretratado con LI y calefacción convencional A pesar de que la madera blanda es más recalcitrante que la madera dura, la mayor digestibilidad de todas las muestras pretratadas se produce con el LI convencional [Emim][OAc] (84 %). Hay que destacar que tras el pretratamiento, el balance de materia indicó que no había apenas pérdida de celulosa en el sólido pretratado, por lo que el rendimiento global del proceso es también el mayor de los obtenidos en toda la tesis doctoral. Por el contrario, la digestibilidad obtenida con los LIPs es muy baja, y en el caso del [Mim][Cl] es menor que el de la madera sin tratar. La ruptura de la lignocelulosa por la efectividad de los LIs ácidos a romper el enlace β-O-4 de la lignina, unido a la fuerte interacción con la celulosa del anión Cl,- ha podido dar lugar a la formación de inhibidores en la superficie de la biomasa (Cox et al., 2011; De Gregorio et al., 2016). Los LIs derivados de la colina dan como resultado digestibilidades de en torno al 54 %. Además, con este tipo de LIs no se degrada la celulosa durante el pretratamiento. En el caso del LI [Amim][Cl], la digestibilidad alcanzada es menor que con [Emim][OAc] lo que coincide con el trabajo realizado por Torr et al. (2016). Los sólidos obtenidos tras la etapa de hidrólisis enzimática (corriente 5) también se caracterizan para determinar la estructura y enlaces de la lignina remanente. Los espectros hsqc (de correlación 13C-1H) de la región aromática y de las cadenas laterales de los sólidos posthidrolizados con algunos de los LIs más ilustrativos se muestran en las Figuras 3.8 y 3.9, respectivamente. Además, se toma como referencia una lignina organosolv de madera de pino y eucalipto obtenidas con una mezcla de etanol/agua al 60 % (en peso) a 200 °C y 50 minutos, basadas en un trabajo anterior del grupo de investigación (Santos, 2017). Pretratamientos en una etapa 69 Figura 3.8: Espectros 2D-NMR de la región aromática de a) lignina organosolv de eucalipto; b) lignina organosolv de pino; c) sólido de eucalipto posthidrolizado, previamente pretratado con [Emim][OAc]; d) sólido de pino posthidrolizado, previamente pretratado con [Emim][OAc]; e) sólido de eucalipto posthidrolizado, previamente pretratado con [Ch][OAc] y f) sólido de pino posthidrolizado, previamente pretratado con [Ch][OAc] Capítulo 3 70 En la Figura 3.8, se observan unidades aromáticas siringilo y guayacilo en el caso de los sólidos posthidrolizados de eucalipto, mientras que en el caso de los sólidos posthidrolizados de pino sólo aparecen unidades guayacilo al ser los anillos mayoritarios (Evtuguin et al., 2001; Crestini et al., 2017). En los sólidos posthidrolizados de madera de eucalipto, se observaron mayores relaciones S/G usando el LI [Emim][OAc], lo cual parece indicar que este LI extrae mayor proporción de unidades G que de unidades S. Esto no ocurre con el sólido posthidrolizado de eucalipto procedente del tratamiento con [Ch][OAc], el cual posee tanto unidades siringilo como guayacilo, por lo que podría no haber selectividad hacia los guayacilo y por ello la digestibilidad de eucalipto pretratado con [Ch][OAc] (69 %) es mayor que la del pino pretratado con este LI (55 %). Los sólidos posthidrolizados de eucalipto y pino pretratados con [Ch][OAc] muestran además estructuras a δC/δH=129,3/7,68 ppm características de C2,6-H2,6, presente en estructuras p-cumarato (Shinde et al., 2018). Esto se atribuye en algunos trabajos a la presencia de acetatos (Del Rio et al., 2008). Sin embargo, en este caso llama la atención no observar la presencia de otros enlaces característicos de la estructura p-cumarato, como C7-H7, C8-H8 y C3,5-H3,5 a δC/δH=143,5/7,52 ppm, δC/δH=117,4/6,45 ppm y δC/δH=122,1/7,14 ppm, lo que podría indicar que estas señales no se atribuyen a estas estructuras (Wen et al., 2014). Asumiendo la presencia de estructuras p-cumarato acetiladas, los aniones pueden proceder del propio LI, lo que puede haber dado lugar a su degradación, su disociación y la acetilación de la madera. Las estructuras identificadas en las ligninas organosolv también poseen estructuras correspondientes a los grupos finales sinapaldehídos (C2,6-H2,6 a δC/δH=105,9/7,03 ppm), grupos finales cinamaldehídos (C8-H8 a δC/δH=126,3/6,76 ppm) y ferulatos (C2-H2 a δC/δH=110,8/7,32 ppm y C6-H6 a δC/δH=122,7/7,16 ppm). En la Figura 3.9 se muestra el espectro de las cadenas laterales de la lignina. Se aprecia una mayor cantidad de estructuras no oxigenadas en los sólidos posthidrolizados obtenidos tras el pretratamiento con [Ch][OAc]. Estas estructuras se deben tanto a las cadenas laterales de la lignina, como a las hemicelulosas acetiladas, confirmando la posible acetilación de la biomasa a causa del anión acetato del LI (Santos et al., 2017). En todas las muestras se observó la presencia de estructuras β-O-4 alquil aril éter (A) en cada una de sus posiciones (Cβ-Hβ a δC/δH= 83,4/4,27 ppm; Cγ-Hγ en subestructuras γ-acetiladas β- O-4’ δC/δH= 63,5/3,83 y 4,30 ppm; Cγ-Hγ en subestructuras γ-hidroxiladas β-O-4’ δC/δH= 59,4/3,40 y 3,72 ppm; Cα-Hα en β-O-4 subestructuras alquil aril éter unidas a enlaces guayacilo a δC/δH= 70,9/4,71 ppm; y Cα-Hα en β-O-4 subestructuras alquil aril éter unidas a unidades siringilo a δC/δH= 71,8/4,83 ppm). Pretratamientos en una etapa 71 Figura 3.9: Espectros 2D-NMR de la cadena lateral de a) lignina organosolv de eucalipto; b) lignina organosolv de pino; c) sólido de eucalipto posthidrolizado, previamente pretratado con [Emim][OAc]; d) sólido de pino posthidrolizado, previamente pretratado con [Emim][OAc]; e) sólido de eucalipto posthidrolizado, previamente pretratado con [Ch][OAc] y f) sólido de pino posthidrolizado, previamente pretratado con [Ch][OAc] Capítulo 3 72 Las estructuras fenilcumaran (B) se encuentran unidas a anillos guayacilo en las muestras de lignina organosolv, así como en el pino posthidrolizado pretratado con [Emim][OAc] (Cβ-Hβ a δC/δH= 53,1/3,45 ppm; Cγ-Hγ a δC/δH= 63,7/3,68 ppm; y Cα-Hα a δC/δH= 86,8/5,43 ppm) mientras que en el resto de muestras no se observaron dichas estructuras. Por último, en las ligninas organosolv procedentes de madera de pino y de eucalipto, así como en los sólidos posthidrolizados de madera de pino y de eucalipto pretratados con [Ch][OAc] aparecen estructuras resinol (C) (Cβ- Hβ a δC/δH= 53,5/3,06 ppm; Cγ-Hγ a δC/δH= 71,0/3,82 y 3,48 ppm; y Cα-Hα a δC/δH= 84,9/4,67 ppm) (Kim & Ralph, 2010; del Río et al., 2012). En la Figura 3.10 se muestran los espectros de FTIR-ATR de los LIs antes y después del pretratamiento de madera de pino y eucalipto. Hay que tener en cuenta que los espectros del LI fresco no contienen agua, pero en el caso de los LIs recuperados, las muestras todavía pueden contener hasta un 15 % de agua. No se aprecian grandes diferencias entre los espectros de LI recuperado del proceso de pretratamiento de madera de pino y los espectros de eucalipto, con ninguno de los LIs ensayados. Figura 3.10: Espectros FTIR-ATR de los LIs empleados: a) [OHEtAm][HCO2], b) [Mim][Cl], c) [Emim][OAc], d) [Amim][Cl], e)[Ch][OAc], f) [Ch][Lys] En el caso del LI [OHEtAm][HCO2], la intensidad de la banda a 1340 cm-1 disminuye tras su utilización y recuperación, lo que se atribuye al estiramiento C-N. La banda a 1589 cm-1 asignada Pretratamientos en una etapa 73 al estiramiento asimétrico C-O-O también disminuye (Papadopoulou et al., 2016). El espectro del [Mim][Cl] también se ve modificado, decreciendo la intensidad de la banda a 1116 cm-1 asignada a la vibración N1-C5-H y el ensanchamiento de la banda a 1450 cm-1 (Haddad et al., 2017). En el caso del LI [Emim][OAc], las bandas atribuidas al LI permanecieron casi inalteradas. La banda a 1394 cm-1 (vibración H-C-H) se desplazó debido a la presencia de agua (Dhumal et al., 2009). Las bandas a 1005 cm-1 y 1174 cm-1 correspondientes al estiramiento C-O y C=O del acetato, respectivamente, se mantuvieron (Weerachanchai & Lee, 2014). Del mismo modo, tampoco se observaron cambios representativos en los espectros del [Amim][Cl]. En el caso del [Ch][OAc], las bandas 1089 cm-1 y 1011 cm-1 correspondientes al estiramiento asimétrico del C3(CH3)N-CH2 disminuyeron su intensidad (Harmon et al., 1986). Otras bandas por el contrario aumentaron su intensidad como la 1487 cm-1 y la 1475 cm-1, correspondientes a la vibración de la colina. Por último, el [Ch][Lys] mantuvo sus bandas representativas como la 1569 cm-1 debida al estiramiento asimétrico del carboxilato C=O, 1475 cm-1 por la vibración C-H, 1091 cm-1 correspondiente al giro de los grupos CHn y 955 cm-1 correspondiente al estiramiento C-N (Tanzi et al., 2014; Campetella et al., 2015; Foulet et al., 2016). Con todo esto, se puede afirmar que los espectros de los LIs [Emim][OAc], [Amim][Cl] y [Ch][Lys] mantienen sus bandas representativas antes y después de su utilización, mientras que en el resto sí que se observan cambios en los enlaces que pueden ser atribuidos a la presencia de impurezas, de agua o a una degradación del LI durante el pretratamiento. Aunque utilizando esta técnica espectroscópica no se puede precisar qué es lo que les ha ocurrido a los LIs en aquellos espectros que han cambiado tras el pretratamiento, permite analizar y comprobar si las bandas representativas de cada LI se mantienen o desaparecen. En la Figura 3.11 se muestran las derivadas de las curvas de degradación térmica del LI antes y después de su utilización y recuperación. Como parámetro de comparación se utilizó la temperatura de pico (Tp), que es la temperatura a la cual la derivada alcanza su mínimo valor. Atendiendo a la Tp, el orden de estabilidad térmica de cada uno de los LIs es el siguiente: [Ch][Lys] (199 °C) < [Ch][OAc] (219 °C) < [OHEtAm][HCO2] (245 °C) < [Emim][OAc] (248 °C) < [Mim][Cl] (265 °C) < [Amim][Cl] (275 °C) En el caso de los LIPs, la Tp ha disminuido del LI fresco a los utilizados en el pretratamiento (Figura 3.11.a y 3.11.b) indicando, bien que los LIs se han degradado o que algunos compuestos se acumulan en el LI; como es el caso de las hemicelulosas, que se degradan en el intervalo de 150- Capítulo 3 74 260 °C (Prado et al., 2016). Además, el balance de materia del proceso indicó que el 32 y el 43 % de las hemicelulosas presentes en el pino y el eucalipto, respectivamente, se extrajeron con [OHEtAm][HCO2], y el 97 y el 98 % de las hemicelulosas del pino y del eucalipto se extrajeron con [Mim][Cl]. Por otro lado, la Tp del LI recuperado (LIR) con eucalipto es menor que con pino, del mismo modo que se extraen más hemicelulosas en ambos casos con eucalipto que con pino. Figura 3.11: Derivadas de los termogramas de los LIs utilizados: a) [OHEtAm][HCO2], b) [Mim][Cl], c) [Emim][OAc], d) [Amim][Cl], e)[Ch][OAc], f) [Ch][Lys] En el caso de los LIs derivados de colina, la Tp ha disminuido del LI fresco a los utilizados en el pretratamiento (Figura 3.11.e y 3.11.f). La curva de [Ch][OAc] sigue manteniendo la misma forma, mientras que las curvas del [Ch][Lys] cambian considerablemente debido a los diferentes pasos de descarboxilación del aminoácido del anión (Bhattacharyya & Shah, 2018). Por el contrario, los LIAs mostraron Tp iguales o mayores a las del LI fresco (Figura 3.11.c y 3.11.d). Las Tp mayores que las del LI fresco obtenidas con el [Amim][Cl] se atribuyen a la presencia de impurezas de lignina, que suele degradar en el intervalo de temperaturas de 260 y 400 °C (Weerachanchai & Lee, 2014; Prado et al., 2016). Pretratamientos en una etapa 75 A modo de ejemplo, se muestra en la Figura 3.12 los termogramas del LI [Mim][Cl] fresco y el recuperado tras el pretratamiento de madera de pino y eucalipto. Figura 3.12: Ejemplo de termograma del [Mim][Cl] En todos los casos, de nuevo hay que tener en cuenta que los LIs recuperados tenían agua, lo que se evidencia por la pérdida de masa a 100-105 °C. 3.3. Pretratamiento con líquidos iónicos con calefacción microondas En este bloque se ha estudiado el proceso de pretratamiento de madera de pino y de eucalipto con el LI [Emim][OAc] a diferentes temperaturas utilizando calefacción por microondas. Se decide utilizar este LI convencional por ser el más estudiado y, por tanto del que más datos se disponen para interpretar los resultados empleando calefacción por microondas. Además, se realiza un ensayo a 80 °C con [Amim][Cl], para compararlo con el [Emim][OAc] ya que también se ha demostrado que el [Amim][Cl] es capaz de pretratar el material lignocelulósico en el apartado 3.2. El procedimiento experimental, la caracterización y los resultados obtenidos se recogen en la Publicación I y se resumen a continuación. 3.3.1. Estructura y procedimiento experimental El planteamiento experimental del pretratamiento con LIs utilizando calefacción por microondas se muestra en la Figura 3.13. En este esquema se recogen las etapas de acondicionamiento, pretratamiento en horno microondas, regeneración, lavado, recuperación del LI e hidrólisis enzimática. De cada una de las etapas se obtienen corrientes sólidas (corrientes 1 y 3) y/o líquidas (corrientes 2, 4, 5 y 6) para analizar y caracterizar. Capítulo 3 76 igura 3.13: Planteamiento experimental del proceso de pretratamiento con líquidos iónicos y calefcción convencional F ig u ra 3 .1 3 : P la n tea m ien to ex p erim en ta l d el p ro ceso d e p retra ta m ien to co n líq u id o s ió n ico s y ca lefa cció n m icro o n d a s. L a s flech a s en d isco n tin u o in d ica n la s co rrien tes a a n a liza r y lo s n ú m ero s in d ica n la s técn ica s d e ca ra cteriza ció n a p lica d a s a ca d a co rrien te Pretratamientos en una etapa 77 La madera se acondiciona previamente, como ya se explicó en el apartado 3.2. A continuación se lleva a cabo el pretratamiento en este caso en un horno microondas Berghof SpeedWave Four. El perfil de calentamiento que se sigue en el horno microondas está compuesto por 4 etapas, según se muestra en la Tabla 3.5. Tabla 3.5: Perfil de temperaturas utilizado en el horno microondas Etapa Temperatura (℃) Tiempo (min) Fusión/calentamiento (I) Hasta Top-10 ℃ 10 Fusión/calentamiento (II) Isoterma a Top-10 ℃ 10 Disolución de la madera en el LI (I) Desde Top-10 ℃ Hasta Top 10 Disolución de la madera en el LI (II) Isoterma a Top ℃ 20 Top= Temperatura de operación El intervalo de temperaturas con el que se ha trabajado con este tipo de calefacción va de los 50 °C a los 150 °C, a un tiempo fijo de 50 minutos. Finalizado el proceso, se regenera la madera como se ha descrito en el Apartado 3.2, obteniendo la primera fracción líquida compuesta principalmente por LI y agua (corriente 2). A continuación, se lava con 70 mL de agua en cada paso, repitiendo esta etapa 5 veces y recogiendo cada una de estas fracciones (corriente 4). Una vez el sólido ha sido lavado, este se recoge para caracterizar (corriente 3) y a continuación se somete a hidrólisis enzimática. Por otra parte, el agua de la fracción líquida se elimina con un rotavapor, recuperando el LI (corriente 5). 3.3.2. Caracterización de las fracciones obtenidas Tras el proceso de pretratamiento con LI, se caracterizaron tanto la fracción sólida como la fracción líquida. La fracción sólida de madera no tratada (corriente 1) y pretratada (corriente 3) se caracteriza de acuerdo a la metodología NREL/TP-510-42618, como se ha descrito en el apartado 3.1.1. Además, para verificar que el sólido se ha lavado correctamente y que no hay presencia de LI en él, se emplea espectroscopía FTIR, comparando los espectros obtenidos en la madera sin tratar, la madera pretratada y el espectro del LI con el que se ha trabajado en el pretratamiento. Para la preparación de la muestra, se mezcla la madera pretratada con KBr en relación 1/200. La mezcla se muele en un mortero de ágata formando la pastilla con una prensa hidráulica de la marca Specac aplicando una presión de 7 toneladas durante 30 segundos. La adquisición del espectro se realiza con 16 barridos en la región del espectro 4.000-400 cm-1, aunque la mayoría de las bandas de caracterización de la madera se observan en la región 2.000-600 cm-1. El software de adquisición y tratamiento del espectro es Spectra Manager. Capítulo 3 78 La fracción líquida, compuesta por LI y agua obtenida tras la etapa de regeneración (corriente 2) y cada uno de los lavados (corriente 4), se cuantifica por HPLC utilizando un detector UV a 235 nm. A continuación, se caracteriza el LI recuperado tras la eliminación del agua en un rotavapor (corriente 5). Para evaluar el contenido en lignina acumulada, las muestras de LIR se analizan por espectroscopía de UV-VIS utilizando un Varian Cary 50 scan UV/VIS. La corriente 6 obtenida después de la hidrólisis enzimática realizada de acuerdo a la norma NREL/TP-5100-63351 se caracteriza tal y como se explica en el apartado 3.1. La descripción detallada de la caracterización realizada, así como la explicación detallada del procedimiento experimenta y los equipos utilizados se muestran en la Publicación I. 3.3.3. Resultados experimentales. Interpretación y discusión En la Figura 3.14 se muestra la composición de la fase sólida obtenida tras el pretratamiento con calefacción microondas. Los valores de recuperación del sólido después del pretratamiento con [Emim][OAc] se encuentran entre el 81 y el 90 %, siendo la muestra sometida a mayor temperatura (150 °C) la que presenta mayor degradación, lo que indica que se debe trabajar a temperaturas inferiores a 150 °C. Por otro lado, se llevó a cabo un ensayo a 80 °C con el LI [Amim][Cl] para compararlo con los resultados con [Emim][OAc] utilizando madera de pino. El rendimiento del sólido recuperado con [Amim][Cl] fue de un 84 %, menor que con el LI [Emim][OAc] donde se recuperó el 91 %. Esto, unido a la elevada viscosidad de [Amim][Cl] y su carácter sólido a temperatura ambiente, hizo que se decidiera seguir planteando el resto de ensayos con [Emim][OAc] (Elgharbawy et al., 2016b). Los ensayos a 50 °C y a 150 °C con calefacción microondas se llevaron a cabo también solo con madera de pino, ya que se comprobó que a temperaturas tan bajas y tan altas, la madera prácticamente no se alteraba o se degradaba en exceso, respectivamente. A las temperaturas de 80 y 120 °C, se comprueba con ambas maderas que la composición del sólido pretratado se ve alterada. La composición relativa en lignina apenas disminuye (con respecto de las maderas sin tratar) con el incremento de la temperatura del pretratamiento. Esto sucede ya que el proceso implica la disolución de la madera, seguida de un paso de regeneración de la madera disuelta sin que se produzca degradación y sin separar ninguna de las fracciones. Además, aunque casi todo el glucano inicial permanece en el sólido pretratado (94 % de glucano introducido), en el caso del eucalipto el sólido pretratado contiene alrededor de un 22 % menos de glucano. La observación del pico de glucosa en el cromatograma del [Emim][OAc] recuperado indica que esta celulosa se degrada por el pretratamiento. La degradación parcial de la celulosa se puede atribuir al efecto del LI en sí mismo, y no al empleo de calefacción microondas que acelera el proceso, ya que este mismo comportamiento se ha observado con calefacción convencional (Brandt et al., 2013). Pretratamientos en una etapa 79 Figura 3.14: Composición y rendimiento del sólido obtenido tras el pretratamiento con LIs con calefacción microondas En la Tabla 3.6 se muestra la recuperación del LI en cada uno de los lavados realizados. El análisis indicó que hacen falta al menos tres lavados (aparte de la etapa de regeneración en la que ya se emplea agua) para recuperar por completo el LI impregnado en el sólido. Tabla 3.6: Recuperación de [Emim][OAc] en cada lavado Cantidad de LI recuperado (%) Lavados Muestra Regeneración 1º 2º 3º 4º 5º MO50_[Emim][OAc]_P 81,2 13,7 2,2 0,1 0 0 MO80[Emim][OAc]_E 74,1 26,2 4,3 0,6 0 0 MO80_[Emim][OAc]_P 72,9 16,2 0,9 0 0 0 MO120_[Emim][OAc]_E 87,3 12,5 2,0 0,2 0 0 MO120_[Emim][OAc]_P 71,0 13,3 4,9 0,5 0 0 MO150_[Emim][OAc]_P 89,3 12,0 3,0 0,4 0,2 0 Capítulo 3 80 Esto equivale a utilizar 13 mL de agua/g de [Emim][OAc] para asegurar la completa eliminación del LI del sólido. La mayoría del LI se recupera en la etapa de regeneración (70-89 % del LI total). Al menos un 89 % del LI introducido se ha recuperado tras eliminar el agua con un rotavapor, lo cual es un poco inferior a las recuperaciones que obtuvo Lozano et al. (2012) quien recuperó entre un 95 % y un 99 % del LI introducido en el proceso. La Figura 3.15 muestra la cantidad de lignina acumulada en el LI recuperado con respecto a la que tiene la madera inicialmente. Los resultados varían entre 0,34 y 10,81 g de lignina, lo cual demuestra la capacidad de extracción del [Emim][OAc] hacia este compuesto, utilizando calefacción por microondas (Ngoc Lan Maia, 2014). Figura 3.15: Lignina acumulada en [Emim][OAc] En el caso de la madera de pino, el incremento de la lignina acumulada es bajo hasta la temperatura de pretratamiento de 120 °C, y aumenta considerablemente a la temperatura de 150 °C (10,81 g de lignina/100 g de lignina introducida). Esto hace creer que parte de la lignina se ha degradado y que también otros compuestos presentes en el [Emim][OAc] recuperado pueden interferir en la medida, incrementando su valor. En el caso del eucalipto, el efecto de la temperatura supone un aumento considerable de la lignina extraída de 80 a 120 °C, lo cual indica que la capacidad de extracción del LI depende en parte del tipo de estructuras que tenga la lignina presente en la madera a pretratar. Pretratamientos en una etapa 81 En la Figura 3.16 se muestra la digestibilidad de la madera de pino y de eucalipto pretratada con LIs y calefacción por microondas a las 72 h. Con ambos tipos de madera, se observa claramente la eficacia del pretratamiento con estos LIs en horno microondas. A iguales condiciones de pretratamiento, los rendimientos finales a glucosa disminuyen más utilizando eucalipto que pino. Por ejemplo, la digestibilidad de MO80 [Emim][OAc]_E es 55 % y su rendimiento 45 %, mientras que en MO80_[Emim][OAc]_P pasa de una digestibilidad del 72 % a un rendimiento del 68 %. Esto indica que la radiación microondas está degradando más la celulosa procedente del eucalipto que la del pino. Figura 3.16: Digestibilidad en glucano y rendimiento en glucosa de eucalipto pretratado con LI y calefacción con microondas Destacar de nuevo que el pretratamiento con el LI [Amim][Cl] presenta digestibilidades elevadas (66 %), aunque no son tan altas como con el [Emim][OAc] (72 %). Como ya se comentó, la degradación de la celulosa en la muestra de pino pretratada con [Emim][OAc] a 150 °C indica que esta temperatura es demasiado elevada. En el caso del eucalipto, el aumento de la temperatura en el pretratamiento da lugar a un incremento de digestibilidad de 55 a 68 %, mientras que en el caso de pino la digestibilidad mejora del 72 al 78 %. De nuevo, la digestibilidad es mayor con pino que con eucalipto como ya se observó con calefacción convencional con este LI (Li et al., 2013a). Capítulo 3 82 3.4. Recuperación y reutilización del líquido iónico El estudio del reciclado del [Emim][OAc] ya se ha realizado en diferentes procesos de pretratamiento de biomasa y Clough et al. (2015) han observado la pérdida de aldehídos durante el pretratamiento (Auxenfans et al., 2012 y 2014). La estabilidad térmica de [Amim][Cl] y sus razonables digestibilidades hacen que se decida optar por estudiar la reutilización de este LI. En este bloque se va a evaluar la capacidad de recuperación del LI [Amim][Cl] y su reutilización con madera de eucalipto y de pino. El procedimiento experimental, la caracterización y los resultados obtenidos se recogen de manera detallada en la Publicación III y a continuación se exponen de forma resumida los resultados más relevantes. 3.4.1. Estructura y procedimiento experimental El planteamiento experimental del proceso de recuperación y reutilización del [Amim][Cl] utilizando calefacción por microondas se muestra en la Figura 3.17. En este esquema se incluyen las etapas de acondicionamiento, pretratamiento en horno microondas, regeneración del sólido rico en celulosa (SRC), separación del sólido rico en lignina (SRL), lavado, recuperación del LI, reutilización del LI y digestibilidad enzimática. De cada una de las etapas se obtienen corrientes sólidas (corrientes 1, 2 y 3) y/o líquidas (corrientes 4, 5 y 6) para analizar y caracterizar. La madera se acondiciona y, a continuación, se lleva a cabo su disolución en horno microondas de acuerdo al apartado 3.3, con la salvedad de que en este caso en lugar de introducir LI y madera, se introduce también una pequeña cantidad de co-solvente, dimetilsulfóxido (DMSO). Una vez terminada la etapa de disolución, se separa mediante centrifugación la fracción sólida (SRL) de la fracción líquida (SRC con LI y DMSO). Para ello, se añaden 1,38 g DMSO/g LI que disminuyen la viscosidad de la mezcla. La fracción rica en lignina (SRL) se lava sucesivas veces, recogiéndose tanto las fracciones de lavado como la fracción sólida (corrientes 2 y 3). Por otro lado, la fracción líquida (formada por madera disuelta + LI) se lleva a una etapa de regeneración adicional (esta vez con metanol) para regenerar la madera disuelta, y a una posterior etapa de lavado. El empleo de metanol permite llevar a cabo la recuperación del LI a menores condiciones de temperatura y presión en el rotavapor. Tanto el SRC como el SRL se llevan a una etapa posterior de hidrólisis enzimática para conocer su digestibilidad. Todos los lavados se recogen (corriente 4) y el LI se recupera, eliminando el metanol y el DMSO mediante destilación a vacío. El LI recuperado (corriente 5) se analiza y se añade, si es necesario, DMSO para partir de las mismas condiciones iniciales y se reutiliza en el ciclo siguiente. El LI se reutiliza 5 veces. Para asegurar que en todos los ciclos las cantidades utilizadas se mantienen, el proceso se repite 5, 4, 3, 2 y 1 vez en el 1º, 2º, 3º, 4º y 5º ciclo, respectivamente. Pretratamientos en una etapa 83 Figura 3.17: Planteamiento experimental del proceso de recuperación y reutilización del LI [Amim][Cl] F ig u ra 3 .1 7 : P la n te a m ie n to e x p er im en ta l d el p ro ce so d e re cu p er a ci ó n y r eu ti li za ci ó n d e L I [A m im ][ C l] . L a s fl ec h a s en d is co n ti n u o in d ic a n l a s co rr ie n te s a a n a li za r, s eñ a la n d o c o n n ú m er o s la s té cn ic a s d e ca ra ct er iz a ci ó n a p li ca d a s a c a d a c o rr ie n te . Capítulo 3 84 3.4.2. Caracterización de las fracciones obtenidas Las fracciones sólidas recuperadas, SRL (corriente 2) y SRC (corriente 3), se pesan y cuantifican para conocer el rendimiento obtenido en cada fracción y en cada ciclo. El SRL se caracteriza de acuerdo a la norma NREL/TP-510-42618. Debido a las pequeñas cantidades de muestra con las que se trabaja, la composición del sólido rico en celulosa (SRC) se determina solamente mediante FTIR. En el caso del SRC, se determina también su cristalinidad al estar mayoritariamente compuesto por celulosa. El desarrollo de la hidrólisis enzimática se ha llevado a cabo de acuerdo a la norma NREL/TP-5100-63351, tal y como se explica en el apartado 3.1 La fracción líquida, compuesta por LI, DMSO y metanol, obtenida tras cada uno de los lavados (corriente 4), se cuantifica por HPLC. El rendimiento del LI recuperado (ƞLIR) se calcula de acuerdo a la Ecuación 3.1. ƞ𝐿𝐼𝑅(%) = 𝑚𝐿𝐼𝑅×( 100−𝑤𝑑𝑣𝑡𝑒(%) 100 ) 𝑚LI0 × 100 [3.1] donde mLIR es la cantidad de LIR obtenido en el rotavapor, wdvte es la fracción de disolvente aún presente en el líquido iónico recuperado tras la destilación y mLI0 es la masa inicial de LI introducida en el proceso. Una vez recuperado el LI, se procede a su caracterización (corriente 5). Para observar cambios en los grupos funcionales del LI se utiliza las espectroscopía de infrarrojo por Transformada de Fourier con Reflectancia Total Atenuada (FTIR/ATR). La resonancia magnética nuclear de protón (1H-RMN) se utiliza para obtener una caracterización más precisa de los LIR en comparación con el LI fresco. Las muestras de LIR se analizan por espectroscopía de UV-VIS para evaluar el contenido en lignina. El estudio de la degradación térmica del [Amim][Cl] reutilizado se lleva a cabo utilizando ensayos termogravimétricos. 3.4.3. Resultados experimentales. Interpretación y discusión A continuación se recogen los resultados del reciclado de [Amim][Cl] tras el procesado de madera de eucalipto. Este mismo estudio empleando madera de pino se muestra en la Publicación III. En ambos casos se extrajeron conclusiones análogas. 3.4.3.1. Gravimetría del proceso: rendimiento de los sólidos recuperados En la Figura 3.18 se muestra la recuperación de SRC y SRL, así como las pérdidas que se han producido a lo largo del proceso. Pretratamientos en una etapa 85 Figura 3.18: Rendimiento de sólidos recuperados con [Amim][Cl] recuperado en cada ciclo para madera de eucalipto La cantidad global de sólidos recuperados se encuentra entre un 81 y un 87 % en los diferentes ciclos de operación. Estas pérdidas son, principalmente, producidas en la operación de filtración (sólido retenido en el filtro húmedo), o al quedarse el sólido impregnado en los diferentes recipientes utilizados. En el SRC existe una gran variación entre los ciclos 2 y 3, mientras que las diferencias entre los demás ciclos son inferiores. Asumiendo que el SRC está formado casi únicamente por holocelulosa, en el primer ciclo se regeneran alrededor del 21 % de las holocelulosas presentes en la biomasa inicial, y en el quinto ciclo, esta regeneración es de apenas un 13 %, es decir, el rendimiento de obtención del SRC disminuye un 36 % tras cinco ciclos. Con respecto al SRL, la madera no disuelta aumenta a medida que se suceden los ciclos de reutilización del líquido iónico. En el primero de los ciclos se obtiene el 66 % de la biomasa introducida, mientras que en el último ciclo se obtiene alrededor de un 78 % de la cantidad de madera inicial, lo que supone un aumento en la cantidad de madera no disuelta de en torno al 15 % del primer al último ciclo, indicando una pérdida de su capacidad de disolución. 3.4.3.2. Composición del sólido rico en lignina recuperado en cada ciclo En la Figura 3.19 se muestra la composición del SRL. El contenido relativo en glucano disminuye significativamente frente al eucalipto sin tratar, viéndose aumentado el porcentaje de lignina y confirmando la eficacia del [Amim][Cl] como disolvente de glucosa (Zhang et al., 2005). Capítulo 3 86 Figura 3.19: Composición del SRL obtenido con [Amim][Cl] recuperado en cada ciclo para madera de eucalipto En este caso, y a diferencia de lo observado con el pino en la Publicación III, el porcentaje en glucano permanece casi inalterado en todos los ciclos, lo que indica que la capacidad de disolución es bastante parecida. 3.4.3.3. Caracterización del sólido rico en celulosa recuperado en cada ciclo En la Figura 3.20 se muestra el espectro de FTIR del SRC. Se puede observar la ausencia de la banda de vibración del anillo aromático de la lignina a 1505 cm-1 en todos los SRC. La ausencia de la banda de flexión C-H en el anillo imidazol a 1165 cm-1 y la de vibración del cloro a 763 cm-1 indica que el lavado permitió eliminar completamente el [Amim][Cl] del SRC (Li et al., 2013b; Liu et al., 2014). A 1383 cm-1 se encuentra la banda correspondiente a la flexión del enlace C-H presente en la celulosa, y relacionado con la cristalinidad del sólido (Carrillo et al., 2004). El aumento de la intensidad de esta banda se debe a la reducción de la cristalinidad de los sólidos, como se comprobó posteriormente con los resultados de difracción de rayos X (DRX) en este mismo apartado. Además, se observan bandas características de las hemicelulosas como a 1731, 1260 y 802 cm-1, correspondientes a los enlaces C=O no conjugados de los xilanos, el estiramiento C-O de arabino-4-o-metilglucuronoxilano y la banda que referencia la presencia de glucomananos (Pandey & Pitman, 2003; Carrillo et al., 2004; Schwanninger et al., 2004). Pretratamientos en una etapa 87 Figura 3.20: Espectros de FTIR del SRC recuperado obtenido con [Amim][Cl] en cada ciclo para madera de eucalipto La disminución de la cristalinidad del SRC mediante difracción de rayos X se muestra en la Figura 3.21. El pico característico de la celulosa I a 2θ≈22º y el de la celulosa amorfa a 2θ≈16,6º no se distinguen, mientras que el pico de la celulosa cristalina se ha desplazado y alargado a 2θ≈20-22º, típico de la celulosa II (Blokhin et al., 2011; Lan et al., 2011). Figura 3.21: Cristalinidad del SRC recuperado en cada ciclo para madera de eucalipto Capítulo 3 88 Ajustando estos picos a curvas gaussianas, la altura a mitad de pico (FWHM) calculada a partir de la gaussiana obtenida sigue el siguiente orden: 6º ciclo < 5º ciclo < 3º ciclo < 4º ciclo < 1º ciclo < 2º ciclo Así se observa una tendencia de aumento de la amorficidad al reutilizar el [Amim][Cl] en sucesivos ciclos (Adel et al., 2011). 3.4.3.4. Recuperación de los disolventes y el LI La determinación de la cantidad de LI, DMSO y metanol determinada por HPLC permite optimizar las etapas de lavado para así poder minimizar el consumo de disolvente. A modo de ejemplo, esta determinación se muestra en la Figura 3.22 para la recuperación de las fracciones de lavado del primer ciclo. Figura 3.22: Recuperación de LI, DMSO y metanol en cada lavado para la madera de eucalipto En este caso particular, la mayoría del LI se recuperó en la etapa de regeneración. Sin embargo, es necesario realizar dos lavados del SRL con un total de 100 mL de antidisolvente para asegurar la correcta eliminación del LI. Al final del proceso de lavado se consiguió recuperar un 99,2 % del LI introducido inicialmente en el proceso y tras la etapa de destilación a vacío, se consigue recuperar un 91,3 %. La dificultad en el lavado de los sólidos pretratados se atribuye a la elevada interacción que existe entre el LI y la madera, ya comentada con los parámetros de Kamlet- Taft en el apartado 3.2. El metanol posee un alto valor de acidez de puentes de hidrógeno (α = 0,98) y una energía molar de transición también elevada (ET N= 0,76) por lo que puede interaccionar con el LI, desplazando los puentes de hidrógeno que se forman entre el LI y la celulosa (Weerachanchai et al., 2014). Pretratamientos en una etapa 89 3.4.3.5. Caracterización del LI recuperado en cada ciclo En la Figura 3.23 se muestran los espectros del [Amim][Cl] de cada uno de los ciclos, junto con el espectro del [Amim][Cl] fresco. Se observa una disminución de la intensidad de la banda a 3382 cm-1, correspondiente a la vibración extensión del enlace O-H debido a la higroscopicidad del LI (Liu et al., 2014; Kathirgamanathan et al., 2015). La disminución de esta banda está relacionada con una pérdida de efectividad del LI, puesto que al igual que disminuye la cantidad de agua que absorbe, también disminuye su capacidad para disolver la celulosa. Los espectros del [Amim][Cl] en cada ciclo no han presentado variaciones sustanciales, permaneciendo todas las bandas existentes a medida que se suceden los ciclos de reutilización. Se produce la aparición, en el [Amim][Cl] recuperado únicamente en el primer ciclo, de una banda a la longitud de onda de 1260 cm-1, que se corresponde con un estiramiento asimétrico C-O- O en los grupos acetilo presentes en las hemicelulosas (Himmelsbach et al., 2002). Sin embargo, la ausencia de esta banda en el posterior LI indica que sí que se ha producido una mayor precipitación de hemicelulosa en el siguiente ciclo, no encontrándose esta banda en el [Amim][Cl] recuperado en el resto de los ciclos. Figura 3.23: Espectros FTIR-ATR del [Amim][Cl] recuperado en cada ciclo para madera de eucalipto Capítulo 3 90 A pesar de que todas las bandas características aparecen en todos los espectros FTIR-ATR, el análisis visual permitió observar que el LI se iba oscureciendo cada vez más a lo largo de los ciclos, como se muestra en la Figura 3.24. Figura 3.24: Fotografías del [Amim][Cl] fresco y recuperado en cada ciclo Para cuantificar el contenido de lignina, se decidió utilizar la espectroscopía de UV-VIS. Debido a la alta absorbancia del líquido iónico en el intervalo de longitudes de onda de 190-380 nm, la determinación de la cantidad de lignina disuelta en éste no ha podido ser realizada a la longitud de onda típica de 280 nm, como realizó Kline et al. (2010) en sus estudios. Además como el LI absorbe en el intervalo de IR medido, la presencia de lignina era enmascarada con FTIR por la elevada señal del LI. En la Figura 3.25 se muestra la evolución del contenido de lignina en el [Amim][Cl] recuperado a lo largo de los ciclos de reutilización. Figura 3.25: Lignina acumulada en el [Amim][Cl] recuperado en cada ciclo para madera de eucalipto LI fresco 1º ciclo 2º ciclo 3º ciclo 4º ciclo 5º ciclo Pretratamientos en una etapa 91 En este caso, y a diferencia de lo ocurrido con FTIR-ATR, se ha podido determinar la presencia de lignina mediante espectroscopía de UV-VIS. Los resultados demuestran que se está acumulando lignina a medida que se suceden los ciclos de reutilización del [Amim][Cl]. Otros autores han determinado la acumulación de lignina con otros LIs, pero con el [Amim][Cl] no había sido estudiado hasta la fecha (Auxenfans et al., 2012; Weerachanchai & Lee, 2014). En la Tabla 3.7 se muestran los parámetros extraídos del análisis termogravimétrico (TGA). La Tp aumenta a medida que se reutiliza el [Amim][Cl]. Aunque parece existir cierta tendencia con el resto de los parámetros determinados, la relación no es tan clara como con la Tp. La acumulación de lignina que se degrada a temperaturas mayores que las del LI podría ser la causa de este incremento (Weerachanchai & Lee, 2014). Tabla 3.7: Temperatura de descomposición, de pérdida de 25 % y 50 % de masa y cenizas a 400 °C para madera de eucalipto Ciclo de recuperación Tp (°C) T25% (°C) T50% (°C) Cenizas a 400ºC (%) LI fresco 276,5 269,0 275,6 3,1 Ciclo 1 280,6 271,7 278,1 2,5 Ciclo 2 287,3 281,2 290,4 4,5 Ciclo 3 289,0 282,9 289,9 4,6 Ciclo 4 298,9 277,3 289,6 3,8 Ciclo 5 302,2 285,1 300,1 5,4 Para poder confirmar la acumulación de compuestos de la madera en el LI, al [Amim][Cl] recuperado en el último ciclo se le añadió agua y H2SO4. Figura 3.26: Fotografías de a) [Amim][Cl] recuperado + agua y b) [Amim][Cl] recuperado + H2SO4 El resultado se observa en la Figura 3.26, donde se visualiza la regeneración de compuestos provenientes de la madera y su posibilidad de recuperarlos mediante precipitación con un antidisolvente. Capítulo 3 92 3.4.3.6. Evaluación de la digestibilidad del SRC y SRL obtenidos con [Amim][Cl] recuperado En la Figura 3.27 se muestra la digestibilidad de la madera de eucalipto pretratada con LIs y calefacción convencional a las 72 h. En el caso del SRC, la digestibilidad es superior al 70 % en todos los casos, debido a su baja cristalinidad. Sin embargo, a medida que pasan los ciclos de recuperación, se observa que esta digestibilidad disminuye. Esto indica que la eficacia del [Amim][Cl] recuperado para pretratar madera se ve disminuida a medida que aumenta el grado de reutilización del LI. En concreto, la digestibilidad del SRC se reduce del 76 al 69 %, lo que indica que el [Amim][Cl] disminuye la capacidad de pretratamiento en un 10 %. Esto podría deberse a la acumulación de lignina que se ha comprobado que existe en el LI a lo largo de los ciclos. Figura 3.27: Digestibilidad en glucano de SRC y SRL pretratado con LI recuperado En el caso del SRL, las digestibilidades son del 25 % y no se observa cambios en sus valores a lo largo de los ciclos de reutilización. 3.5. Estudio de la accesibilidad del sólido a través de técnicas microscópicas Con los pretratamiento realizados, se plantea la necesidad de emplear técnicas de visualización, como la microscopía electrónica de barrido (SEM) y la microscopía de fluorescencia confocal (MFC), para aportar información que ayude a la comparación de los procesos de pretratamiento y para comprender por qué unos pretratamientos son más efectivos que otros de cara a la hidrólisis enzimática. En este apartado se han aplicado técnicas de análisis de imagen digital para el procesado de imágenes y estudio de sus parámetros más representativos para relacionarlos con la digestibilidad. El estudio de accesibilidad completo se describe en la Publicación IV. Pretratamientos en una etapa 93 3.5.1. Cálculo de la distribución de tamaños de partícula, dimensión fractal y lacunaridad a través de microscopía electrónica de barrido La morfología de las muestras se analizó mediante SEM, utilizando el microscopio electrónico de barrido Jeol JSM 6400. Para poder observar las muestras, se les recubre con una capa de oro con el fin de hacer a las muestras conductoras. El voltaje de operación utilizado es de 20 kV. Los análisis se realizan en las instalaciones del Centro Nacional de Microscopía Electrónica de la UCM. A partir de las imágenes de SEM de 100 aumentos se determina la distribución de tamaños de partícula. Se utilizan 2 procedimientos diferentes, seleccionando el más adecuado en cada muestra. En el primero de ellos, se utiliza el software Matlab aplicando la función imbinarize (para binarizar la imagen), seguida de la función de conteo de partículas bwpropfilt, también de Matlab, o Analyze Analyze particles del software ImageJ. En el segundo de los casos, se aplica una segmentación controlada de objetos que se tocan para hacer un conteo correcto. Además, todos los resultados se chequean manualmente para evitar un conteo erróneo debido a partículas que tocan. Las partículas que se cuentan, se filtran a un intervalo de 10-250 pixeles y posteriormente se escalan a µm. De manera complementaria, se calcula tanto la dimensión fractal (Db) como la lacunaridad (Ʌ) de las muestras. El concepto de dimensión fractal se aplica a objetos o texturas no tradicionales para definir el grado de complejidad de una imagen (Mandelbrot & Pignoni, 1983). La lacunaridad es un parámetro introducido por Mandelbrot and Pignoni (1983) como complemento a la dimensión fractal, y es un índice que indica el grado de homogeneidad de una imagen desde un punto de vista multiescalar. Ambos parámetros se proponen como complemento a las imágenes para cuantificar la complejidad que se observa en las imágenes de SEM, así como su homogeneidad. En el caso de estructuras complejas, como las partículas de madera, el análisis y cálculo de la dimensión fractal se realiza dividiendo el conjunto de la imagen en piezas más pequeñas llamadas cajas, y analizando cada pieza de manera individual. Al tamaño de la caja elegido se le llama calibre (ε). La caja de calibre (ε) se va desplazando por el conjunto de toda la imagen de acuerdo a la Figura 3.28. Capítulo 3 94 Figura 3.28: Ejemplo de conteo de cajas en imagen binarizada de madera de pino pretratada De este modo, la dimensión fractal se calcula según la Ecuación 3.2: Db = lim ε→0 [ lnNε lnε−1 ] [3.2] donde N es el número de partes que conforman la figura y ε es el calibre de caja empleado. La lacunaridad, por su parte, se calcula de acuerdo a la Ecuación 3.3: Ʌε.g = (CVε.g)2 = ( σε.g με.g ) 2 [3.3] definida como el coeficiente de variación (CV) obtenido mediante el cociente de la desviación estándar (σ) dividido por la media (μ) de píxeles que hay en cada caja. De nuevo, ε es el calibre de la caja y g hace referencia a cada una de las orientaciones posibles que puede tener cada una de las cajas para el cálculo de la distribución de los píxeles por caja (Mandelbrot, 1977; Mandelbrot & Pignoni, 1983; Karperien, 2013). Para calcular Db y Ʌ se utilizan de manera directa las imágenes en escala de grises. Para ello, se asume que las imágenes en escala de grises existen en un pseudo-espacio en 3D donde cada píxel no se cuenta como 1 o como 0, sino a un nivel de intensidad en la escala de 0-255. Este tipo de adaptación del cálculo de Db y Ʌ se utiliza sobre todo en el análisis de texturas, como las que se observan en las imágenes de 400 aumentos en adelante en las que toda la imagen debe estar compuesta por la superficie de la partícula de madera. Así pues, se introduce una tercera dimensión en el cálculo llamada I (diferencia de intensidad) que hace referencia a la diferencia entre el píxel de máxima intensidad y el de mínima intensidad, de acuerdo a la Ecuación 3.4. Ii.j.ε = 1 + Intensidad máxima de píxel(i.j.ε) − Intensidad mínima de píxel(i.j.ε) [3.4] donde i y j es la posición del píxel en la imagen. De este modo, utilizando el método de variación de volumen diferencial, se cumple la siguiente aproximación. Pretratamientos en una etapa 95 Vi.j.ε~Ii.j.εε2 De manera que la dimensión fractal se calcula de acuerdo a la Ecuación 3.5. Db.gris = 3 − ( (lim ε→0 lnVε ln(1/ε) ) 2 ) [3.5] El cálculo de Db y Ʌ se realiza utilizando el complemento de Fiji Fraclac. 3.5.2. Cálculo de la relación holocelulosa/lignina superficial a partir de microscopía de fluorescencia confocal Para observar las diferencias de composición superficial de la madera pretratada, se utiliza la microscopía de fluorescencia confocal. De manera previa a la observación en el microscopio, las muestras se tiñen con una disolución al 0,1 % (v) de blanco de calcoflúor durante 5 minutos, de manera que se coloreen selectivamente los enlaces β-1-4 de los polisacáridos (celulosa y hemicelulosa). Las muestras lavadas y secas se preparan utilizando el medio de montaje Prolong® Gold (ThermoFisher Scientific). A continuación, las muestras se observan en un microscopio confocal modelo Leica SP- 2 AOBS con una lente de 60 aumentos y una longitud de paso en el eje Z de 2 μm. Para excitar las muestras, se emplea un láser a 405 nm. El intervalo de longitudes de onda de emisión es de 428- 480 nm para la holocelulosa (celulosa + hemicelulosa) y de 547-658 nm para la lignina. Para poder comparar la saturación de imágenes, se fijan los intervalos de longitud de onda de emisión y recepción, así como el estenopeico. Todas las muestras se analizaron en las Instalaciones técnicas del Centro de Citometría y Microscopía de Fluorescencia de la Universidad Complutense de Madrid. Para determinar la lignina superficial una vez adquiridas las imágenes, se procesan usando el software abierto Fiji-ImageJ siguiendo el esquema de la Figura 3.29. Figura 3.29: Procesado de imágenes para la determinación de la lignina superficial Capítulo 3 96 Para ello, se combinan ambos canales y cada una de las proyecciones en el eje Z, para dar la proyección en el eje Z combinada de máxima intensidad. Esta proyección se vuelve a dividir en la máxima proyección de cada uno de los canales. A continuación, se corrige el brillo y el contraste (si se considera necesario) de cada canal y se binariza utilizando el método de Otsu (Otsu, 1979). Finalmente, se utiliza la herramienta del histograma para cuantificar el número de píxeles de cada imagen. 3.5.3. Resultados experimentales. Interpretación y discusión En la Tabla 3.8 se muestran algunas de las imágenes adquiridas al microscopio con cada sólido pretratado por autohidrólisis de las dos maderas. Cabe destacar que, en el caso de las imágenes de MFC, en azul se representa la holocelulosa y en verde la lignina. Tabla 3.8: Imágenes al microscopio de muestras pretratadas por autohidrólisis Madera Técnica AH150 AH175 AH200 Eucalipto SEM MFC Pino SEM MFC Llevando a cabo la AH a 150 °C (condiciones suaves), la superficie es bastante lisa y no se observan agujeros/poros. Se visualiza tanto holocelulosa (probablemente hemicelulosa), como lignina. Operando a 175 °C, parte de la hemicelulosa se extrae, por lo que las estructuras, que siguen siendo bastante lisas, ahora tienen más lignina (en verde) en su superficie. Bajo condiciones severas Pretratamientos en una etapa 97 (200 °C), el eucalipto pretratado se presenta como una aglomeración de partículas más pequeñas, con mayor porosidad. Aunque de manera muy tenue, en la superficie de las partículas sólo se observa celulosa, puesto que la hemicelulosa se ha eliminado por completo, como se adelantó en el apartado 3.1.2. En el caso de la madera de pino, el comportamiento es opuesto, ya que no hay presencia de celulosa y prácticamente toda la superficie está compuesta por lignina. Este comportamiento está en consonancia con los resultados de digestibilidad obtenidos, y valida la teoría de que la lignina de madera blanda causa la inhibición de las enzimas durante la hidrólisis (Nitsos et al., 2016). La teoría de la inhibición de las enzimas con lignina de madera blanda ha sido ampliamente estudiada. A través de ensayos de inhibición enzimática en los que se lleva a cabo la sacarificación en presencia de lignina procedente de madera dura y blanda, se ha podido comprobar que su comportamiento es distinto (Rahikainen et al., 2013). Li et al. (2014) también han demostrado mediante la observación de imágenes de SEM que los pretratamientos de AH producen la formación de pseudolignina, que posteriormente da lugar a la inhibición enzimática. La aplicación de técnicas de MFC para la visualización y posterior análisis de la composición superficial de los sólidos, permite complementar los resultados que se obtienen mediante ensayos de inhibición y cuantificar la composición superficial, extrayendo conclusiones análogas. En la Tabla 3.9 se muestran algunas de las imágenes adquiridas al microscopio con los sólidos más característicos pretratados con LIs utilizando calefacción convencional. De nuevo cabe mencionar que en el caso de las imágenes de MFC, en azul se representa la holocelulosa y en verde la lignina. En los sólidos pretratados con el LI [Mim][Cl], la morfología tanto del pino como del eucalipto muestra partículas aglomeradas, grandes y sin porosidad visible. En el caso del pino, la estructura es muy lisa y la superficie de los sólidos presenta tanto celulosa como lignina. La morfología de los sólidos pretratados con LIAs es rugosa, como observó Sun et al. (2011), con mucha porosidad, ocasionada por la disolución de la celulosa y posterior regeneración en el proceso. La rugosidad y porosidad se han considerado dos de los factores clave para mejorar la digestibilidad del sólido, como se explicará más en detalle en el apartado 3.5.6. Las muestras pretratadas con LIs derivados del catión colina presentan mucha rugosidad. Además, y a diferencia de las muestras pretratadas con [Emim][OAc], las partículas son más grandes. La superficie de las muestras de madera de eucalipto pretratada con [Ch][OAc] posee una gran cantidad de lignina, lo cual indica la extracción de hemicelulosas. Sin embargo, en la madera de pino pretratada con [Ch][OAc] sigue habiendo hemicelulosas presentes en el sólido, como se ha discutido en el apartado 3.2.3, y además existe lignina en la superficie, observada al ver tanto la imagen mezcla de ambos canales (más brillante que el resto por la acoplación de los mismos) como la imagen de cada canal por separado. Capítulo 3 98 Tabla 3.9: Imágenes al microscopio de muestras pretratadas con LIs y calefacción convencional Madera Técnicas [Mim][Cl] [Emim][OAc] [Ch][OAc] Eucalipto SEM MFC Pino SEM MFC En conclusión, las imágenes de SEM permiten visualizar los cambios producidos en la madera tras el pretratamiento, y las imágenes de MFC, cambios en la química superficial de la biomasa. Además, mientras lo sólidos pretratados procedentes de la madera blanda se ve afectada por la presencia de lignina en la superficie, la presencia de lignina en la superficie de madera dura no dificulta la accesibilidad de las enzimas, siendo en este caso la deconstrucción de la biomasa y el tamaño de partícula los factores que más afectan a la digestibilidad del material. 3.5.4. Validación del estudio de la accesibilidad del sólido a través de técnicas microscópicas En la Publicación IV se aplica el análisis de imagen digital para la cuantificación de las imágenes obtenidas en muestras de madera de pino y eucalipto pretratadas por autohidrólisis y LI. Principalmente se determinaron 4 parámetros: - Distribución de tamaños de partícula y diámetro equivalente a partir de las imágenes de SEM de 100 aumentos. - Dimensión fractal (Db) y lacunaridad (Ʌ) a partir de las imágenes de SEM de 400 aumentos. Pretratamientos en una etapa 99 - Proporción holocelulosa/lignina a partir de las imágenes de MFC. En la Figura 3.30 se muestra la matriz de correlaciones lineales de estos parámetros y el de digestibilidad enzimática. Figura 3.30: Matriz de correlaciones del análisis de imagen digital de pino y eucalipto autohidrolizados y pretratados con [Emim][OAc] con calefacción por microondas La determinación de estos parámetros demuestra que existe cierta relación entre la dimensión fractal de las muestras y la digestibilidad enzimática independientemente de la madera (pino o eucalipto) y del pretratamiento (AH o pretratamiento con LI con calefacción por microondas). Los valores de Db se encuentra en el intervalo de 2,65-2,74 que, aunque parece un intervalo pequeño, es suficiente para encontrar diferencias entre muestras (Risović et al., 2008; Chi et al., 2017). Valores de Db elevados suponen estructuras complejas, como son las que presentan más rugosidad. Además, cuanto mayor es la complejidad de la textura, el material tiene más rugosidad y es más accesible, por lo que su digestibilidad es mayor. Aunque de manera menos evidente, la lacunaridad y la digestibilidad también están relacionados. La lacunaridad indica el grado de heterogeneidad del material. Así pues, los materiales homogéneos que menor Ʌ tienen, son los que presentan mayor digestibilidad. Además, existe una tendencia entre Db y Ʌ. Los resultados mostraron que, cuanto mayor es la complejidad de los sólidos, más homogéneos son. Capítulo 3 100 Por el contrario, ni la distribución del tamaño de partículas (a través del parámetro de la mediana del diámetro equivalente) ni los ratios holocelulosa/lignina superficial, guardan una tendencia con la digestibilidad independiente del material (Kadić et al., 2014). Esto está en concordancia con lo comentado en los apartados 3.2 y 3.3, ya que la lignina de pino inhibe a las enzimas en mayor grado que la de eucalipto. Sin embargo, Chabbert et al. (2018) sí que encontraron cierta tendencia mediante MFC con la digestibilidad. Las diferencias aplicadas en el proceso experimental, en el que sus muestras fueron pretratadas mediante otros métodos, pueden explicar estos resultados. 3.6. Evaluación de los pretratamientos en una etapa A raíz de los resultados obtenidos, se ha podido comprobar que los pretratamientos de autohidrólisis son efectivos para extraer hemicelulosas tanto en madera de pino como de eucalipto. En las condiciones más favorables, se ha conseguido extraer un 69,7 % de las hemicelulosas del eucalipto y un 68,8 % de las hemicelulosas del pino, en forma de oligosacáridos. Sin embargo, los rendimientos finales a glucosa en el eucalipto y el pino son de 49 % y 9 %, respectivamente. El empleo de pretratamientos con LIs, por el contrario, no extrae hemicelulosa, pero permite alcanzar rendimientos finales a glucosa de hasta un 84 % en pino y un 70 % en eucalipto, ambos empleando el LI [Emim][OAc]. El empleo de calefacción por microondas con este LI reduce los tiempos de pretratamientos, logrando pasar de pretratamientos de 6 horas de duración a 50 minutos. En el caso de la madera de eucalipto, aparte del empleo de [Emim][OAc], resulta interesante el empleo de LIs derivados del catión colina como el [Ch][OAc] porque, aunque da como resultado rendimientos inferiores que con [Emim][OAc], (57 % vs 70 %), podría ser optimizado y tiene la ventaja de ser menos tóxico y más biocompatible de cara a la etapa de hidrólisis enzimática. La reutilización del LI es un parámetro a tener en cuenta especialmente en los líquidos iónicos más costosos y tóxicos. El empleo de metodologías como la utilizada en este trabajo para medir la capacidad de reutilización del LI, permite evaluar cuanto se ha recuperado, no solo en términos de eficacia en la etapa de hidrólisis enzimática, sino en su capacidad de disolución de la biomasa del LI en cada ciclo, dando mayor detalle a lo que le ocurre al LI en cada ciclo. Por último, el empleo de técnicas microscópicas para visualizar la madera pretratada ha permitido identificar la dimensión fractal Db como un parámetro determinante en la eficacia del pretratamiento de cara a la accesibilidad del material. Cabe destacar que el empleo de la microscopía de fluorescencia confocal ha puesto de manifiesto que la madera dura (eucalipto) y la blanda (pino) responden de manera distinta al pretratamiento de autohidrólisis, a diferencia de lo que ocurre en el pretratamiento con LIs donde ambas maderas reaccionan de modo similar, lo cual ayuda a la comprensión de las tendencias opuestas encontradas en la digestibilidad. Capítulo 4: PRETRATAMIENTOS MULTIETAPA: AUTOHIDRÓLISIS + LÍQUIDOS IÓNICOS 4. Pretratamientos multietapa: Autohidrólisis + líquidos iónicos 103 4. PRETRATAMIENTOS MULTIETAPA: AUTOHIDRÓLISIS + LÍQUIDOS IÓNICOS En este capítulo se estudian los pretratamientos multietapa realizados en la tesis doctoral. Se ha llevado a cabo una secuencia de dos pretratamientos en madera de pino y de eucalipto: una primera etapa de autohidrólisis que pretende la extracción de hemicelulosas de la madera, seguida de un pretratamiento con líquidos iónicos con calefacción por microondas para aumentar la accesibilidad del material en la etapa de hidrólisis enzimática posterior. La descripción de la combinación de pretratamientos realizada con madera de eucalipto y con madera de pino se expone en las Publicaciones V y VI, respectivamente. 4.1. Estructura y procedimiento experimental El esquema del planteamiento experimental que se ha llevado a cabo en este capítulo se muestra en la Figura 4.1. En dicho esquema se recogen las etapas de acondicionamiento, autohidrólisis (AH), pretratamiento con LI, regeneración, lavado, recuperación del LI e hidrólisis enzimática. De cada una de las etapas se obtienen corrientes sólidas (corrientes 1 y 3) y/o líquidas (corrientes 2, 4, 5 y 6) para analizar y determinar su composición. La madera se acondiciona y a continuación se introduce en el reactor Parr modelo 4567 para llevar a cabo la AH. Los pretratamientos de AH se realizan a diferentes condiciones de severidad (S0), aplicando la metodología explicada en el apartado 3.1 Capítulo 4 104 Figura 4.1: Esquema del proceso de combinación de pretratamientos AH+LI F ig u ra 4 .1 : E sq u em a d el p ro ceso d e co m b in a ció n d e p retra ta m ien to s A H + L I. L a s flech a s en d isco n tin u o rep resen ta n la s co rrien tes a a n a liza r, in d ica n d o co n n ú m ero s la s técn ica s d e ca ra cteriza ció n a p lica d a s a ca d a co rrien te Pretratamientos multietapa: Autohidrólisis + líquidos iónicos 105 Los sólidos obtenidos se reacondicionan moliéndolos y tamizándolos a un tamaño de partícula <150 μm y se introducen en el horno microondas Berghof SpeedWave Four, junto con el LI [Emim][OAc] siguiendo la metodología explicada en el apartado 3.3. Las condiciones de operación empleadas son las que se recogen en la Tabla 4.1. Tabla 4.1: Condiciones de operación en ensayos de autohidrólisis + LI Madera Etapas Nomenclatura Autohidrólisis Líquido iónico Temperatura (°C) Tiempo (min) Factor de severidad Temperatura (°C) Pino 150 30 2,95 50 AH150LI50P 80 AH150LI80P 120 AH150LI120P 150 AH150LI150P 175 60 3,99 50 AH175LI50P 80 AH175LI80P 120 AH175LI120P 150 AH175LI150P 200 90 4,90 50 AH200LI50P 80 AH200LI80P 120 AH200LI120P 150 AH200LI150P Eucalipto 150 30 2,95 80 AH150LI80E 120 AH150LI120E 175 30 3,69 80 AH175LI80P 120 AH175LI120P 200 30 4,42 80 AH200LI80E 120 AH200LI120E Como se puede observar, con madera de eucalipto no se muestran los resultados de los ensayos a baja temperatura en la etapa del LI (50 °C) ni a alta temperatura (150 °C). Los ensayos a baja temperatura dieron como resultado un pretratamiento casi inapreciable y a temperatura muy alta dieron lugar a la degradación del sólido, por lo que se decidió descartar estos experimentos. Finalizado el proceso, se lleva a cabo la etapa de regeneración de la madera pretratada del mismo modo que se ha descrito en el apartado 3.3. A continuación, se lava con 70 mL de agua en Capítulo 4 106 cada paso, repitiendo esta etapa 5 veces y recogiendo cada una de estas fracciones (corriente 4). Después de cada lavado se mide el volumen del mismo para estudiar la curva de recuperación del LI y analizar su concentración. 4.2. Caracterización de las fracciones obtenidas La fracción líquida, compuesta por LI y agua tras la etapa de regeneración (corriente 2) y cada uno de los lavados (corriente 4) se analiza por HPLC. El agua contenida en el lavado se elimina con un rotavapor, y el LI recuperado (corriente 5) se caracteriza para conocer el contenido de lignina mediante espectroscopía de UV/VIS. La fracción sólida de madera pretratada (corriente 3) se caracteriza de acuerdo con la norma NREL/TP-510-42618. Para verificar que el sólido pretratado final se ha lavado correctamente y que no hay presencia de LI, se analiza por FTIR en un espectrofotómetro Jasco 4700 siguiendo el procedimiento explicado en el apartado 3.3. A continuación, los sólidos pretratados se hidrolizan enzimáticamente para determinar su digestibilidad (corriente 6) como se explicó en el apartado 3.1. Los sólidos pretratados también se caracterizan en lo que a morfología (SEM) y composición superficial (MFC) se refiere. Además, se calcula la dimensión fractal (Db) y la lacunaridad (Ʌ) de los sólidos pretratados de madera de pino y de eucalipto. En el caso de la madera de pino se utilizan imágenes de 1000 aumentos de 300x300 píxeles que se ajustan en brillo y contraste. En el caso de la madera de eucalipto se parte de imágenes de 400 aumentos de un tamaño de 128x128 píxeles que son binarizadas utilizando un umbral local de tipo Bernsen con un radio de 15. La descripción detallada de la caracterización realizada se recoge en las Publicaciones V y VI. 4.3. Resultados experimentales. Interpretación y discusión En la Figura 4.2 se muestra la recuperación de los sólidos obtenidos con madera de pino y de eucalipto, así como su composición química. A continuación, se explica el comportamiento de ambas maderas y se comparan entre sí. El rendimiento del sólido recuperado en el caso del pino (Figura 4.2.a) muestra cómo las condiciones de operación en la AH presentan mayor influencia que el pretratamiento con LI. El efecto de la etapa de AH supone una disminución del rendimiento del sólido recuperado de un 20 % en condiciones de AH suaves (AH150LI80P y AH150LI120P), de un 35 % en condiciones de AH intermedias (AH175LI80P y AH175LI120P) y de un 36 % en condiciones de AH severas (AH200LI80P y AH200LI120P). Esto puede deberse a que la AH provoca una variación de la composición, mientras que los LIs pretratan sin degradar el material, lo que hace que la composición en esta etapa no se vea alterada, como ya se comentó en el Capítulo 3 (Elgharbawy et al., 2016b). Pretratamientos multietapa: Autohidrólisis + líquidos iónicos 107 Figura 4.2: Composición y recuperación de madera pretratada mediante el proceso multietapa con a) pino y b) eucalipto En el caso de la madera de eucalipto (Figura 4.2.b), se observa también una disminución del sólido recuperado al aumentar la intensidad del tratamiento, sobre todo en las condiciones más severas (AH200LI120E), en donde sólo se recupera un 55 % de la madera inicial. Comparando los Capítulo 4 108 resultados entre ambas maderas, en todas las condiciones se recupera menos sólido pretratado procedente de la madera de eucalipto que de pino, lo que vuelve a poner de manifiesto que el pino es menos reactivo a los pretratamientos que el eucalipto. La composición en lignina en el sólido pretratado procedente de madera de pino viene marcada por la severidad de la etapa de AH. A condiciones suaves de AH (S0 = 2,95), la composición de lignina está en torno al 36 %, debido a la eliminación moderada de hemicelulosas (manano principalmente). A condiciones intermedias (S0 = 3,99), la composición en lignina varía entre un 34 y 36 %. A condiciones severas (S0=4,90), la total eliminación de hemicelulosas y la excesiva degradación de celulosa y lignina hace que la composición en lignina se incremente hasta un 43- 45 % (Clough et al., 2015; Nitsos et al., 2016). En el caso de la composición en lignina de los sólidos recuperados procedentes de madera de eucalipto, la única diferencia notable se observa entre las muestras AH200LI80E y AH200L120E, con una diferencia de un 6 % en la composición en lignina entre ellas. La composición en glucano en los sólidos pretratados de madera de pino es ≥48 %, lo cual es superior a la del pino sin tratar (~43 %). Esto es debido a la eliminación de hemicelulosas en la etapa de AH. Sin embargo en términos globales, en las muestras autohidrolizadas a condiciones severas (S0 = 4,90), se pierden cantidades considerables de glucano (25 %), lo que demuestra de nuevo que estas condiciones pueden estar eliminando glucano estructural (como ya se vio en el capítulo anterior). En el caso de las muestras de eucalipto, los mayores contenidos en glucano se producen en las muestras AH175LI80E (66,8 %) y AH175LI120E (61,2 %), lo cual va a contribuir a un incremento en el rendimiento final hacia glucosa. Por último y como se mostraba en la Tabla 4.1, se hicieron ensayos con madera de pino utilizando [Emim][OAc] a baja temperatura (AH150LI50P, AH175LI50P y AH200LI50P) y a alta temperatura (AH150LI150P, AH175LI150P y AH200LI150P), para ver la influencia de trabajar a condiciones muy suaves y muy severas. Los ensayos a baja temperatura dieron como resultado un pretratamiento casi inapreciable (en comparación con la madera autohidrolizada en una etapa). Por otro lado, las muestras de pino pretratadas con LI a temperatura muy alta, dieron lugar a la degradación de hemicelulosas y a un aumento de grupos acetilo en el sólido, promovido probablemente por la descomposición del LI y la liberación de acetatos del anión (Clough et al., 2015; Rashid et al., 2016; Kandhola et al., 2017). Este comportamiento es similar al observado en el Capítulo 3. La recuperación del LI en cada etapa de lavado se muestra en la Tabla 4.2. Para ambas maderas, la recuperación de LI se da mayoritariamente en la etapa de regeneración, obteniendo un 72-82 % en esta etapa. En ambos casos, se comprueba que hacen falta al menos 3 lavados para asegurar la completa eliminación del LI, lo que equivale a un consumo de 10,5 mL de Pretratamientos multietapa: Autohidrólisis + líquidos iónicos 109 agua/g [Emim][OAc] introducido. La recuperación global del [Emim][OAc] ha estado, en todos los casos, por encima del 95 %. Una vez eliminada el agua del lavado, más de un 95 % (aproximadamente) del LI introducido se ha recuperado. Tabla 4.2: Recuperación de [Emim][OAc] en cada lavado en pretratamiento multietapa Cantidad de LI recuperado (%) Lavados Muestra Regeneración 1º 2º 3º 4º 5º AH150LI80P 74,1 17,2 3,3 0,5 0,2 0 AH150LI120P 71,6 24,4 4,1 0,7 0 0 AH175LI180P 79,9 16,4 1,3 0,8 0,2 0 AH175LI20P 81,6 18,7 1,6 0,4 0,1 0 AH200LI80P 73,3 18,5 5,5 1,0 0,2 0 AH200LI120P 77,2 24,1 4,9 1,1 0 0 AH150LI80E 80,6 15,9 1,6 0,2 0 0 AH150LI120E 76,6 20,5 3,6 0,5 0 0 AH175LI80E 77,6 16,4 2,3 0,03 0 0 AH175LI120E 77,6 20,9 2,3 0,4 0 0 AH200LI80E 78,7 18,4 1,7 0,5 0 0 AH200LI120E 80,7 15,4 0,8 0,4 0 0 En la Figura 4.3 se muestra la lignina que se acumula en [Emim][OAc] recuperado. Los resultados se expresan como cantidad de lignina que se acumula por cada 100 mg de [Emim][OAc] y como cantidad de lignina que se acumula con respecto de la lignina introducida. El máximo porcentaje de lignina que se acumula en el líquido iónico recuperado al pretratar madera de pino, hasta un 6 %, es considerablemente inferior al de la lignina que se acumula en el pretratamiento de madera de eucalipto, hasta un 70 %, lo que refleja la diferente reactividad de la lignina de madera autohidrolizada procedente de madera dura y blanda, con el [Emim][OAc]. En las muestras de [Emim][OAc] recuperado procedentes del pretratamiento de madera de eucalipto, la cantidad de lignina acumulada aumenta con la intensidad del pretratamiento multietapa. Sin embargo, en las muestras de [Emim][OAc] recuperado procedente del pretratamiento de madera de pino no se observa el mismo comportamiento, destacando la baja extracción de lignina que se ha producido a condiciones de AH intermedias (S0 = 3,99). La posible degradación de algunos componentes, la recondensación y alteración química de la lignina durante la etapa de autohidrólisis, y factores físicos (tamaño y forma de la biomasa autohidrolizada), puede haber limitado la Capítulo 4 110 extracción de lignina de madera de pino a estas condiciones (Lai et al., 2014; Wang et al., 2017; Das et al., 2019). Figura 4.3: Lignina acumulada en el LI recuperado del proceso multietapa Los resultados de lignina acumulada en el [Emim][OAc] recuperado están relacionados con la composición de los sólidos recuperados tras la combinación de pretratamientos. No obstante, el balance global de las muestras AH200LI80E y AH200LI120E indica una sobreestimación de la lignina tras la combinación de pretratamientos. Al utilizarse condiciones tan severas, y al ser la madera de eucalipto menos recalcitrante que la de pino, se atribuye esta diferencia a la formación de pseudo-lignina, compuestos de degradación que se cuantifican como lignina insoluble y dan lugar a sobreestimaciones en el balance de materia del sólido (Shinde et al., 2018). En la Figura 4.4 se muestra la digestibilidad a 72 h de la madera pretratada con la combinación de pretratamientos de AH + LI. Los resultados están expresados en g de glucano/100 g glucano introducido en la etapa de hidrólisis enzimática. En la madera de pino, el proceso multietapa es efectivo en las muestras autohidrolizadas a condiciones suaves (AH150LI80P y AH150LI120P), obteniendo unas digestibilidades de 73 y 78 %, respectivamente. En estas dos muestras, al aumentar la temperatura del pretratamiento con LI, la digestibilidad del material aumenta. Comparando las muestras AH150LI80P y AH150LI120P con sus homólogos de eucalipto a condiciones de AH suaves, la digestibilidad del pino es mayor que la del eucalipto. El incremento de la digestibilidad con madera de pino ya había sido previamente Pretratamientos multietapa: Autohidrólisis + líquidos iónicos 111 demostrado con calefacción convencional hasta 130 °C (Trinh et al., 2015; Torr et al., 2016). Los resultados obtenidos con la madera de pino autohidrolizada a condiciones intermedias (AH175LI80P y AH175LI120P) y severas (AH200LI120P) dan como resultado digestibilidades inferiores al 70 %, lo que indica que el pretratamiento de AH está obstaculizando el posterior pretratamiento con LI que pretende incrementar la digestibilidad enzimática del sólido. Además, las temperaturas en el pretratamiento con LI superiores a 80 °C producen una disminución de la digestibilidad de los sólidos, probablemente debido a la acumulación de productos que inactiven las enzimas e impidan la unión efectiva con el centro activo (Sun et al., 2014). Los resultados obtenidos a condiciones intermedias de AH (AH175LI80P y AH175LI120P) fueron los más bajos de los probados, con digestibilidades de 44 y 40 % para AH175LI80P y AH175LI120P, respectivamente; lo cual coincide con la baja extracción de lignina que se obtuvo en estas muestras. Figura 4.4: Digestibilidad y rendimiento de madera pretratada mediante el proceso multietapa En la madera de eucalipto, las diferencias observadas entre los resultados obtenidos en la madera no tratada y la pretratada, confirman la eficacia de la combinación de pretratamientos de AH + LI, del mismo modo que cada uno de los pretratamientos funcionó correctamente de manera individual (Lynam & Coronella, 2014; Elgharbawy et al., 2016b; Sun et al., 2016c). Con las muestras tratadas bajo las condiciones de AH más suaves (AH150LI80E y AH150LI120E), se consiguen unas digestibilidades del 68 y 74 %, mientras que con las muestras autohidrolizadas a condiciones severas (AH200LI80E y AH200LI120E) las digestibilidades son superiores (86 y 81 %). En las muestras autohidrolizadas a condiciones intermedias y severas, el incremento de Capítulo 4 112 temperatura de 80 a 120 °C en la etapa de pretratamiento con LI no supone un incremento de la digestibilidad, por lo que se podría realizar el pretratamiento con [Emim][OAc] a 80 °C en lugar de a 120 °C. Las muestras con las que mayores digestibilidades se alcanzan son las autohidrolizadas a condiciones intermedias (S0=3,69), en particular la muestra AH175LI120E, alcanzando un valor de 86 %. En este caso, y a diferencia de lo observado con el pino, la combinación de pretratamientos potencia la mejora de la digestibilidad enzimática, por lo que parece que no hay productos que inactiven las enzimas. En el siguiente apartado se va a analizar la morfología y composición superficial de los sólidos para explicar los resultados obtenidos. 4.4. Estudio de la accesibilidad del sólido a través de técnicas microscópicas En la Tabla 4.3 se muestran a modo de resumen algunas imágenes adquiridas al microscopio para algunos de los sólidos de eucalipto y pino pretratados con procesos multietapa. Como el efecto de la temperatura del LI no produce diferencias visuales observables, sólo se incluyen imágenes de muestras autohidrolizadas a los distintos factores de severidad empleados, seguidos por un pretratamiento con [Emim][OAc] a 80 °C. Las imágenes de MFC, representan la holocelulosa superficial en azul y la lignina superficial en verde. Las imágenes de SEM de muestras pretratadas de madera de pino tienen estructuras bastante diferentes entre sí. Las muestras autohidrolizadas a condiciones suaves (como AH150LI80P) posee una estructura esponjosa con muchos poros visibles, parecida a los que se observaban en pretratamientos sólo realizados con [Emim][OAc]. Al aumentar el factor de severidad en la etapa de AH (S0=3,99 y 4,90), las estructuras se vuelven más redondeadas, más lisas y con menos porosidad. Precisamente estas muestras (AH175LI80P y AH200LI80P) han presentado un cambio de tendencia en la digestibilidad del sólido, y muestran morfologías diferentes a las de las muestras pretratadas con pretratamientos de AH o LI en una etapa, mostrada en el Capítulo 3. En las imágenes de MFC de muestras pretratadas de madera de pino, cabe destacar que la lignina emite con mucha intensidad en la muestra AH175LI80P, lo que evidencia que existe mayor cantidad de lignina en la superficie. Esto podría justificar tanto la baja extracción de lignina con [Emim][OAc], que se mostraba en AH175LI80P y AH175LI120P, como sus bajas digestibilidades (44 y 40 %, respectivamente). Al aumentar la severidad de la autohidrólisis, se observa en la muestra AH200LI80P que la intensidad del canal de la lignina de la superficie disminuye, del mismo modo que la digestibilidad de esta muestra (72 %) aumenta y la extracción de lignina fue mayor. En las imágenes de SEM de muestras pretratadas de madera de eucalipto, se han observado en todos los casos estructuras homogéneas, con bastante porosidad y fusionadas entre sí (Sun et al., Pretratamientos multietapa: Autohidrólisis + líquidos iónicos 113 2011). Cabe destacar, en la muestra AH200LI80E, que la superficie de las partículas es más rugosa y se produce aglomeración de partículas más pequeñas. Tabla 4.3: Imágenes al microscopio de muestras de pino y eucalipto pretratadas con procesos multietapa Madera Técnica AH150LI80 AH175LI80 AH200LI80 Pino SEM MFC Eucalipto SEM MFC Las imágenes de MFC de muestras pretratadas de madera de eucalipto presentan diferencias significativas entre ellas. En la muestra AH150LI80E, la estructura morfológica está alterada, y la celulosa y la lignina aparecen entremezcladas. En la muestra AH175LI80E se observa bastante lignina en la superficie. Esta lignina no parece haber inhibido las enzimas, ya que la digestibilidad de esta muestra es muy alta (84 %). A condiciones de AH severas (AH200LI80E), las imágenes son mucho más oscuras y solo parte de la muestra emite a las longitudes de onda fijadas. La comparación de las imágenes de SEM entre muestras pretratadas de madera de pino y de eucalipto equivalentes, también muestra diferencias significativas en algunos casos. Cabe destacar la mayor porosidad que se observa en la muestra AH150LI80P, lo cual coincide con mayor digestibilidad, en comparación con la muestra AH150LI80E. Por el contrario, a condiciones severas Capítulo 4 114 de AH, la muestra AH200LI80E posee mayor porosidad y también mayor digestibilidad que la AH200LI80P. La comparación de las imágenes de MFC permite observar diferencias principalmente entre las muestras AH200LI80P y AH200LI80E. Mientras que la muestra AH200LI80E ya se ha comentado que a la longitud de onda de emisión de la lignina casi no emite fluorescencia, la lignina de la muestra AH200LI80P sí que emite fluorescencia y se observa en la imagen, lo que puede indicar que las alteraciones que ha experimentado la lignina de pino y de eucalipto con el pretratamiento han sido diferentes, debido a su diferente estructura de partida (Chabbert et al., 2018). La aplicación de técnicas de análisis digital permite correlacionar parámetros característicos de la imagen (Db, Ʌ y relación superficial holocelulosa/lignina) entre sí, y con otras variables del pretratamiento (digestibilidad y relación holocelulosa/lignina total). En la Figura 4.5 se muestran los resultados para los experimentos realizados con madera de pino que se exponen en la Publicación VI, mientras que los resultados con madera de eucalipto se recogen en la Publicación V. Figura 4.5: Matriz de correlaciones del pretratamiento de pino multietapa, resultado del análisis de imagen digital Los valores de Db obtenidos se encuentran en el intervalo de 2,54 a 2,71, en consonancia con otros autores que aplican este parámetro a otros biocompuestos (Alvarez et al., 2013; Utrilla- Pretratamientos multietapa: Autohidrólisis + líquidos iónicos 115 Coello et al., 2013). Los resultados muestran la misma tendencia que la obtenida en los pretratamientos individuales de AH y LIs (Capítulo 3), en donde Db aumenta cuando la complejidad de la textura aumenta (al formarse rugosidades) e implica un aumento de la digestibilidad del sólido. No obstante, la muestra AH200LI50P con una Db de 2,66 no sigue esa tendencia, por lo que deben existir otros factores que también afecten a la digestibilidad enzimática, como es el caso de la lignina superficial que se explica a continuación. La lacunaridad (Ʌ), obtenida como medida de la heterogeneidad en todas las muestras, se encuentra por debajo de 0,0053, lo cual indica que las muestras son homogéneas (Utrilla-Coello et al., 2013; Liu & Ostadhassan, 2017). Así pues, se corrobora lo expuesto en el Capítulo 3, donde la alta digestibilidad de las muestras discutidas en los pretratamientos individuales de AH y LI corresponde con menores valores de Ʌ. De nuevo se observa la tendencia mencionada en la que muestras homogéneas con Ʌ bajas dan lugar a elevadas digestibilidades. Sin embargo, los errores en el cálculo de este parámetro se deben tener en cuenta, así como la desviación de la tendencia de la muestra AH200LI50P que tiene un comportamiento diferente. Empleando la MFC se comprueba que la muestra con menor digestibilidad (AH200LI50P) es la que menor relación holocelulosa/lignina superficial posee. De este modo, se confirma que existe una relación entre la cantidad de lignina superficial en la madera de pino y la digestibilidad del sólido, de forma que a menor lignina observada en la superficie del sólido, mayor es la digestibilidad del mismo. Además, se comprueba que la proporción holocelulosa/lignina superficial y la composición del sólido pretratado calculada a partir de la norma NREL/TP-510-42618 no están relacionadas, por lo que la composición en la superficie es diferente a la composición total. 4.5. Evaluación de los pretratamientos multietapa La combinación de pretratamientos en madera de pino ha supuesto como máximo una digestibilidad del 79 %, en la muestra autohidrolizada a 150 °C durante 30 minutos, seguido de un pretratamiento con [Emim][OAc] a 120 °C. El rendimiento en glucosa final del proceso ha sido del 70 %. El incremento en la severidad del tratamiento de autohidrólisis ha limitado la eficacia del pretratamiento posterior con el [Emim][OAc], disminuyendo la digestibilidad del sólido final. En la madera de eucalipto, la combinación de pretratamientos multietapa ha dado lugar a una digestibilidad máxima del 86 % en la muestra autohidrolizada a 175 °C durante 30 minutos, seguido de un pretratamiento con [Emim][OAc] a 120 °C. En las muestras autohidrolizadas por encima de 175 ºC la digestibilidad disminuye. El máximo rendimiento en glucosa final obtenido ha sido de un 70 %, en la muestra autohidrolizada a 175 °C durante 30 minutos, seguido de un pretratamiento con [Emim][OAc] a 80 °C. En la Figura 4.6 se ha planteado un esquema donde se indica el balance de materia de la combinación de AH + LI de esta muestra (AH175LI80E) que es representativa de la combinación de pretratamientos más favorable de los estudiados. En este Capítulo 4 116 proceso en el que se combinan los pretratamientos de AH y LI, se consiguen extraer 15 g de oligosacáridos por cada 100 g de madera, 39,6 g de glucosa y 19,26 g de lignina impurificada con celulosa y hemicelulosa en pequeñas cantidades. *Hace referencia a corrientes calculadas por diferencia, no determinadas experimentalmente. Figura 4.6: Balance de materia por 100 g de eucalipto sin tratar combinando los pretratamientos de autohidrólisis y líquido iónico Este proceso multietapa se corresponde con la obtención de una corriente rica en hemicelulosas que contiene un 64 % de las hemicelulosas en forma de oligosacáridos, una corriente rica en glucosa con el 70 % de la glucosa procedente de la madera y una corriente sólida rica en lignina que contiene el 70 % de la lignina inicial. Se puede afirmar que la combinación de un pretratamiento de AH seguido de uno con LI parece funcionar con la madera de eucalipto. Esta combinación podría optimizarse para maximizar las fracciones valorizables. Capítulo 5: PROCESO “ONE-POT”: PRETRATAMIENTO + HIDRÓLISIS ENZIMÁTICA 5. Proceso “one-pot”: Pretratamiento + hidrólisis enzimática 119 5. PROCESO “ONE-POT”: PRETRATAMIENTO + HIDRÓLISIS ENZIMÁTICA En este capítulo se estudia un pretratamiento con líquidos iónicos (LIs) seguido de una etapa de hidrólisis enzimática en un mismo equipo, eliminando la etapa de lavado, proceso conocido como “one-pot”. Para ello, se van a utilizar LIs próticos (LIPs) de síntesis sencilla, a partir de reactivos de bajo coste. Uno de los aspectos clave del proceso “one-pot” es la biocompatibilidad del LI empleado con el microorganismo que se utiliza en la etapa de fermentación. Por ello, se realizan unos ensayos previos para seleccionar el sistema madera-LI prótico-levadura más favorable. A continuación, aquel sistema que presenta mejores resultados, se escoge para realizar una prueba de concepto y evaluar los aspectos clave a tener en cuenta en esta configuración de proceso, de cara a la etapa de fermentación, para la obtención de bisaboleno como producto final. Este trabajo fue realizado durante la estancia realizada en el The Joint BioEnergy Institute (JBEI) (Emeryville, EE.UU), bajo la dirección de Dr. Seema Singh y Dr. Blake Simmons. La descripción detallada de los ensayos planteados con maderas de eucalipto y pino, la discusión de los resultados, y las conclusiones obtenidas de la prueba de concepto se recogen en la Publicación VII. 5.1. Estructura y procedimiento experimental El planteamiento experimental llevado a cabo se muestra en la Figura 5.1, la cual recoge las etapas de acondicionamiento, ensayos previos y prueba de concepto del “one-pot” de pretratamiento con LIs + hidrólisis enzimática, seguido de fermentación hacia el producto final, bisaboleno. Capítulo 5 120 Figura 5.1: Planteamiento experimental del proceso “one-pot” F ig u ra 5 .1 : P la n tea m ien to ex p erim en ta l d el p ro ceso “ o n e-p o t” . L a s flech a s en d isco n tin u o in d ica n la s co rrien tes a a n a liza r id en tifica n d o co n n ú m ero s la s técn ica s d e ca ra cteriza ció n a p lica d a s a ca d a co rrien te Proceso “one-pot”: Pretratamiento + hidrólisis enzimática 121 Como en todos los procesos estudiados en capítulos anteriores, de cada una de las etapas se obtienen corrientes sólidas (corriente 2 y 5) y/o líquidas (corrientes 1, 3, 4, 6 y 7) para analizar y caracterizar. Como material de partida se utiliza madera de Eucalyptus globulus y Pinus radiata procedentes del laboratorio del Profesor G.L. Gresham (Laboratorio Nacional de Idaho) para los ensayos previos, y madera de Eucalyptus globulus procedente del Centro Nacional de Investigación y Tecnología Agraria y Alimentaria (CIFOR-INIA) para la prueba de concepto. Además, se utiliza el complejo enzimático de celulasa (Cellic® CTec3) y hemicelulasa (Cellic® HTec3) suministrado por Novozymes, Norte América (Franklinton, NC), para llevar a cabo la hidrólisis enzimática. Las levaduras utilizadas en este estudio son Rhodosporidium toruloides BIS3 (JBEI Id: JBx 065244) capaz de producir bisaboleno y Saccharomyces cerevisiae BY4742 capaz de consumir hexosas y pentosas, que ya se ha demostrado que es biocompatible con el LI [Ch][Lys] (Xu et al., 2016a). Ambas cepas son suministradas por el registro de cepas del JBEI. Los LIPs que se han utilizado se muestran en la Tabla 5.1. Tabla 5.1: Suministrador y pureza de los líquidos iónicos utilizados Líquido iónico Fórmula Procedencia Lisinato de colina [Ch][Lys] Advanced Biofuels and Bioproducts (Sun et al., 2014a) Acetato de 2-hidroxietilamonio [OHEtAm][OAc] Síntesis propia Acetato de bis (2-hidroxietilamonio) [b-OHEtAm][OAc] Síntesis propia Acetato de trietilamonio [TEtAm][OAc] Síntesis propia La síntesis de [OHEtAm][OAc], [b-OHEtAm][OAc] y [TEtAm][OAc] se lleva a cabo de manera directa, haciendo reaccionar el ácido de Brönsted (ácido acético glaciar en todos los casos) con la base de Brönsted de amonio correspondiente, en cantidades equimolares, según muestra el esquema de la Figura 5.2. Este proceso se realiza en un matraz de 1 L de vidrio de fondo redondo agitado magnéticamente y sumergido en un baño de hielo abierto a la atmósfera, en el cual se adiciona gota a gota, y de manera equimolar, el ácido sobre la base. Tras completar la adición, se enjuaga la bureta con etanol para asegurar la completa adición del ácido. A continuación, se calienta la solución resultante a 70 °C durante 24 h y se valida la síntesis mediante 1H-RMN (corriente 1). Capítulo 5 122 a) b) c) Figura 5.2: Reacciones de síntesis de los LIs próticos a) [OHEtAm][OAc], b) [b-OHEtAm][OAc] y c) [TEtAm][OAc] Para los ensayos previos, las maderas de eucalipto y de pino se acondicionan moliendo y tamizando a un tamaño de partícula de 425-500 µm (Sun et al., 2017). Se introducen 100 mg de madera, al 10 % en peso, en viales de vidrio de 15 mL durante 6 h en un baño de aceite calentado a 150 °C y con agitación magnética. En el caso del LI [TEtAm][OAc], el pretratamiento se lleva a cabo a baja temperatura (30 °C) durante 7 días, para evitar su degradación (Shmukler et al., 2016). Una vez terminado el pretratamiento, las muestras se enfrían hasta temperatura ambiente. A continuación, la mezcla de pretratamiento se diluye hasta una concentración de LI del 11,02±0,2 % en peso y el pH de la mezcla se ajusta a 5 con la base de Brönsted (trietilamina) o el ácido correspondiente (ácido acético glacial). En el mismo vial en donde se ha llevado a cabo el pretratamiento, se añade el complejo enzimático Cellic CTec3/Cellic HTec3, en relación 9:1 en volumen, utilizando una dosis de 0,25 mL de cóctel/g glucano. A partir de este experimento, se calculará la digestibilidad enzimática del proceso (corriente 4), lo cual permitirá escoger la madera, el LI y la levadura que ofrezca resultados más favorables para el planteamiento de un proceso “one- pot”. Seleccionado el sistema madera-LI-levadura, se plantea el proceso esta vez con 400 mg de madera acondicionada (tamaño de partícula <150 µm). El planteamiento experimental es el mismo, pero en este caso las condiciones de ensayo se llevan a cabo con y sin ajuste de pH entre la etapa de pretratamiento y la de sacarificación. Finalizada la hidrólisis enzimática, se separan las fracciones sólida y líquida por centrifugación. La fracción sólida se lava y liofiliza para su caracterización (corriente 5) y las fracciones líquidas, con y sin ajuste de pH, se caracterizan (corriente 6) y se Proceso “one-pot”: Pretratamiento + hidrólisis enzimática 123 separan para evaluar la biocompatibilidad del proceso con el microorganismo seleccionado: Rhodosporidium toruloides. El crecimiento de Rhodosporidium toruloides se lleva a cabo en hidrolizados que contienen 1,25, 2,5, 5, 7,5 y 10 % (en peso) de líquido iónico con ajuste de pH a 7,5 o sin ajuste de pH en medio sintético definido (SD). Dicho medio se prepara mezclando una base nitrogenada de levadura sin aminoácidos (BD Difco, Sparks, MD) y completando la mezcla de acuerdo a las especificaciones del fabricante (Sunrise Science, San Diego, CA). El crecimiento se lleva a cabo a 200 rpm, 30 °C y durante 96 h. En cada experimento se utiliza un volumen de 10 mL, al cual se le adicionan 2 mL de dodecano para formar un sistema bifásico que captura el bisaboleno (producto de la fermentación). 5.2. Caracterización de las fracciones obtenidas La composición de la fracción sólida de madera no tratada y pretratada (corrientes 2 y 5) se caracterizan de acuerdo a la norma NREL/TP-510-42618 (Sluiter, 2011; Ibáñez & Bauer, 2014). Además, se utiliza la pirólisis de gases masas (Py-GC/MS) para determinar los cambios estructurales en los monómeros de lignina, obtenidos en la fracción sólida una vez finalizado el proceso “one- pot” (Papa et al., 2012; Shi et al., 2016). También se analiza la morfología de este sólido utilizando microscopía electrónica de barrido (SEM). Los LIs próticos empleados se caracterizan para estudiar su biocompatibilidad (corriente 3) con las levaduras. Para ello, se parte de las cepas, Rhodosporidium toruloides y Saccharomyces cerevisiae, almacenadas en platos de agar a 4 °C. Las colonias aisladas de cada levadura se utilizan para inocular el cultivo de 5 mL en medio YPD (10 g/L extracto de levadura, 20 g/L peptona y 20 g/L dextrosa), en un incubador orbital a 200 rpm, 30 °C y 12 h. Seguidamente, ambas levaduras se resuspenden en medio fresco durante 4 h y a continuación se vuelven a resuspender para evaluar su biocompatibilidad en los LIs. El estudio de crecimiento para S. cerevisiae y R. toruloides se hace partiendo inicialmente de una DO600 de 0,15 y de 0,20, respectivamente. Los medios de crecimiento se preparan a partir de disoluciones sintéticas que contienen 16,67, 11,11, 5,55, 2,78, 1,11 y 0,56 % (en peso) de LI/agua, que se mezclan en relación en peso de 9:1 con medio de cultivo YPD concentrado 10 veces (100 g/L de extracto de levadura, 200 g/L de peptona y 200 g/L de dextrosa). El crecimiento de las levaduras a lo largo del tiempo se monitoriza con un Tecan MicroPlate Reader (Mennëdorf, Suiza), precalentado a 30 °C. Los análisis realizados con R. toruloides se llevan a cabo a pH 7,5 (óptimo del crecimiento de R. toruloides) utilizando NaOH y sin ajuste de pH. Sin embargo, con S. cerevisiae los ensayos solo se realizan sin ajuste de pH porque se ha comprobado bibliográficamente que esta levadura puede funcionar en un amplio intervalo de pHs (Sun et al., 2016a). La determinación de azúcares en la fase líquida (corrientes 4 y 6) y el consumo de azúcares y compuestos fenólicos en los hidrolizados obtenidos en la prueba de concepto (corrientes 6 y 7) se Capítulo 5 124 lleva a cabo por HPLC (Eudes et al., 2014; Rodriguez et al., 2017). Por último, la detección del bisaboleno en la corriente 7 se realiza mediante un cromatógrafo de gases acoplado a un espectrómetro de masas (Özaydın et al., 2013; Yaegashi et al., 2017) La descripción detallada de la caracterización de cada corriente se recoge en la Publicación VII. 5.3. Resultados experimentales. Interpretación y discusión En este apartado se van a exponer los resultados de los ensayos previos realizados con ambas maderas, así como la prueba de concepto realizada con la combinación de madera-LI-levadura que mejores resultados ofrece. 5.3.1. Ensayos previos: selección del sistema madera-LI-levadura En la Figura 5.3 se muestran los resultados de digestibilidad obtenidos tras el proceso “one- pot”. Las digestibilidades obtenidas con madera dura pretratadas con LIPs son considerablemente superiores a las obtenidas con madera blanda en todos los casos excepto con el LI [TEtAm][OAc]. Este comportamiento es análogo al observado en los pretratamientos de una etapa con el LIP [Mim][Cl] tal y como se explicó en el apartado 3.3. Las digestibilidades empleando el proceso “one- pot” son menores a 45 % en todos los experimentos realizados, valores inferiores a los obtenidos en los pretratamientos en una etapa con LIs apróticos y multietapa. Ello se debe probablemente a que los ensayos en este caso se han realizado con un tamaño de partícula tres veces superior al utilizado en los otros pretratamientos estudiados. La digestibilidad del pino pretratado con [Ch][Lys] es, aproximadamente, tres veces superior a la del pino sin tratar (22 g/100 g de glucano), y en el caso del eucalipto, llega a ser, aproximadamente, seis veces superior a la digestibilidad del eucalipto sin tratar (44 g/100 g de glucano). La mayor efectividad del [Ch][Lys] para pretratar eucalipto que pino ya se demostró en los pretratamientos en una etapa, y recientemente se ha podido confirmar en bibliografía (Dutta et al., 2018). El LI prótico con el que se obtienen mayores digestibilidades de la madera pretratada de los ensayados es [OHEtAm][OAc], con 17 g/100 g glucano en el pino y 42 g/100 g glucano en el eucalipto, debido probablemente a la presencia del anión acetato (Sun et al., 2017). Proceso “one-pot”: Pretratamiento + hidrólisis enzimática 125 Figura 5.3: Digestibilidad enzimática de pino y de eucalipto tras el proceso “one-pot” La mayor ventaja del proceso “one-pot” radica en la eliminación de la etapa de lavado, aunque sea necesario llevar a cabo una etapa intermedia de ajuste del pH (Xu et al., 2016b). En la Tabla 5.2 se recogen las proporciones de ácido o base que fue necesario añadir para ajustar el pH al óptimo para llevar a cabo la hidrólisis enzimática. Tabla 5.2: Condiciones de ajuste de pH en el proceso “one-pot” LI pH inicial pH final Relación ácido/base*:LI [Ch][Lys] 10,6-10,7 5,0-5,3 2,23-2,60 [OHEtAm][OAc] 5,3-5,8 5,2-5,3 0-0,02 [b-OHEtAm][OAc] 6,4-7,0 5,0-5,3 0,26-0,29 [TEtAm][OAc] 4,7-4,8 5,0-5,1 0,25-0,28 * El ácido empleado es ácido acético glacial para [Ch][Lys], [OHEtAm][OAc] y [b-OHEtAm][OAc]; la base empleada es trietilamina para [TEtAm][OAc] Como el LI [OHEtAm][OAc] y el [TEtAm][OAc] ya tienen un pH muy cercano a 5, no hace falta llevar a cabo esa etapa de ajuste de pH durante el proceso “one-pot”. Sin embargo, los LIs derivados basados en aminoácidos en su mayoría son muy básicos y necesitan incluir una etapa de ajuste de pH para evitar la inactivación de las enzimas en el medio (Xu et al., 2016b; Sun et al., 2017). El efecto de la toxicidad del LI de cara a una posterior etapa de fermentación, se puede observar comparando las curvas de crecimiento de S. cerevisiae y R. toruloides en presencia de Capítulo 5 126 [Ch][Lys], [OHEtAm][OAc], [b-OHEtAm][OAc] y [TEA][OAc]. En la Figura 5.4 se muestra el crecimiento de las levaduras a 5 % en peso de cada LI. En la Figura 5.4.a y 5.4.c donde se muestra el crecimiento de R. toruloides y S. cerevisiae sin ajuste de pH, la biocompatibilidad de [b- OHEtAm][OAc] > [OHEtAm][OAc] > [TEtAm][OAc] > [Ch][Lys]. En el caso de la Figura 5.4.b, el bajo pH inicial que proporcionan el [TEtAm][OAc] se compensa con el ajuste y el orden de biocompatibilidad es: [b-OHEtAm][OAc] > [TEtAm][OAc] > [OHEtAm][OAc] > [Ch][Lys]. Al comparar el crecimiento de R. toruloides con y sin ajuste de pH (Figuras 5.4.a y 5.4.b), con todos los LIs excepto con [b-OHEtAm][OAc], ajustar el pH favorece el crecimiento. En el caso de [b- OHEtAm][OAc], el pH de la disolución (~6,4 - 7) ya es muy cercano al óptimo, lo cual implica que el crecimiento de R. toruloides sin ajuste ya es favorable. Figura 5.4: Crecimiento de levaduras en presencia de LI al 5 % en peso: a) R. toruloides sin ajuste de pH; b) R. toruloides con ajuste de pH (7,5) y c) S. cerevisiae sin ajuste de pH En el diseño de un proceso completo “one-pot” incluyendo la etapa de fermentación, hay que encontrar un LI con características adicionales más allá de mejorar la digestibilidad del material a tratar. Así pues, el LI debe ser biocompatible con las enzimas en la etapa de sacarificación y con las levaduras en la etapa de fermentación (Xu et al., 2016c). Adicionalmente, habría que buscar la minimización o la eliminación de la etapa de ajuste de pH. Las digestibilidades más elevadas (> 40 %) se obtienen utilizando los LI [OHEtAm][OAc] y [Ch][Lys] (Figura 5.3). Además, con el LI [OHEtAm][OAc] no hace falta incluir una etapa de ajuste de pH entre pretratamiento y sacarificación. Sin embargo, el mayor crecimiento de las levaduras se consigue cuando se utiliza el [B2-OHEtAm][OAc], pero las digestibilidades son < 20 %. Aunque en menor medida que con el [B2-OHEtAm][OAc], ambas levaduras crecen en presencia de [OHEtAm][OAc], pero no lo hacen en presencia de [Ch][Lys]. Estos resultados implican tomar una solución de compromiso donde se asegure el crecimiento de las levaduras sin comprometer la eficacia del pretratamiento. Por ello, para realizar una prueba de concepto del Proceso “one-pot”: Pretratamiento + hidrólisis enzimática 127 proceso “one-pot”, se ha decidido escoger el LI [OHEtAm][OAc] con la madera de eucalipto, puesto que, a pesar de que la biocompatibilidad es menor que la del [B2-OHEtAm][OAc], mejora significativamente la digestibilidad enzimática del eucalipto, evita llevar a cabo una etapa de ajuste de pH y se observa cierto crecimiento en levaduras. La levadura que se eligió fue R. toruloides, en base a trabajos previos del grupo de investigación, debido a su capacidad de producción de bisaboleno y para comprobar su viabilidad en presencia de compuestos fenólicos (Yaegashi et al., 2017; Sundstrom et al., 2018). 5.3.2. Prueba de concepto: Proceso “one-pot” seguido de fermentación con Rhodosporidium toruloides En el proceso “one-pot” realizado con madera de eucalipto (<150 µm) y el LI [OHEtAm][OAc] se obtuvo una digestibilidad final de 75 g/100 g glucosa inicial. La digestibilidad alcanzada con este LI es mayor a la que se obtuvo con el LI [Emim][OAc] con calefacción convencional (70 g/100 g glucosa inicial). Esto verifica que el LIP [OHEtAm][OAc] mejora su eficacia a menores tamaños de partícula, evitando problemas difusionales (Sun et al., 2013). La digestibilidad alcanzada coincide con su rendimiento al no existir pérdidas, ya que no hay etapa de lavado. Por tanto, el proceso “one-pot” con [OHEtAm][OAc] consigue las mayores digestibilidades del eucalipto pretratado de todas las combinaciones probadas en esta tesis. La digestibilidad obtenida con eucalipto y [OHEtAm][OAc] supone una concentración de glucosa de 4,55 g/L y 0,71 g/L de xilosa. Estas concentraciones están lejos de las requeridas para poder llevar a cabo un paso de fermentación en condiciones de sustrato favorables para la levadura y supone una de las principales desventajas de cara a la implementación de un proceso “one-pot”, tal y como está planteado. Las alternativas futuras que solventen este problema deben ir orientadas a disminuir la dilución que hay que hacer entre el pretratamiento y la etapa de hidrólisis enzimática para reducir la concentración del LI. Algunos autores han intentado resolver este problema llevando a cabo el pretratamiento con disoluciones de LI y agua al 10 % en peso (de LI), aunque esta solución sacrifica la ventaja de tener un sistema con baja o nula presión de vapor como el que proporcionan los LIs puros (Xu et al., 2016a). En otros casos, el uso de materias primas menos recalcitrantes permite introducir mayores relaciones sólido:líquido en el pretratamiento (Sun et al., 2017). En la Figura 5.5 se muestra la composición del sólido obtenido una vez finalizado el proceso “one-pot” y en comparación con eucalipto sin tratar. La cantidad de sólido recuperado disminuye considerablemente como consecuencia, principalmente, de la hidrólisis de la celulosa en forma de glucosa a la fase líquida. Sólo un 11 % del glucano que contiene inicialmente el eucalipto se mantiene en este sólido. La suma de la fracción sólida y líquida permite identificar el 70,8 % de los compuestos, mientras que el 30 % restante se atribuye a la despolimerización de la lignina, y a los azúcares que se hayan arrastrado al lavar el sólido antes de su caracterización. Las hemicelulosas Capítulo 5 128 del sólido (glucuronoxilano acetilado) también han sido extraídas casi por completo, ya que el sólido recuperado tras el proceso “one-pot” sólo tiene un 2 % del xilano y un 8 % de los grupos acetilo que tenía inicialmente (Isikgor & Becer, 2015). Aunque el sólido está compuesto mayoritariamente de lignina (69 %), solo se ha recuperado un 68 % de la que inicialmente tenía el eucalipto. Figura 5.5: Caracterización del sólido obtenido tras el “one-pot” con eucalipto y [OHEtAm][OAc] Los principales compuestos derivados de la lignina obtenidos del análisis de Py-GC/MS se resumen en la Tabla 5.3. Dicho análisis mostró que casi todos los derivados de la lignina que se encuentran en la madera de eucalipto sin tratar se encuentran también en los sólidos autohidrolizados. Los derivados de la lignina 2-metoxi-4-metilfenol, siringilacetona, 2-alil-3,6- dimetoxifenilmetanol y 4’-hidroxi-3’,5’-dimetoxiacetofenona sólo se encontraron en el eucalipto sin tratar, mientras que los derivados de la lignina 3-metoxi-catecol, 1-(3-hidroxi-4-metoxifeniletanona) y 4-(1E) 3-hidroxi-1-propenil-2-metoxifenol sólo se encontraron en el sólido obtenido tras el proceso “one-pot”. Hay que mencionar que la proporción de unidades fenólicas tipo siringilo (S)/ guayacilo (G) obtenidas a través de Py-GC/MS de eucalipto sin tratar fue de 3,94, mientras que esta relación en el sólido obtenido tras el “one-pot” fue de 2,89. Este resultado indica que el contenido relativo de unidades guayacilo aumenta, lo que implica que se extraen en mayor proporción unidades siringilo que guayacilo. Precisamente esto último podría también ser una de las razones por las que [OHEtAm][OAc] no es tan efectivo en la madera de pino, que prácticamente no contiene grupos siringilo (Crestini et al., 2017). Proceso “one-pot”: Pretratamiento + hidrólisis enzimática 129 Tabla 5.3: Productos derivados de la pirólisis de la lignina y tiempos de retención encontrados en las muestras de eucalipto sin tratar y tras el “one- pot” con [OHEtAm][OAc] *Hace referencia a la muestra en la que se encuentra cada derivado de la lignina Derivados de lignina (Py-GC/MS) Tiempo de retención (min) Tipo de unidad Muestra* 2-metoxifenol 5,65 G Eucalipto, “One-pot” Creosol 6,85 G Eucalipto , “One-pot” 2-metoxi-4-metilfenol 9,46 G Eucalipto 3-metoxicatecol 9,54 G “One-pot” 4-vinilguayacol 11,69 G Eucalipto , “One-pot” 2,6-dimetoxyfenol 12,82 S Eucalipto , “One-pot” Metil éster de ácido vanílico 14,14 G Eucalipto , “One-pot” 2-metoxi-5-propenil-E-fenol 14,26 G Eucalipto , “One-pot” 2-metoxi-4-vinilfenol 15,05 G Eucalipto , “One-pot” 2-metoxi-4-propenilfenol (trans) 15,17 G Eucalipto , “One-pot” α-etil-p-metoxi benzilalcohol 15,33 G Eucalipto , “One-pot” 1-(3-hidroxi-4-metoxifeniletanona) 15,90 G “One-pot” 4-3-hidroxi-1-propenil-2-metoxifenol 16,58 G Eucalipto , “One-pot” 4-vinil-2,6-dimetoxifenol 17,13 S Eucalipto , “One-pot” 4-alil-2,6- dimetoxifenol 17,64 S Eucalipto , “One-pot” 4-propenil-trans 2,6- dimetoxifenol 18,35 S Eucalipto , “One-pot” Siringil aldehído 18,55 S Eucalipto , “One-pot” 2-alil-1,4-dimetoxi-3-metilbenceno 18,77 S Eucalipto , “One-pot” 4-2-propenil-2,6-dimetoxifenol 19,04 S Eucalipto , “One-pot” Siringol-4-propenil (trans) 19,42 S Eucalipto , “One-pot” 4-((1E)-3-hidroxi)-1-propenil-2-metoxifenol 19,60 S “One-pot” Siringil acetona 19,85 S Eucalipto 2-alil-3,6-dimetoxifenilmetanol 20,19 S Eucalipto 4’-hidroxi-3’,5’-dimetoxiacetofenona 20,51 S Eucalipto Relación S/G Pino sin tratar 3,94 Relación S/G tras el “one-pot” 2,89 Capítulo 5 130 En la Figura 5.6 se muestra una imagen de SEM del sólido antes y después de proceso “one- pot”. Las partículas de eucalipto son uniformes, compactas y sin porosidad visible (Figura 5.6.a), mientras que los sólidos obtenidos tras la sacarificación (Figura 5.6.b) están formados por una aglomeración de partículas más pequeñas. Así, la morfología del sólido se ve alterada por el ataque del LIP, así como por la etapa de hidrólisis enzimática. a) b) Figura 5.6: Morfología de los sólido obtenidos a) antes del proceso “one-pot” y b) después del proceso “one-pot” con eucalipto y [OHEtAm][OAc] La fase líquida que contiene una concentración de azúcares de 5,26 g/L, un 10 % de [2- HEA][OAc] y monómeros de lignina será el hidrolizado de partida que se va a emplear como sustrato para el crecimiento de R. toruloides y fermentación. En la Figura 5.7 se muestran las curvas de crecimiento de R. toruloides de hidrolizados que contienen el LI sin ajuste y con ajuste de pH 7,5. En la Figura 5.7.a, donde el crecimiento se prueba en hidrolizados sin ajuste de pH a 5 % de [OHEtAm][OAc], prácticamente no se observa crecimiento. Por otro lado, en la Figura 5.7.b, donde se ajusta el pH, sí que se observa crecimiento en los hidrolizados que contienen un 5 % de [OHEtAm][OAc]. Ello concuerda con el crecimiento que se observó en el hidrolizado y en el ensayo con [OHEtAm][OAc] y agua mostrado anteriormente en la Figura 5.4.b (DO600nm= 0,22). Todo ello indica que en el hidrolizado no existen otros compuestos aparte del [OHEtAm][OAc] que inhiban el crecimiento de R. toruloides. El crecimiento de R. toruloides está limitado por la baja concentración de sustrato (< 5 g/L), y la presencia de LI. Los resultados sugieren que el efecto negativo que causa el [OHEtAm][OAc], prevalece sobre la falta de sustrato, ya que se muestran mayores crecimientos al 5 % de [2- HEA][OAc] (y baja concentración de sustrato) que al 7,5 % de [2-HEA][OAc]. La influencia del pH, sobre todo a bajas concentraciones de [OHEtAm][OAc] (≤ 2,5 %), resultó ser baja (DO600nm ≈0,38 al 2,5 % de [OHEtAm][OAc], con y sin ajuste de pH) y se demostró que R. toruloides crece incluso a concentraciones de sustrato < 1 g/L. Proceso “one-pot”: Pretratamiento + hidrólisis enzimática 131 Figura 5.7: Crecimiento de levadura R. toruloides en hidrolizados obtenidos tras el “one-pot” a) sin ajuste de pH; b) con ajuste de pH 7,5 Tanto los ensayos de la Figura 5.4 como como los de la Figura 5.7, se llevaron a cabo en placas de 96 pocillos de geometría cilíndrica con un volumen de muestra de 200 µL. El análisis en placas de pocillos, ha permitido seleccionar de manera comparativa el mejor LI. No obstante, el crecimiento se ve afectado por la transferencia de oxígeno o la agitación (Bommareddy et al., 2015). En la Figura 5.8 se muestra el crecimiento en los tubos de ensayo donde se realizó la fermentación. Figura 5.8: Crecimiento de levaduras en hidrolizados obtenidos tras el “one-pot” en tubos de fermentación Capítulo 5 132 En el experimento realizado con un 5 % de [OHEtAm][OAc], contrasta el crecimiento observado en tubos de ensayo (DO600nm= 1,69) en comparación con el crecimiento que se obtuvo en placas de pocillo (DO600nm= 0,22), lo que confirma que la geometría y la agitación de las placas de pocillo están limitando el crecimiento, y esos datos deben tomarse como una referencia comparativa, ya que R. toruloides presenta en medios que contienen hasta un 7,5 % de [OHEtAm][OAc]a pH 7,5 (DO600nm= 0,37). En cuanto al consumo de sustrato, se observa en la Figura 5.9 que R. toruloides consume todos los azúcares presentes en el hidrolizado (2,6 g/L de azúcares formados por pentosas y hexosas) en las muestras que contienen un 5 % de [OHEtAm][OAc] a pH 7,5. Figura 5.9: Efecto del ajuste de pH en el consumo de azucares por R. toruloides en presencia de [OHEtAm][OAc] Sin embargo, los hidrolizados con 5 % de [OHEtAm][OAc] sin ajuste de pH consumen sólo un 30 % de los azúcares. El crecimiento limitado observado en la Figura 5.9 (DO600nm=0,39 a las 48 h) verifica que en ese caso se produce inhibición debida al pH del medio, que impide a R. toruloides consumir gran parte de los azúcares presentes en el hidrolizado. En el caso de los hidrolizados que contienen un 7,5 % de [OHEtAm][OAc], hay 3,9 g/L de azúcares. Sin embargo, R. toruloides sólo consume un 20 % del total, confirmando en este caso que la presencia de LI a concentraciones ≥ 7,5 % limitan el crecimiento y por extensión, el consumo de sustrato. En el resto de las condiciones probadas, R. toruloides no fue capaz de consumir sustrato. Además de azúcares, en los hidrolizados se comprobó que existían, en muy pequeñas cantidades (<20 mg/L), compuestos derivados de lignina. En la Figura 5.10 se muestran los Proceso “one-pot”: Pretratamiento + hidrólisis enzimática 133 cromatogramas de los hidrolizados que contienen un 5 % de [OHEtAm][OAc] sin (Figura 5.11.a) y con ajuste de pH (Figura 5.11.b) obtenidos por HPLC. Los monómeros de lignina detectados por HPLC fueron ácido vanílico (tr= 7,5 min, λ= 280 nm), ácido 4-cumárico (tr= 8,4 min, λ= 310 nm), vainillina (tr= 9,8 min, λ= 280 nm) y ácido benzoico (tr= 10,54 min, λ= 254 nm), todos en muy bajas concentraciones (< 10 mg/L). En la muestra al 5% de [OHEtAm][OAc] sin ajuste de pH se consumieron ácido 4-cumárico, ácido benzoico y vainillina. En la muestra al 5 % de LI con ajuste de pH, además de consumirse los monómeros citados, en mayor medida, también se consumió el ácido vanílico, como ya se ha expuesto en literatura en otro tipo de pretratamientos (Yaegashi et al., 2017; Sundstrom et al., 2018). El consumo de compuestos derivados de la lignina por parte de R. toruloides a bajas concentraciones, los convierte en sustrato, en lugar de ser compuestos tóxicos como ocurre en otros sistemas de fermentación. Figura 5.10: Cromatogramas del consumo de monómeros de lignina por R. toruloides en hidrolizados en presencia de [OHEtAm][OAc] Las condiciones más favorables de las ensayadas sólo produjeron 120 µg/L de bisaboleno como producto final, como consecuencia de los problemas de inhibición y la falta de sustrato, así como la no optimización de esta etapa. Los resultados demuestran la dificultad que ofrece la integración de un proceso “one-pot” de pretratamiento + hidrólisis enzimática seguido de una etapa Capítulo 5 134 de fermentación, pero muestran información relevante en relación a la biocompatibilidad de los hidrolizados. 5.4. Evaluación del proceso “one-pot” y evaluación conjunta de todos los procesos estudiados A modo de resumen de este apartado y de evaluación conjunta de todos los resultados obtenidos en este trabajo, a continuación se va a exponer un análisis global de los resultados más relevantes obtenidos, con el objetivo de seleccionar aquellos procesos que maximicen la digestibilidad del sólido con cada una de las maderas. 5.4.1. Evaluación del proceso “one-pot” A raíz de los ensayos realizados se demuestra que existen configuraciones de proceso alternativas, como el proceso “one-pot”, que podrían reducir el número de etapas de proceso, así como disminuir los requerimientos de agua de lavado. En este capítulo se ha comprobado que el empleo de los LIPs [OHEtAm][OAc], [b- OHEtAm][OAc] y [TEtAm][OAc] no son adecuados para realizar un proceso “one-pot” con madera de pino. Sin embargo, se ha conseguido desarrollar un proceso de pretratamiento + hidrólisis enzimática sin etapa de lavado con el LI [2-HEA][OAc] y madera de eucalipto, en el que se consiguen digestibilidades del 75 % de glucano. Además, como en este proceso no hay etapas de lavado, se realiza tanto el proceso de pretratamiento como el de sacarificación en el mismo equipo. De este modo, se consigue anular la pérdida de sólidos, llegando a un rendimiento de glucosa del 75 %. El proceso “one-pot” aporta los rendimientos a glucosa más altos conseguidos en toda la tesis. Además, el empleo del LI [OHEtAm][OAc] elimina el inconveniente de utilizar una etapa de ajuste intermedio al pH óptimo de las enzimas, como ocurre con otros LIs como [Ch][Lys]. Finalmente, se ha comprobado el crecimiento de R. toruloides en hidrolizados con hasta un 5 % de [OHEtAm][OAc], así como el consumo de la totalidad de los azúcares y de algunos monómeros derivados de la lignina. 5.4.2. Evaluación de los procesos en una etapa, multietapa y “one-pot” En la Figura 5.11 se recoge un resumen de las respuestas más significativas estudiadas con los experimentos más favorables. Los principales parámetros de evaluación han sido la digestibilidad y el rendimiento de glucosa. Proceso “one-pot”: Pretratamiento + hidrólisis enzimática 135 AH175E: Autohidrólisis de eucalipto a S0= 3,69; AH200E: Autohidrólisis de eucalipto a S0= 4,42; AH175P: Autohidrólisis de pino a S0= 3,99; [Emim][OAc]_P: disolución de pino con [Emim][OAc] a 120 C durante 6 h; [Emim][OAc]_E: disolución de eucalipto con [Emim][OAc] a 120 C durante 6 h; [Amim][Cl]_P: disolución de pino con [Amim][Cl] a 120 C durante 6 h; [Amim][Cl]_E: disolución de eucalipto con [Amim][Cl] a 120 C durante 6 h; [Ch][OAc]_P: disolución de pino con [Ch][OAc] a 120 C durante 6 h; [Ch][OAc]_E: disolución de eucalipto con [Ch][OAc] a 120 C durante 6 h; [Ch][Lys]_P: disolución de pino con [Ch][Lys] a 120 C durante 6 h; [Ch][Lys]_E: disolución de eucalipto con [Ch][Lys] a 120 C durante 6 h; [Emim][OAc]_P_MO: disolución de pino con [Emim][OAc] a 120 C durante 50 min con calefacción por microondas; [Emim][OAc]_E_MO: disolución de eucalipto con [Emim][OAc] a 120 C durante 50 min con calefacción por microondas; AH175LI80E: autohidrólisis de eucalipto a S0= 3,69 seguido de disolución con [Emim][OAc] a 80 C durante 50 min con calefacción por microondas; AH175LI80P: autohidrólisis de pino a S0= 3,99 seguido de disolución con [Emim][OAc] a 120 C durante 50 min con calefacción por microondas; [OHEtAm][OAc]_E: Proceso “one-pot” de eucalipto por disolución con [OHEtAm][OAc] a 150 °C y 6 h. Figura 5.11: Resumen de las variables estudiadas en las muestras más representativas Capítulo 5 136 Adicionalmente, como factores secundarios de la evaluación se tuvo en cuenta la capacidad de reutilización del disolvente, la no degradación de las fracciones de hemicelulosa y lignina y el consumo de agua de lavado. Todos los factores se representan en una escala porcentual, indicando el valor máximo de forma positiva y el valor mínimo como un efecto negativo. Los mejores resultados con madera de eucalipto se han alcanzado con el proceso “one-pot” y el LI [OHEtAm][OAc] a 150 °C durante 6 h, obteniendo un rendimiento a glucosa del 75 % ([OHEtAm][OAc]_E). Hay que tener en cuenta que el LI en este proceso no se va a poder recuperar de manera sencilla, al estar en la fase líquida junto a la glucosa. Sin embargo, este proceso reduce considerablemente el consumo de agua de la etapa de lavado. Con la madera de pino, el mayor rendimiento a glucosa, 84 %, se consigue con un proceso en una etapa con el LI [Emim][OAc] a 120 °C durante 6 h y calefacción convencional ([Emim][OAc]_P). A pesar de que se requieren grandes consumos de agua para lavar el sólido pretratado, en el sólido posthidrolizado se conservan parte de las hemicelulosas (78 %) y de la lignina (85 %) presente en la madera de partida. Capítulo 6: CONCLUSIONES / CONCLUSIONS Conclusiones/Conclusions 139 6. CONCLUSIONES A partir de los resultados obtenidos en el presente trabajo de investigación y de su discusión se pueden extraer las siguientes conclusiones. Pretratamientos en una etapa Autohidrólisis - El pretratamiento de autohidrólisis permite la extracción de hemicelulosas en madera de eucalipto y madera de pino. Las condiciones para el pretratamiento que dan lugar a la máxima recuperación de hemicelulosas en forma de oligosacáridos son 175 °C y 30 min (S0 = 3,69), obteniendo un 69,7 % de las hemicelulosas con la madera de eucalipto y 175 °C y 60 min (S0 = 3,99), obteniendo un 68,8 % de las hemicelulosas con la madera de pino. - La digestibilidad de la madera de pino autohidrolizada disminuye con el aumento del factor de severidad. A las condiciones de S0 = 3,99, (175 °C, 60 min), la digestibilidad es del 19 %, lo que equivale a un rendimiento en glucosa final del 9 %. La baja digestibilidad del pino autohidrolizado se debe a la deposición de lignina en la superficie del sólido pretratado. - La digestibilidad del eucalipto tras la etapa de autohidrólisis aumenta con el factor de severidad. La digestibilidad obtenida es de un 53 % cuando se opera con un factor de severidad de S0 = 3,69, alcanzando un rendimiento final de glucosa del 49 %. Capítulo 6 140 Pretratamiento con líquidos iónicos y calefacción convencional - Los LIs muestran diferente reactividad y capacidad de pretratamiento en función del tipo de madera que se pretrate. - La digestibilidad de eucalipto pretratado con [Mim][Cl] alcanza hasta un 48 %, pero la degradación durante el pretratamiento da lugar a un rendimiento final en glucosa del 21 %. - Los LIs derivados del catión colina pueden ser considerados una alternativa más sostenible a los LIs convencionales en el caso del pretratamiento de madera de eucalipto, al conseguir digestibilidades intermedias entre las obtenidas con los LIs apróticos [Amim][Cl] (53 %) y [Emim][OAc] (80 %). - Los LIs apróticos son más efectivos con madera de pino que con madera de eucalipto. La mayor digestibilidad de madera pretratada conseguida en esta tesis, 84 %, se obtiene con el [Emim][OAc] a partir de madera de pino. Pretratamiento con líquidos iónicos con calefacción por microondas - El empleo de calefacción por microondas permite reducir los tiempos de pretratamiento con LIs de 6 h a 50 min. Las digestibilidades a 120 °C con el LI [Emim][OAc], aunque son ligeramente inferiores a las logradas con calefacción convencional, alcanzan los valores de 68 % y 78 % en madera de eucalipto y pino, respectivamente. - El pretratamiento con [Emim][OAc] y calefacción por microondas degrada más la celulosa de la madera de eucalipto que la de pino. - Se ha comprobado en madera de pino que el empleo de temperaturas mayores de 120 °C da lugar a elevadas digestibilidades del sólido (84 %), pero también produce un descenso del rendimiento final a glucosa de hasta un 75 % debido a la degradación de la celulosa en el pretratamiento. A temperaturas iguales o inferiores a 120 °C, la degradación de la celulosa es igual o menor al 5 %. Recuperación y reutilización del [Amim][Cl] - La capacidad del LI para disolver la madera disminuye a lo largo de 5 ciclos de recuperación y reutilización. El rendimiento del sólido rico en celulosa recuperado va siendo menor al aumentar el número de ciclos de reutilización del LI, y la fase amorfa de este sólido aumenta con los ciclos de reutilización del LI. - La acumulación de lignina en el LI y probablemente de celulosa son las causas que provocan la disminución de la capacidad de disolución del LI en los ciclos sucesivos. Conclusiones/Conclusions 141 - En cada ciclo se recupera hasta un 96 % del [Amim][Cl] introducido en el proceso. Aplicación de técnicas microscópicas para la determinación de la accesibilidad - La cuantificación de imágenes con parámetros numéricos como la mediana de diámetro equivalente o la relación holocelulosa/lignina en la superficie de la biomasa, muestran una relación con las condiciones de pretratamiento y el tipo de madera empleada. - Los sólidos pretratados con elevadas dimensiones fractales (Db > 2,72) son más accesibles para las enzimas por su rugosidad y porosidad. Los sólidos pretratados con lacunaridades bajas (Ʌ < 0,07) son más homogéneos y más accesibles para las enzimas. Pretratamientos multietapa de autohidrólisis + LI con calefacción por microondas - El efecto negativo que provoca la etapa de autohidrólisis en madera de pino limita la digestibilidad del sólido tras la etapa de pretratamiento con LI. Los pretratamientos multietapa en la madera de pino dan lugar a morfologías compactas y con poca porosidad. - La combinación más favorable del proceso multietapa con madera de eucalipto consiste en un pretratamiento de autohidrólisis con un factor de severidad S0 = 3,69 y un pretratamiento con el LI [Emim][OAc] en horno microondas a 80 °C. Con estas condiciones, se consigue la máxima digestibilidad del sólido pretratado. Se obtienen 39,6 g de glucosa/100 g de madera a la vez que se extraen 15,09 g oligosacáridos/100 g eucalipto y 19,3 g lignina/100 g de eucalipto se conservan en el sólido posthidrolizado. Proceso “one-pot” con LIs próticos - El proceso “one-pot” puede representar una alternativa a los procesos de pretratamiento convencionales al reducir el número de etapas de proceso, así como disminuir los requerimientos de agua de lavado. - Los LI próticos empleados son efectivos pretratando la madera de eucalipto pero no la madera de pino. Entre el pretratamiento y la hidrólisis enzimática es necesario llevar a cabo una etapa de ajuste de pH en todos los casos menos con el [2-HEA][OAc], lo cual evita añadir una etapa adicional al proceso. - De los sistemas probados, el más favorable es el formado por madera de eucalipto, [2- HEA][OAc] y R. toruloides, con el que se consiguen hidrolizados con un rendimiento de glucosa del 75 %. El sólido obtenido tras el proceso “one-pot” está compuesto en un 69 % Capítulo 6 142 por lignina, eliminándose selectivamente las unidades siringilo. El hidrolizado más favorable al crecimiento de la levadura R. toruloides contiene como máximo un 5 % de [2- HEA][OAc]. Evaluación de los procesos en una etapa, multietapa y “one-pot” - La evaluación de todos los procesos en conjunto se ha resumido valorando como factores principales la digestibilidad del sólido y el rendimiento a glucosa. Además, se han valorado otros factores como la recuperación de hemicelulosa, lignina o agua de lavado. El resumen de todos estos factores con las muestra más favorables se expone en la Figura 5.11. - Con madera de eucalipto, el mayor rendimiento final en glucosa obtenido, un 75 %, se ha conseguido mediante el proceso “one-pot” con el LI [OHEtAm][OAc] a 150 °C durante 6 h. - El mayor rendimiento final en glucosa alcanzado en el presente trabajo fue de un 84 % para la madera de pino, en un proceso en una etapa con el LI [Emim][OAc] utilizando calefacción convencional a 120 °C y 6 h. Conclusiones/Conclusions 143 6-. CONCLUSIONS The conclusions drawn from the results and discussions of this research work are presented below. Single step pretreatments Autohydrolysis - Autohydrolysis enhances hemicellulose extraction in both, eucalyptus and pine wood. Pretreatment conditions that maximize hemicellulose extraction in the form of oligosaccharides are: 175 °C and 30 min (S0=3.69), recovering 69.7 % of total hemicelluloses from eucalyptus, and 175 °C and 60 min (S0=3.99), recovering 68.8 % of total hemicelluloses from pine. - Pine wood digestibility decreases as the severity factor increases. Using S0=3.99 (175 °C and 60 min), pine wood yielded a digestibility of 19 %, which results in a final glucose yield of 9 %. Low digestibilities obtained for pine are attributed to lignin deposition on particles surface. - Eucalyptus wood digestibility after autohydrolysis increases as the severity factor increases. A digestibility of 53 % is achieved using a severity factor of S0=3.69, which results in a final glucose yield of 49 %. Capítulo 6 144 Ionic liquids pretreatment using conventional heating - Ionic liquids (ILs) show different reactivity and pretreatment efficiency depending on the pretreated wood species. - The protic IL [Mim][Cl] increases digestibility of eucalyptus pretreated wood up to 48 %, but wood degradation occurs during pretreatment, as final glucose yield is only of 21 %. - Choline-derived ILs can be considered a more sustainable alternative to the conventional aprotic ILs when using eucalyptus wood as raw material, as the digestibility obtained employing [Ch][OAc] is 69 %, an intermediate value between the obtained with [Amim][Cl] (53 %) and [Emim][OAc] (80 %). - Aprotic ILs are more effective at pretreating pine than eucalyptus wood. The highest digestibility obtained among all the experiments, 84 %, is achieved using pine wood as raw material, and [Emim][OAc]. - Protic and choline-derived ILs lead to lower Tp values in the recovered ILs than in fresh ILs. Ionic liquids pretreatment using microwave heating - Microwave heating reduces pretreatment times from 6 hours to 50 minutes. Eucalyptus and pine wood digestibilities are 68 % and 78 %, respectively, when employing [Emim][OAc] at 120 °C. These results were lower than the obtained using conventional heating. - [Emim][OAc] pretreatment using microwave heating degrades more cellulose in eucalyptus than in pine wood. - Pretreatment temperatures above 120 °C lead to high digestibilities (84 %) for pine, but final glucose yield is reduced down to 75 % due to cellulose degradation. Temperatures equal or below 120 °C only degrades 5 % of the initial cellulose present in the wood. Recovery and reuse of [Amim][Cl] - Recovery and reuse capacity of [Amim][Cl] is decreased from the first to the fifth cycle. The cellulose rich-solid yield decreases when the number of IL recovery cycles increases, as well as amorphicity of this solid increases with the increasing IL recovery grade. - Lignin accumulation in the recovered IL (and probably cellulose) is the main reason for the decrease in the dissolution capacity of the IL in forthcoming cycles. - Up to 96 % of initial [Amim][Cl] used in the process can be recovered. Conclusiones/Conclusions 145 Application of microscopy techniques in biomass accessibility determination - Images quantification with numerical parameters such as median equivalent diameter and biomass holocellulose/lignin surface ratio show a relation between pretreatment conditions and the wood species employed in the pretreatment. - Pretreated solids with high fractal dimensions (Db>2,72) are more accessible to enzymes due to their rugosity and porosity. Pretreated solids with low lacunarities (Ʌ<0,07) are more homogeneous and more accessible to enzymes. Multi-step pretreatments combining autohydrolysis + ionic liquid pretreatment using microwave heating - Multi-step pretreatments with pine wood are limited by the negative effect that autohydrolysis causes over the pretreated solid, hindering the enzymatic hydrolysis step even when including an IL pretreatment. Multi-step pretreatments in pine results in compact solid morphologies with low porosity. - The most favorable conditions tested are obtained for eucalyptus multi-step pretreatments using autohydrolysis at a severity factor S0=3.69, followed by [Emim][OAc] pretreatment under microwave heating. Under these conditions digestibility is maximized and 39.6 g glucose/100 g untreated eucalyptus are obtained, 15.1 g oligosaccharides/100 g untreated eucalyptus are also extracted and 19.3 g lignin/100 g untreated eucalyptus remain in the saccharified solid. One-pot process using protic ionic liquids - One-pot approaches are an alternative to conventional pretreatments due the reduction in number of steps as well as water consumption. - Tested protic ILs are more effective pretreatment solvents for eucalyptus than for pine wood. Between the pretreatment and enzymatic hydrolysis step, a pH adjustment step was necessary in all the cases, except for the IL [OHEtAm][OAc], which does not require additional steps in the process. - Among all the systems screened, the most favorable system was formed by eucalyptus wood, [OHEtAm][OAc] and R. toruloides achieving a glucose yield of 75 %. The post- hydrolyzed solids contains 69 % of lignin, removing selectively syringil units. The most Capítulo 6 146 favorable hydrolysate for a subsequent fermentation step with R. toruloides, is obtained for a maximum amount of 5 % of [OHEtAm][OAc]. Single step, multi-step and one-pot process evaluation - The overall evaluation was developed considering as main factors pretreated solid digestibility and glucose consumption. Additionally, hemicellulose and lignin recovery as well as water consumption and solvents reuse has been taken into account. Results of the most representative samples are exposed in Figure 5.11. - The highest glucose yield achieved with eucalyptus wood in this work was 75 %, using a “one-pot” process with the IL [OHEtAm][OAc] at 150 °C for 6 h. - The highest glucose yield achieved in this work was 84 %, employing pine wood in a single step with the IL [Emim][OAc] using conventional heating for 120 C and 6 h. Capítulo 7: NOMENCLATURA Nomenclatura 149 7. NOMENCLATURA [Amim][Cl] Cloruro de 1-alil-3-metilimidazolio [Bmim][Cl] Cloruro de 1-butil-3-metilimidazolio [b-OHEtAm][OAc] Acetato de bis (2-hidroxietilamonio) [Ch][AA] Aminoácidos de colina [Ch][CA] Carboxilato de colina [Ch][Lys] Lisinato de colina [Ch][OAc] Acetato de colina [Dmim][DMPO4] Dimetilfosfato de 1, 3-dimetilimidazolio [Emim][DEPO4] Dietilfosfato de 1-etil-3-metilimidazolio [Emim][HSO4] Hidrogenosulfato de 1-etil-3-metilimidazolio [Emim][OAc] Acetato de 1-etil-3-metilimidazolio [Emim][Tf2N] Bis(trifluorometilsulfonil)imida de 1-etil-3-metilimidazolio [Mim][Cl] Cloruro de 1-metilimidazolio [OHEtAm][HCO2] Formato de 2-hidroxietilamonio Capítulo 7 150 [OHEtAm][OAc] Acetato de 2-hidroxietilamonio [TBA][R-Pro] R-prolinato de tetrabutilamonio [TBA][S-Pro] S-prolinato de tetrabutilamonio [TBA][t-4OH-LPro] Trans-4-hidroxi-L-prolinato de tetrabutilamonio [TBP][S-Pro] S-prolinato de tetrabutilfosfonio [TEtAm][OAc] Acetato de trietilamonio A Alquil aril éter/ Aspergillus AH Autohidrólisis AH150E Madera de eucalipto sometida a autohidrólisis a 150ºC durante 30 minutos AH150LI120E Madera de eucalipto sometida a autohidrólisis a 150ºC durante 30 minutos y pretratada con [Emim][OAc] a 120ºC durante 50 minutos AH150LI120P Madera de pino sometida a autohidrólisis a 150ºC durante 30 minutos y pretratada con [Emim][OAc] a 120ºC durante 50 minutos AH150LI150P Madera de pino sometida a autohidrólisis a 150ºC durante 30 minutos y pretratada con [Emim][OAc] a 150ºC durante 50 minutos AH150LI50P Madera de pino sometida a autohidrólisis a 150ºC durante 30 minutos y pretratada con [Emim][OAc] a 50ºC durante 50 minutos AH150LI80E Madera de eucalipto sometida a autohidrólisis a 150ºC durante 30 minutos y pretratada con [Emim][OAc] a 80ºC durante 50 minutos AH150LI80P Madera de pino sometida a autohidrólisis a 150ºC durante 60 minutos y pretratada con [Emim][OAc] a 80ºC durante 50 minutos AH150P Madera de pino sometida a autohidrólisis a 150ºC durante 30 minutos AH175E Madera de eucalipto sometida a autohidrólisis a 175ºC durante 30 minutos AH175LI120P Madera de pino sometida a autohidrólisis a 175ºC durante 60 minutos y pretratada con [Emim][OAc] a 120ºC durante 50 minutos AH175LI120E Madera de eucalipto sometida a autohidrólisis a 175ºC durante 30 minutos y pretratada con [Emim][OAc] a 120ºC durante 50 minutos Nomenclatura 151 AH175LI150P Madera de pino sometida a autohidrólisis a 175ºC durante 60 minutos y pretratada con [Emim][OAc] a 150ºC durante 50 minutos AH175LI50P Madera de pino sometida a autohidrólisis a 175ºC durante 60 minutos y pretratada con [Emim][OAc] a 50ºC durante 50 minutos AH175LI80P Madera de pino sometida a autohidrólisis a 175ºC durante 60 minutos y pretratada con [Emim][OAc] a 80ºC durante 50 minutos AH175LI80E Madera de eucalipto sometida a autohidrólisis a 175ºC durante 30 minutos y pretratada con [Emim][OAc] a 80ºC durante 50 minutos AH175P Madera de pino sometida a autohidrólisis a 175ºC durante 60 minutos AH200E Madera de eucalipto sometida a autohidrólisis a 200ºC durante 30 minutos AH200LI120E Madera de eucalipto sometida a autohidrólisis a 200ºC durante 30 minutos y pretratada con [Emim][OAc] a 120ºC durante 50 minutos AH200LI120P Madera de pino sometida a autohidrólisis a 200ºC durante 90 minutos y pretratada con [Emim][OAc] a 120ºC durante 50 minutos AH200LI150P Madera de pino sometida a autohidrólisis a 200ºC durante 90 minutos y pretratada con [Emim][OAc] a 150ºC durante 50 minutos AH200LI50P Madera de pino sometida a autohidrólisis a 200ºC durante 90 minutos y pretratada con [Emim][OAc] a 50ºC durante 50 minutos AH200LI80E Madera de eucalipto sometida a autohidrólisis a 200ºC durante 30 minutos y pretratada con [Emim][OAc] a 80ºC durante 50 minutos AH200LI80P Madera de pino sometida a autohidrólisis a 200ºC durante 90 minutos y pretratada con [Emim][OAc] a 80ºC durante 50 minutos AH200P Madera de pino sometida a autohidrólisis a 200ºC durante 90 minutos B Estructuras fenilcumaran C Estructuras resinol CIFOR-INIA Centro de Investigación forestal. Instituto Nacional de Investigación y Tecnología Agraria CV Coeficiente de variación Capítulo 7 152 Db Dimensión fractal DCT Disolución casi total DMSO Dimetilsulfóxido DMSO-d6 Dimetilsulfóxido deuterado DO Demanda de oxígeno DP Disolución parcial DRX Difracción de rayos x Dvte Disolvente E Madera de eucalipto Eqvte Equivalente ES Enzima-sustrato ET N Energía molar de transición FTIR Espectroscopía de infrarrojo por Transformada de Fourier FTIR-ATR Espectroscopía de infrarrojo por Transformada de Fourier con reflectancia total atenuada FWHM Altura a mitad de pico G Guayacilo H p-hidroxifenilo HE hidrólisis enzimática HPLC Cromatografía en fase líquida de alta resolución Ii,j,ε Diferencia de intensidad en las posiciones i, j y ε JBEI The Joint BioEnergy Institute LI Líquido iónico LIA Líquido iónico aprótico LIP Líquido iónico prótico LIR Líquido iónico recuperado m Masa (g) MFC Microscopía de fluorescencia confocal Nomenclatura 153 MO50_[Emim][OAc]_P Madera de pino pretratada con [Emim][OAc] a 50 °C en horno microondas durante 50 minutos MO80[Emim][OAc]_E Madera de eucalipto pretratada con [Emim][OAc] a 80 °C en horno microondas durante 50 minutos MO80_[Emim][OAc]_P Madera de pino pretratada con [Emim][OAc] a 80 °C en horno microondas durante 50 minutos MO120_[Emim][OAc]_E Madera de eucalipto pretratada con [Emim][OAc] a 120 °C en horno microondas durante 50 minutos MO120_[Emim][OAc]_P Madera de pino pretratada con [Emim][OAc] a 120 °C en horno microondas durante 50 minutos MO150_[Emim][OAc]_P Madera de eucalipto pretratada con [Emim][OAc] a 150 °C en horno microondas durante 50 minutos NREL National Renewable Energies Laboratories p Posición para en estructuras fenólicas P Madera de pino Py/GC-MS Pirólisis de cromatografía de gases acoplada a espectroscopía de masas RMN Resonancia magnética nuclear S Siringilo SD Medio sintético definido SEM Microscopía electrónica de barrido So Factor de severidad SRC Sólido rico en celulosa SRL Sólido rico en lignina T Trichoderma T0 Temperatura inicial (ºC) T25% Temperatura de degradación al 25% (ºC) T50% Temperatura de degradación al 50% (ºC) TGA Análisis termogravimétrico Capítulo 7 154 Tonset Temperatura de inicio de la degradación Top Temperatura de operación (ºC) Tp Temperatura de pico (ºC) tr Tiempo de retención UV/VIS Espectroscopía de ultravioleta/visible v Volumen Vi,j,ε Volumen diferencial de dimensiones i, j y ε YPD Medio sintético con 10 g/L de extracto de levadura, 20 g/L peptona y 20 g/L dextrosa Nomenclatura 155 Símbolos griegos α Dador de puentes de hidrógeno o acidez de puentes de hidrógeno β Aceptor de puentes de hidrógeno o basicidad de puentes de hidrógeno δC/δH Desplazamientos químicos de carbono y protón (ppm) ε Calibre de la caja Ʌ Lacunaridad λ Longitud de onda µ Media estadística ηLIR Rendimiento del líquido iónico recuperado π* polarizabilidad o dipolaridad σ Desviación estándar Capítulo 8: BIBLIOGRAFÍA Bibliografía 159 8. BIBLIOGRAFÍA A. Sluiter, B.H., R.Ruiz, C.Scarlata, J.Sluitr, D.Templeton and D. Crocker. 2011. Determination of Structural carbohydrates and lignin in biomass. NREL. Abushammala, H., Goldsztayn, R., Leao, A., Laborie, M.-P. 2016. Combining steam explosion with 1-ethyl-3-methylimidazlium acetate treatment of wood yields lignin-coated cellulose nanocrystals of high aspect ratio. Cellulose, 23(3), 1813-1823. Achinivu, E.C. 2014. Isolation and Recovery of Lignin from Lignocellulosic Biomass Using Recyclable Protic Ionic Liquids (PILs) for a Cost-Effective Biomass Processing Technique. North Carolina State University. Adel, A.M., El-Wahab, Z.H.A., Ibrahim, A.A., Al-Shemy, M.T. 2011. Characterization of microcrystalline cellulose prepared from lignocellulosic materials. Part II: Physicochemical properties. Carbohydr. Polym., 83(2), 676-687. Akhtar, N., Gupta, K., Goyal, D., Goyal, A. 2016. Recent advances in pretreatment technologies for efficient hydrolysis of lignocellulosic biomass. Environmental Progress & Sustainable Energy, 35(2), 489-511. Akhtar, N., Gupta, K., Goyal, D., Goyal, A. 2015. Recent advances in pretreatment technologies for efficient hydrolysis of lignocellulosic biomass. Environ. Prog. Sustain. Energy, 35(2), 489- 511. Alvarez, A., Passé-Coutrin, N., Gaspard, S. 2013. Determination of the textural characteristics of carbon samples using scanning electronic microscopy images: comparison with mercury porosimetry data. Adsorption, 19(2-4), 841-850. Alvira, P., Tomás-Pejó, E., Ballesteros, M., Negro, M.J. 2010. Pretreatment technologies for an efficient bioethanol production process based on enzymatic hydrolysis: A review. Bioresource Technology, 101(13), 4851-4861. An, S., Li, W., Liu, Q., Xia, Y., Zhang, T., Huang, F., Lin, Q., Chen, L. 2019. Combined dilute hydrochloric acid and alkaline wet oxidation pretreatment to improve sugar recovery of corn stover. Bioresource Technology, 271, 283-288. An, Y.-X., Zong, M.-H., Wu, H., Li, N. 2015. Pretreatment of lignocellulosic biomass with renewable cholinium ionic liquids: Biomass fractionation, enzymatic digestion and ionic liquid reuse. Bioresource Technology, 192, 165-171. Capítulo 8 160 Arenas-Cárdenas, P., López-López, A., Moeller-Chávez, G.E., León-Becerril, E. 2016. Current Pretreatments of Lignocellulosic Residues in the Production of Bioethanol. Waste and Biomass Valorization, 8(1), 161-181. Arenas-Cárdenas, P., López-López, A., Moeller-Chávez, G.E., León-Becerril, E. 2017. Current Pretreatments of Lignocellulosic Residues in the Production of Bioethanol. Waste and Biomass Valorization, 8(1), 161-181. Asakawa, A., Kohara, M., Sasaki, C., Asada, C., Nakamura, Y. 2015. Comparison of choline acetate ionic liquid pretreatment with various pretreatments for enhancing the enzymatic saccharification of sugarcane bagasse. Industrial Crops and Products, 71, 147-152. Asakawa, A., Oka, T., Sasaki, C., Asada, C., Nakamura, Y. 2016. Cholinium ionic liquid/cosolvent pretreatment for enhancing enzymatic saccharification of sugarcane bagasse. Industrial Crops and Products, 86, 113-119. Asmadi, M., Kawamoto, H., Saka, S. 2017. Characteristics of softwood and hardwood pyrolysis in an ampoule reactor. Journal of Analytical and Applied Pyrolysis, 124, 523-535. ASPAPEL. 2018. La bioindustria circular del papel y su descarrbonización, Vol. 2019. Austen Angell, C., Ansari, Y., Zhao, Z. 2012. Ionic Liquids: Past, present and future. Faraday Discussions, 154(0), 9-27. Auxenfans, T., Buchoux, S., Djellab, K., Avondo, C., Husson, E., Sarazin, C. 2012. Mild pretreatment and enzymatic saccharification of cellulose with recycled ionic liquids towards one-batch process. Carbohydrate Polymers, 90(2), 805-813. Auxenfans, T., Buchoux, S., Larcher, D., Husson, G., Husson, E., Sarazin, C. 2014. Enzymatic saccharification and structural properties of industrial wood sawdust: Recycled ionic liquids pretreatments. Energy Conversion and Management, 88, 1094-1103. Auxenfans, T., Husson, E., Sarazin, C. 2017. Simultaneous pretreatment and enzymatic saccharification of (ligno) celluloses in aqueous-ionic liquid media: A compromise. Biochemical Engineering Journal, 117, Part A, 77-86. Badgujar, K.C., Bhanage, B.M. 2015. Factors governing dissolution process of lignocellulosic biomass in ionic liquid: Current status, overview and challenges. Bioresource Technology, 178, 2-18. Balat, M. 2011. Production of bioethanol from lignocellulosic materials via the biochemical pathway: A review. Energy conversion and management, 52(2), 858-875. Barsett, H., Ebringerová, A., Harding, S., Heinze, T., Hromádková, Z., Muzzarelli, C., Muzzraelli, R., Paulsen, B., Elseoud, O. 2005. Polysaccharides I: Structure, characterisation and use. Springer Science & Business Media. Baruah, J., Nath, B., Sharma, R., Kumar, S., Deka, R., Baruah, D., Kalita, E. 2018. Recent Trends in the Pretreatment of Lignocellulosic Biomass for Value-Added Products. Front. Energy Resources, 6, 141. Bauer, F., Coenen, L., Hansen, T., McCormick, K., Palgan, Y.V. 2017. Technological innovation systems for biorefineries: a review of the literature. Biofuels, Bioproducts and Biorefining, 11(3), 534-548. Behera, S., Arora, R., Nandhagopal, N., Kumar, S. 2014. Importance of chemical pretreatment for bioconversion of lignocellulosic biomass. Renewable and Sustainable Energy Reviews, 36, 91-106. Bhattacharyya, S., Shah, F.U. 2018. Thermal stability of choline based amino acid ionic liquids. Journal of Molecular Liquids, 266, 597-602. Bhutto, A.W., Qureshi, K., Harijan, K., Abro, R., Abbas, T., Bazmi, A.A., Karim, S., Yu, G. 2017. Insight into progress in pre-treatment of lignocellulosic biomass. Energy, 122, 724-745. Biorefinery, Energy, Environmental, and Economics (E3) Handbook, (Eds.) 1997. U.S.D.o.E.O.o.E.E. and, R. Energy. http://infohouse.p2ric.org/ref/36/35057.htm. Blokhin, A.V., Voitkevich, O.V., Kabo, G.J., Paulechka, Y.U., Shishonok, M.V., Kabo, A.G., Simirsky, V.V. 2011. Thermodynamic properties of plant biomass components. Heat capacity, combustion energy, and gasification equilibria of cellulose. Journal of Chemical and Engineering Data, 56(9), 3523-3531. Bibliografía 161 Bommareddy, R.R., Sabra, W., Maheshwari, G., Zeng, A.-P. 2015. Metabolic network analysis and experimental study of lipid production in Rhodosporidium toruloides grown on single and mixed substrates. Microbial Cell Factories, 14(1), 1. Brandt, A., Grasvik, J., Hallett, J.P., Welton, T. 2013. Deconstruction of lignocellulosic biomass with ionic liquids. Green Chemistry, 15(3), 550-583. Brandt, A., Gschwend, F., Fennell, P., Lammens, T., Tan, B., Weale, J., Hallett, J. 2017. An economically viable ionic liquid for the pretreatment of lignocellulosic biomass. Green Chemistry, 19, 3078-3102. Campetella, M., Bodo, E., Caminiti, R., Martino, A., D’Apuzzo, F., Lupi, S., Gontrani, L. 2015. Interaction and dynamics of ionic liquids based on choline and amino acid anions. The Journal of Chemical Physics, 142(23), 234502. Carrillo, F., Colom, X., Sunol, J.J., Saurina, J. 2004. Structural FTIR analysis and thermal characterisation of lyocell and viscose-type fibres. European Polymer Journal, 40(9), 2229- 2234. Carvalheiro, F., Duarte, L., Gírio, F., Moniz, P. 2016. Hydrothermal/Liquid Hot Water Pretreatment (Autohydrolysis): A Multipurpose Process for Biomass Upgrading. in: Biomass Fractionation Technologies for a Lignocellulosic Feedstock Based Biorefinery, Elsevier, pp. 315-347. Casas, A., Oliet, M., Alonso, M.V., Santos, T.M., Rodriguez, F. 2013. Dissolution of Pinus radiata and Eucalyptus Globulus Woods in 1-Allyl-3-methylimidazolium Chloride for Cellulose or Lignin Regeneration. Industrial & Engineering Chemistry Research, 52(10), 3628-3636. Castañar, L., Sistaré, E., Virgili, A., Williamson, R.T., Parella, T. 2015. Suppression of phase and amplitude J (HH) modulations in HSQC experiments. Magnetic Resonance in Chemistry, 53(2), 115-119. Castro, M.C., Rodríguez, H., Arce, A., Soto, A. 2014. Mixtures of Ethanol and the Ionic Liquid 1- Ethyl-3-methylimidazolium Acetate for the Fractionated Solubility of Biopolymers of Lignocellulosic Biomass. Industrial & Engineering Chemistry Research, 53(29), 11850- 11861. Cebreiros, F., Ferrari, M.D., Lareo, C. 2017. Combined autohydrolysis and alkali pretreatments for cellulose enzymatic hydrolysis of Eucalyptus grandis wood. Biomass Conversion and Biorefinery, 1-10. Cerasoli, S., Caldeira, M., Pereira, J., Caudullo, G., de Rigo, D. 2016. Eucalyptus globulus and other eucalypts in Europe: distribution, habitat, usage and threats. European Atlas of Forest Tree Species. Luxembourg: Publishing Office of the EU, e01b5bb. Chabbert, B., Terryn, C., Herbaut, M., Vaidya, A., Habrant, A., Paës, G., Donaldson, L. 2018. Fluorescence techniques can reveal cell wall organization and predict saccharification in pretreated wood biomass. Industrial Crops and Products, 123, 84-92. Chakar, F.S., Ragauskas, A.J. 2004. Review of current and future softwood kraft lignin process chemistry. Industrial Crops and Products, 20(2), 131-141. Chen, H., Liu, J., Chang, X., Chen, D., Xue, Y., Liu, P., Lin, H., Han, S. 2017. A review on the pretreatment of lignocellulose for high-value chemicals. Fuel Processing Technology, 160, 196-206. Cherubini, F. 2010. The biorefinery concept: using biomass instead of oil for producing energy and chemicals. Energy Conversion and Management, 51(7), 1412-1421. Cherubini, F., Jungmeier, G., Mandl, M., Philips, C., Wellisch, M., Jrgensen, H., Skiadas, I., Boniface, L., Dohy, M., Pouet, J.-C. 2007. IEA Bioenergy Task 42 on Biorefineries: Co- production of fuels, chemicals, power and materials from biomass. IEA bioenergy task. pp. 1-37. Cherubini, F., Jungmeier, G., Wellisch, M., Willke, T., Skiadas, I., Van Ree, R., de Jong, E. 2009. Toward a common classification approach for biorefinery systems. Biofuels, Bioproducts and Biorefining, 3(5), 534-546. Chi, C., Hui, Z., Liu, M., Zhang, S., Gong, Y. 2017. Effect of Acetic Acid Pretreatment on Wood Pore Structure and Fractal Dimension. BioResources, 12(2), 3905-3917. Clough, M.T., Geyer, K., Hunt, P.A., Son, S., Vagt, U., Welton, T. 2015. Ionic liquids: not always innocent solvents for cellulose. Green Chemistry, 17(1), 231-243. Capítulo 8 162 Cox, B.J., Ekerdt, J.G. 2013. Pretreatment of yellow pine in an acidic ionic liquid: Extraction of hemicellulose and lignin to facilitate enzymatic digestion. Bioresource Technology, 134, 59-65. Cox, B.J., Jia, S., Zhang, Z.C., Ekerdt, J.G. 2011. Catalytic degradation of lignin model compounds in acidic imidazolium based ionic liquids: Hammett acidity and anion effects. Polymer Degradation and Stability, 96(4), 426-431. Crestini, C., Lange, H., Sette, M., Argyropoulos, D.S. 2017. On the structure of softwood kraft lignin. Green Chemistry, 19(17), 4104-4121. Das, P., Stoffel, R.B., Area, M.C., Ragauskas, A.J. 2019. Effects of one-step alkaline and two-step alkaline/dilute acid and alkaline/steam explosion pretreatments on the structure of isolated pine lignin. Biomass and Bioenergy, 120, 350-358. Datta, S., Holmes, B., Park, J.I., Chen, Z., Dibble, D.C., Hadi, M., Blanch, H.W., Simmons, B.A., Sapra, R. 2010. Ionic liquid tolerant hyperthermophilic cellulases for biomass pretreatment and hydrolysis. Green Chemistry, 12(2), 338-345. De Corato, U., De Bari, I., Viola, E., Pugliese, M. 2018. Assessing the main opportunities of integrated biorefining from agro-bioenergy co/by-products and agroindustrial residues into high-value added products associated to some emerging markets: A review. Renewable and Sustainable Energy Reviews, 88, 326-346. De Gregorio, G.F., Weber, C.C., Gräsvik, J., Welton, T., Brandt, A., Hallett, J.P. 2016. Mechanistic insights into lignin depolymerisation in acidic ionic liquids. Green Chemistry, 18(20), 5456- 5465. de Jong, E., Jungmeier, G. 2015. Chapter 1 - Biorefinery Concepts in Comparison to Petrochemical Refineries. in: Industrial Biorefineries & White Biotechnology, (Eds.) A. Pandey, R. Höfer, M. Taherzadeh, K.M. Nampoothiri, C. Larroche, Elsevier. Amsterdam, pp. 3-33. Deb, S., Labafzadeh, S.R., Liimatainen, U., Parviainen, A., Hauru, L.K.J., Azhar, S., Lawoko, M., Kulomaa, T., Kakko, T., Fiskari, J., Borrega, M., Sixta, H., Kilpelainen, I., King, A.W.T. 2016. Application of mild autohydrolysis to facilitate the dissolution of wood chips in direct- dissolution solvents. Green Chemistry, 18, 3286-3294. Del Rio, J.C., Rencoret, J., Marques, G., Gutiérrez, A., Ibarra, D., Santos, J.I., Jiménez-Barbero, J.s., Zhang, L., Martínez, A.n.T. 2008. Highly acylated (acetylated and/or p-coumaroylated) native lignins from diverse herbaceous plants. Journal of Agricultural and Food Chemistry, 56(20), 9525-9534. del Río, J.C., Rencoret, J., Prinsen, P., Martínez, Á.T., Ralph, J., Gutiérrez, A. 2012. Structural Characterization of Wheat Straw Lignin as Revealed by Analytical Pyrolysis, 2D-NMR, and Reductive Cleavage Methods. Journal of Agricultural and Food Chemistry, 60(23), 5922-5935. Dessbesell, L., Xu, C., Pulkki, R., Leitch, M., Mahmood, N. 2016. Forest biomass supply chain optimization for a biorefinery aiming to produce high-value bio-based materials and chemicals from lignin and forestry residues: a review of literature. Canadian Journal of Forest Research, 47(3), 277-288. Dhumal, N.R., Kim, H.J., Kiefer, J. 2009. Molecular Interactions in 1-Ethyl-3-methylimidazolium Acetate Ion Pair: A Density Functional Study. The Journal of Physical Chemistry A, 113(38), 10397-10404. Dibble, D.C., Li, C., Sun, L., George, A., Cheng, A., Cetinkol, O.P., Benke, P., Holmes, B.M., Singh, S., Simmons, B.A. 2011. A facile method for the recovery of ionic liquid and lignin from biomass pretreatment. Green Chemistry, 13(11), 3255-3264. Doherty, T.V., Mora-Pale, M., Foley, S.E., Linhardt, R.J., Dordick, J.S. 2010. Ionic liquid solvent properties as predictors of lignocellulose pretreatment efficacy. Green Chemistry, 12(11), 1967-1975. Doherty, W.O., Mousavioun, P., Fellows, C.M. 2011. Value-adding to cellulosic ethanol: Lignin polymers. Industrial Crops and Products, 33(2), 259-276. Domínguez de María, P. 2014. Recent trends in (ligno)cellulose dissolution using neoteric solvents: switchable, distillable and bio-based ionic liquids. Journal of Chemical Technology & Biotechnology, 89(1), 11-18. Bibliografía 163 Dutta, T., Isern, N.G., Sun, J., Wang, E., Hull, S., Cort, J.R., Simmons, B.A., Singh, S. 2017. Survey of Lignin-Structure Changes and Depolymerization during Ionic Liquid Pretreatment. ACS Sustainable Chemistry and Engineering, 5(11), 10116-10127. Dutta, T., Papa, G., Wang, E., Sun, J., Isern, N.G., Cort, J.R., Simmons, B.A., Singh, S. 2018. Characterization of lignin streams during bionic liquid-based pretreatment from grass, hardwood and softwood. ACS Sustainable Chemistry & Engineering. Ek, M., Gellerstedt, G., Henriksson, G. 2009. Wood chemistry and biotechnology. Walter de Gruyter. El Seoud, O.A., Koschella, A., Fidale, L.C., Dorn, S., Heinze, T. 2007. Applications of ionic liquids in carbohydrate chemistry: a window of opportunities. Biomacromolecules, 8(9), 2629- 2647. Elgharbawy, A.A., Alam, M.Z., Moniruzzaman, M., Goto, M. 2016. Ionic liquid pretreatment as emerging approaches for enhanced enzymatic hydrolysis of lignocellulosic biomass. Biochemical engineering journal, 109, 252-267. Erythropel, H.C., Zimmerman, J.B., de Winter, T.M., Petitjean, L., Melnikov, F., Lam, C.H., Lounsbury, A.W., Mellor, K.E., Jankovic, N.Z., Tu, Q., Pincus, L.N., Falinski, M.M., Shi, W., Coish, P., Plata, D.L., Anastas, P.T. 2018. The Green ChemisTREE: 20 years after taking root with the 12 principles. Green Chemistry, 20(9), 1929-1961. Eudes, A., Juminaga, D., Baidoo, E.E.K., Collins, F.W., Keasling, J.D., Loqué, D. 2014. Correction: Production of hydroxycinnamoyl anthranilates from glucose in Escherichia coli. Microbial Cell Factories, 13, 8-8. Evtuguin, D.V., Neto, C.P., Silva, A.M.S., Domingues, P.M., Amado, F.M.L., Robert, D., Faix, O. 2001. Comprehensive Study on the Chemical Structure of Dioxane Lignin from Plantation Eucalyptus globulus Wood. Journal of Agricultural and Food Chemistry, 49(9), 4252-4261. Fang, Z., Smith, R.L., Qi, X. 2014. Production of Biofuels and Chemicals with Ionic Liquids. Dordrecht, Heidelberg, New York, London: Springer. FAO. 2016. GLOBAL FOREST PRODUCTS FACTS AND FIGURES, Vol. 2019, Food and Agriculture Organization of United Nations. Foulet, A., Ghanem, O.B., El-Harbawi, M., Lévêque, J.-M., Mutalib, M.I.A., Yin, C.-Y. 2016. Understanding the physical properties, toxicities and anti-microbial activities of choline- amino acid-based salts: Low-toxic variants of ionic liquids. Journal of Molecular Liquids, 221, 133-138. Gao, J., Zheng, C., Tan, T., Liu, S., Ji, H. 2018. Enhanced saccharification of rice straw using combined ultra-high pressure and ionic liquid microemulsion pretreatments. 3 Biotech, 8(4), 208. García-Domínguez, M.T., García-Domínguez, J.C., Feria, M.J., Gómez-Lozano, D.M., López, F., Díaz, M.J. 2013. Furfural production from Eucalyptus globulus: Optimizing by using neural fuzzy models. Chemical. Engineering Journal, 221, 185-192. George, A., Brandt, A., Zahari, S., Klein-Marcuschamer, D., Parthasarathi, R., Sun, N., Sathitsuksanoh, N., Shi, J., Stavila, V., Tran, K., Singh, S., Holmes, B.M., Welton, T., Simmons, B., Hallett, J. 2014. Design of low-cost ionic liquids for lignocellulosic biomass pretreatment. Green Chemistry, 17(3), 1728-1734. Ghosh, B., Bhattacharya, D., Mukhopadhyay, M. 2018. Pre-Treatment of Lignocellulose for the Production of Biofuels. Principles and Applications of Fermentation Technology, 307-350. Gunny, A.A.N., Arbain, D., Edwin Gumba, R., Jong, B.C., Jamal, P. 2014. Potential halophilic cellulases for in situ enzymatic saccharification of ionic liquids pretreated lignocelluloses. Bioresource Technology, 155, 177-181. Haddad, B., Mokhtar, D., Goussem, M., Belarbi, E.-h., Villemin, D., Bresson, S., Rahmouni, M., Dhumal, N.R., Kim, H.J., Kiefer, J. 2017. Influence of methyl and propyl groups on the vibrational spectra of two imidazolium ionic liquids and their non-ionic precursors. Journal of Molecular Structure, 1134, 582-590. Hämäläinen, S., Näyhä, A., Pesonen, H.-L. 2011. Forest biorefineries–A business opportunity for the Finnish forest cluster. Journal of Cleaner Production, 19(16), 1884-1891. Hamidah, U., Arakawa, T., H’ng, Y.Y., Nakagawa-izumi, A., Kishino, M. 2018. Recycled ionic liquid 1-ethyl-3-methylimidazolium acetate pretreatment for enhancing enzymatic Capítulo 8 164 saccharification of softwood without cellulose regeneration. Journal of Wood Science, 64(2), 149-156. Hao, Y., Peng, J., Hu, S., Li, J., Zhai, M. 2010. Thermal decomposition of allyl-imidazolium-based ionic liquid studied by TGA–MS analysis and DFT calculations. Thermochimica Acta, 501(1), 78-83. Harmon, K.M., Avci, G.F., Thiel, A.C. 1986. Hydrogen bonding: Part 21. Infrared spectral study of the high-temperature phases of choline bromide and choline iodide. Journal of Molecular Structure, 145(1), 83-91. Hassan, S.S., Williams, G.A., Jaiswal, A.K. 2019. Moving towards the second generation of lignocellulosic biorefineries in the EU: Drivers, challenges, and opportunities. Renewable and Sustainable Energy Reviews, 101, 590-599. Hauru, L.K., Hummel, M., King, A.W., Kilpeläinen, I., Sixta, H. 2012. Role of solvent parameters in the regeneration of cellulose from ionic liquid solutions. Biomacromolecules, 13(9), 2896-2905. Hauru, L.K.J., Ma, Y., Hummel, M., Alekhina, M., King, A.W.T., Kilpelainen, I., Penttila, P.A., Serimaa, R., Sixta, H. 2013. Enhancement of ionic liquid-aided fractionation of birchwood. Part 1: autohydrolysis pretreatment. RSC Advances, 3(37), 16365-16373. Haykir, N.I., Bahcegul, E., Bicak, N., Bakir, U. 2013. Pretreatment of cotton stalk with ionic liquids including 2-hydroxy ethyl ammonium formate to enhance biomass digestibility. Industrial Crops and Products, 41, 430-436. Heggset, E.B., Syverud, K., Øyaas, K. 2016. Novel pretreatment pathways for dissolution of lignocellulosic biomass based on ionic liquid and low temperature alkaline treatment. Biomass and Bioenergy, 93, 194-200. Heikkinen, S., Toikka, M.M., Karhunen, P.T., Kilpeläinen, I.A. 2003. Quantitative 2D HSQC (Q- HSQC) via suppression of J-dependence of polarization transfer in NMR spectroscopy: application to wood lignin. Journal of the American Chemical Society, 125(14), 4362-4367. Hendriks, A.T.W.M., Zeeman, G. 2009. Pretreatments to enhance the digestibility of lignocellulosic biomass. Bioresource Technology, 100(1), 10-18. Himmelsbach, D.S., Khalili, S., Akin, D.E. 2002. The use of FT‐IR microspectroscopic mapping to study the effects of enzymatic retting of flax (Linum usitatissimum L) stems. Journal of the Science of Food and Agriculture, 82(7), 685-696. Hofer, R. 2009. Sustainable solutions for modern economies. Royal Society of Chemistry. Hou, Q., Ju, M., Li, W., Liu, L., Chen, Y., Yang, Q. 2017. Pretreatment of lignocellulosic biomass with ionic liquids and ionic liquid-based solvent systems. Molecules, 22(3), 490. Huang, K., Wu, R., Cao, Y., Li, H., Wang, J. 2013. Recycling and Reuse of Ionic Liquid in Homogeneous Cellulose Acetylation. Chinese Journal of Chemical Engineering, 21(5), 577-584. Ibáñez, A.B., Bauer, S. 2014. Downscaled method using glass microfiber filters for the determination of Klason lignin and structural carbohydrates. Biomass and Bioenergy, 68, 75-81. IDAE. 2011. Plan de Energías Renovables (PER) 2011–2020. Isikgor, F.H., Becer, C.R. 2015. Lignocellulosic biomass: a sustainable platform for the production of bio-based chemicals and polymers. Polymer Chemistry, 6(25), 4497-4559. Jiang, L.-Q., Fang, Z., Li, X.-K., Luo, J., Fan, S.-P. 2013. Combination of dilute acid and ionic liquid pretreatments of sugarcane bagasse for glucose by enzymatic hydrolysis. Process Biochemistry, 48(12), 1942-1946. Kadić, A., Palmqvist, B., Lidén, G. 2014. Effects of agitation on particle-size distribution and enzymatic hydrolysis of pretreated spruce and giant reed. Biotechnology for Biofuels, 7(77). Kamiya, N., Matsushita, Y., Hanaki, M., Nakashima, K., Narita, M., Goto, M., Takahashi, H. 2008. Enzymatic in situ saccharification of cellulose in aqueous-ionic liquid media. Biotechnology Letters, 30(6), 1037-1040. Kamm, B., Gruber, P.R., Kamm, M. 2007. Biorefineries–industrial processes and products. Ullmann's Encyclopedia of Industrial Chemistry. Kandhola, G., Djioleu, A., Carrier, D.J., Kim, J.-W. 2017. Pretreatments for Enhanced Enzymatic Hydrolysis of Pinewood: a Review. BioEnergy Research, 10(4), 1138-1154. Bibliografía 165 Karimi, K., Taherzadeh, M.J. 2016. A critical review of analytical methods in pretreatment of lignocelluloses: Composition, imaging, and crystallinity. Bioresource Technology, 200, 1008-1018. Karperien, A. 2013. FracLac for ImageJ, 2017, FracLac. https://imagej.nih.gov/ij/plugins/fraclac/FLHelp/Introduction.htm. Kassaye, S., Pant, K.K., Jain, S. 2016. Synergistic effect of ionic liquid and dilute sulphuric acid in the hydrolysis of microcrystalline cellulose. Fuel Processing Technology, 148, 289-294. Kathirgamanathan, K., Grigsby, W.J., Al-Hakkak, J., Edmonds, N.R. 2015. Two-Dimensional FTIR as a Tool to Study the Chemical Interactions within Cellulose-Ionic Liquid Solutions. International Journal of Polymer Science. Kim, H., Ralph, J. 2010. Solution-state 2D NMR of ball-milled plant cell wall gels in DMSO- d(6)/pyridine-d(5). Organic & Biomolecular Chemistry, 8, 576-591. Kim, S., Park, J.M., Seo, J.-W., Kim, C.H. 2012. Sequential acid-/alkali-pretreatment of empty palm fruit bunch fiber. Bioresource Technology, 109, 229-233. Klein‐Marcuschamer, D., Simmons, B.A., Blanch, H.W. 2011. Techno‐economic analysis of a lignocellulosic ethanol biorefinery with ionic liquid pre‐treatment. Biofuels, Bioproducts and Biorefining, 5(5), 562-569. Kline, L.M., Hayes, D.G., Womac, A.R., Labbe, N. 2010. Simplified determination of lignin content in hard and soft woods via UV-spectrophotometric analysis of biomass dissolved in ionic liquids. BioResources, 5(3), 1366-1383. Kostamo, A., Kukkonen, J.V.K. 2003. Removal of resin acids and sterols from pulp mill effluents by activated sludge treatment. Water Research, 37(12), 2813-2820. Kumar, P., Barrett, D.M., Delwiche, M.J., Stroeve, P. 2009. Methods for Pretreatment of Lignocellulosic Biomass for Efficient Hydrolysis and Biofuel Production. Industrial & Engineering Chemistry Research, 48(8), 3713-3729. Lai, C., Tu, M., Shi, Z., Zheng, K., Olmos, L.G., Yu, S. 2014. Contrasting effects of hardwood and softwood organosolv lignins on enzymatic hydrolysis of lignocellulose. Bioresource Technology, 163, 320-327. Lan, W., Liu, C.-F., Yue, F.-X., Sun, R.-C., Kennedy, J.F. 2011. Ultrasound-assisted dissolution of cellulose in ionic liquid. Carbohydr. Polym., 86(2), 672-677. Lee, J.W., Kim, J.Y., Jang, H.M., Lee, M.W., Park, J.M. 2015. Sequential dilute acid and alkali pretreatment of corn stover: Sugar recovery efficiency and structural characterization. Bioresource Technology, 182, 296-301. Lee, S.H., Doherty, T.V., Linhardt, R.J., Dordick, J.S. 2009. Ionic liquid-mediated selective extraction of lignin from wood leading to enhanced enzymatic cellulose hydrolysis. Biotechnology and Bioengineering, 102(5), 1368-1376. Li, C., Knierim, B., Manisseri, C., Arora, R., Scheller, H.V., Auer, M., Vogel, K.P., Simmons, B.A., Singh, S. 2010. Comparison of dilute acid and ionic liquid pretreatment of switchgrass: Biomass recalcitrance, delignification and enzymatic saccharification. Bioresource Technology, 101(13), 4900-4906. Li, C., Sun, L., Simmons, B.A., Singh, S. 2013a. Comparing the Recalcitrance of Eucalyptus, Pine, and Switchgrass Using Ionic Liquid and Dilute Acid Pretreatments. BioEnergy Research, 6(1), 14-23. Li, H., Pu, Y., Kumar, R., Ragauskas, A.J., Wyman, C.E. 2014. Investigation of lignin deposition on cellulose during hydrothermal pretreatment, its effect on cellulose hydrolysis, and underlying mechanisms. Biotechnology and bioengineering, 111(3), 485-492. Li, W.-Z., Ju, M.-T., Wang, Y.-N., Liu, L., Jiang, Y. 2013b. Separation and recovery of cellulose from Zoysia japonica by 1-allyl-3-methylimidazolium chloride. Carbohydrate Polymers, 92(1), 228-235. Lin, N., Dufresne, A. 2014. Nanocellulose in biomedicine: Current status and future prospect. European Polymer Journal, 59, 302-325. Liu, C.G., Qin, J.C., Liu, L.Y., Jin, B.W., Bai, F.W. 2016. Combination of Ionic Liquid and Instant Catapult Steam Explosion Pretreatments for Enhanced Enzymatic Digestibility of Rice Straw. ACS Sustainable Chemistry and Engineering, 4(2), 577-582. Capítulo 8 166 Liu, K., Ostadhassan, M. 2017. Quantification of the microstructures of Bakken shale reservoirs using multi-fractal and lacunarity analysis. Journal of Natural Gas Science and Engineering, 39, 62-71. Liu, L., Ju, M., Li, W., Jiang, Y. 2014. Cellulose extraction from Zoysia japonica pretreated by alumina-doped MgO in AMIMCl. Carbohydrate Polymers, 113, 1-8. Lozano, P., Bernal, B., Recio, I., Belleville, M.-P. 2012. A cyclic process for full enzymatic saccharification of pretreated cellulose with full recovery and reuse of the ionic liquid 1- butyl-3-methylimidazolium chloride. Green Chem., 14(9), 2631-2637. Lynam, J.G., Coronella, C.J. 2014. Glycerol as an ionic liquid co-solvent for pretreatment of rice hulls to enhance glucose and xylose yield. Bioresour. Technol., 166, 471-478. MacAskill, J., Suckling, I., Lloyd, J., Manley-Harris, M. 2018. Unravelling the effect of pretreatment severity on the balance of cellulose accessibility and substrate composition on enzymatic digestibility of steam-pretreated softwood. Biomass and Bioenergy, 109, 284-290. Mai, N.L., Ahn, K., Koo, Y.-M. 2014. Methods for recovery of ionic liquids—A review. Process Biochem., 49(5), 872-881. Maitan-Alfenas, G.P., Visser, E.M., Guimarães, V.M. 2015. Enzymatic hydrolysis of lignocellulosic biomass: converting food waste in valuable products. Current Opinion in Food Science, 1, 44-49. Mandelbrot, B. 1977. Fractals, Form, Chance and Dimension WH Freeman and Co. Mandelbrot, B.B., Pignoni, R. 1983. The fractal geometry of nature. WH freeman New York. Martín-Sampedro, R., Fillat, Ú., Ibarra, D., Eugenio, M.E. 2015. Use of new endophytic fungi as pretreatment to enhance enzymatic saccharification of Eucalyptus globulus. Bioresource Technology, 196, 383-390. Mead, D.J. 2013. Sustainable management of Pinus radiata plantations. Food and agriculture organization of the United nations (FAO). Menrad, K., Klein, A., Kurka, S. 2009. Interest of industrial actors in biorefinery concepts in Europe. Biofuels, Bioproducts and Biorefining, 3(3), 384-394. Mussatto, S.I. 2016. Biomass fractionation technologies for a lignocellulosic feedstock based biorefinery. Elsevier. Nargotra, P., Sharma, V., Gupta, M., Kour, S., Bajaj, B.K. 2018. Application of ionic liquid and alkali pretreatment for enhancing saccharification of sunflower stalk biomass for potential biofuel-ethanol production. Bioresource Technology, 267, 560-568. Ngoc Lan Maia, S.H.H., Yoon-Mo Koo. 2014. Efficient pretreatment of lignocellulose in ionic liquids/co-solvent forenzymatic hydrolysis enhancement into fermentable sugars. Process Biochemistry, 49, 1144-1151. Nguyen, T.-A.D., Kim, K.-R., Han, S.J., Cho, H.Y., Kim, J.W., Park, S.M., Park, J.C., Sim, S.J. 2010. Pretreatment of rice straw with ammonia and ionic liquid for lignocellulose conversion to fermentable sugars. Bioresource Technology, 101(19), 7432-7438. Ninomiya, K., Inoue, K., Aomori, Y., Ohnishi, A., Ogino, C., Shimizu, N., Takahashi, K. 2015a. Characterization of fractionated biomass component and recovered ionic liquid during repeated process of cholinium ionic liquid-assisted pretreatment and fractionation. Chemical Engineering Journal, 259, 323-329. Ninomiya, K., Kohori, A., Tatsumi, M., Osawa, K., Endo, T., Kakuchi, R., Ogino, C., Shimizu, N., Takahashi, K. 2015b. Ionic liquid/ultrasound pretreatment and in situ enzymatic saccharification of bagasse using biocompatible cholinium ionic liquid. Bioresource Technology, 176, 169-174. Ninomiya, K., Yamauchi, T., Kobayashi, M., Ogino, C., Shimizu, N., Takahashi, K. 2013. Cholinium carboxylate ionic liquids for pretreatment of lignocellulosic materials to enhance subsequent enzymatic saccharification. Biochemical Engineering Journal, 71, 25-29. Nishimura, H., Kamiya, A., Nagata, T., Katahira, M., Watanabe, T. 2018. Direct evidence for α ether linkage between lignin and carbohydrates in wood cell walls. Scientific Reports, 8(6538). Nitsos, C.K., Choli-Papadopoulou, T., Matis, K.A., Triantafyllidis, K.S. 2016. Optimization of Hydrothermal Pretreatment of Hardwood and Softwood Lignocellulosic Residues for Selective Hemicellulose Recovery and Improved Cellulose Enzymatic Hydrolysis. ACS Sustainable Chemistry and Engineering, 4(9), 4529-4544. Bibliografía 167 NREL: Definición de biorrefinería. http://www.nrel.gov/biomass/biorefinery.html (Último acceso: 20/05/2016). Octave, S., Thomas, D. 2009. Biorefinery: Toward an industrial metabolism. Biochimie, 91(6), 659- 664. Olivier-Bourbigou, H., Magna, L., Morvan, D. 2010. Ionic liquids and catalysis: recent progress from knowledge to applications. Applied Catalysis A: General, 373(1), 1-56. Otsu, N. 1979. A threshold selection method from gray-level histograms. IEEE Transactions on Systems, Management, and Cybernetics, 9(1), 62-66. Özaydın, B., Burd, H., Lee, T.S., Keasling, J.D. 2013. Carotenoid-based phenotypic screen of the yeast deletion collection reveals new genes with roles in isoprenoid production. Metabolic Engineering, 15, 174-183. Pandey, K.K., Pitman, A.J. 2003. FTIR studies of the changes in wood chemistry following decay by brown-rot and white-rot fungi. International Biodeterioration & Biodegradation, 52(3), 151-160. Papa, G., Varanasi, P., Sun, L., Cheng, G., Stavila, V., Holmes, B., Simmons, B.A., Adani, F., Singh, S. 2012. Exploring the effect of different plant lignin content and composition on ionic liquid pretreatment efficiency and enzymatic saccharification of Eucalyptus globulus L. mutants. Bioresource technology, 117, 352-359. Papadopoulou, A.A., Tzani, A., Alivertis, D., Katsoura, M.H., Polydera, A.C., Detsi, A., Stamatis, H. 2016. Hydroxyl ammonium ionic liquids as media for biocatalytic oxidations. Green Chemistry, 18(4), 1147-1158. Parthasarathi, R., Romero, R.A., Redondo, A., Gnanakaran, S. 2011. Theoretical study of the remarkably diverse linkages in lignin. The Journal of Physical Chemistry Letters, 2(20), 2660-2666. Peleteiro, S., Rivas, S., Alonso, J.L., Santos, V., Parajó, J.C. 2015. Utilization of ionic liquids in lignocellulose biorefineries as agents for separation, derivatization, fractionation, or pretreatment. Journal of Agricultural and Food Chemistry, 63(37), 8093-8102. Pielhop, T., Larrazábal, G.O., Studer, M.H., Brethauer, S., Seidel, C.-M., von Rohr, P.R. 2015. Lignin repolymerisation in spruce autohydrolysis pretreatment increases cellulase deactivation. Green Chemistry, 17(6), 3521-3532. Pin, T.C., Nakasu, P.Y.S., Mattedi, S., Rabelo, S.C., Costa, A.C. 2019. Screening of protic ionic liquids for sugarcane bagasse pretreatment. Fuel, 235, 1506-1514. Prado, R., Erdocia, X., Labidi, J. 2014. Ionic liquid application for lignin extraction and depolymerization. International Journal of Condensed Matter, Advanced Materials, and Superconductivity Research, 13(4), 263. Prado, R., Erdocia, X., Labidi, J. 2016. Study of the influence of reutilization ionic liquid on lignin extraction. Journal of Cleaner Production, 111, Part A, 125-132. Ragauskas, A.J., Williams, C.K., Davison, B.H., Britovsek, G., Cairney, J., Eckert, C.A. 2006. The path forward for biofuels and biomaterials. Science, 311. Raghavi, S., Sindhu, R., Binod, P., Gnansounou, E., Pandey, A. 2016. Development of a novel sequential pretreatment strategy for the production of bioethanol from sugarcane trash. Bioresource Technology, 199, 202-210. Rahikainen, J.L., Martin-Sampedro, R., Heikkinen, H., Rovio, S., Marjamaa, K., Tamminen, T., Rojas, O.J., Kruus, K. 2013. Inhibitory effect of lignin during cellulose bioconversion: The effect of lignin chemistry on non-productive enzyme adsorption. Bioresource Technology, 133, 270-278. Rashid, T., Kait, C.F., Regupathi, I., Murugesan, T. 2016. Dissolution of kraft lignin using Protic Ionic Liquids and characterization. Industrial Crops and Products, 84, 284-293. Ravindran, R., Jaiswal, S., Abu-Ghannam, N., Jaiswal, A.K. 2017. Two-step sequential pretreatment for the enhanced enzymatic hydrolysis of coffee spent waste. Bioresource Technology, 239, 276-284. Risović, D., Poljaček, S.M., Furić, K., Gojo, M. 2008. Inferring fractal dimension of rough/porous surfaces—A comparison of SEM image analysis and electrochemical impedance spectroscopy methods. Applied Surface Science, 255(5), 3063-3070. http://www.nrel.gov/biomass/biorefinery.html Capítulo 8 168 Rocha, E.G.A., Pin, T.C., Rabelo, S.C., Costa, A.C. 2017. Evaluation of the use of protic ionic liquids on biomass fractionation. Fuel, 206, 145-154. Rodriguez, A., Salvachúa, D., Katahira, R., Black, B.A., Cleveland, N.S., Reed, M., Smith, H., Baidoo, E.E.K., Keasling, J.D., Simmons, B.A., Beckham, G.T., Gladden, J.M. 2017. Base- Catalyzed Depolymerization of Solid Lignin-Rich Streams Enables Microbial Conversion. ACS Sustainable Chemistry & Engineering, 5(9), 8171-8180. Romaní, A., Garrote, G., López, F., Parajó, J.C. 2011. Eucalyptus globulus wood fractionation by autohydrolysis and organosolv delignification. Bioresource Technology, 102(10), 5896- 5904. Romaní, A., Ruiz, H.A., Pereira, F.B., Teixeira, J.A., Domingues, L. 2014. Integrated approach for effective bioethanol production using whole slurry from autohydrolyzed Eucalyptus globulus wood at high-solid loadings. Fuel, 135, 482-491. Santos, J.I., Fillat, Ú., Martín-Sampedro, R., Eugenio, M.E., Negro, M.J., Ballesteros, I., Rodríguez, A., Ibarra, D. 2017. Evaluation of lignins from side-streams generated in an olive tree pruning-based biorefinery: Bioethanol production and alkaline pulping. International Journal of Biological Macromolecules, 105, 238-251. Santos, T.M. 2017. Autohidrólisis y deslignificación organosolv de madera de Pinus radiata para la recuperación de hemicelulosas y lignina con aprovechamiento de la fracción celulósica por vía enzimática. Tesis Doctoral, Universidad Complutense de Madrid. Santos, T.M., Alonso, M.V., Oliet, M., Domínguez, J.C., Rigual, V., Rodriguez, F. 2018. Effect of autohydrolysis on Pinus radiata wood for hemicellulose extraction. Carbohydrate Polymers, 194, 285-293. Schwanninger, M., Rodrigues, J.C., Pereira, H., Hinterstoisser, B. 2004. Effects of short-time vibratory ball milling on the shape of FT-IR spectra of wood and cellulose. Vib. Spectrosc, 36(1), 23-40. Schwarze, F.W. 2007. Wood decay under the microscope. fungal biology reviews, 21(4), 133-170. Scopus. 2019. Keywords search. Shi, J., Gladden, J.M., Sathitsuksanoh, N., Kambam, P., Sandoval, L., Mitra, D., Zhang, S., George, A., Singer, S.W., Simmons, B.A., Singh, S. 2013. One-pot ionic liquid pretreatment and saccharification of switchgrass. Green Chemistry, 15(9), 2579-2589. Shi, J., Pattathil, S., Parthasarathi, R., Anderson, N.A., Im Kim, J., Venketachalam, S., Hahn, M.G., Chapple, C., Simmons, B.A., Singh, S. 2016. Impact of engineered lignin composition on biomass recalcitrance and ionic liquid pretreatment efficiency. Green Chemistry, 18(18), 4884-4895. Shill, K., Padmanabhan, S., Xin, Q., Prausnitz, J.M., Clark, D.S., Blanch, H.W. 2011. Ionic liquid pretreatment of cellulosic biomass: enzymatic hydrolysis and ionic liquid recycle. Biotechnology Bioengineering, 108(3), 511-20. Shinde, S.D., Meng, X., Kumar, R., Ragauskas, A.J. 2018. Recent advances in understanding the pseudo-lignin formation in a lignocellulosic biorefinery. Green Chemistry. Shmukler, L.E., Gruzdev, M.S., Kudryakova, N.O., Fadeeva, Y.A., Kolker, A.M., Safonova, L.P. 2016. Thermal behavior and electrochemistry of protic ionic liquids based on triethylamine with different acids. RSC Advances, 6(111), 109664-109671. Shuddhodana, Mohnot, D., Biswas, R., Bisaria, V.S. 2016. Chapter 23 - Enzymatic Hydrolysis of Lignocellulosic Residues A2 - Mussatto, Solange I. in: Biomass Fractionation Technologies for a Lignocellulosic Feedstock Based Biorefinery, Elsevier. Amsterdam, pp. 543-560. Silva-Fernandes, T., Duarte, L.C., Carvalheiro, F., Marques, S., Loureiro-Dias, M.C., Fonseca, C., Gírio, F. 2015. Biorefining strategy for maximal monosaccharide recovery from three different feedstocks: Eucalyptus residues, wheat straw and olive tree pruning. Bioresource Technology, 183, 203-212. Singh, R., Krishna, B.B., Kumar, J., Bhaskar, T. 2016. Opportunities for utilization of non- conventional energy sources for biomass pretreatment. Bioresour. Technol., 199, 398-407. Sjostrom, E. 1993. Wood chemistry: fundamentals and applications. Gulf professional publishing. Sluiter, A., Hames, B., Ruiz, R., Scarlata, C., Sluiter, J., Templeton, D. 2006. Determination of sugars, byproducts, and degradation products in liquid fraction process samples. NREL. Speight, J.G. 2011. The biofuels handbook. Royal Society of Chemistry. Bibliografía 169 Stand, L.I.M., Valencia, G.E., Benavides, A.E. 2018. Trend analysis of research results in biodiesel production from 2009 to 2016/Análisis de tendencia de la los resultados de investigación en producción de biodiésel de 2009 a 2016. Prospectiva, 16(2), 89-93. Stern, T., Ledl, C., Braun, M., Hesser, F., Schwarzbauer, P. 2015. Biorefineries' impacts on the Austrian forest sector: A system dynamics approach. Technological Forecasting and Social Change, 91, 311-326. Su, Z., Yu, Y., Liang, C., Li, L., Yu, S. 2013. Properties of chitosan-immobilized cellulase in ionic liquid. Biotechnology and Applied Biochemistry, 60(2), 231-235. Sun, J., Konda, N.V.S.N.M., Shi, J., Parthasarathi, R., Dutta, T., Xu, F., Scown, C.D., Simmons, B.A., Singh, S. 2016a. CO2 enabled process integration for the production of cellulosic ethanol using bionic liquids. Energy & Environmental Science, 9(9), 2822-2834. Sun, J., Konda, S., Ramakrishnan, P., Dutta, T., Valiev, M., Xu, F., Simmons, B., Singh, S. 2017. One-pot integrated biofuel production using low-cost biocompatible protic ionic liquids. Green Chemistry, 19(13), 3152-3163. Sun, N., Liu, H., Sathitsuksanoh, N., Stavila, V., Sawant, M., Bonito, A., Tran, K., George, A., Sale, K.L., Singh, S. 2013. Production and extraction of sugars from switchgrass hydrolyzed in ionic liquids. Biotechnology for biofuels, 6(1), 1. Sun, N., Parthasarathi, R., Socha, A.M., Shi, J., Zhang, S., Stavila, V., Sale, K.L., Simmons, B.A., Singh, S. 2014. Understanding pretreatment efficacy of four cholinium and imidazolium ionic liquids by chemistry and computation. Green Chemistry, 16(5), 2546-2557. Sun, N., Rahman, M., Qin, Y., Maxim, M.L., Rodriguez, H., Rogers, R.D. 2009. Complete dissolution and partial delignification of wood in the ionic liquid 1-ethyl-3- methylimidazolium acetate. Green Chemistry, 11(5), 646-655. Sun, N., Rodriguez, H., Rahman, M., Rogers, R.D. 2011. Where are ionic liquid strategies most suited in the pursuit of chemicals and energy from lignocellulosic biomass? Chemical Communications, 47(5), 1405-1421. Sun, S., Sun, S., Cao, X., Sun, R. 2016b. The role of pretreatment in improving the enzymatic hydrolysis of lignocellulosic materials. Bioresource Technology, 199, 49-58. Sun, X., Sun, X., Zhang, F. 2016c. Combined pretreatment of lignocellulosic biomass by solid base (calcined Na2SiO3) and ionic liquid for enhanced enzymatic saccharification. RSC Advances, 6(101), 99455-99466. Sun, Y., Cheng, J. 2002. Hydrolysis of lignocellulosic materials for ethanol production: a review. Bioresource technology, 83(1), 1-11. Sundstrom, E., Yaegashi, J., Yan, J., Masson, F., Papa, G., Rodriguez, A., Mirsiaghi, M., Liang, L., He, Q., Tanjore, D. 2018. Demonstrating a separation-free process coupling ionic liquid pretreatment, saccharification, and fermentation with Rhodosporidium toruloides to produce advanced biofuels. Green Chemistry. Swatloski, R.P., Spear, S.K., Holbrey, J.D., Rogers, R.D. 2002. Dissolution of cellose with ionic liquids. Journal of the American Chemical Society, 124(18), 4974-4975. Tanzi, L., Benassi, P., Nardone, M., Ramondo, F. 2014. Vibrations of Bioionic Liquids by Ab Initio Molecular Dynamics and Vibrational Spectroscopy. The Journal of Physical Chemistry A, 118(51), 12229-12240. Torr, K.M., Love, K.T., Simmons, B.A., Hill, S.J. 2016. Structural features affecting the enzymatic digestibility of pine wood pretreated with ionic liquids. Biotechnology and Bioengineering, 113(3), 540-549. Trinh, L.T.P., Lee, Y.J., Lee, J.-W., Bae, H.-J., Lee, H.-J. 2013. Recovery of an ionic liquid [BMIM]Cl from a hydrolysate of lignocellulosic biomass using electrodialysis. Separation and Purification Technology, 120, 86-91. Trinh, L.T.P., Lee, Y.J., Lee, J.-W., Lee, H.-J. 2015. Characterization of ionic liquid pretreatment and the bioconversion of pretreated mixed softwood biomass. Biomass and Bioenergy, 81, 1-8. Ungurean, M., Csanádi, Z., Gubicza, L., Péter, F. 2014. An Integrated Process of Ionic Liquid Pretreatment and Enzymatic Hydrolysis of Lignocellulosic Biomass with Immobilised Cellulase. BioResources, 9(4), 6100-6116. Capítulo 8 170 Utrilla-Coello, R.G., Bello-Pérez, L.A., Vernon-Carter, E.J., Rodriguez, E., Alvarez-Ramirez, J. 2013. Microstructure of retrograded starch: Quantification from lacunarity analysis of SEM micrographs. Journal of Food Engineering, 116(4), 775-781. Van Dyk, J.S., Pletschke, B.I. 2012. A review of lignocellulose bioconversion using enzymatic hydrolysis and synergistic cooperation between enzymes factors affecting enzymes, conversion and synergy. Biotechnology Advances, 30(6), 1458-80. van Osch, D.J., Kollau, L.J., van den Bruinhorst, A., Asikainen, S., Rocha, M.A., Kroon, M.C. 2017. Ionic liquids and deep eutectic solvents for lignocellulosic biomass fractionation. Physical Chemistry Chemical Physics, 19(4), 2636-2665. van Spronsen, J., Cardoso, M.A.T., Witkamp, G.-J., de Jong, W., Kroon, M.C. 2011. Separation and recovery of the constituents from lignocellulosic biomass by using ionic liquids and acetic acid as co-solvents for mild hydrolysis. Chemical Engineering and Processing: Process Intensification, 50(2), 196-199. Vanholme, R., Demedts, B., Morreel, K., Ralph, J., Boerjan, W. 2010. Lignin Biosynthesis and Structure. Plant Physiology, 153(3), 895-905. Wahlström, R. 2014. Enzymatic hydrolysis of cellulose in aqueous ionic liquids. VTT Technical Research Centre of Finland. Wahlström, R.M., Suurnäkki, A. 2015. Enzymatic hydrolysis of lignocellulosic polysaccharides in the presence of ionic liquids. Green Chemistry, 17(2), 694-714. Waldron, K.W. 2010. Bioalcohol production: biochemical conversion of lignocellulosic biomass. Elsevier. Wang, H., Ben, H., Ruan, H., Zhang, L., Pu, Y., Feng, M., Ragauskas, A.J., Yang, B. 2017. Effects of Lignin Structure on Hydrodeoxygenation Reactivity of Pine Wood Lignin to Valuable Chemicals. ACS Sustainable Chemistry & Engineering, 5(2), 1824-1830. Weerachanchai, P., Lee, J.-M. 2014. Recyclability of an ionic liquid for biomass pretreatment. Bioresource Technology, 169, 336-343. Weerachanchai, P., Lim, K.H., Lee, J.-M. 2014. Influence of organic solvent on the separation of an ionic liquid from a lignin–ionic liquid mixture. Bioresource Technology, 156(0), 404-407. Wen, J.-L., Yuan, T.-Q., Sun, S.-L., Xu, F., Sun, R.-C. 2014. Understanding the chemical transformations of lignin during ionic liquid pretreatment. Green Chemistry, 16(1), 181- 190. Windeisen, E., Wegener, G. 2009. Capítulo.4: Wood. en: Sustainable Solutions for Modern Economies, pp. 300-338. Xu, F., Sun, J., Konda, N.M., Shi, J., Dutta, T., Scown, C.D., Simmons, B.A., Singh, S. 2016a. Transforming biomass conversion with ionic liquids: process intensification and the development of a high-gravity, one-pot process for the production of cellulosic ethanol. Energy & Environmental Science, 9(3), 1042-1049. Xu, J., He, B., Wu, B., Wang, B., Wang, C., Hu, L. 2014. An ionic liquid tolerant cellulase derived from chemically polluted microhabitats and its application in in situ saccharification of rice straw. Bioresource Technology, 157, 166-173. Xu, J., Liu, B., Hou, H., Hu, J. 2017. Pretreatment of eucalyptus with recycled ionic liquids for low- cost biorefinery. Bioresource Technology, 234, 406-414. Xu, J., Liu, X., He, J., Hu, L., Dai, B., Wu, B. 2015a. Enzymatic in situ saccharification of rice straw in aqueous-ionic liquid media using encapsulated Trichoderma aureoviride cellulase. Journal of Chemical Technology and Biotechnology, 90(1), 57-63. Xu, J., Sheng, Z., Wang, X., Liu, X., Xia, J., Xiong, P., He, B. 2016b. Enhancement in ionic liquid tolerance of cellulase immobilized on PEGylated graphene oxide nanosheets: Application in saccharification of lignocellulose. Bioresource Technology, 200, 1060-1064. Xu, J., Wu, B., Hu, L., Wu, Z., Xu, N., Dai, B., He, J. 2015b. Enzymatic in situ saccharification of lignocellulose in a compatible ionic liquid-cellulase system. Chemical Engineering Journal, 267, 163-169. Xu, J., Xiong, P., He, B. 2016c. Advances in improving the performance of cellulase in ionic liquids for lignocellulose biorefinery. Bioresource Technology, 200, 961-970. Bibliografía 171 Xue-Dan Hou, J.X., Ning Li Min-Hua Zong. 2015. Effect of anion Structures on Cholinium Ionic Liquids pretreatment of Rice Sraw and the subsequent Enzymatic Hydrolysis. Biotechnology and Bioengineering, 112, 65-73. Yaegashi, J., Kirby, J., Ito, M., Sun, J., Dutta, T., Mirsiaghi, M., Sundstrom, E.R., Rodriguez, A., Baidoo, E., Tanjore, D. 2017. Rhodosporidium toruloides: a new platform organism for conversion of lignocellulose into terpene biofuels and bioproducts. Biotechnology for biofuels, 10(1), 241. Yildiz, S.Y., Oner, E.T. 2014. Mannan as a Promising Bioactive Material for Drug Nanocarrier Systems. in: Application of Nanotechnology in Drug Delivery, (Ed.) A.D. Sezer, InTech. Rijeka, pp. Ch. 09. Zhang, H., Wu, J., Zhang, J., He, J. 2005. 1-Allyl-3-methylimidazolium Chloride Room Temperature Ionic Liquid:  A New and Powerful Nonderivatizing Solvent for Cellulose. Macromolecules, 38(20), 8272-8277. Zhao, H., Jones, C.L., Baker, G.A., Xia, S., Olubajo, O., Person, V.N. 2009. Regenerating cellulose from ionic liquids for an accelerated enzymatic hydrolysis. Journal of Biotechnology, 139(1), 47-54. ANEXO I: LISTADO DE LÍQUIDOS IÓNICOS Y ESTRUCTURA QUÍMICA Anexos 175 LISTADO DE LÍQUIDOS IÓNICOS Y ESTRUCTURA QUÍMICA [Amim][Cl] Cloruro de 1-alil-3- metilimidazolio [b-OHEtAm][OAc] Acetato de bis (2- hidroxietilamonio) [Ch][Lys] Lisinato de colina [Ch][OAc] Acetato de colina 176 [Dmim][DMPO4] Dimetilfosfato de 1, 3- dimetilimidazolio [Emim][DEPO4] Dietilfosfato de 1-etil-3- metilimidazolio [Emim][HSO4] Hidrogenosulfato de 1- etil-3-metilimidazolio [Emim][OAc] Acetato de 1-etil-3- metilimidazolio [Emim][Tf2N] Bis(trifluorometilsulfoni l)imida de 1-etil-3- metilimidazolio [Mim][Cl] Cloruro de 1- metilimidazolio Anexos 177 [OHEtAm][HCO2] Formato de 2- hidroxietilamonio [OHEtAm][OAc] Acetato de 2- hidroxietilamonio [TBA][L-Pro] L-prolinato de tetrabutilamonio [TBA][R-Pro] R-prolinato de tetrabutilamonio [TBA][t-4OH-L-Pro] Trans-4-hidroxi-L- prolinato de tetrabutilamonio [TBP][S-Pro] L-prolinato de tetrabutilfosfonio [TEtAm][OAc] Acetato de trietilamonio ANEXO II: PUBLICACIONES PUBLICACIÓN I “Evaluation of hardwood and softwood fractionation using autohydrolysis and ionic liquid microwave pretreatment” Victoria Rigual, Tamara M. Santos, Juan C. Domínguez, M. Virginia. Alonso, Mercedes Oliet y Francisco Rodriguez Biomass and Bioenergy, 2018, 117, 190-197 Contents lists available at ScienceDirect Biomass and Bioenergy journal homepage: www.elsevier.com/locate/biombioe Research paper Evaluation of hardwood and softwood fractionation using autohydrolysis and ionic liquid microwave pretreatment Victoria Rigual∗, Tamara M. Santos, Juan Carlos Domínguez, M. Virginia Alonso, Mercedes Oliet, Francisco Rodriguez Department of Chemical Engineering and Materials, Faculty of Chemistry, Complutense University of Madrid. Avda. Complutense s/n, 28040, Madrid, Spain A R T I C L E I N F O Keywords: Softwood Hardwood Autohydrolyis Microwave ionic liquid Confocal fluorescence microscopy Biomass pretreatment A B S T R A C T Differences in hardwood and softwood, elucidates their behaviour against pretreatments varies. In this work, microwave ionic liquid (IL) and autohydrolysis (AH) pretreatments were applied to Eucalyptus globulus (as a model of hardwood) and Pinus radiata (as a model of softwood). The comparison between hardwood and soft- wood of microwave ionic liquid (IL) and autohydrolysis (AH) were evaluated in terms of chemical composition of pretreated solids, liquid by streams composition (hemicellulose and lignin extraction) and, substrates enzy- matic digestibility. Furthermore, micrographs using scanning electron microscopy (SEM) and confocal fluores- cence microscopy supported results obtained. In this study, it has been demonstrated that autohydrolysis pre- treatment effectiveness, through maximizing enzymatic digestibility, is opposite in hardwood (73 g glucan/100 g glucan introduced at severe conditions) and softwood (10 g/100 g glucan). IL pretreatment has been especially effective in softwood with higher digestibilities (78 g glucan/100 g glucan introduced) than those obtained in hardwood (68 g glucan/100 g glucan introduced). Confocal fluorescence microscopy images, together with SEM images have resulted to be a clarifying technique to explain enzymatic digestibility results. Final sugars yields after the whole process have shown that low solid yields recoveries obtained in AH treatments have considerably worsened final glucose production, mainly in softwood. IL microwave pretreatment have resulted in higher glucose yields in softwood than in hardwood. 1. Introduction Despite recent advances, the biorefinery of lignocellulosic biomass is still a challenge [1]. There is a bottle-neck of the process in the conversion of complex carbohydrates to fermentable sugars [2]. Lig- nocellulose is a recalcitrance non-uniform three-dimensional structure, that requires pretreatment processes to deconstruct the linkages and disrupt the structure [3]. Pretreatment technologies can constitute up to 40% of the total processing costs of lignocellulosic biomass conversion [4]. Autohydrolysis (AH) has been described as an inexpensive, en- vironmentally friendly and easy-handle process to selectively remove hemicellulose with low cellulose and lignin degradation [5,6]. AH only uses water as reactive, which results in the water autoionization to- wards acid hydronium ions (H3O+). Oligosaccharides obtained in the liquid phase are value added products used in food and pharmaceutical industries [7]. Ionic liquids (ILs) are effective biomass solvents that reduce recalcitrance, enabling deconstruction, and disruption of the lignin and hemicellulose network [8]. The non-flammability, high chemical and thermal stability, and negligible vapour pressure are some of the advantages against other pretreatment processes [9]. ILs also reduce cellulose crystallinity, increasing its accessibility and favouring high glucose conversions [10–12]. ILs are good microwave absorbers, enhancing biomass conversion processes and fastening the heating rate [13,14] However, the use of microwave must be assessed to avoid very severe conditions that may produce degradation [15]. Some studies have already been reported comparing softwoods and hardwoods pretreatments [16–19]. Pretreatments are, in general, more effective in hardwoods than in softwoods [19]. In this work, a comparison between hardwood and softwood pre- https://doi.org/10.1016/j.biombioe.2018.07.014 Received 20 November 2017; Received in revised form 4 June 2018; Accepted 20 July 2018 Abbreviations: IL, Ionic liquid; [Emim][OAc], 1-ethyl-3-methylimidazolium acetate; AH, autohydrolysis; OS, oligosaccharides ∗ Corresponding author. E-mail addresses: vicrigua@ucm.es, rigual.victoria@gmail.com (V. Rigual). Biomass and Bioenergy 117 (2018) 190–197 Available online 10 August 2018 0961-9534/ © 2018 Elsevier Ltd. All rights reserved. T http://www.sciencedirect.com/science/journal/09619534 https://www.elsevier.com/locate/biombioe https://doi.org/10.1016/j.biombioe.2018.07.014 https://doi.org/10.1016/j.biombioe.2018.07.014 mailto:vicrigua@ucm.es mailto:rigual.victoria@gmail.com https://doi.org/10.1016/j.biombioe.2018.07.014 http://crossmark.crossref.org/dialog/?doi=10.1016/j.biombioe.2018.07.014&domain=pdf treatment efficiency are performed. Autohydrolysis and IL microwave pretreatments at different severity conditions are carried out. Chemical compositions of solid and liquid fractions for each pretreatment were determined. Solids enzymatic hydrolysis digestibilities were evaluated, compared and, supported with morphological structures obtained from scanning electron microscopy and confocal fluorescence laser micro- scopy techniques, in each case. In this study, identical enzyme cocktails, loading levels, and analytical methods were employed in order to offer an overview of the sugars and by-products obtained. 2. Material and methods 2.1. Materials and reagents The ionic liquid 1-ethyl-3-methylimidazolium acetate ([Emim] [OAc],> 95%, Iolitec GmbH) was employed for wood dissolution. Eucalyptus globulus and Pinus radiata sawdust were provided by the National Institute for Agronomic Research (CIFOR-INIA). Organosolv lignin from E. globulus and P. radiata wood were used as reference materials, for UV/VIS measurements (explained below). Sulfuric acid was used to precipitate lignin from the mixture. The enzymatic cocktail Accellerase 1500®, containing 70mg of protein/mL of dissolution was donated by Dupont Industrial BioSciences and used in the enzymatic hydrolysis step. 2.2. Experimental 2.2.1. Autohydrolysis (AH) pretreatment AH pretreatments were developed in a 450mL stainless steel pres- sure reactor (Parr Instrument Company, model 4567). The reactor was fitted with a four blade turbine impeller working at 150 rpm. Before pretreatment, wood chips were milled and sieved to obtain a particle size between 0.3 and 2mm. Eucalyptus and pine wood sawdust were mixed with deionized water in a liquid:solid ratio of 8:1 and 10:1 (g water: g dry biomass), respectively [20,21]. Mild autohydrolysis were developed at 150 °C for 30min for both woods (AH150E for eucalyptus wood and AH150P for pine wood); intermediate, at 175 °C for 30 and 60min in the case of eucalyptus and pine wood (AH175E and AH175P); and severe autohydrolysis conditions were performed at 200 °C for 30 and 90min (AH200E and AH200P). The AH severity factor (S0) was calculated [22]. 2.2.2. IL microwave pretreatment Eucalyptus and pine were milled and sieved to obtain particles with sizes< 150 μm. Extractives were removed using acetone and water to avoid foam formation that may cause dissolution problems [23,24]. 0.8 g extractives free samples of eucalyptus and pine wood were mixed with 20 g of [Emim][OAc]. Samples were heated under microwave ir- radiation in a Berghof SpeedWave Four microwave oven, using a two- step programme detailed in a previous work [25]. Operation tem- peratures were 80 °C (IL80E for eucalyptus and IL80P for pine wood) and 120 °C (IL120E and IL120P) in a total time of 50min. Afterwards, 50mL of deionized water was added to precipitate the wood dissolved in the process. The solution was stirred for 10min in a water bath at 40 °C, and was subsequently filtered under a vacuum to obtain the pretreated wood. Pretreated samples were washed 5-fold with 70mL of deionized water. 2.2.3. Enzymatic hydrolysis The enzymatic hydrolysis was carried out in an orbital incubator at 150 rpm and at 50 °C. 1% (w/w) pretreated wood (< 150 μm) was suspended in 50mM citrate buffer (pH 5.0) containing 0.002% of sodium azide in a working volume of 8mL. Accellerase 1500 enzymatic cocktail in a dosage of 0.25mL/g glucan was added. Aliquots of 150 μL were periodically taken at 3, 6, 12, 24, 48 and 72 h, and centrifuged to stop the enzymatic reaction. 2.3. Analytical methods 2.3.1. Chemical characterization of biomass samples Biomass compositions, before and after pretreatments, were de- termined according to the NREL/TP-510-42618 methodology adapted to small quantities of samples [26,27]. The acid-soluble lignin amount was determined using a Varian Cary 50 UV-VIS spectrometer at 205 and 240 nm, with an absorptivity of 110 and 12 L g-1 cm-1 for eucalyptus and pine wood, respectively. Sugars in the hydrolysate were determined by HPLC, neutralizing with CaCO3 and, filtering under 0.45 μm before the analysis, using a 300×7.8mm Casbosep-CHO 682 column with Micro-Guard cartridges (BioRad, Life Science Group Hercules, Ca) at 80 °C, using water as mobile phase, and a flow rate of 0.4mL/min. 2.3.2. Morphology of biomass samples A Jeol JSM 6400 scanning electron microscope (SEM) was em- ployed to observe the surfaces of untreated and pretreated samples. A Gold sputtering onto the sample surface was used to impart electrical conductivity. The operation voltage of the SEM was 20 kV. Analysis were developed in the technical facilities of the Spanish National Centre for Electron Microscopy. A Leica SP-2 AOBS confocal laser microscope was used to visualize supramolecular structure changes. A laser at 405 nm was used to excite samples for fluorescence visualization. Wavelength emission ranges were 428–480 nm and 547–658 nm for hollocellulose and lignin (au- tofluorescent) respectively. β-1-4 polysaccharides linkages were dyed using 0.1% Calcofluor white stain [3]. Images were acquired at a step size of 2 μm and were combined into a z-axis max projection using Image-J software. Analysis were developed in the technical facilities of the Centre for Cytometry and Fluorescence Microscopy of the Com- plutense University of Madrid (UCM). 2.3.3. Chemical composition of autohydrolysis liquors Monomeric sugars, organic acids and furans obtained from auto- hydrolysis liquors were directly measured by HPLC using a refractive index detector, according to the NREL/TP-510-42623 procedure [28]. The above mentioned operating conditions were used for sugars de- termination. Organic acids and furans were determined, using a Rezex ROA-Organic Acid H+ (8%) 300 × 7.8 mm column at 60 °C, with a mobile phase (0.005 M H2SO4) eluted at 0.6mL/min. 2.3.4. Chemical composition of liquid by-stream obtained after microwave IL pretreatment Recovered IL in the washing fractions was quantified by HPLC equipped with a UV detector measuring at 235 nm. The Eclipse Plus C18 4.6×100mm column was operated using a mixture of acetoni- trile/water 50/50 % (v/v) as mobile phase with a flow rate of 1mL/min and at 30 °C. The lignin content accumulated in the IL was analyzed by UV/VIS spectroscopy using a Varian Cary 50 scan UV/VIS spectrophotometer. A rotary evaporator was used to remove the water and recover the IL. Samples were diluted in 0.1 N NaOH and filtered, to measure the ab- sorbance at 280 nm [29]. Water content in the recovered ionic liquid was determined using a thermobalance and mass was corrected. The total dissolved lignin concentration was obtained from the reference curve of eucalyptus and pine wood organosolv lignin samples. Re- ference samples used were prepared. The absorbance measurement of V. Rigual et al. Biomass and Bioenergy 117 (2018) 190–197 191 each sample was subtracted of a blank sample prepared with the mi- crowave heated ionic liquid at the working temperature. 2.3.5. Enzymatic digestibility Glucose and xylose were determined by HPLC as explained above (section 2.3.3). Glucan and xylan digestibilities were defined as the percentage of glucan and xylan converted into glucose or xylose during enzymatic hydrolysis, according to the NREL/TP-5100-63351 proce- dure [30]. 3. Results and discussion 3.1. Autohydrolysis of eucalyptus and pine wood Solid recovery and compositional analysis are shown in Table 1. A decrease of the solid recovery is produced from mild conditions to se- vere conditions. The relative composition of both woods as compared to the untreated sample, varies in most experiments, indicating specific biomass fractions are selectively removed [31]. Tendencies followed in eucalyptus and pine wood were similar. Lignin content is increased between untreated and AH wood as a consequence of hemicelluloses extraction in the case of AH150 and AH175 samples, and as a result of hemicellulose and cellulose degradation in AH200 samples. 3.2. Eucalyptus and pine autohydrolyzed liquors Liquid fractions composition obtained after AH are shown in Table 2. Due to the presence of extractives, small amounts of phenolics, resin acids, fatty acids, juvabiones, alcohols, alkanes and esters may be present [32–34]. As a consequence of the higher liquid: solid ratio employed in AH of pine, concentrations are slightly lower than in eu- calyptus. Xylo-oligosaccharides are the OS present at higher con- centration in eucalyptus liquors, resulting from the solubilisation of xylan, the major component of the E. globulus hemicelluloses. Mannan- oligosaccharides are the OS present at higher concentration in P. radiata liquors, resulting from the solubilisation of mannan. In contrast, gluco- oligosaccharides have not been largely obtained in any of the woods although higher values are observed in pine wood due to the higher content of non-structural glucan in softwood (due to glucomannan composition) [35,36]. This statement confirms the limited removal of structural glucan by AH [31]. Eucalyptus at AH mild conditions (AH150E), yields 7.5% of total xylan recovered in form of xylo-oligo- saccharides, while pine at AH mild conditions, yields 15.8% of total mannan recovered in form of mannan-oligosaccharides. At inter- mediate AH conditions of eucalyptus wood 69.7% of total xylan is re- covered in form of xylo-oligosaccharides, while 68.8% of total mannan coming from pine wood is obtained as mannan-oligosaccharides. Severe operating conditions of eucalyptus and pine wood (AH200E and AH200P) leads to nor xylo-oligosaccharides, in the case of eucalyptus, and 2.0% of total mannan in form of mannan-oligosaccharides, in the case of pine, obtained. Monomeric sugars, mainly glucose are increased at severe operating conditions as a result of cellulose extraction in the liquid phase [18]. The main acids and furans detected were formic acid, acetic acid, furfural and hydroxymetylfurfural (HMF). As acetic acid depends on the acetylation level of hemicellulose and treatment severity (mainly tem- perature), acetic acid production of eucalyptus wood is superior at se- vere conditions. Furfural (formed via xylose/arabinose dehydration) production is also higher in eucalyptus due to the higher content of xylose and arabinose. As HMF is formed via glucose/galactose/man- nose dehydration, the high HMF production in eucalyptus may be at- tributed to gluco-oligosaccharides dehydration and the already men- tioned differences used in the liquid-to-solid-ratio. Finally, formic acid, produced via furfural and HMF degradation, is also formed in both woods at severe operating conditions [18]. From a biorefinery perspective, intermediate and severe conditions would generate value added products. Although mild conditions may elucidate behaviours in softwood similar to hardwood, the need of higher severity factors to obtain comparable results between soft and hardwood evince the more recalcitrance structure of softwood, which requires stronger pretreatment conditions [17]. Table 1 Solid recovery and compositional analysis of untreated and pretreated wood samples. ± Standard deviations values. Sample name Severity factor Solid Yield COMPOSITION (g/100 g oven dry weight pretreated biomass) Lignin Glucan Xylan Galactan Arabinan Mannan Acetate % % % % % % % % Untreated eucalyptus 0 100 26.17 ± 0.19 49.13 ± 1.27 17.45 ± 0.34 0.60 ± 0.10 0.35 ± 0.18 0.80 ± 0.40 3.96 ± 0.11 AH150E 2.95 92.95 29.05 ± 1.70 51.94 ± 1.29 16.85 ± 0.42 0.81 ± 0.15 0.20 ± 0.20 0.70 ± 0.70 5.66 ± 0.13 AH175E 3.69 74.73 37.22 ± 0.58 60.96 ± 1.75 4.77 ± 0.02 0 0 0 1.90 ± 1.59 AH200E 4.42 68.72 62.16 ± 0.75 37.45 ± 1.06 0 0 0 0 0 Untreated pine 0 100 33.00 ± 0.76 41.52 ± 0.21 4.20 ± 0.25 2.10 ± 0.30 0.82 ± 0.15 11.37 ± 0.05 2.53 ± 0.40 AH150P 2.95 91.66 34.72 ± 0.18 44.29 ± 1.09 6.78 ± 0.24 3.67 ± 0.42 0.13 ± 0.01 10.40 ± 0.11 1.61 ± 0.01 AH175P 3.99 76.80 36.45 ± 1.59 54.12 ± 0.48 2.49 ± 0.00 0.26 ± 0.02 0 4.30 ± 0.12 0.62 ± 0.14 AH200P 4.90 66.58 46.60 ± 0.43 49.72 ± 0.63 0.44 ± 0.03 0.13 ± 0.13 0 0.11 ± 0.11 0.34 ± 0.02 Untreated eucalyptus 100 27.21 ± 0.20 51.08 ± 1.32 18.14 ± 0.35 0.62 ± 0.10 0.18 ± 0.18 0.41 ± 0.41 4.12 ± 0.11 IL80E 88.46 30.54 ± 0.41 45.44 ± 0.20 13.25 ± 0.53 1.35 ± 0.22 0 3.52 ± 1.76 6.62 ± 0.04 IL120E 86.67 28.14 ± 0.89 45.89 ± 1.43 14.22 ± 0.67 0.87 ± 0.17 0 3.79 ± 1.90 4.62 ± 0.99 Untreated pine 100 34.09 ± 0.79 42.89 ± 0.22 4.34 ± 0.26 2.17 ± 0.31 0.85 ± 0.16 11.75 ± 0.05 2.61 ± 0.41 IL80P 90.88 33.16 ± 0.54 44.31 ± 0.67 5.80 ± 0.44 2.61 ± 0.43 0.94 ± 0.05 11.36 ± 0.71 1.65 ± 0.24 IL120P 89.81 31.55 ± 0.59 45.22 ± 0.12 5.52 ± 0.31 2.33 ± 0.52 1.20 ± 0.01 11.70 ± 0.64 1.15 ± 0.00 V. Rigual et al. Biomass and Bioenergy 117 (2018) 190–197 192 3.3. Microwave IL pretreatment of eucalyptus and pine wood Solid recovery values after microwave ionic liquid treatment are from 86 to 90%, as shown in Table 1. Chemical composition of the pretreated solid is partially altered with the IL microwave treatment. Lignin content slightly decreases with the increase of the temperature. This happens because of biomass dissolution without degrading com- ponents and a subsequent precipitation step without fractional se- paration. Additionally, while the glucan present in untreated pine al- most completely remain in the pretreated solid (around 94%) there is a considerable decrease (around 22%) of the glucan present in eu- calyptus. Partial cellulose degradation has been produced in hardwood (supported by the detection of sugars in the IL stream by HPLC). The partial degradation of the cellulose may be attributed to the effect of the IL itself, or the employment of microwave heating that accelerate process but also may produce partial degradation under harsher con- ditions [15]. 3.3.1. Liquid by-stream of microwave IL pretreatment Washing fractions analysis showed that at least 3 washing steps are needed to recover the IL impregnated in the solid. Most of the IL (62–85 g of IL/100 g of initial IL introduced) was recovered in the re- generation step. No IL was detected in the 4th and 5th wash. That, would sum 13mL of water needed per g of IL used in the process. At least, 96% of the initial IL introduced in the process was located in the washing steps, which is in accordance with other authors results in si- milar processes and our previous work [36,37]. The recovered IL after the vacuum distillation step was 91–95%. The pH of the diluted IL was in the range 7.38–7.69, and the lowest pH corresponded with the lowest solid yield obtained (86.67%). The amount of lignin present in the recovered IL (Figure S1) is noticeable compared to the initial lignin introduced in the process (1.20–10.81 g of lignin per 100 g of lignin introduced). The effect of microwave heating, has been found to enhance lignin extraction in other materials such as rice straw [38]. Eucalyptus lignin content in the IL strongly increases with the increase of the temperature in the IL microwave pretreatment. However, lignin content in the recovered IL in pine remains almost unaltered. Differences between both woods may indicate that [Emim] [OAc] delignification capacity depends on the kind of lignin to extract. 3.3.2. Enzymatic digestibility and pretreated solids morphology Glucan digestibilities during enzymatic hydrolysis of samples sub- jected to AH or IL treatment is shown in Fig. 1. Untreated eucalyptus wood obtains the lowest glucan digestibility (15 g hydrolysed glucan/100 g glucan introduced). There is a gradual increment for AH150E, AH175E and AH200E, justified by the physical disruption of lignocellulose and reduction in hemicellulose content [5,39]. Tendencies in autohydrolyzed pine wood are the opposite to eu- calyptus, resulting in a decrease of glucan digestibility when the auto- hydrolysis severity factor increases. Softwood is commonly known as a “worst-case scenario” for biomass pretreatment. In this case, auto- hydrolysis worsen glucan digestibility in comparison to untreated pine wood. Partial solubilisation and depolymerisation of lignin and its re- condensation and deposition on the external surface is one of the main causes observed for enzymes inhibition after autohydrolysis treatments [18]. Evaluating the effect of temperature variations on IL microwave pretreatments, glucan digestibility of eucalyptus pretreated wood is enhanced to 55 g/100 g and 68 g/100 g with the addition of the IL pretreatment for experiments IL80E and IL120E, respectively. In the case of P. radiata wood, the increase of temperature does not cause a relevant increase in glucan digestibility that is only enhanced to 72 g/100 g and 78 g/100 g for experiments IL80P and IL120P, respec- tively. These results are attributed to the decrease of DP of cellulose, due to the effect of IL and microwave heating. Delignification of soft- wood has not been as high as in hardwood (when the IL pretreatment temperature increases), so the effect of the IL has mainly affected cel- lulose dissolution. The highest digestibility is obtained in IL120P sample. These results contrast to the observed by Li et al., where Table 2 Composition of the hydrolysates obtained from autohydrolysis. V. Rigual et al. Biomass and Bioenergy 117 (2018) 190–197 193 Fig. 1. Glucan digestibility in a) untreated eucalyptus and eucalyptus AH samples; b) untreated pine and pine AH samples; c) untreated eucalyptus and eucalyptus IL microwave pretreated samples; d) untreated pine and pines IL microwave pretreated samples. Bars denote standard deviations. Fig. 2. SEM micrographs of a) Untreated eucalyptus, b) AH175E, c) AH200E and d) IL120E. V. Rigual et al. Biomass and Bioenergy 117 (2018) 190–197 194 Fig. 3. Confocal fluorescence microscopy images of a) Untreated eucalyptus, b) AH175E, c) AH200E and d) IL120E. Fig. 4. SEM micrographs of a) Untreated pine, b) AH175P, c) AH200P and d) IL120P. V. Rigual et al. Biomass and Bioenergy 117 (2018) 190–197 195 eucalyptus yielded higher (93%) digestibility than pine (62%) (in that case using conventional heating) [17]. In a global comparison, eu- calyptus highest digestibility is obtained at severe AH conditions, while the pine highest digestibility is obtained with IL120P treatment. Xylan digestibility was also quantified in the case of eucalyptus (Figure S2 from the supplementary material), and the highest values were for IL80E and IL120E, respectively. In the case of pine wood, xylose and mannose were not obtained. Figs. 2–5 show the most representatives SEM and confocal micro- scope samples. The overall surface of untreated eucalyptus (Fig. 2a) and pine wood (Fig. 4a) was smoothed with no visible pores in the structure. Hemicelluloses (in blue) dyed with calcofluor are observed (Figs. 3a and 5a) in the surface as well as lignin (in green). After intermediate autohydrolysis conditions (AH175E and AH175P), eucalyptus suffers an alteration of its native structures and hemicellulose is removed, re- maining lignin in the surface. In the case of pine wood, morphology remains unaltered (Fig. 4b) and the decrease of hemicellulose compo- sition is not clearly observed on the surface structure (Fig. 5b). At se- vere operating conditions (AH200P and AH200E), eucalyptus solids are observed as an agglomeration of smaller particles (Fig. 2c). The surface is observed to be rough and large number of pores are visualized. Due to the opacity, AH200 samples are more difficult to be observed in the confocal microscope (Fig. 3c). However, autofluorescent lignin (green) is barely observed and cellulose is only appreciated. This behaviour contrasts with the pine wood structures observed. In AH200P, cellulose is not visible and almost the complete surface of the particle is com- pound by lignin (Fig. 5c). The already mentioned lignin recondensation of softwood is in this way proved, probably causing enzymatic sac- charification low yields [18]. IL microwave treated samples (Figs. 2d and 4d), in both cases, have a relatively homogeneous macrostructure with more porosity, due to partial dissolution and regeneration of wood fibres in a fusing structure, where structural modification of lignin and hollocellulose have been produced (Figs. 3d and 5d) [3,40]. 3.4. Process mass balance Figure S3 and Table S1 from the supplementary material sum- marizes the calculations developed in the whole process. The highest glucose yields are obtained with microwave ionic liquid pretreatments. It is highlighted that under severe AH conditions (specially AH200E), the partial degradation of wood results in strong decrease of glucose yields. Although glucan digestibility is in this case the highest of eu- calyptus samples, the final glucose yield is lower (20.8 g of glucose/ 100 g of untreated eucalyptus wood) than in AH175E sample (27 g of glucose/100 g of untreated eucalyptus wood). The highest eucalyptus glucose yield is obtained in the sample IL120E. Despite of the recalci- trance of softwood and low cellulose content of the raw material in relation to hardwood, the highest glucose production is obtained with pine wood (IL120P) with 35.4 g of glucose/100 g of initial pine wood. From a biorefinery perspective, autohydrolysis pretreatment is performed at higher particle sizes (experiments run at particle sizes between 0.3 and 2mm); while the high viscosity of ILs forces this pretreatment efficacy to the operation at lower particle sizes (dust). The milling energy requirements before the pretreatment should be con- sidered to determine the most appropriate pretreatment. Additionally, IL microwave pretreatment will require an extra pre-conditioning step to remove extractives. According to the quantity of solvent employed, consumption is considerably higher in IL pretreatments than the water employed in autohydrolysis due to the low solid:liquid ratios employed in IL pretreatments. Furthermore, water washing volumes are also much higher, as well as water washing employed. Note that overall mass balance has evinced that lignin has been overestimated in the case of experiments developed with auto- hydrolized at 200 °C eucalyptus wood, probably due to “pseudo-lignin”. Thus, by-products such as lignin-furfural that cannot be distinguished from lignin increases the lignin value in the mass balance [41]. Fig. 5. Confocal fluorescence microscopy images of a) Untreated pine, b) AH175P, c) AH200P and d) IL120P. V. Rigual et al. Biomass and Bioenergy 117 (2018) 190–197 196 4. Conclusions Microwave ionic liquid and autohydrolysis pretreatments of hard- wood (E. globulus) vs softwood (P. radiata) have been studied. Autohydrolysis enhanced enzymatic digestibility of eucalyptus, while pine digestibility decreased due to lignin deposition in the external surface as observed in confocal fluorescence microscopy images. IL microwave pretreatment slightly affected biomass composition, al- though lignin extraction was considerably higher in eucalyptus (up to 11 g of lignin/100 g of initial lignin) than in softwood (less than 2 g/ 100 g). The high effectivity of IL microwave pretreatment in both bio- masses was proved. The high effectivity of IL microwave pretreatment in softwood (digestibility of 78%, 35.4 g of glucose/100 g of pine) over hardwood become IL in an alternative technology for this kind of re- calcitrant biomasses. Acknowledgements This work was performed thanks to the financial support of the “Ministerio de Economía y Competitividad” under the funded project CTQ2013-42006-R and the contract BES-2014-067788. Appendix A. Supplementary data Supplementary data related to this article can be found at https:// doi.org/10.1016/j.biombioe.2018.07.014. References [1] Z. Liu, X. Sun, M. Hao, C. Huang, Z. Xue, T. Mu, Preparation and characterization of regenerated cellulose from ionic liquid using different methods, Carbohydr. Polym. 117 (2015) 99–105. [2] A.A.N. Gunny, D. Arbain, E.M. Nashef, P. Jamal, Applicability evaluation of Deep Eutectic Solvents-Cellulase system for lignocellulose hydrolysis, Bioresour. Technol. 181 (2015) 297–302. [3] J.A. Pérez-Pimienta, A. Vargas-Tah, K.M. López-Ortega, Y.N. Medina-López, J.A. Mendoza-Pérez, S. Avila, S. Singh, B.A. Simmons, I. Loaces, A. Martinez, Sequential enzymatic saccharification and fermentation of ionic liquid and orga- nosolv pretreated agave bagasse for ethanol production, Bioresour. Technol. 225 (2017) 191–198. [4] S. Raghavi, R. Sindhu, P. Binod, E. Gnansounou, A. Pandey, Development of a novel sequential pretreatment strategy for the production of bioethanol from sugarcane trash, Bioresour. Technol. 199 (2016) 202–210. [5] T. Silva-Fernandes, L.C. Duarte, F. Carvalheiro, S. Marques, M.C. Loureiro-Dias, C. Fonseca, F. Gírio, Biorefining strategy for maximal monosaccharide recovery from three different feedstocks: Eucalyptus residues, wheat straw and olive tree pruning, Bioresour. Technol. 183 (2015) 203–212. [6] J. Liu, M. Li, X. Luo, L. Chen, L. Huang, Effect of hot-water extraction (HWE) se- verity on bleached pulp based biorefinery performance of eucalyptus during the HWE–Kraft–ECF bleaching process, Bioresour. Technol. 181 (2015) 183–190. [7] A.F.A. Carvalho, P.d.O. Neto, D.F. da Silva, G.M. Pastore, Xylo-oligosaccharides from lignocellulosic materials: chemical structure, health benefits and production by chemical and enzymatic hydrolysis, Food Res. Int. 51 (1) (2013) 75–85. [8] J.G. Lynam, C.J. Coronella, Glycerol as an ionic liquid co-solvent for pretreatment of rice hulls to enhance glucose and xylose yield, Bioresour. Technol. 166 (2014) 471–478. [9] C. Zhu, R.M. Richardson, K.D. Potter, A.F. Koutsomitopoulou, J.S. Van Duijneveldt, S.R. Vincent, N.D. Wanasekara, S.J. Eichhorn, S.S. Rahatekar, High modulus re- generated cellulose fibers spun from a low molecular weight microcrystalline cel- lulose solution, ACS Sustain. Chem. Eng. 4 (9) (2016) 4545–4553. [10] J. Xu, P. Xiong, B. He, Advances in improving the performance of cellulase in ionic liquids for lignocellulose biorefinery, Bioresour. Technol. 200 (2016) 961–970. [11] R. Singh, B.B. Krishna, J. Kumar, T. Bhaskar, Opportunities for utilization of non- conventional energy sources for biomass pretreatment, Bioresour. Technol. 199 (2016) 398–407. [12] A.A. Elgharbawy, M.Z. Alam, M. Moniruzzaman, M. Goto, Ionic liquid pretreatment as emerging approaches for enhanced enzymatic hydrolysis of lignocellulosic bio- mass, Biochem. Eng. J. 109 (2016) 252–267. [13] Z. Fang, Production of Biofuels and Chemicals with Microwave, Springer, 2015. [14] K.C. Badgujar, B.M. Bhanage, Factors governing dissolution process of lig- nocellulosic biomass in ionic liquid: current status, overview and challenges, Bioresour. Technol. 178 (0) (2015) 2–18. [15] A. Brandt, J. Grasvik, J.P. Hallett, T. Welton, Deconstruction of lignocellulosic biomass with ionic liquids, Green Chem. 15 (3) (2013) 550–583. [16] C. Asada, C. Sasaki, T. Hirano, Y. Nakamura, Chemical characteristics and enzy- matic saccharification of lignocellulosic biomass treated using high-temperature saturated steam: comparison of softwood and hardwood, Bioresour. Technol. 182 (2015) 245–250. [17] C. Li, L. Sun, B.A. Simmons, S. Singh, Comparing the recalcitrance of Eucalyptus, pine, and switchgrass using ionic liquid and dilute acid pretreatments, Bioenerg. Res. 6 (1) (2013) 14–23. [18] C.K. Nitsos, T. Choli-Papadopoulou, K.A. Matis, K.S. Triantafyllidis, Optimization of hydrothermal pretreatment of hardwood and softwood lignocellulosic residues for selective hemicellulose recovery and improved cellulose enzymatic hydrolysis, ACS Sustain. Chem. Eng. 4 (9) (2016) 4529–4544. [19] L. Zhang, Y. Pu, J.R. Cort, A.J. Ragauskas, B. Yang, Revealing the molecular structural transformation of hardwood and softwood in dilute acid flowthrough pretreatment, ACS Sustain. Chem. Eng. 4 (12) (2016) 6618–6628. [20] M.T. García-Domínguez, J.C. García-Domínguez, M.J. Feria, D.M. Gómez-Lozano, F. López, M.J. Díaz, Furfural production from Eucalyptus globulus: optimizing by using neural fuzzy models, Chem. Eng. J. 221 (2013) 185–192. [21] A. Romaní, H.A. Ruiz, F.B. Pereira, J.A. Teixeira, L. Domingues, Integrated ap- proach for effective bioethanol production using whole slurry from autohydrolyzed Eucalyptus globulus wood at high-solid loadings, Fuel 135 (2014) 482–491. [22] R.P. Overend, E. Chornet, J.A. Gascoigne, Fractionation of lignocellulosics by steam-aqueous pretreatments [and discussion], Philos. Trans. R. Soc. Lond. Ser. A- Math. Phys. Eng. Sci. 321 (1561) (1987) 523. [23] A. Casas, M. Oliet, M.V. Alonso, T.M. Santos, F. Rodriguez, Dissolution of Pinus radiata and Eucalyptus globulus woods in 1-Allyl-3-methylimidazolium chloride for cellulose or lignin regeneration, Ind. Eng. Chem. Res. 52 (10) (2013) 3628–3636. [24] A. Kostamo, J.V.K. Kukkonen, Removal of resin acids and sterols from pulp mill effluents by activated sludge treatment, Water Res. 37 (12) (2003) 2813–2820. [25] A. Casas, M.V. Alonso, M. Oliet, T.M. Santos, F. Rodriguez, Characterization of cellulose regenerated from solutions of pine and eucalyptus woods in 1-allyl-3- methilimidazolium chloride, Carbohydr. Polym. 92 (2) (2013) 1946–1952. [26] B.H.A. Sluiter, R. Ruiz, C. Scarlata, J. Sluitr, D. Templeton, D. Crocker, Determination of Structural Carbohydrates and Lignin in Biomass, NREL, 2011. [27] A.B. Ibáñez, S. Bauer, Downscaled method using glass microfiber filters for the determination of Klason lignin and structural carbohydrates, Biomass Bioenergy 68 (2014) 75–81. [28] M. Resch, J. Baker, S. Decker, Low Solids Enzymatic Saccharification of Lignocellulosic Biomass, NREL Laboratory Analytical Procedure, 2015. [29] A. Procentese, E. Johnson, V. Orr, A. Garruto Campanile, J.A. Wood, A. Marzocchella, L. Rehmann, Deep eutectic solvent pretreatment and subsequent saccharification of corncob, Bioresour. Technol. 192 (2015) 31–36. [30] M. Ochoa-Villarreal, E. Aispuro-Hernández, I. Vargas-Arispuro, M.Á. Martínez- Téllez, Plant Cell wall Polymers: Function, Structure and Biological Activity of Their Derivatives, Polymerization, InTech, 2012. [31] S.Y. Yildiz, E.T. Oner, Mannan as a promising bioactive material for drug nano- carrier systems, in: A.D. Sezer (Ed.), Application of Nanotechnology in Drug Delivery, InTech, Rijeka, 2014p. Ch. 09. [32] M.J. González-Muñoz, R. Alvarez, V. Santos, J.C. Parajó, Production of hemi- cellulosic sugars from Pinus pinaster wood by sequential steps of aqueous extraction and acid hydrolysis, Wood Sci. Technol. 46 (1) (2012) 271–285. [33] J. Rangel, M. Hornus, F.E. Felissia, M.C. Area, Hydrothermal treatment of eu- calyptus sawdust for a forest biorefinery, Cellul. Chem. Technol. 50 (5–6) (2016) 521–528. [34] E. Conde, W. Fang, J. Hemming, S. Willför, A. Moure, H. Domínguez, J.C. Parajó, Water-soluble components of Pinus pinaster wood, BioResources 8 (2) (2013) 2047–2063. [35] A. Romaní, G. Garrote, F. López, J.C. Parajó, Eucalyptus globulus wood fractionation by autohydrolysis and organosolv delignification, Bioresour. Technol. 102 (10) (2011) 5896–5904. [36] P. Lozano, B. Bernal, I. Recio, M.-P. Belleville, A cyclic process for full enzymatic saccharification of pretreated cellulose with full recovery and reuse of the ionic liquid 1-butyl-3-methylimidazolium chloride, Green Chem. 14 (9) (2012) 2631–2637. [37] V. Rigual, T.M. Santos, J.C. Domínguez, M.V. Alonso, M. Oliet, F. Rodriguez, Recovery and reuse of 1-Allyl-3-methylimidazolium chloride in the fractionation of Pinus radiata wood, ACS Sustain. Chem. Eng. 5 (3) (2017) 2384–2392. [38] S.H.H. Ngoc Lan Maia, Yoon-Mo Koo, Efficient pretreatment of lignocellulose in ionic liquids/co-solvent forenzymatic hydrolysis enhancement into fermentable sugars, Process Biochem. 49 (2014) 1144–1151. [39] P. Alvira, E. Tomás-Pejó, M. Ballesteros, M.J. Negro, Pretreatment technologies for an efficient bioethanol production process based on enzymatic hydrolysis: a review, Bioresour. Technol 101 (13) (2010) 4851–4861. [40] N. Sun, H. Rodríguez, M. Rahman, R.D. Rogers, Where are ionic liquid strategies most suited in the pursuit of chemicals and energy from lignocellulosic biomass? Chem. Commun. 47 (5) (2011) 1405–1421. [41] W. Huijgen, A. Smit, P. De Wild, H. Den Uil, Fractionation of wheat straw by pre- hydrolysis, organosolv delignification and enzymatic hydrolysis for production of sugars and lignin, Bioresour. Technol 114 (2012) 389–398. V. Rigual et al. Biomass and Bioenergy 117 (2018) 190–197 197 https://doi.org/10.1016/j.biombioe.2018.07.014 https://doi.org/10.1016/j.biombioe.2018.07.014 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref1 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref1 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref1 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref2 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref2 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref2 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref3 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref3 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref3 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref3 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref3 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref4 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref4 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref4 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref5 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref5 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref5 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref5 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref6 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref6 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref6 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref7 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref7 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref7 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref8 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref8 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref8 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref9 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref9 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref9 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref9 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref10 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref10 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref11 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref11 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref11 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref12 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref12 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref12 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref13 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref14 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref14 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref14 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref15 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref15 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref16 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref16 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref16 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref16 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref17 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref17 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref17 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref18 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref18 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref18 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref18 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref19 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref19 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref19 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref20 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref20 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref20 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref21 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref21 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref21 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref22 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref22 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref22 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref23 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref23 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref23 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref24 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref24 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref25 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref25 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref25 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref26 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref26 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref27 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref27 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref27 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref28 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref28 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref29 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref29 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref29 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref30 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref30 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref30 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref31 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref31 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref31 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref32 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref32 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref32 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref33 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref33 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref33 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref34 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref34 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref34 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref35 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref35 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref35 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref36 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref36 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref36 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref36 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref37 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref37 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref37 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref38 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref38 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref38 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref39 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref39 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref39 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref40 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref40 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref40 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref41 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref41 http://refhub.elsevier.com/S0961-9534(18)30183-1/sref41 S1 Supplementary data Title: Evaluation of hardwood and softwood fractionation using autohydrolysis and ionic liquid microwave pretreatments Authors: Victoria Rigual, Tamara M. Santos, Juan C. Domínguez, M. Virginia Alonso, Mercedes Oliet and Francisco Rodriguez Affiliation and address: Department of Chemical Engineering, Faculty of Chemistry, Complutense University of Madrid, Avda. Complutense s/n, 28040 Madrid, Spain. Corresponding author: Victoria Rigual E-mail: vicrigua@ucm.es ; Tel: +34913948505 mailto:vicrigua@ucm.es S2 Figure S1: Lignin content in the recovered IL IL80E IL120E IL80P IL120P 0.00 0.01 0.02 0.03 0.04 0.05 0.06 wt% (g lignin/100g IL) w t% ( g l ig n in /1 0 0 g I L ) 0 2 4 6 8 10 12 Accumulated lignin (g lignin IL/100g lignin introduced) A c c u m u la te d l ig n in (g l ig n in I L /1 0 0 g l ig n in i n tr o d u c e d ) S3 Figure S2: Xylan digestibility in a) untreated eucalyptus and AH eucalyptus samples, b) untreated eucalyptus and IL microwave pretreated eucalyptus sample a) b) S4 a) Wood H2O Accel. 1500® AH0 AH1 AH3 AH2 AH4 AH5 Wood H2O Accel. 1500® IL0 IL1 IL3 IL2 IL4 IL5 IL b) Wood H2O Accel. 1500® AH0 AH1 AH3 AH2 AH4 AH5 Wood H2O Accel. 1500® IL0 IL1 IL3 IL2 IL4 IL5 IL Figure S3: Scheme of the process; a) IL b) autohydrolysis S5 T a b le S 1 : M as s b al an ce s IL 8 0 P IL 0 IL 1 IL 2 IL 3 IL 4 IL 5 In it ia l H 2O w a sh in g IL li q u id b y- st re a m IL p re tr ea te d so lid En zy m a ti c h yd ro ly za te En zy m a ti c h yd ro ly ze d so lid G lu ca n ( g ) 4 2 .8 9 0 .0 0 - 4 0 .2 7 0 .0 0 1 1 .2 1 X yl a n +g a la ct a n +a ra b in a n +m a n n a n +a ce ta te ( g ) 2 1 .7 2 0 .0 0 - 2 0 .3 3 0 .0 0 2 0 .3 3 Li g n in ( g ) 3 4 .0 9 0 .0 0 0 .4 1 3 0 .1 3 0 .0 0 3 0 .1 3 G lu co se ( g ) 0 .0 0 0 .0 0 - 0 .0 0 3 2 .2 9 0 .0 0 X yl o se +g a la ct o se +a ra b in o se +m a n n o se ( g ) 0 .0 0 0 .0 0 - 0 .0 0 0 .0 0 0 .0 0 O lig o sa cc h a ri d es ( g ) 0 .0 0 0 .0 0 - 0 .0 0 0 .0 0 0 .0 0 A cc el le ra se 1 50 0® ( m L) 0 .0 0 0 .0 0 0 .0 0 0 .0 0 1 0 .0 7 0 .0 0 W a te r (L ) 0 .0 0 3 2 .5 0 3 2 .5 0 0 .0 0 1 0 .0 0 0 .0 0 Em im O A c (g ) 2 5 0 0 .0 0 0 .0 0 2 1 3 8 .1 1 0 .0 0 0 .0 0 0 .0 0 IL 1 2 0 P IL 0 IL 1 IL 2 IL 3 IL 4 IL 5 In it ia l H 2O w a sh in g IL li q u id b y- st re a m IL p re tr ea te d so lid En zy m a ti c h yd ro ly za te En zy m a ti c h yd ro ly ze d so lid G lu ca n ( g ) 4 2 .8 9 0 .0 0 - 4 0 .6 1 0 .0 0 8 .7 8 X yl a n +g a la ct a n +a ra b in a n +m a n n a n +a ce ta te ( g ) 2 1 .7 2 0 .0 0 - 1 9 .6 7 0 .0 0 1 9 .6 7 Li g n in ( g ) 3 4 .0 9 0 .0 0 0 .4 3 2 8 .3 3 0 .0 0 2 8 .3 3 G lu co se ( g ) 0 .0 0 0 .0 0 - 0 .0 0 3 5 .3 6 0 .0 0 X yl o se +g a la ct o se +a ra b in o se +m a n n o se ( g ) 0 .0 0 0 .0 0 - 0 .0 0 0 .0 0 0 .0 0 O lig o sa cc h a ri d es ( g ) 0 .0 0 0 .0 0 - 0 .0 0 0 .0 0 0 .0 0 A cc el le ra se 1 50 0® ( m L) 0 .0 0 0 .0 0 0 .0 0 0 .0 0 1 0 .1 5 0 .0 0 W a te r (L ) 0 .0 0 3 2 .5 0 3 2 .5 0 0 .0 0 1 0 .0 0 0 .0 0 Em im O A c (g ) 2 5 0 0 .0 0 0 .0 0 2 0 8 4 .8 7 0 .0 0 0 .0 0 0 .0 0 S6 IL8 0 E IL0 IL1 IL2 IL3 IL4 IL5 In itia l H 2 O w a sh in g IL liq u id b y-strea m IL p retrea ted so lid En zym a tic h yd ro lyza te En zym a tic h yd ro lyzed so lid G lu ca n (g ) 5 1 .08 0 .0 0 - 4 0 .19 0 .0 0 1 8 .01 X yla n +g a la cta n +a ra b in a n +m a n n a n +a ceta te (g ) 2 4 .07 0 .0 0 - 2 1 .88 0 .0 0 8 .7 1 Lig n in (g ) 2 7 .21 0 .0 0 0 .6 4 2 7 .01 0 .0 0 2 7 .01 G lu co se (g ) 0 .0 0 0 .0 0 - 0 .0 0 2 4 .64 0 .0 0 X ylo se+g a la cto se+a ra b in o se+m a n n o se (g ) 0 .0 0 0 .0 0 - 0 .0 0 3 .4 2 0 .0 0 O lig o sa cch a rid es (g ) 0 .0 0 0 .0 0 - 0 .0 0 0 .0 0 0 .0 0 A ccellera se 1 5 0 0® (m L) 0 .0 0 0 .0 0 0 .0 0 0 .0 0 1 0 .05 0 .0 0 W a ter (L) 0 .0 0 3 2 .50 3 2 .50 0 .0 0 1 0 .00 0 .0 0 Em im O A c (g ) 2 5 0 0 .0 0 0 .0 0 2 2 7 7 .2 0 0 .0 0 0 .0 0 0 .0 0 IL1 2 0 E IL0 IL1 IL2 IL3 IL4 IL5 In itia l H 2 O w a sh in g IL liq u id b y-strea m IL p retrea ted so lid En zym a tic h yd ro lyza te En zym a tic h yd ro lyzed so lid G lu ca n (g ) 5 1 .08 0 .0 0 - 3 9 .78 0 .0 0 1 2 .87 X yla n +g a la cta n +a ra b in a n +m a n n a n +a ceta te (g ) 2 4 .07 0 .0 0 - 2 0 .38 0 .0 0 7 .7 5 Lig n in (g ) 2 7 .21 0 .0 0 2 .9 4 2 4 .39 0 .0 0 2 4 .39 G lu co se (g ) 0 .0 0 0 .0 0 - 0 .0 0 2 9 .89 0 .0 0 X ylo se+g a la cto se+a ra b in o se+m a n n o se (g ) 0 .0 0 0 .0 0 - 0 .0 0 5 .2 0 0 .0 0 O lig o sa cch a rid es (g ) 0 .0 0 0 .0 0 - 0 .0 0 0 .0 0 0 .0 0 A ccellera se 1 5 0 0® (m L) 0 .0 0 0 .0 0 0 .0 0 0 .0 0 9 .9 4 0 .0 0 W a ter (L) 0 .0 0 3 2 .50 3 2 .50 0 .0 0 1 0 .00 0 .0 0 Em im O A c (g ) 2 5 0 0 .0 0 0 .0 0 2 3 5 6 .0 6 0 .0 0 0 .0 0 0 .0 0 S7 A H 1 5 0 P A H 0 A H 1 A H 2 A H 3 A H 4 A H 5 In it ia l H 2O w a sh in g IL li q u id b y- st re a m IL p re tr ea te d so lid En zy m a ti c h yd ro ly za te En zy m a ti c h yd ro ly ze d so lid G lu ca n ( g ) 4 1 .5 2 0 .0 0 0 .0 0 4 0 .6 0 0 .0 0 2 8 .5 8 X yl a n +g a la ct a n +a ra b in a n +m a n n a n +a ce ta te ( g ) 2 1 .0 2 0 .0 0 0 .0 0 2 0 .7 1 0 .0 0 2 0 .7 1 Li g n in ( g ) 3 3 .0 0 0 .0 0 - 3 1 .8 2 0 .0 0 3 1 .8 2 G lu co se ( g ) 0 .0 0 0 .0 0 0 .0 0 0 .0 0 1 3 .3 6 0 .0 0 X yl o se +g a la ct o se +a ra b in o se +m a n n o se ( g ) 0 .0 0 0 .0 0 0 .9 0 0 .0 0 0 .0 0 0 .0 0 O lig o sa cc h a ri d es ( g ) 0 .0 0 0 .0 0 4 .2 0 0 .0 0 0 .0 0 0 .0 0 A cc el le ra se 1 50 0® ( m L) 0 .0 0 0 .0 0 0 .0 0 0 .0 0 1 0 .1 5 0 .0 0 W a te r (L ) 1 .0 0 2 .0 0 3 .0 0 0 .0 0 1 0 .0 0 0 .0 0 A H 1 7 5 P A H 0 A H 1 A H 2 A H 3 A H 4 A H 5 In it ia l H 2O w a sh in g IL li q u id b y- st re a m IL p re tr ea te d so lid En zy m a ti c h yd ro ly za te En zy m a ti c h yd ro ly ze d so lid G lu ca n ( g ) 4 1 .5 2 0 .0 0 0 .0 0 4 1 .5 7 0 .0 0 3 3 .7 9 X yl a n +g a la ct a n +a ra b in a n +m a n n a n +a ce ta te ( g ) 2 1 .0 2 0 .0 0 0 .0 0 5 .8 9 0 .0 0 5 .8 9 Li g n in ( g ) 3 3 .0 0 0 .0 0 - 3 5 .7 9 0 .0 0 3 5 .7 9 G lu co se ( g ) 0 .0 0 0 .0 0 0 .0 8 0 .0 0 8 .6 4 0 .0 0 X yl o se +g a la ct o se +a ra b in o se +m a n n o se ( g ) 0 .0 0 0 .0 0 3 .8 8 0 .0 0 0 .0 0 0 .0 0 O lig o sa cc h a ri d es ( g ) 0 .0 0 0 .0 0 1 6 .0 4 0 .0 0 0 .0 0 0 .0 0 A cc el le ra se 1 50 0® ( m L) 0 .0 0 0 .0 0 0 .0 0 0 .0 0 1 0 .3 9 0 .0 0 W a te r (L ) 1 .0 0 2 .0 0 3 .0 0 0 .0 0 1 0 .0 0 0 .0 0 S8 A H 2 0 0 P A H 0 A H 1 A H 2 A H 3 A H 4 A H 5 In itia l H 2 O w a sh in g IL liq u id b y-strea m IL p retrea ted so lid En zym a tic h yd ro lyza te En zym a tic h yd ro lyzed so lid G lu ca n (g ) 4 1 .52 0 .0 0 0 .0 0 3 3 .10 0 .0 0 2 9 .65 X yla n +g a la cta n +a ra b in a n +m a n n a n +a ceta te (g ) 2 1 .02 0 .0 0 0 .0 0 0 .6 7 0 .0 0 0 .6 7 Lig n in (g ) 3 3 .00 0 .0 0 - 3 1 .02 0 .0 0 3 1 .02 G lu co se (g ) 0 .0 0 0 .0 0 1 .6 1 0 .0 0 3 .8 3 0 .0 0 X ylo se+g a la cto se+a ra b in o se+m a n n o se (g ) 0 .0 0 0 .0 0 2 .2 0 0 .0 0 0 .0 0 0 .0 0 O lig o sa cch a rid es (g ) 0 .0 0 0 .0 0 0 .9 0 0 .0 0 0 .0 0 0 .0 0 A ccellera se 1 5 0 0® (m L) 0 .0 0 0 .0 0 0 .0 0 0 .0 0 8 .2 8 0 .0 0 W a ter (L) 1 .0 0 2 .0 0 3 .0 0 0 .0 0 1 0 .00 0 .0 0 A H 1 5 0 E A H 0 A H 1 A H 2 A H 3 A H 4 A H 5 In itia l H 2 O w a sh in g IL liq u id b y-strea m IL p retrea ted so lid En zym a tic h yd ro lyza te En zym a tic h yd ro lyzed so lid G lu ca n (g ) 4 9 .13 0 .0 0 0 .0 0 4 8 .28 0 .0 0 3 3 .33 X yla n +g a la cta n +a ra b in a n +m a n n a n +a ceta te (g ) 2 3 .16 0 .0 0 0 .0 0 2 2 .51 0 .0 0 2 2 .51 Lig n in (g ) 2 6 .17 0 .0 0 - 2 7 .00 0 .0 0 2 7 .00 G lu co se (g ) 0 .0 0 0 .0 0 0 .0 5 0 .0 0 1 6 .60 0 .0 0 X ylo se+g a la cto se+a ra b in o se+m a n n o se (g ) 0 .0 0 0 .0 0 0 .0 6 0 .0 0 3 .0 6 0 .0 0 O lig o sa cch a rid es (g ) 0 .0 0 0 .0 0 2 .1 0 0 .0 0 0 .0 0 0 .0 0 A ccellera se 1 5 0 0® (m L) 0 .0 0 0 .0 0 0 .0 0 0 .0 0 1 2 .07 0 .0 0 W a ter (L) 0 .8 0 1 .6 0 2 .4 0 0 .0 0 1 0 .00 0 .0 0 S9 A H 1 7 5 E A H 0 A H 1 A H 2 A H 3 A H 4 A H 5 In it ia l H 2O w a sh in g IL li q u id b y- st re a m IL p re tr ea te d so lid En zy m a ti c h yd ro ly za te En zy m a ti c h yd ro ly ze d so lid G lu ca n ( g ) 4 9 .1 3 0 .0 0 0 .0 0 4 5 .5 6 0 .0 0 2 1 .2 6 X yl a n +g a la ct a n +a ra b in a n +m a n n a n +a ce ta te ( g ) 2 3 .1 6 0 .0 0 0 .0 0 5 .2 5 0 .0 0 5 .2 5 Li g n in ( g ) 2 6 .1 7 0 .0 0 - 2 7 .8 2 0 .0 0 2 7 .8 2 G lu co se ( g ) 0 .0 0 0 .0 0 0 .3 4 0 .0 0 2 7 .0 0 0 .0 0 X yl o se +g a la ct o se +a ra b in o se +m a n n o se ( g ) 0 .0 0 0 .0 0 2 .1 2 0 .0 0 2 .0 7 0 .0 0 O lig o sa cc h a ri d es ( g ) 0 .0 0 0 .0 0 1 5 .1 0 0 .0 0 0 .0 0 0 .0 0 A cc el le ra se 1 50 0® ( m L) 0 .0 0 0 .0 0 0 .0 0 0 .0 0 1 1 .3 9 0 .0 0 W a te r (L ) 0 .8 0 1 .6 0 2 .4 0 0 .0 0 1 0 .0 0 0 .0 0 A H 2 0 0 E A H 0 A H 1 A H 2 A H 3 A H 4 A H 5 In it ia l H 2O w a sh in g IL li q u id b y- st re a m IL p re tr ea te d so lid En zy m a ti c h yd ro ly za te En zy m a ti c h yd ro ly ze d so lid G lu ca n ( g ) 4 9 .1 3 0 .0 0 0 .0 0 2 5 .7 3 0 .0 0 7 .0 2 X yl a n +g a la ct a n +a ra b in a n +m a n n a n +a ce ta te ( g ) 2 3 .1 6 0 .0 0 0 .0 0 0 .0 0 0 .0 0 0 .0 0 Li g n in ( g ) 2 6 .1 7 0 .0 0 - 4 2 .7 2 * 0 .0 0 4 2 .7 2 G lu co se ( g ) 0 .0 0 0 .0 0 1 .2 7 0 .0 0 2 0 .7 9 0 .0 0 X yl o se +g a la ct o se +a ra b in o se +m a n n o se ( g ) 0 .0 0 0 .0 0 0 .0 0 0 .0 0 0 .0 0 0 .0 0 O lig o sa cc h a ri d es ( g ) 0 .0 0 0 .0 0 0 .8 0 0 .0 0 0 .0 0 0 .0 0 A cc el le ra se 1 50 0® ( m L) 0 .0 0 0 .0 0 0 .0 0 0 .0 0 6 .4 3 0 .0 0 W a te r (L ) 0 .8 0 1 .6 0 2 .4 0 0 .0 0 1 0 .0 0 0 .0 0 -: n o t d et er m in ed ; * o v er es ti m at e PUBLICACIÓN II “Protic, aprotic and choline-derived ionic liquids: towards enhancing the accessibility of hardwood and softwood” Victoria Rigual, Antonio Ovejero-Pérez, Sandra Rivas, Juan C. Domínguez, M. Virginia Alonso, Mercedes Oliet y Francisco Rodriguez Enviado a ACS Sustainable Chemistry and Engineering 1 Protic, aprotic and choline-derived ionic liquids: towards enhancing the accessibility of hardwood and softwood Victoria Rigual*, Antonio Ovejero-Pérez, Sandra Rivas, Juan C. Domínguez, M. Virginia Alonso, Mercedes Oliet and Francisco Rodriguez Department of Chemical Engineering, Faculty of Chemistry, Complutense University of Madrid, Avda. Complutense s/n, 28040 Madrid, Spain. *E-mail: vicrigua@ucm.es; Tel: +34913948505 Keywords: Choline ionic liquids, Ionic liquids, Protic Ionic liquids, 2D-NMR, Saccharification Synopsis: Protic, aprotic and choline derived ionic liquids can disrupt biomass structure, making cellulose more accesible to enzymes, increasing enzymatic digestibility. ABSTRACT A side-by-side comparison of softwood (pine) vs. hardwood (eucalyptus) pretreatment using 2 protic, 2 aprotic and 2 choline-derived ionic liquids is proposed. While the protic ionic liquid [OHEtAm][HCO2] lead to alkali lignin dissolution at 30 °C after 1 h, the lack of interactions with the whole cell wall limit the biomass disruption. On the contrary, the protic ionic liquid [Mim][Cl] produces a catalytic effect that extracts almost all the hemicelluloses, and partially lignin. Remarkable digestibilities are obtained with [Ch][OAc] in eucalyptus (69 %), while in pine, nor protic, neither choline-derive IL tested look like to be a real “greener” alternative to conventional ILs such as [Emim][OAc] (the highest digestibility, 84 %). Solid morphology revealed a smoothed surface in [Mim][Cl]_P sample, and confocal fluorescence microscopy enhanced to distinguish surface holocellulose and lignin, highlighting differences in the accessibility of hardwood vs. softwood in the presence of surface lignin. 2D-NMR showed that lignin obtained after saccharification of [Ch][OAc] samples were acylated, an indicative of the dissociation of the ionic liquid and its degradation. Finally, TGA and spectroscopy techniques evince the difficulties to 2 recover the ionic liquid and conclude a work that offers an outlook of strengths and weaknesses in the pairs IL-biomass studied. INTRODUCTION Lignocellulosic biomass is considered one of the main alternatives to face the gradual depletion of fossil fuels. 1 In particular, woody biomass is called to become in raw material in biorefineries. 2 Softwood and hardwood are promising raw materials, due to their availability (pine softwood is the most abundant wood in the whole Northern Hemisphere, while eucalyptus is cultivated in template regions covering 1.3 million of hectares in Europe in 2016), fast growth, and adaptability. 3-6 In order to exploit the maximum value of softwood and hardwood, the fractionation of its constituents through green and sustainable processes is essential. 1 Pretreatment comprise up to 40 % of the total cost of processing. 7 Among all the alternatives (steam explosion, autohydrolysis, alkali pretreatment, etc.), ionic liquids (ILs) are a promising and not-yet implemented pretreatment, offering the deconstruction of lignocellulosic biomass. ILs potential advantages in this field of application are their task-specific properties, wide range of polarities, and their capacity to dissolve a variety of chemicals, including polymers. 8 Imidazolium based ILs such as [Emim][OAc] or [Amim][Cl] are conventional ILs employed for biomass pretreatment that has been shown to effectively decrease the recalcitrance of softwood and hardwood, enhancing saccharification. 7, 9-10 This approach included lignocellulose deconstruction through partial cellulose dissolution, and cellulose precipitation with the aid of an antisolvent 11. Although at the beginning ILs were treated as “green solvents”, the risk assessment of these conventional ILs through effluents or soil, and the complexity for their synthesis arises concerns about their “green” character. 8 Another aspect that must be overcome is the cost of ILs. [Emim][OAc] bulk production price has been estimated in 2017 in 20-101 $·kg-1 .12 In this context, the substitution of conventional ILs by emerging alternative ILs is being studied in a wide range of biomasses. Two of the most emerging groups of ILs are protic ionic liquids (PILs) and choline derived ILs or so-called “bionic liquids” (BILs). In the biorefinery context, some PILs have been proved to fractionate biomass and extract lignin due to the acidic behavior of these solvents. 13 They have been considered “greener” than conventional ILs due to the easiness of synthesis, the availability of starting materials and the biodegradability. 14 The PIL 2-hydroxyethylammonium formate ([OHEtAm][HCO2]) is an example of Brönsted acidic IL. 14 [OHEtAm][HCO2] has been effectively proved in cotton stalk pretreatment, and is easy to synthetize by the combination of an amine and formic acid present in nature, with acidic hydrogens in cation and anion. 15 Another kind of PIL is [Mim][Cl], a Brönsted acidic ionic liquid with acidic hydrogen in the cation, formed by the proton transfer from a Brönsted acid to a Brönsted base. 14 The acidic behavior of this solvent enhanced pretreatment of yellow pine for further saccharification. 16 3 Choline derived ILs strongest advantage is its friendliness towards cellulases and yeasts. Choline lysinate ([Ch][Lys]) was the prime example of biocompatible IL, firstly proved as effective pretreatment solvent in herbaceous biomass and more recently in hardwood and softwood. 9, 17-18 Another checked (in herbaceous and softwood) example of effective choline IL is choline acetate ([Ch][OAc]) that benefits from the effectivity of acetate anion for biomass pretreatment combined with a “greener” cation. 19 In view of the powerful alternative and advantages of emerging ILs, this work aims to study mechanisms that govern pretreatment of biomass with PILs and choline derived, in comparison to conventional ILs. ILs pretreatment is carried out over a model of softwood (pine) and hardwood (eucalyptus) to offer a side-by-side comparison of different kind of ILs in different wood biomasses. The enzymatic digestibility is the mandatory aspect to overcome, but other outstanding characteristics, such as thermal degradation and recovery of the IL are also discussed. A deep study of biomass morpholology is performed to elucidate aspects that govern the pretreatment in every kind of IL. Additionally, the most promising saccharified by-stream samples are further characterized using 2D-NMR. MATERIALS AND METHODS Materials and reagents The ILs 2-hydroxyethylammonium formate ([OHEtAm][HCO2], >97 %, Iolitec GmbH), 1-methylimidazolium chloride ([Mim][Cl], 98 %, Iolitec GmbH), 1-ethyl-3- methylimidazolium acetate ([Emim][OAc], 95 %, Iolitec GmbH), 1-allyl-3- methylimidazolium chloride ([Amim][Cl], >98 %, Iolitec GmbH), choline acetate ([Ch][OAc], 98 %, Iolitec GmbH) and choline lysinate ([Ch][Lys], 95 %, Iolitec GmbH) were employed for wood dissolution. Pinus radiata (P) and Eucalyptus globulus (E) sawdust were provided by the Instituto Nacional de Investigación y Tecnología Agraria (CIFOR - INIA). Accellerase 1500® enzyme cocktail (protein content: 70 mg of protein/mL cocktail) was kindly provided by DuPont Industrial BioSciences. Biomass pretreatment Due to the high viscosity of ILs, pine and eucalyptus were milled and sieved at a particle size <150 µm. Extractives were previously removed to avoid foam formation using acetone and water. 20 Before the pretreatment, a screening of the six ILs chosen was utilized using both woods, alkali lignin, and microcrystalline cellulose. Samples were placed in crystal vials and heated progressively from 30 to 150 ℃ in a thermostatic block (Labnet Accublock). Qualitative analysis of the dissolution rate of every sample was visually/microscopically checked, using an Axio A1 optical microscope. Samples were periodically checked every 1 h and every 24 h, increasing the temperature 20 ℃ if after 24 h, complete 4 dissolution/degradation had not been observed. If dissolution was confirmed, water was added as antisolvent to observe if sample precipitation is produced. According to screening results, 4 % (w/w) of pine/eucalyptus wood and IL were mixed in capped glass pressure tubes of 25 mL and pretreated using a heating block Barnstead Steam RS9000 at 200 rpm, for 6 h, at 120 ℃. Afterwards, water was added as antisolvent up to a mixture of 40 % (w/w) of IL/deionized water, at 40 ℃ and stirred for 10 min to favor regeneration. Solid and liquid fractions were separated by vacuum filtration and washed with deionized water until the complete removal of the IL. Pretreated Biomass, by-streams and IL characterization The detailed analysis employed to characterize biomass (compositional analysis, SEM, confocal fluorescence microscopy, digestibility), by-streams (2D-NMR) and ILs (FTIR and TGA) are described in the Supporting Information document. RESULTS AND DISCUSSION In order to describe results obtained, ILs are going to be classified into three groups: PILs are [OHEtAm][HCO2] and [Mim][Cl]; conventional ILs are the aprotic ILs (AILs) [Emim][OAc] and [Amim][Cl]; and choline derived ILs are [Ch][OAc] and [Ch][Lys]. Effect of the IL on the pretreated biomass composition Screening results highlighted the following observations: alkali lignin dissolution at 30 C was achieved using [OHEtAm][HCO2] after 1h (Figure S1). The increase of temperature up to 150 ℃ resulted in some mixtures ([Mim][Cl] and [Ch][OAc] with cellulose) into the formation of bubbles and black samples, which may be caused by the degradation of biomass and/or the IL. The dissolution of pine and eucalyptus with [Ch][Lys] after 110 ℃ lead to an increase of viscosity in the mixtures. Finally, vapor pressure was experimentally confirmed at 150 ℃ in mixtures of [OHEtAm][HCO2], [Ch][OAc] and [Ch][Lys], which suggest the utilization of lower temperatures with these ILs. The recovery yield and composition of untreated/pretreated wood is shown in Table 1. In the case of PILs, solid recovery in softwood is similar to hardwood. However, solid recoveries are very different between [OHEtAm][HCO2] (83-85 %) and [Mim][Cl] (40-42 %), for softwood and hardwood respectively. Although both ILs are classified as Brönsted acidic ionic liquids, their - Kamlet-Taft parameters, as an indicative of cellulose dissolution and subsequent regeneration are very different, with values of 0.43 and 0.31 for [OHEtAm][HCO2] and [Mim][Cl], respectively. 21 According to screening, [OHEtAm][HCO2] could be a potential solvent for lignin, but the lack of interactions with the whole cell wall limit the network biomass disruption present in wood, evincing the need of interaction of both, cellulose and lignin to ensure biomass pretreatment with IL. 5 Table 1. Solid recovery and compositional analysis of: A) Untreated and pretreated pine samples and; B) Untreated and pretreated eucalyptus samples 1A). Untreated/pretreated pine composition (g/100g ODW[a] biomass) Solid yield (%) Lignin (%) Glucan (%) Xylan (%) Galactan (%) Arabinan (%) Mannan (%) Acetyl groups (%) Untreated pine 100.0 34.09±0.79 42.89±0.22 4.34±0.26 2.17±0.31 0.85±0.16 11.75±0.05 2.61±0.41 [OHEtAm][HCO2] 82.9 34.63±0.37 50.69±0.77 4.49±0.71 1.77±0.02 0.00±0.00 11.27±0.36 0.24±0.11 [Mim][Cl] 41.6 46.45±1.01 52.79±0.78 0.00±0.00 0.00±0.00 0.00±0.00 1.22±0.05 0.27±0.01 [Emim][OAc] 89.2 32.46±2.22 48.30±1.53 3.22±0.04 1.77±0.09 0.89±0.00 11.61±0.23 1.45±0.17 [Amim][Cl] 84.3 33.10±0.21 47.39±0.67 5.00±0.10 1.94±0.22 0.87±0.00 10.88±0.07 1.99±0.09 [Ch][OAc] 81.7 36.99±2.10 50.00±1.08 2.66±0.29 1.45±0.05 0.88±0.05 10.63±0.11 0.58±0.05 [Ch][Lys] 84.7 31.77±1.03 51.69±0.11 2.83±0.31 1.74±0.07 0.78±0.39 10.95±0.38 0.00±0.00 1B). Untreated/pretreated eucalyptus composition (g/100g ODW[a] biomass) Solid yield (%) Lignin (%) Glucan (%) Xylan (%) Galactan (%) Arabinan (%) Mannan (%) Acetyl groups (%) Untreated eucalyptus 100 27.21±0.20 51.08±1.32 18.14±0.35 0.62±0.10 0.18±0.18 0.41±0.41 4.12±0.11 [OHEtAm][HCO2] 85.1 30.21±0.40 49.67±0.35 13.31±0.52 0.89±0.05 0.00±0.00 0.42±0.00 1.19±0.09 [Mim][Cl] 39.8 32.48±0.13 64.19±0.71 0.19±0.10 0.19±0.09 0.00±0.00 0.00±0.00 0.86±0.02 [Emim][OAc] 75.9 26.50±0.96 56.70±1.49 11.49±0.74 0.40±0.06 0.10±0.10 0.93±0.10 1.45±0.09 [Amim][Cl] 85.7 27.16±0.66 47.53±1.00 14.18±0.25 1.05±0.14 0.24±0.01 0.88±0.04 4.93±0.29 [Ch][OAc] 75.3 29.24±0.31 55.81±0.80 7.68±0.13 0.41±0.04 0.00±0.00 0.73±0.00 0.91±0.04 [Ch][Lys] 69.5 23.70±1.86 69.25±0.44 6.39±0.10 0.49±0.03 0.27±0.13 1.11±0.05 0.00±0.00 [a] ODW: Oven dry weight. In the case of aprotic ILs (AILs), [Emim][OAc] degrades more hardwood (solid recovery of 76 %) than softwood (solid recovery of 89 %). This behavior was already observed in a previous work with microwave heating and shorter pretreatment times, although differences were not so big (in that case solid recoveries were 87 % and 90 % for eucalyptus and pine, respectively). 22 However, solid recoveries after pretreatment with [Amim][Cl] are slightly higher in hardwood (86 %) than in softwood (84 %). This may be attributed to differences in the anion reactivity to cellulose 6 dissolution, as Kamlet-Taft  value for [Emim][OAc] (1.14) is higher than for [Amim][Cl] (0.83) and cellulose content in eucalyptus is higher than in pine. 23 As occurred with [Emim][OAc], solid recoveries values are lower in hardwood than in softwood with choline ILs, evincing the high interaction with cellulose in both cases ( values of 1.22 and 1.30 for [Ch][OAc] and [Ch][Lys], respectively). Additionally, the overall differences observed between hardwood and softwood may also be caused by the kind of lignin extracted with each IL (discussed below). With respect to compositional analysis, [OHEtAm][HCO2], with hydrogens in the cation and the anion should have a breakage effect in the biomass stronger than [Mim][Cl]. However, the comparison show that [OHEtAm][HCO2] barely affects composition in softwood, and slightly in hardwood. Lignin relative content in [OHEtAm][HCO2]_P is similar to untreated pine (34 % in both cases), with an increment of glucan relative content in [OHEtAm][HCO2]_P (51 % vs. 43 % in untreated pine) due to partial hemicellulose extraction. In the case of eucalyptus, hemicellulose extraction is higher than in pine (hemicellulose relative content of 16 % in [OHEtAm][HCO2]_E with respect to 23 % in untreated eucalyptus), causing higher lignin relative content. On the contrary, [Mim][Cl] effect is completely catalytic, causing the degradation of biomass components. 16 Apart from the low solid recovery values obtained for both woods, hemicellulose extraction was almost complete in both woods (hemicellulose relative content of 1.5 % and 1.2 % for [Mim][Cl]_P and [Mim][Cl]_E, respectively). Furthermore, lignin was also extracted, more in hardwood than in softwood (lignin extraction of 43 % for [Mim][Cl]_P and 52 % for [Mim][Cl]_E). Conventional ILs have been widely reported to produce partial lignin and hemicellulose extraction without degrading cellulose. 24-25 Results of this work support the literature, as cellulose content in [Emim][OAc]_P and [Amim][Cl]_P is similar to untreated pine, while lignin is slightly lower, and also hemicellulose content. In the case of eucalyptus samples, results follow the same tendency, although xylan removal is higher. Biomass compositional analysis is different with [Ch][OAc] and [Ch][Lys]. [Ch][OAc] extracts for pine and eucalyptus 11 % and 19 % of lignin; 39 % and 69 % of hemicellulose; and 5 % and 18 % of cellulose, respectively. This demonstrated the potentiality of this IL for biomass pretreatment of hardwood and softwood towards hemicellulose extraction, keeping the majority of cellulose and lignin fractions. Other authors have already employed sugarcane bagasse, switchgrass, pine, rice straw, and in this work, the effectivity of this IL is also confirmed with eucalyptus. 19, 23, 26-27 Finally, [Ch][Lys] extracts 21.6 % and 40 % of lignin; 37 % and 76 % of hemicellulose; and non and 5 % of cellulose in pine and eucalyptus wood, respectively. Dutta et al recently observed this delignification effect in hardwood and softwood, using a mixture of IL and water for the pretreatment. In this work, the capacity of the IL itself to interact with lignin without the need of a water solution has been additionally checked. 17 Glucan digestibility of pine and eucalyptus is shown in Figure 1. In the case of pine, digestibility increases following the order: 7 [Mim][Cl]_P < Untreated pine < [OHEtAm][HCO2]_P < [Ch][Lys]_P < [Ch][OAc]_P < [Amim][Cl]_P < [Emim][OAc]_P Figure 1. Enzymatic hydrolysis of a) Pine untreated and pretreated samples and b) Eucalyptus untreated and pretreated samples. Results are expressed according to Eq. S1 In the case of eucalyptus, the digestibility order is: Untreated eucalyptus< [OHEtAm][HCO2]_E < [Mim][Cl]_E < [Amim][Cl]_E < [Ch][Lys]_E < [Ch][OAc]_E < [Emim][OAc]_E Digestibility order is different in softwood and hardwood, evincing different reactivity of each IL with the kind of wood. Despite of the high recalcitrance of softwood in relation to hardwood, it is observed that the highest digestibility is produced with the combination of conventional IL and softwood [Emim][OAc]_P, obtaining 84 % of digestibility at 72 h 22. Softwood digestibility with PILs is very low, being in the case of [OHEtAm][HCO2]_P similar to the untreated pine 72 h digestibility (26.3 and 28.5 for untreated and [OHEtAm][HCO2]_P). Furthermore, initial rates are similar, verifying that this IL is not appropriate for softwood pretreatment. The effect of [Mim][Cl] produces digestibilities even lower than in the untreated pine. Acidic ionic liquids are effectives breaking down the lignin β-O-4 linkage. Additionally, Cl- anion increase the cleavage due to the strong interaction formed with cellulose. Both effects lead to lignin breakage and possibly to the formation of inhibitors in the surface of biomass. 28,29 Choline derived ILs have digestibilities of around 54 %. In the case of AILs, results with [Amim][Cl] are lower than with [Emim][OAc], as was observed by Torr et al (2015) 24. Concerning hardwood results, [OHEtAm][HCO2] produces a slight improvement on eucalyptus digestibility, but very low in comparison with AILs. The effect of [Mim][Cl]_E yields a digestibility of 47.8 %, indicating that the negative effect produced in pine wood with this IL, is not observed in eucalyptus. [Ch][OAc]_E also yield a digestibility of 69 % only smaller than the obtained in [Emim][OAc]_E(digestibility of 80.4 % at 72 h), indicating the potential of this IL to really become in an alternative “greener” IL for hardwood pretreatment. The high digestibilities obtained with [Emim][OAc] and [Ch][OAc] indicate the 8 favorable effect of employing acetate anions. The comparison between results obtained with softwood vs. hardwood shows higher digestibilities obtained with hardwood than with softwood in all the PILs and choline ILs ([OHEtAm][HCO2], [Mim][Cl], [Ch][OAc], and [Ch][Lys]). However, AILs are more effective with softwood. These results suggest that dissolution effect is different in every IL. In PILs, the presence of acidic ions may produce a catalytic effect during the pretreatment, which in hardwood produce the disruption of the network and lignin depolymerisation through the β-O-4 ether linkage while in softwood enhances the lignin recondensation reactions, limiting enzymes accessibility. 29-30 Differences in lignin in hardwood and softwood may lead to the effective disruption of hardwood, enhancing enzyme accessibility without enzymes inactivation. 2 Evaluating glucan rates, in the case of softwood [Mim][Cl]_P reaction rate is negligible after 12 h, [Amim][Cl]_P after 24 h, and the rest after 48 h. In the case of hardwood, untreated eucalyptus, [OHEtAm][HCO2]_E, [Amim][Cl]_E, [Ch][OAc]_E and [Emim][OAc]_E are negligible after 48 h; while, [Ch][Lys]_E and mainly [Mim][Cl]_E continue reacting at 72 h, which slow down considerably the process of glucose production from a biorefinery perspective. Some SEM micrographs of the most representative samples obtained are shown in Figure S2 in the supporting information in order to support digestibility results. In the case of PILs, the morphology differs from the obtained with AILs and choline ILs. [OHEtAm][HCO2] pretreated samples offer a structure similar to the obtained in untreated pine and eucalyptus. In the case of [Mim][Cl] pretreated samples, morphology in pine and eucalyptus differs among themselves. [Mim][Cl]_P sample (Figure S2b) does not present visible holes, with bigger particle sizes than untreated pine and smoothed, which may difficult enzymes accessibility inside the material. [Mim][Cl]_E sample (Figure S2f) particle size is smaller than in case of pine and holes are not appreciated neither. In the case of AILs, surface is rough with a lot of visible holes and porosity caused by the already explained in literature dissolution and regeneration of wood fibers. 31-32 More holes and porosity were observed in [Emim][OAc] samples (Figures S2c and S2g) than with [Amim][Cl] samples. In choline ILs, the high rugosity and the presence of holes is noticeable, with more twists and turns in the material. However, particles are bigger than in the case of AILs. Characteristic morphology was observed in [Ch][OAc]_E, with high rugosity in the surface. The use of CFM showed differences in the distribution of holocellulose and lignin in particles surface. It is important to remark that although the merged channels image may reveal some visual appreciations, the splitted, corrected and thresholded images are necessary to interpret results, as colocalization occurs in most of the images (Figures S3 and S4). 33-34 Furthermore, holocellulose (in green) is due to hemicellulose and cellulose in all the cases, excepting [Mim][Cl]; where most hemicelluloses were removed according to Table 1. Comparing binary-thresholded images, it is 9 observed that there are more pixels attributed to lignin than to holocellulose in all the samples, excepting untreated pine, untreated eucalyptus, and [OHEtAm][HCO2]_E. Lignin surface ratios are presented in Figure 2a. Figure 2. a) Ratio surface lignin and enzymatic hydrolysis and; b) ratio surface lignin vs total lignin in pine (blue) and eucalyptus (red) The size of every bubble is proportional to the enzymatic hydrolysis digestibility obtained. Ratio values in pine are in the range of 0.48 for untreated pine and 0.70 for [Mim][Cl]_P. Tendency observed in pine indicates that low ratios of surface lignin favors enzyme accessibility, while high presence of surface lignin difficult enzyme accessibility. This is supported by the negative effect that softwood lignin has over cellulases, blocking the lignocellulose accessibility. 35 Untreated pine wood and [OHEtAm][HCO2]_P are not morphologically affected by the pretreatment, and they have low lignin ratios in contrast to the low digestibility obtained. This is attributed to the lack of the alteration of the cell wall, so hemicellulose is present in the external part of the cell wall, and this holocellulose is mainly attributed to hemicellulose. In rest of the samples, the deconstruction observed with SEM analysis enhances the accessibility. In the case of eucalyptus samples, higher quantities of lignin are observed on the surface of most of the samples. However, in contrast to the 10 observed in pine, the presence of higher lignin ratios on the surface, do not limit saccharification. In fact, the highest lignin content on the surface is attributed to [Emim][OAc]_E sample, with a lignin ratio of 0.89 and the highest digestibility. The opposite tendency obtained in eucalyptus samples (in comparison to the observed in pine samples) is explained by differences of hardwood and softwood lignin chemistry that in this case does not inhibit enzymatic digestibility. 36 In Figure 2b surface lignin ratio was plotted against lignin ratio (result of Table 1) to demonstrate that lignin distribution in the whole particle, and in the surface are not linearly correlated. Characterization of lignin-rich solid obtained after enzymatic saccharification The HSQC spectra (1-bond 13C-1H correlation) of the whole “post-hydrolysis” cell wall samples are shown in Figures S5 and S6 in the supporting information, with the purpose of characterizing the lignin that was not removed from the solid. Organosolv lignin (as a model), [Emim][OAc], and [Ch][OAc] “post hydrolysed” cell wall samples are depicted, as they are those with the highest digestibilities of each type of IL. PILs were not studied due to the low digestibility provided in comparison with the other kind of ILs. According to literature, pine wood lignin presents 20 % of β-O-4 linkages (of 100 total interunit linkages), while eucalyptus β-O-4 interunit linkages is of 80 %. Pine wood lignin is mainly formed by guayacil units, while eucalyptus lignin is formed by H/S/G ratio of 2/84/14. 37-38 Figure S5 frame the aromatic region of both pine and eucalyptus lignin, compound mainly by G units in the case of softwood and S and G units in hardwood. In the case of both [Ch][OAc]_P and [Ch][OAc]_E post-hydrolysed samples, the presence of acylated C2,6-H2,6 in p-coumarate (PCE) structures is observed (δC/ δH = 129.3/7.68 ppm). 39 Higher ratios of PCE2,6 were found in [Ch][OAc]_E sample than in [Ch][OAc]_P sample. The presence of acylated groups is explained in some works by the presence of acetates. 40 In this case, acetate anions are present in the ionic liquid, and the dissociation may lead to the IL degradation and biomass acylation. The S/G ratios obtained in the three eucalyptus samples were [Emim][OAc]_E>[Ch][OAc]_E>Org_E. The higher S/G ratio obtained with [Emim][OAc]_E evinces the G units removal during the pretreatment with this IL, that also favors the material digestibility. This selective G lignin extraction can be one of the reasons why [Emim][OAc]_P digestibility is higher than [Emim][OAc]_E, as pine lignin is formed by G units. Side chain regions of pine and eucalyptus saccharified cell wall samples are observed in Figure S6. The non-oxygenated aliphatic region (δC ~ 50 – 0 ppm) shows the α/β-CH2- groups, γ-CH3 in n-propyl side chains of lignin, and acetyl hemicelluloses. 41 In this case, higher amounts of non- oxygenated compounds were observed in [Ch][OAc]_P and [Ch][OAc]_E samples. Organosolv samples had lower amounts of non-oxygenated aliphatic compounds, and the non-oxygenated aliphatic compounds present in saccharified cell wall samples may be attributed mainly to lignin, but also acylated hemicelluloses. The presence of β-O-4 interunit linkages (Cβ-Hβ at δC/ δH = 11 83.4/4.27 ppm; Cγ−Hγ in γ-hydroxylated β-O-4′ substructures at δC/ δH = 59.4/3.40 and 3.72 ppm; Cγ−Hγ in γ-acylated β-O-4′ substructures at δC/ δH = 63.5/3.83 and 4.30 ppm; Cα−Hα in β-O-4′ substructures (A) linked to a G-unit at δC/ δH = 70.9/4.71 ppm; and Cα−Hα in β-O-4′ substructures (A) linked to a S-unit at δC/ δH = 71.8/4.83 ppm) are observed in all the samples. 40 However, phenylcoumaran substructures (Cγ−Hγ at δC/ δH = 63.4/3.7 ppm) were only observed in the standard Org_P. Resinol substructures (Cβ−Hβ in β−β′ resinol substructures at δC/ δH = 53.5/3.05 ppm and; Cγ−Hγ in β−β′ resinol substructures at δC/ δH = 71.0/3.81 and 4.17 ppm) were observed in Org_P, Org_E, [Ch][OAc]_P and [Ch][OAc]_E. 40, 42 More resinol structures were observed in [Ch][OAc]_P than in [Ch][OAc]_E, which may support the previous statement about modification of softwood lignin, as one of the causes that limit enzyme digestibility. 12 The decreased intensities in β-β’ linkages indicates that these linkages were modified, as it is unlikely that these C- C bonds have been truly broken and more likely they have been chemically altered. 12 A substantial amount of carbohydrates were also appreciated, suggesting that they may be still bound to the lignin. 12 Recovery capacity of ILs In order to evaluate the recyclability of ILs, FTIR-ATR spectra of fresh and recovered IL is shown in Figure S7. It must be highlighted that in all the cases recovered ILs still contain relative amounts of water (up to 15 %). All the spectra obtained of the recovered IL used with pine and eucalyptus are analogous, so the recovery of the IL is independent of the hardwood/softwood kind of biomass. [OHEtAm][HCO2] spectra decreases the intensity of 1340 cm-1 band, assigned to C-N stretching, as well as 1589 cm-1 band (assigned to C-O-O asymmetric stretching), keeping all the bands. 43 [Mim][Cl] FTIR-ATR is also modified, decreasing the intensity of 1116 cm-1 band (assigned to N1-C5-H rock vibration) and broadening the band observed at 1450 cm-1 (CH3(C2) scissor vibration). 44 The spectra observed for [Emim][OAc] remained unaltered and all the bands observed in the fresh IL were present in the recovered IL. The band at 1394 cm-1 (H-C-H rock vibration) was shifted due to the presence of water. 45-46 Although in different experiments the alteration of acetate anion is demonstrated, in this case bands that confirm the presence of the acetate anion are kept (1005 cm-1 correspond to C-O stretching in the acetate anion and 1174cm-1 corresponds to C=O stretching of the acetate anion). 47 The spectra comparison of fresh and recovered [Amim][Cl] evinces that all the bands are still present and no apparent differences in the chemical structure of this IL are observed, as was previously observed in other pretreatments. 48 In the case of [Ch][OAc], there are some bands that decrease the intensity (at 1089 cm-1 CHn rocking and twisting, and at 1011 cm-1 asymmetric stretching C3(CH3)N-CH2). 49-50 Other bands of the recovered [Ch][OAc] spectra increases the intensity at 1487 cm-1 CH3 scissor vibration of choline and at 1475 cm-1 C-H stretching vibration of CH3. [Ch][Lys] spectra of raw and recovered IL keep the representative bands (1569 cm-1 due to the asymmetric stretching of the carboxylate C=O; 12 1475 cm-1 corresponding to C-H stretching vibration of CH3; 1399 cm-1 corresponding to CO2 symmetric stretching; 1091 cm-1 corresponding to CHn rocking/twisting, and 955 cm-1 corresponding to C-N stretching). 50 Thermal decomposition temperature of every fresh and recovered IL is shown in Figure 3 and in the Supplementary material (Figure S8). The decomposition temperature of fresh ILs increase following the order: [Ch][Lys] (199 ℃) < [Ch][OAc] (219 ℃) < [OHEtAm][HCO2] (245 ℃) < [Emim][OAc] (248 ℃) < [Mim][Cl] (265 ℃) < [Amim][Cl] (275 ℃). Certain amounts of water are present in fresh and recovered IL (evinced by the loss of weight below 100 ℃ in all the cases). Recovered PILs decreased their decomposition temperatures around 20 ℃, indicating the structure has changed or some compounds are being accumulated in the IL. Hemicelluloses degradation occurs in the range of temperatures of 150 - 260 ℃, so the accumulation of these compounds may be one of the causes of the decrease of decomposition temperatures. 51 Furthermore, the global mass balance evinced that 32 and 43 % of the initial hemicelluloses were removed from [OHEtAm][HCO2]_P and [OHEtAm][HCO2]_E, and 97 and 98 % of the initial hemicelluloses present in [Mim][Cl]_P and [Mim][Cl]_E were also extracted. Note that Tpeak of recovered IL used with eucalyptus was lower than the one used with pine, in the same way that hemicelluloses extracted with eucalyptus are higher than with pine. However, taking into account the FTIR-ATR results of PILs, IL degradation or partial evaporation has also occurred. Choline based ILs show Tpeak lower with the recovered IL than with the fresh IL. In the case of [Ch][OAc], Tpeak is lower, but the curve keeps the same sigmoid shape. In the case of [Ch][Lys], the DTG curve aspect change considerably. In this case, the IL degrade at different steps due to the decarboxylation of aminoacid. 68 While Tpeak in the fresh IL are 199, 227, 253, 351 and 422 ℃, the degradation produced at 227 ℃ does not appear in the recovered IL, indicating that degradation step does not occur probably due to the degradation of this compound during the pretreatment. AILs from the other side showed decomposition temperatures similar or higher in the case of [Emim][OAc] and [Amim][Cl], respectively. The increase of degradation temperature produced in the recovered [Amim][Cl] is attributed to the presence of lignin impurities (lignin degradation occurs in the range 260 - 400 ℃) accumulated in the ionic liquid. 51-53 The results obtained are in accordance with the bands observed in the FTIR-ATR spectra, which keep all the representative bands. As a matter of fact, the higher biocompatibility of PILs and choline derived ILs contrast with low thermal stabilities. 52 Biomass accumulation occurred in all the cases and may be solved by the precipitation of its constituents with the aid of an antisolvent, but the degradation observed will alter the pretreatment if the IL pretends to be reused. 47 13 Figure 3. Tdecomp of raw and recovered ILs ACKNOWLEDGEMENTS This work was performed thanks to the financial support of the “Ministerio de Economía y Competitividad“, under the funded project CTQ2017-88623-R, the contract BES-2014-067788 and the “Juan de la Cierva” postdoctoral grant FJCI-2015-23765. Authors thank DuPont Industrial Biosciences for the donation of enzymatic cocktails. Graphical artwork was done thank to the work of Oscar Notario. 2D-NMR experiments were developed in the technical facilities of Nuclear Magnetic Resonance and Electron Spin of the Complutense University of Madrid. REFERENCES 1. Liu, Y.; Chen, W.; Xia, Q.; Guo, B.; Wang, Q.; Liu, S.; Liu, Y.; Li, J.; Yu, H., Efficient Cleavage of Lignin-Carbohydrate Complexes and Ultrafast Extraction of Lignin Oligomers from Wood Biomass by Microwave-Assisted Treatment with Deep Eutectic Solvent. ChemSusChem 2017, 10 (8), 1692-1700. 2. Nitsos, C. K.; Choli-Papadopoulou, T.; Matis, K. A.; Triantafyllidis, K. S., Optimization of Hydrothermal Pretreatment of Hardwood and Softwood Lignocellulosic Residues for Selective Hemicellulose Recovery and Improved Cellulose Enzymatic Hydrolysis. ACS Sustain. Chem. Eng. 2016, 4 (9), 4529-4544. 3. Wang, H. L.; Ben, H. X.; Ruan, H.; Zhang, L. B.; Pu, Y. Q.; Feng, M. Q.; Ragauskas, A. J.; Yang, B., Effects of Lignin Structure on Hydrodeoxygenation Reactivity of Pine Wood Lignin to Valuable Chemicals. ACS Sustain. Chem. Eng. 2017, 5 (2), 1824-1830. 4. Kandhola, G.; Djioleu, A.; Carrier, D. J.; Kim, J. W., Pretreatments for Enhanced Enzymatic Hydrolysis of Pinewood: a Review. Bioenergy Res 2017, 10 (4), 1138-1154. 5. Carrillo, I.; Vidal, C.; Elissetche, J. P.; Mendonça, R. T., Wood anatomical and chemical properties related to the pulpability of Eucalyptus globulus: a review. South. Forests 2017, 80 (1), 1-8. 6. Cerasoli, S.; Caldeira, M.; Pereira, J.; Caudullo, G.; de Rigo, D., Eucalyptus globulus and other eucalypts in Europe: distribution, habitat, usage and threats. European Atlas of Forest Tree Species. Luxembourg: Publishing Office of the EU 2016, e01b5bb. 14 7. Elgharbawy, A. A.; Alam, M. Z.; Moniruzzaman, M.; Goto, M., Ionic liquid pretreatment as emerging approaches for enhanced enzymatic hydrolysis of lignocellulosic biomass. Biochem. Eng. J. 2016, 109, 252-267. 8. da Costa Lopes, A. M.; Lukasik, R. M., Separation and Recovery of a Hemicellulose- Derived Sugar Produced from the Hydrolysis of Biomass by an Acidic Ionic Liquid. ChemSusChem 2018, 11 (6), 1099-1107. 9. Sun, J.; Konda, N. V. S. N. M.; Shi, J.; Parthasarathi, R.; Dutta, T.; Xu, F.; Scown, C. D.; Simmons, B. A.; Singh, S., CO2 enabled process integration for the production of cellulosic ethanol using bionic liquids. Energ. Environ. Sci. 2016, 9 (9), 2822-2834. 10. Fang, Z., Production of Biofuels and Chemicals with Ionic Liquids.. Dordrecht, Heidelberg, New York, Springer: London, 2014. 11. Morales-delaRosa, S.; Campos-Martin, J. M.; Fierro, J. L., Complete chemical hydrolysis of cellulose into fermentable sugars through ionic liquids and antisolvent pretreatments. ChemSusChem 2014, 7 (12), 3467-75. 12. Brandt, A.; Gschwend, F.; Fennell, P.; Lammens, T.; Tan, B.; Weale, J.; Hallett, J., An economically viable ionic liquid for the pretreatment of lignocellulosic biomass. Green Chem. 2017, 19, 3078-3102. 13. Thierry, M.; Majira, A.; Pégot, B.; Cezard, L.; Bourdreux, F.; Clément, G.; Perreau, F.; Boutet‐Mercey, S.; Diter, P.; Vo‐Thanh, G., Imidazolium‐Based Ionic Liquids as Efficient Reagents for the C− O Bond Cleavage of Lignin. ChemSusChem 2018, 11 (2), 439-448. 14. Amarasekara, A. S., Acidic Ionic Liquids. Chem. Rev. 2016, 116 (10), 6133-83. 15. Haykir, N. I.; Bahcegul, E.; Bicak, N.; Bakir, U., Pretreatment of cotton stalk with ionic liquids including 2-hydroxy ethyl ammonium formate to enhance biomass digestibility. Ind. Crops Prod. 2013, 41 (0), 430-436. 16. Cox, B. J.; Ekerdt, J. G., Pretreatment of yellow pine in an acidic ionic liquid: extraction of hemicellulose and lignin to facilitate enzymatic digestion. Bioresour. Technol. 2013, 134, 59-65. 17. Dutta, T.; Papa, G.; Wang, E.; Sun, J.; Isern, N. G.; Cort, J. R.; Simmons, B. A.; Singh, S., Characterization of Lignin Streams during Bionic Liquid-Based Pretreatment from Grass, Hardwood, and Softwood. ACS Sustain. Chem. Eng. 2018, 6 (3), 3079-3090. 18. Sun, J.; Konda, N. V. S. N. M.; Parthasarathi, R.; Dutta, T.; Valiev, M.; Xu, F.; Simmons, B. A.; Singh, S., One-pot integrated biofuel production using low-cost biocompatible protic ionic liquids. Green Chem. 2017, 19 (13), 3152-3163. 19. Xue-Dan Hou, J. X., Ning Li Min-Hua Zong, Effect of anion Structures on Cholinium Ionic Liquids pretreatment of Rice Sraw and the subsequent Enzymatic Hydrolysis. Biotechnol. Bioeng. 2015, 112, 65-73. 20. Casas, A.; Oliet, M.; Alonso, M. V.; Santos, T. M.; Rodriguez, F., Dissolution of Pinus radiata and Eucalyptus Globulus Woods in 1-Allyl-3-methylimidazolium Chloride for Cellulose or Lignin Regeneration. Ind. Eng. Chem. Res. 2013, 52 (10), 3628-3636. 21. Hauru, L. K.; Hummel, M.; King, A. W.; Kilpelainen, I.; Sixta, H., Role of solvent parameters in the regeneration of cellulose from ionic liquid solutions. Biomacromolecules 2012, 13 (9), 2896-905. 22. Rigual, V.; Santos, T. M.; Dominguez, J. C.; Alonso, M. V.; Oliet, M.; Rodriguez, F., Evaluation of hardwood and softwood fractionation using autohydrolysis and ionic liquid microwave pretreatment. Biomass Bioenerg. 2018, 117, 190-197. 23. Sun, N.; Parthasarathi, R.; Socha, A. M.; Shi, J.; Zhang, S.; Stavila, V.; Sale, K. L.; Simmons, B. A.; Singh, S., Understanding pretreatment efficacy of four cholinium and imidazolium ionic liquids by chemistry and computation. Green Chem. 2014, 16 (5), 2546-2557. 24. Torr, K. M.; Love, K. T.; Simmons, B. A.; Hill, S. J., Structural features affecting the enzymatic digestibility of pine wood pretreated with ionic liquids. Biotechnol. Bioeng. 2016, 113 (3), 540-9. 25. Li, C. L.; Sun, L.; Simmons, B. A.; Singh, S., Comparing the Recalcitrance of Eucalyptus, Pine, and Switchgrass Using Ionic Liquid and Dilute Acid Pretreatments. Bioenergy Res 2013, 6 (1), 14-23. 26. Ninomiya, K.; Inoue, K.; Aomori, Y.; Ohnishi, A.; Ogino, C.; Shimizu, N.; Takahashi, K., Characterization of fractionated biomass component and recovered ionic liquid during repeated 15 process of cholinium ionic liquid-assisted pretreatment and fractionation. Chem. Eng. J. 2015, 259, 323-329. 27. Cheng, F.; Wang, H.; Chatel, G.; Gurau, G.; Rogers, R. D., Facile pulping of lignocellulosic biomass using choline acetate. Bioresour. Technol. 2014, 164, 394-401. 28. Cox, B. J.; Jia, S. Y.; Zhang, Z. C.; Ekerdt, J. G., Catalytic degradation of lignin model compounds in acidic imidazolium based ionic liquids: Hammett acidity and anion effects. Polym. Degrad. Stab. 2011, 96 (4), 426-431. 29. De Gregorio, G. F.; Weber, C. C.; Gräsvik, J.; Welton, T.; Brandt, A.; Hallett, J. P., Mechanistic insights into lignin depolymerisation in acidic ionic liquids. Green Chem. 2016, 18 (20), 5456-5465. 30. Zhang, L. B.; Pu, Y. Q.; Cort, J. R.; Ragauskas, A. J.; Yang, B., Revealing the Molecular Structural Transformation of Hardwood and Softwood in Dilute Acid Flowthrough Pretreatment. ACS Sustain. Chem. Eng. 2016, 4 (12), 6618-6628. 31. Rigual, V.; Santos, T. M.; Dominguez, J. C.; Alonso, M. V.; Oliet, M.; Rodriguez, F., Combining autohydrolysis and ionic liquid microwave treatment to enhance enzymatic hydrolysis of Eucalyptus globulus wood. Bioresour. Technol. 2018, 251, 197-203. 32. Sun, N.; Rodriguez, H.; Rahman, M.; Rogers, R. D., Where are ionic liquid strategies most suited in the pursuit of chemicals and energy from lignocellulosic biomass? Chem. Commun. 2011, 47 (5), 1405-1421. 33. French, A. P.; Mills, S.; Swarup, R.; Bennett, M. J.; Pridmore, T. P., Colocalization of fluorescent markers in confocal microscope images of plant cells. Nat. Protoc. 2008, 3 (4), 619-28. 34. Donaldson, L.; Vaidya, A., Visualising recalcitrance by colocalisation of cellulase, lignin and cellulose in pretreated pine biomass using fluorescence microscopy. Sci. Rep. 2017, 7, 44386. 35. Shi, J.; Pattathil, S.; Parthasarathi, R.; Anderson, N. A.; Kim, J. I.; Venketachalam, S.; Hahn, M. G.; Chapple, C.; Simmons, B. A.; Singh, S., Impact of engineered lignin composition on biomass recalcitrance and ionic liquid pretreatment efficiency. Green Chem. 2016, 18 (18), 4884-4895. 36. Lai, C.; Tu, M.; Shi, Z.; Zheng, K.; Olmos, L. G.; Yu, S., Contrasting effects of hardwood and softwood organosolv lignins on enzymatic hydrolysis of lignocellulose. Bioresour. Technol. 2014, 163, 320-7. 37. Crestini, C.; Lange, H.; Sette, M.; Argyropoulos, D. S., On the structure of softwood kraft lignin. Green Chem. 2017, 19 (17), 4104-4121. 38. Evtuguin, D. V.; Neto, C. P.; Silva, A. M.; Domingues, P. M.; Amado, F. M.; Robert, D.; Faix, O., Comprehensive study on the chemical structure of dioxane lignin from plantation Eucalyptus globulus wood. J. Agric. Food. Chem. 2001, 49 (9), 4252-61. 39. Shinde, S. D.; Meng, X. Z.; Kumar, R.; Ragauskas, A. J., Recent advances in understanding the pseudo-lignin formation in a lignocellulosic biorefinery. Green Chem. 2018, 20 (10), 2192-2205. 40. del Rio, J. C.; Rencoret, J.; Prinsen, P.; Martinez, A. T.; Ralph, J.; Gutierrez, A., Structural characterization of wheat straw lignin as revealed by analytical pyrolysis, 2D-NMR, and reductive cleavage methods. J. Agric. Food. Chem. 2012, 60 (23), 5922-35. 41. Santos, J. I.; Fillat, U.; Martin-Sampedro, R.; Eugenio, M. E.; Negro, M. J.; Ballesteros, I.; Rodriguez, A.; Ibarra, D., Evaluation of lignins from side-streams generated in an olive tree pruning- based biorefinery: Bioethanol production and alkaline pulping. Int. J. Biol. Macromol. 2017, 105 (Pt 1), 238-251. 42. Kim, H.; Ralph, J., Solution-state 2D NMR of ball-milled plant cell wall gels in DMSO- d(6)/pyridine-d(5). Org. Biomol. Chem. 2010, 8 (3), 576-91. 43. Papadopoulou, A. A.; Tzani, A.; Alivertis, D.; Katsoura, M. H.; Polydera, A. C.; Detsi, A.; Stamatis, H., Hydroxyl ammonium ionic liquids as media for biocatalytic oxidations. Green Chem. 2016, 18 (4), 1147-1158. 44. Haddad, B.; Mokhtar, D.; Goussem, M.; Belarbi, E. H.; Villemin, D.; Bresson, S.; Rahmouni, M.; Dhumal, N. R.; Kim, H. J.; Kiefer, J., Influence of methyl and propyl groups on the vibrational spectra of two imidazolium ionic liquids and their non-ionic precursors. J. Mol. Struct. 2017, 1134, 582-590. 45. Dhumal, N. R.; Kim, H. J.; Kiefer, J., Molecular interactions in 1-ethyl-3- methylimidazolium acetate ion pair: a density functional study. J. Phys. Chem. A 2009, 113 (38), 10397-404. 16 46. Jeon, Y.; Sung, J.; Seo, C.; Lim, H.; Cheong, H.; Kang, M.; Moon, B.; Ouchi, Y.; Kim, D., Structures of ionic liquids with different anions studied by infrared vibration spectroscopy. J. Phys. Chem. B 2008, 112 (15), 4735-40. 47. Weerachanchai, P.; Lim, K. H.; Lee, J. M., Influence of organic solvent on the separation of an ionic liquid from a lignin-ionic liquid mixture. Bioresour. Technol. 2014, 156 (0), 404-7. 48. Rigual, V.; Santos, T. M.; Domínguez, J. C.; Alonso, M. V.; Oliet, M.; Rodriguez, F., Recovery and Reuse of 1-Allyl-3-methylimidazolium Chloride in the Fractionation of Pinus radiata Wood. ACS Sustain. Chem. Eng. 2017, 5 (3), 2384-2392. 49. Harmon, K. M.; Avci, G. F.; Thiel, A. C., Hydrogen bonding. J. Mol. Struct. 1986, 145 (1- 2), 83-91. 50. Campetella, M.; Bodo, E.; Caminiti, R.; Martino, A.; D'Apuzzo, F.; Lupi, S.; Gontrani, L., Interaction and dynamics of ionic liquids based on choline and amino acid anions. J. Chem. Phys. 2015, 142 (23), 234502. 51. Prado, R.; Erdocia, X.; Labidi, J., Study of the influence of reutilization ionic liquid on lignin extraction. J Clean Prod 2016, 111, 125-132. 52. Bhattacharyya, S.; Shah, F. U., Thermal stability of choline based amino acid ionic liquids. J. Mol. Liq. 2018, 266, 597-602. 53. Weerachanchai, P.; Lee, J. M., Recyclability of an ionic liquid for biomass pretreatment. Bioresour. Technol. 2014, 169, 336-343. S1 Supporting information Protic, aprotic and choline-derived ionic liquids: towards enhancing the accessibility of hardwood and softwood Authors: Victoria Rigual, Antonio Ovejero-Pérez, Sandra Rivas, Juan C. Domínguez, M. Virginia Alonso, Mercedes Oliet and Francisco Rodriguez Affiliation and address: Department of Chemical Engineering and Materials, Faculty of Chemistry, Complutense University of Madrid, Avda. Complutense s/n, 28040 Madrid, Spain. Corresponding author: Victoria Rigual E-mail: vicrigua@ucm.es ; Tel: (+34)913948505 Number of pages: 14 Number of Figures: 8 mailto:vicrigua@ucm.es S2 DETAILED MATERIALS AND METHODS Pretreated biomass characterization Compositional analysis Untreated/pretreated biomass composition was measured following the National Renewable Energy Laboratory methodology, adapted to small quantities of samples (50 mg of biomass) 1-2. A more detailed description of the methodology employed and HPLC methods used in this analysis can be found in our previous work 3. Enzymatic accessibility Biomass accessibility to enzymes of the pretreated samples was measured according to the National Renewable Energy procedure and specifications marked by the supplier 4. Low solid loading (1 % w/w) was employed to saccharify samples in an IKA KS 4000I orbital incubator at 50 °C and 150 rpm, as described before 3. Glucan digestibility was calculated applying the glucose- to-glucan correction according to Eq. S1. 𝐺𝑙𝑢𝑐𝑎𝑛 𝑑𝑖𝑔𝑒𝑠𝑡𝑖𝑏𝑖𝑙𝑖𝑡𝑦(%) = 𝐺𝑙𝑢𝑐𝑎𝑛 ℎ𝑦𝑑𝑟𝑜𝑙𝑦𝑠𝑒𝑑 (𝑔) 𝐺𝑙𝑢𝑐𝑎𝑛 𝑖𝑛 𝑢𝑛𝑡𝑟𝑒𝑎𝑡𝑒𝑑/𝑝𝑟𝑒𝑡𝑟𝑒𝑎𝑡𝑒𝑑 𝑤𝑜𝑜𝑑 (𝑔) (S1) Solid morphology Scanning electron microscopy (SEM) images of untreated and pretreated biomass samples were acquired in a Jeol JSM 6400 at an operation voltage of 20 kV. A magnification of x1000 was employed to capture images. Analyses were developed in the technical facilities of the Spanish National Centre for Electron Microscopy. Confocal fluorescence microscopy (CFM) was also employed to observe variations in the holocellulose (cellulose+hemicellulose) and lignin ratios in the biomass surface as previously described 5. Detection and emission wavelength ranges and pinhole were fixed, and every sample was stored and dried under the same conditions and during the same time. All the images were acquired in the technical facilities of the Centre for Cytometry and Fluorescence Microscopy of the Complutense University of Madrid. Samples were obtained in the two-coupled channels, resulting in a two colours Z-stack combination of images. Characterization of lignin-rich solid obtained after enzymatic saccharification Samples obtained after the enzymatic saccharification were recovered, washed, and stored for further analysis. Eucalyptus and pine saccharified by-stream samples, previously pretreated with [Emim][OAc] and [Ch][OAc], were analyzed to determine lignin structure and hemicelluloses. These two ILs were selected after mass balance analysis and calculation of every sample (high lignin content, high digestibility and low solid degradation in the pretreatment). Additionally, pine and S3 eucalyptus organosolv lignin samples were used as reference. These lignins were obtained with a mixture of 60 % (w/w) ethanol/water at 200 °C for 50 min, using sulfuric acid to precipitate lignin. Organosolv pretreatments were carried out in a Parr steel pressure reactor (Parr Instrument Company, model 4567) 6. 2D-NMR (hsqc) spectroscopy Saccharified samples were analyzed by 2D HSQC NMR according to Heikkinen et al 7. 15- 25 mg of sample were solubilized in 500 μL of DMSO-d6 and sonicate for 210 min at room temperature in NMR tubes, producing homogeneous solutions. The presence of polysaccharides in the saccharified cell wall samples, resulted in samples remained largely in gel-state (due to the high viscosity and the presence of insoluble components), which enhanced rapid relaxation times 8. Spectra were acquired using a Bruker AVANCE III spectrometer (Bruker BioSpin, Rheinstetten, Germany) operating at 700.17 MHz proton frequency and with a 5 mm TXI cryopprobe. 13C-1H correlation spectra were measured using a perfect_hsqc (hetero-nuclear single quantum coherence) pulse program. This experiment provides a phase-sensitive, gradient 2D HSQC spectrum using adiabatic pulses for inversion and refocusing. Experiments were carried out at 25 °C with the following parameters: Spectral width of 16 ppm in F2 (1H) dimension with 2048 data points (TD1) and 240 ppm in F1 (13C) dimension with 192 data points (TD2); scan number (SN) of 8; interscan delay (D1) of 20 s; and acquisition time of 8 h. Chemical shifts were referenced to the central DMSO peak at 39.5/2.50 ppm (δC/ δH). Before processing, squared sine apodization (90 °) in f2 and f1 was applied. Assignments of the spectra were fixed based on literature 9-11. The NMR data acquisition software employed was Bruker’s Topspin 3.5 (Bruker BioSpin, Rheinstetten, Germany), and processing was performed using Mestrenova software. Ionic liquids characterization After pretreatment, washing fractions were distilled to remove water using a Buchi R-114. Recovered IL samples were sealed and stored at room temperature for analysis. Thermogravimetrical analysis Fresh and recovered ILs were examined in terms of thermal decomposition behavior using a TG Mettler Toledo instrument TGA/SDTA851e. Each sample was initially heated from 30 to 105 ℃ to remove remaining water, and kept the temperature for 30 min. Consequently, samples were heated at a heating rate of 10 ℃/min up to 800 ℃. Experiments were performed using dry S4 nitrogen at a flow rate of 20 mL/min in PCA/Sapphire crucibles (Mettler Toledo) of 70 µL. 7-9 mg of IL were employed in each experiment. Tpeak has been chosen as the comparison parameter to determine differences between fresh and recovered IL, although maximum operation temperature must be lower (Tonset or Tstart) 12. FTIR-ATR spectroscopy Attenuated total reflectance Fourier Transform infrared (FTIR/ATR) spectra of every raw and recovered IL were acquired in a Jasco 4700 spectrometer equipped with a Golden Gate TM diamond accessory Specac 10542. Samples were acquired in the range 4000-400 cm-1 and cut to the most representative bands and normalized. 64 scans were employed for analysis with a resolution of 4 cm-1. Enzymatic accessibility A dosage of 0.25 mL of Accellerase 1500® enzyme cocktail/g glucan in the untreated or pretreated samples was suspended in 50 mM citrate buffer of pH 5.0 containing 0.002 % of sodium azide. Every 3, 6, 12, 24, 48 and 72 h, 150 μL aliquots were collected and centrifuged. Figure S1: Result of the screening [OHEtAm][HCO2]_lignin alkali and regeneration S5 Figure S2. SEM micrographs (x1000) of a) untreated pine; b) [Mim][Cl]_P; c) [Emim][OAc]_P; d) [Ch][OAc]_P;e) untreated eucalyptus; f) [Mim][Cl]_E; g) [Emim][OAc]_E; h) [Ch][OAc]_E Figure S3: CFM images of a) Untreated pine; b) [OHEtAm][HCO2]_P; c) [Mim][Cl]_P; d) [Emim][OAc]_P; e) [Amim][Cl]_P; f) [Ch][OAc]_P and g) [Ch][Lys]_P. Below every image holocellulose thresholded channel (on the left, green channel) and lignin channel (on the right, red channel). S6 Figure S4: CFM images of a) Untreated eucalyptus; b) [OHEtAm][HCO2]_E; c) [Mim][Cl]_E; d) [Emim][OAc]_E; e) [Amim][Cl]_E; f) [Ch][OAc]_E and g) [Ch][Lys]_E. Below every image holocellulose thresholded channel (on the left, green channel) and lignin channel (on the right, red channel). S7 Figure S5: 2D-NMR lignin aromatic region of a) Organosolv pine lignin; b) [Emim][OAc]_P saccharified biomass; c) [Ch][OAc]_P sacchariefied biomass; d) Organosolv eucalyptus lignin; e) [Emim][OAc]_E saccharified biomass and; f) [Ch][OAc]_E sacchariefied biomass S8 Figure S6: 2D-NMR lignin side chain, polysaccharide and non-oxygenated aliphatic regions of a) Organosolv pine lignin; b) [Emim][OAc]_P saccharified biomass; c) [Ch][OAc]_P saccharified biomass; d) Organosolv eucalyptus lignin; e) [Emim][OAc]_E saccharified biomass and; f) [Ch][OAc]_E saccharified biomass S9 Figure S7: FTIR-ATR spectra of a) [OHEtAm][HCO2]; b) [Mim][Cl]; c) [Emim][OAc]; d) [Amim][Cl]; e) [Ch][OAc] and f) [Ch][Lys]. S10 Figure S8: DTg curves of a) [OHEtAm][HCO2]; b) [Mim][Cl]; c) [Emim][OAc]; d) [Amim][Cl]; e) [Ch][OAc] and f) [Ch][Lys] 150 300 450 600 750 0.0 0.2 0.4 0.6 0.8 1.0 150 300 450 600 750 0.0 0.2 0.4 0.6 0.8 1.0 150 300 450 600 750 0.0 0.2 0.4 0.6 0.8 1.0 150 300 450 600 750 0.0 0.2 0.4 0.6 0.8 1.0 150 300 450 600 750 0.0 0.2 0.4 0.6 0.8 1.0 150 300 450 600 750 0.0 0.2 0.4 0.6 0.8 1.0 DTG [OHEtAm][HCO2] DTG [OHEtAm][HCO2]_E DTG [OHEtAm][HCO2]_P 1 s t d e ri v a ti v e (m g ·s -1 ) Sample temperature (ºC) f) d)d) b) e) c) DTG [Mim][Cl] DTG [Mim][Cl]_E DTG [Mim][Cl]_P 1 s t d e ri v a ti v e (m g ·s -1 ) Sample temperature (ºC) a) DTG [Emim][OAc] DTG [Emim][OAc]_E DTG [Emim][OAc]_P 1 s t d e ri v a ti v e (m g ·s -1 ) Sample temperature (ºC) DTG [Ch][OAc] DTG [Ch][OAc]_E DTG [Ch][OAc]_P 1 s t d e ri v a ti v e (m g ·s -1 ) Sample temperature (ºC) DTG [Ch][Lys] DTG [Ch][Lys]_E DTG [Ch][Lys]_P 1 s t d e ri v a ti v e (m g ·s -1 ) Sample temperature (ºC) DTG [Amim][Cl] DTG [Amim][Cl]_E DTG [Amim][Cl]_P 1 s t d e ri v a ti v e (m g ·s -1 ) Sample temperature (ºC) S11 1. Ibanez, A. B.; Bauer, S., Downscaled method using glass microfiber filters for the determination of Klason lignin and structural carbohydrates. Biomass Bioenerg. 2014, 68 (0), 75- 81. 2. A. Sluiter, B. H., R.Ruiz, C.Scarlata, J.Sluitr, D.Templeton and D. Crocker, Determination of Structural carbohydrates and lignin in biomass. NREL Laboratory Analytical Procedure 2011. 3. Rigual, V.; Santos, T. M.; Dominguez, J. C.; Alonso, M. V.; Oliet, M.; Rodriguez, F., Evaluation of hardwood and softwood fractionation using autohydrolysis and ionic liquid microwave pretreatment. Biomass Bioenerg. 2018, 117, 190-197. 4. Resch, M.; Baker, J.; Decker, S., Low solids enzymatic saccharification of lignocellulosic biomass. NREL Laboratory Analytical Procedure 2015. 5. Rigual, V.; Santos, T. M.; Domínguez, J. C.; Alonso, M. V.; Oliet, M.; Rodriguez, F., Combining autohydrolysis and ionic liquid microwave treatment to enhance enzymatic hydrolysis of Eucalyptus globulus wood. Bioresource technology 2018, 251, 197-203. 6. Meneses, T. M. S. Autohidrólisis y deslignificación organosolv de madera de Pinus radiata para la recuperación de hemicelulosas y lignina con aprovechamiento de la fracción celulósica por vía enzimática. Universidad Complutense de Madrid, 2017. 7. Heikkinen, S.; Toikka, M. M.; Karhunen, P. T.; Kilpelainen, I. A., Quantitative 2D HSQC (Q-HSQC) via suppression of J-dependence of polarization transfer in NMR spectroscopy: application to wood lignin. J. Am. Chem. Soc. 2003, 125 (14), 4362-7. 8. Kim, H.; Ralph, J., Solution-state 2D NMR of ball-milled plant cell wall gels in DMSO- d(6)/pyridine-d(5). Org. Biomol. Chem. 2010, 8 (3), 576-91. 9. del Rio, J. C.; Rencoret, J.; Prinsen, P.; Martinez, A. T.; Ralph, J.; Gutierrez, A., Structural characterization of wheat straw lignin as revealed by analytical pyrolysis, 2D-NMR, and reductive cleavage methods. J. Agric. Food. Chem. 2012, 60 (23), 5922-35. 10. Castanar, L.; Sistare, E.; Virgili, A.; Williamson, R. T.; Parella, T., Suppression of phase and amplitude J(HH) modulations in HSQC experiments. Magn. Reson. Chem. 2015, 53 (2), 115- 119. 11. Rico, A.; Rencoret, J.; del Rio, J. C.; Martinez, A. T.; Gutierrez, A., In-Depth 2D NMR Study of Lignin Modification During Pretreatment of Eucalyptus Wood with Laccase and Mediators. Bioenergy Res 2015, 8 (1), 211-230. 12. Eshetu, G. G.; Jeong, S.; Pandard, P.; Lecocq, A.; Marlair, G.; Passerini, S., Comprehensive Insights into the Thermal Stability, Biodegradability, and Combustion Chemistry of Pyrrolidinium- Based Ionic Liquids. ChemSusChem 2017, 10 (15), 3146-3159. PUBLICACIÓN III “Recovery and reuse of 1‑allyl-3- methylimidazolium chloride in the fractionation of Pinus radiata wood” Victoria Rigual, Tamara M. Santos, Juan C. Domínguez, M. Virginia Alonso, Mercedes Oliet y Francisco Rodriguez ACS Sustainable Chemistry and Engineering, 2017, 5(3), 2384-2392 Recovery and Reuse of 1‑Allyl-3-methylimidazolium Chloride in the Fractionation of Pinus radiata Wood Victoria Rigual,* Tamara M. Santos, Juan Carlos Domínguez, M. Virginia Alonso, Mercedes Oliet, and Francisco Rodriguez Department of Chemical Engineering, Faculty of Chemistry, Complutense University of Madrid, Avda. Complutense s/n, 28040 Madrid, Spain *S Supporting Information ABSTRACT: Ionic liquids are expected to be potential solvents for biorefinery purposes. However, the high cost has limited widespread implementation, making their recycling essential to the process. Herein, the capacity to recover and reuse the ionic liquid 1-allyl-3- methylimidazolium chloride ([Amim][Cl]) in the fractionation of Pinus radiata wood using microwave heating was evaluated. In the proposed process, cellulose- and lignin-rich materials were obtained. The ionic liquid used was recovered and reused in the forthcoming cycles up to five times. Characterization of [Amim][Cl] recovered demonstrated that although the structure of the ionic liquid remained unaltered, some woody fractions were accumulated. As a consequence, a reduction of the fractionation yields was obtained, and variations of the composition and amorphicity of the fractions recovered were produced. Through a global mass balance, this study gives an overall approximation of the concentration of the main components and the characteristic streams involved in successive cycles. KEYWORDS: Ionic liquids, Recycling, Biomass fractionation, Pine wood, Microwave heating ■ INTRODUCTION Ionic liquids (ILs) are effective solvents for a variety of materials, including polymeric compounds.1 Their negligible vapor pressure, nonflammability, and high chemical and thermal stability, among other features, make them potential solvents for a wide range of polymers.2 Lignocellulosic biomass is the most available plant material on our planet and the most sustainable feedstock for the production of biomass-based energy.3 Biomass softwood resources are expected to be part of the biorefinery feedstock base. However, their high recalcitrance makes them difficult to fractionate compared to grass or hardwood biomass.4 Fractionation and pretreatment technologies can account for up to 40% of the total processing costs of lignocellulosic biomass conversion.5 Current implemented technologies are high energy input (liquid hot water pretreatments), excessive degradation of biomass releasing inhibitors (steam explosion), or a few effective treatments in softwoods (alkaline pretreat- ments).5 Ionic liquids based biomass fractionation technologies have emerged as potential solvents in lignocellulosic biomass pretreatment processes. Although at commercial scale ILs are still a challenging task, their advantages have made them an alternative for the future of biorefineries.6 In a fractionation approach, ILs reduce recalcitrance enabling deconstruction and disruption of the lignin and hemicellulose network.7 In a pretreatment approach, the cellulose crystallinity reduction increases cellulose accessibility, which favors high conversions from cellulose to glucose.5,8 Furthermore, their low volatilities make possible treatments of biomass at atmospheric pressure, with nonodorous and safer liquors.9 Several imidazolium ILs, such as [Emim][OAc], [Bmim]- [Cl], [Amim][Cl], or [Emim][DEP], among others, have been proven to dissolve cellulose and lignocellulosic biomass.10,11 [Emim][OAc] is considered the most suitable IL for biomass fractionation to date, but also one of the most expensive.12 Although most chloride salts are difficult to handle due to the high melting point and viscosity, [Amim][Cl] was found to be the most effective to dissolve wood chips. In particular, microwave heating has been effectively used as a method to accelerate dissolution in other chemical processes and has already been checked by Casas et al. as an efficient method to dissolve pine wood with [Amim][Cl].13,14 Cellulose dissolution in lignocellulosic materials with [Amim][Cl] is due to the disruption of hydrogen bonding and the coordination of chloride ions with the hydroxyl groups of cellulose. In addition, the double bond in the side chain of [Amim][Cl] facilitates a reduction in viscosity and an increase in dissolution capacity compared to [Bmim][Cl].15 Finally, [Amim][Cl] Tonset (273 Received: November 10, 2016 Revised: January 22, 2017 Published: January 30, 2017 Research Article pubs.acs.org/journal/ascecg © 2017 American Chemical Society 2384 DOI: 10.1021/acssuschemeng.6b02723 ACS Sustainable Chem. Eng. 2017, 5, 2384−2392 pubs.acs.org/journal/ascecg http://dx.doi.org/10.1021/acssuschemeng.6b02723 °C) is higher than [Emim][OAc] (216 °C), enabling higher working temperature conditions without thermal degradation.16 After biomass dissolution, an antisolvent is commonly employed to regenerate the biomass dissolved. Water, methanol, ethanol, or isopropanol are some commonly reported antisolvents studied to date.17 Some former studies have already pointed out Kamleft−Taft and polarity scale (ET N) as adequate parameters to describe solvent properties (Table 1). Focusing on lignin solubility, Weerachanchai et al. observed that water does not precipitate lignin from an IL−lignin dissolution in contrast with alcohols such as methanol, ethanol, or isopropanol.18 Regarding cellulose regeneration, a large β value (>0.8) is recquired to effectively dissolve cellulose.19 Methanol has higher values of α (hydrogen-bond acidity) and ET N (normalized molar transition energies) than other alcohols, which indicates a high hydrogen bond donor, and a moderate polarity of the antisolvent interacting strongly with the IL and faster displacing the interactions between the IL and cellulose, leading to quick cellulose regeneration.18,20 Despite the advantages of ILs, their high cost compared to the substrate processed, and their toxicity limit large-scale application. Techno-economical analysis suggests ionic liquid recycling as a mandatory requirement.22 Many methods are used to recover ILs. The conventional method consists of evaporating the antisolvent used under a vacuum with the subsequent intensive energy demand of the process.22,23 Another alternative is the use of aqueous kosmotropic salt solutions to form a three-phase system, but high salt-rich fractions may cause environmental problems.24 Recently, Lynam et al. have reported a membrane distillation method to separate water and the ionic liquid that could be used in biomass pretreatment processes.25 Several studies have demonstrated the feasibility of recovering [Emim][OAc] under vacuum distillation and reusing it several times.26 One of the factors that should be assessed is the reactivity of the IL with the biomass, as Clough et al. observed the degradation of the hydroxyalkyl chain and the loss of formaldehyde units in the IL [Emim][OAc].27 ILs that are not reactive with biomass are assessed to avoid limitations in their recycling. In this work, the feasibility of the recovery and reuse of [Amim][Cl] used for Pinus radiata wood fractionation, to obtain a cellulose and a lignin rich material is studied. Solids yields for each cycle are studied. The recovered IL was analyzed and correlated with the biomass yields. Furthermore, the solids obtained in the fractionation process with fresh and recovered IL were analyzed to determine its chemical composition, crystallinity, and functional groups. The concentration of involved components in each stream was determined in a mass balance approach. ■ EXPERIMENTAL SECTION Materials and Reagents. The ionic liquid 1-allyl-3-methyl- imidazolium chloride ([Amim][Cl], >98%, Iolitec GmbH) was employed for wood dissolution. Previously, [Amim][Cl] was dried in a vacuum oven for 24 h at 60 °C. Pinus radiata chips were provided by the Instituto Nacional de Investigacioń y Tecnologıá Agraria y Alimentaria (INIA). The biomass samples were characterized in terms of moisture content and chemical composition following the NREL methodology.28 Pinus radiata chips were milled and sieved to obtain particle sizes <150 μm. Extractives were removed by extraction, first with acetone and then with water, in a Soxhlet extractor. Dimethyl sulfoxide (DMSO, ≥ 99.9%, Panreac) and dry methanol (MeOH, max 0.005% water, Panreac) were employed as the cosolvent and antisolvent, respectively. In addition, microcrystalline cellulose (Avicel PH-101) was used as a reference material in the character- ization of the regenerated cellulose. Organosolv lignin from Pinus radiata wood was also used as a reference material for the UV−vis measurements, obtained with a mixture of 60% (w/w) ethanol/water at 200 °C for 50 min (process developed in a Parr reactor model 4567). Sulfuric acid was used to precipitate the lignin from the mixture. Experimental Procedure. The scheme of the experimental process is shown in Scheme 1. Pinus radiata wood (0.8 g) was mixed with 20 g of [Amim][Cl] and 2.6 g of DMSO (stream 1). Samples were heated under microwave radiation in a Berghof SpeedWave Four microwave oven, using a two-step program, following the procedure reported in a previous work.13 Once the dissolution step was finished, the cosolvent DMSO (25 mL) was added to the mixture (stream 2), and centrifuged at 6000 rpm for 10 min to separate the liquid phase formed by [Amim][Cl] and the dissolved wood, and the solid phase formed by the lignin rich material (LRM). Afterward, methanol (200 mL) was added to the liquid phase to precipitate a cellulose-rich material (CRM) (stream 3). The solution was stirred for 10 min in a water bath at 40 °C. This solution was filtered under a vacuum to obtain the above-mentioned CRM (stream 10). A crucial step to guarantee the maximum [Amim][Cl] recovery yield and avoid [Amim][Cl] residues in the CRM and LRM is the washing step. Both solids, the CRM and the lignin rich material (LRM) were washed several times to ensure the [Amim][Cl] removal. The CRM was first washed with the 200 mL of methanol used in the Table 1. Kamleft−Taft and Polarity Scale (ET N) Parameters9,17,20,21 Kamleft−Taft Parameters α β π* ET N AmimCl 0.46 0.83 1.17 >0.5 methanol 0.98 0.66 0.60 0.76 ethanol 0.86 0.75 0.54 0.65 isopropanol 0.76 0.84 0.48 0.57 water 1.17 0.47 1.09 1.00 Scheme 1. Experimental Process ACS Sustainable Chemistry & Engineering Research Article DOI: 10.1021/acssuschemeng.6b02723 ACS Sustainable Chem. Eng. 2017, 5, 2384−2392 2385 http://dx.doi.org/10.1021/acssuschemeng.6b02723 regeneration step and then with 200 mL of water (not included in the scheme). The LRM was washed two consecutive times with methanol (50 mL per washing) (streams 4 and 5). LRM (stream 8) and CRM (stream 10) were dried in a vacuum oven at 60 °C. All the washing fractions containing [Amim][Cl], DMSO, and methanol were analyzed by HPLC. The above process was repeated five times, using the IL recovered in the previous cycle. To ensure that the [Amim][Cl] used in all the cycles was completely derived from the previous cycle, the process was repeated 6, 5, 4, 3, and 2 times in the first, second, third, fourth, and fifth cycle, respectively. The washing fractions (streams 6, 7, and 9) were collected (stream 11) and distilled in a vacuum rotary evaporator. Assuming that residual DMSO may remain in the IL-rich fraction, an [Amim][Cl]/DMSO ratio of 90/10 (wt %/wt %) in the recovered ionic liquid was fixed. Thus, the DMSO amount was added to maintain the [Amim][Cl]/ DMSO ratio and reused in the next cycle (stream 12). CRM and LRM yields were obtained according to eqs 1 and 2, respectively: = ∑ ∑ ×= = ⎛ ⎝ ⎜⎜ ⎞ ⎠ ⎟⎟yield CRM (wt%) mass CRM mass raw wood 100n i n i ncycle 1 1 (1) = ∑ ∑ ×= = ⎛ ⎝ ⎜⎜ ⎞ ⎠ ⎟⎟yield LRM (wt%) mass LRM mass raw wood 100n i n i ncycle 1 1 (2) Additionally, nonlocated wood fractions will be defined according to eq 3: = − −yield nonlocated (wt%) 100 yield LRM yield CRM (3) [Amim][Cl] Characterization. An Agilent 1260 HPLC was used to determine the [Amim][Cl] concentration in the washing and distillate fractions. The analysis was developed using a Hi-Plex Ca column and an RI detector with a flow rate of 0.6 mL/min at 40 °C. Water was used as mobile phase, and samples were diluted at least 20 times before injecting. Attenuated total reflectance Fourier Transform infrared (FTIR/ ATR) spectra of the [Amim][Cl] samples were acquired using a Jasco 4700 spectrometer operating with a Golden Gate TM diamond accessory Specac 10542. FTIR/ATR spectra were measured between 3500 and 700 cm−1 using 64 scans and a resolution of 4 cm−1. The Spectra manager software was used for instrument management. [Amim][Cl] was also studied by proton nuclear magnetic resonance (1H NMR) in H2O-d6, which was performed in a Bruker DPX 300 MHz using 64 scans. For the analysis, 10−20 mg of each sample were diluted in 1 mL of solvent. The lignin content in the recovered [Amim][Cl] was analyzed by UV/vis spectroscopy using a Varian Cary 50 scan UV/vis spectrophotometer. Due to the strong absorbance of [Amim][Cl] in the UV region, the samples were measured at 440 nm.29 The recovered [Amim][Cl] was diluted 40 times with 0.1 N NaOH and filtered before measuring the absorbance.30 The total dissolved lignin concentration was obtained from the reference curve of the Organosolv lignin samples prepared. The lignin content in the recovered [Amim][Cl] was calculated from eq 4. = × ⎛ ⎝⎜ ⎞ ⎠⎟lignin in [Amim][Cl](wt%) mass of lignin in IL total mass of IL 100 (4) The thermal decomposition behavior of the reused [Amim][Cl] was studied and compared to a mixture of fresh [Amim][Cl]/DMSO 90/ 10 (wt %/wt %) using a TG/DSC1Mettler Toledo instrument. Each sample was heated from 40 to 105 °C to remove water and maintained for 30 min. Then, the sample was heated to 800 °C at a heating rate of 10 °C/min. Dry nitrogen was added at a flow rate of 50 mL/min and the experiments were performed in PCA/Sapphire crucibles (Mettler Toledo) of 70 μL. For each experiment, 7−9 mg of sample were used. The characteristic parameters of the TG and DTG curves were determined. CRM and LRM Characterization. FTIR spectra of CRM and LRM were recorded in a Jasco 4700 spectrometer between 2000 and 600 cm−1. Samples (1 mg) were mixed with 200 mg of KBr in an agate mortar; the resultant mixtures were pressed at 7 tonnes for 30 s and analyzed using 16 scans at 4 cm−1 of resolution. X-ray diffraction (XRD) patterns of the CRM were measured on an X’Pert Pro MPD System. The diffracted intensity of Cu Kα radiation (45 kV and 40 mA) was analyzed in a 2θ range between 5° and 50° and a step size of 0.03. All spectra were subjected to baseline correction using X’Pert HighScore Plus software. The chemical composition of LRM was calculated based on the National Renewable Energy Laboratory methodology adapted to small quantities of samples.31 LRM was treated in a two-step acid hydrolysis, resulting in an insoluble residue compound by acid-insoluble lignin and ashes and a liquid fraction containing soluble lignin and sugars. The acid insoluble lignin was gravimetrically measured after drying the residue at 105 °C for 12 h. The amount of acid-soluble lignin was determined from the UV absorbance of the hydrolysate at 240 nm and an absorptivity of 12 L·g−1·cm−1. Sugars in the hydrolysate were determined by HPLC using an RI detector. A Carbosep CHO-682 column was run at 80 °C with water as the mobile phase and a flow rate of 0.4 mL/min. The anhydrous correction to hexoses (162/180) and pentoses (132/150) were applied to each sugar concentration, respectively. Sugar recovery standards (SRS) correction was applied following the NREL/TP-510-42618 procedure. ■ RESULTS AND DISCUSSION This paper focuses on studying the ability of the [Amim][Cl] to be reused. The IL characterization and the evolution of the treatment efficiency along the reuse cycles is discussed. Furthermore, the properties of the solids obtained (CRM and LRM) as a consequence of the IL reuse grade are also examined. Recovery of [Amim][Cl]. To determine the quantity of solvent required to wash each solid and recover the majority of the IL used in the two-step process, a washing methodology was used. Figure 1 shows that [Amim][Cl] remains in both solids (CRM and LRM). In the case of CRM, most of the [Amim][Cl] is removed in the wash with methanol (82−84% of the initial IL introduced). In the case of LRM, recovery of [Amim][Cl] is observed in both washings. Although most of the [Amim][Cl] impregned into the LRM is recovered in the first wash (9−11%), there is a small quantity of [Amim][Cl] (1.3−1.7%) that is recovered in the second wash. A sequential washing was also performed with [Bmim][Cl] and cellulose by Lozano et al, who also used HPLC analysis to determine the Figure 1. [Amim][Cl] recovery. ACS Sustainable Chemistry & Engineering Research Article DOI: 10.1021/acssuschemeng.6b02723 ACS Sustainable Chem. Eng. 2017, 5, 2384−2392 2386 http://dx.doi.org/10.1021/acssuschemeng.6b02723 total [Bmim][Cl] recovered in the washing step, with IL recovery yields in the range of 95−99%.32 Previous FTIR analyses of CRM showed that, on some occasions, just a wash with methanol was not sufficient to completely remove the IL retained into the CRM. The recovery is divided into two steps. In the first step, methanol is completely removed. In the second step, conditions must be more severe to remove most of the DMSO added. After distillation, the final recovery indicates that around 92− 96% of the [Amim][Cl] introduced in this step is recovered. The high viscosity of [Amim][Cl] (2090 mPa s at 25 °C) difficult higher recovery yields in the distillation stage.33 The recovered IL analyzed by HPLC, did not show the presence of glucose, indicating that no cellulose to glucose degradation was produced during the pretreatment. To verify the negligible vapor pressure of [Amim][Cl], distilled fractions were analyzed by HPLC, showing no evidence of the presence of [Amim][Cl] in the distilled fractions. Efficiency of Recovered [Amim][Cl] in Biomass Fractionation. The relation between CRM, LRM, and nonlocated yields calculated according to eqs 1, 2, and 3, and the reuse grade is shown in Figure 2. There is a clear linear correlation between the CRM, LRM, and nonlocated fractions yields and the IL reuse grade. The LRM results show an incremental loss of the IL dissolving efficacy in consecutive cycles; the efficiency to dissolve pine wood decreases approximately 34% from the first to the fifth cycle. As a consequence of a decrease of the dissolution capacity, the decrease of the CRM yield indicates that regeneration rate is smaller. Nonlocated fractions represent 20−26% of the initial biomass introduced in the process, indicating that not all of the dissolved components are regenerated, and they presumably are being accumulated into the IL, as described below. Chemical Structure of Recovered [Amim][Cl]. The FTIR spectra of fresh and recovered [Amim][Cl] in successive cycles between 3500 and 700 cm−1 is shown in Figure 3. In the case of fresh [Amim][Cl], DMSO in a ratio of 90/10 (wt %/wt %) was added to maintain the proportion. Some bands variations were detected in the range of 3500− 3200 cm−1, indicating traces of water were accumulated in [Amim][Cl] after the recovery process.34 Typical cation bands appear at 3028 cm−1 (shoulder corresponding to the CH stretching vibration), 2963 cm−1 (CH2HCH asymmetric stretching), 1644 cm−1 (CC stretching vibration in allyl), 1577 cm−1 (asymmetric CH3(N)CN), 1424 cm−1 (CH2/CH bending vibration in side chain), 1335 cm−1 (CN stretching vibration), 1288 cm−1 (CH rocking vibration), 1166 cm−1 CH (bending vibration in imidazole ring), and 1048 cm−1 (CH3N stretching).35−37 A chloride anion vibration band appears at 760 cm−1.38 No relative differences were observed in these bands, which apparently indicates no thermal degradation of the ionic liquid occurred after the dissolution and recovery Figure 2. CRM, LRM, and nonlocated fractions yields. Figure 3. FTIR/ATR spectra of fresh and recovered [Amim][Cl]. Figure 4. XRD patterns of CRM. Figure 5. FTIR spectra of CRM. ACS Sustainable Chemistry & Engineering Research Article DOI: 10.1021/acssuschemeng.6b02723 ACS Sustainable Chem. Eng. 2017, 5, 2384−2392 2387 http://dx.doi.org/10.1021/acssuschemeng.6b02723 treatments. This assumption was also observed by Chin et al, using [Amim][Cl] in homogeneous cellulose acetylation.39 A typical DMSO band is observed at 951 cm−1 for all the spectra. Height is similar for all the samples, verifying the proportion of 90/10 (wt %/wt %) [Amim][Cl]/DMSO used in the process. 1H NMR spectra of fresh and recovered IL are shown in Figure S1. It is observed how peaks at chemical shifts of 8.65, 7.40, 5.91, 5.38, 4.75, and 3.82 correspond to the hydrogen atoms of cation position nos. 7 (1H), 6 and 5 (1H), 2 (1H), 1 (2H), 3 (1H), and 4 (3H), respectively. Peak at chemical shift of 2.65 corresponds to hydrogen atoms of DMSO position nos. 8 and 9 (3H). No other additional peaks are observed in any of the spectra, confirming no degradation of the IL. These results are in accordance with those obtained by Xu et al.11 In this case the presence of water is not observed.11 Biomass Accumulation. UV−vis analysis was performed to determine the lignin concentration in the recovered [Amim][Cl]. Portions of 0.02, 0.04, 0.07, 0.10, and 0.14 g of lignin per 100 g of IL are accumulated in the first, second, third, fourth, and fifth cycles, respectively. Lignin accumulated in the first cycle remains, and additional quantities are summed in consecutive cycles. This pattern has also been observed with other ionic liquids used for biomass pretreatment such us [Emim][OAc], or [BMIM][MeSO4]. 30,40 The addition of sulfuric acid solution to the recovered [Amim][Cl], following the strategy used to separate black liquor derived from the pulping process, results in dark particles precipitation (Figure S2a from the Supporting Information).41,42 Moreover, the addition of water results in clearer particles precipitation (Figure S2b from Supporting Information) similar to the polysaccharides precipitated in the regeneration step, visually confirming that biomass accumulation in the IL is being produced. TG characteristic parameters are shown in Table S1 (from the Supporting Information). Tonset is 255.5 °C for fresh [Amim][Cl], which is lower than reported (273 °C) due to the DMSO content.43,44 In all cases, temperatures are slightly lower in the first cycle and increase gradually up to the fifth cycle. Weerachanchai et al. already observed this behavior in the recovery of [Emim][OAc], attributing the change of degrada- tion temperature to lignin accumulation.30 In this case, the linear correlation observed for all TG characteristic parameters with nonlocated fractions yields (R2 = 0.98−0.99), elucidate that not only lignin is the cause of higher degradation temperatures, but also carbohydrates. The T50% for organosolv lignin and Tonset for Avicel PH-101, both used as reference materials, are 376.5 and 311 °C, both higher than those obtained in the IL. Characterization of CRM. The process developed to obtain a CRM was already detailed by Casas et al., who deeply characterized the regenerated solid to confirm that this is composed by polysaccharides.13 XRD diffraction reveals the amorphicity of all CRM solids (Figure 4). According to Seagal’s method, the crystallinity index must be calculated as a ratio between amorphous cellulose at approximately 2θ = 16.6° and crystalline cellulose at approximately 2θ = 22°.45 However, both characteristic peaks of cellulose I and the reflection peak at 2θ = 34.5° only appear in the untreated pine wood. In all CRM samples, a shift in the crystalline cellulose peak is produced from 2θ ≈ 22 to 20. This new peak is typical of cellulose II.46,47 The width of this amorphous peak at half-height is higher when the reuse grade increases (from the first to the fifth cycle, fwhm values are 5.79°, 6.49°, 7.08°, 7.04°, and 7.25°, respectively). The peak at 2θ = 34.5° shifts to 2θ = 35.8°. The XRD data reveals that CRM amorphicity is higher, when the [Amim][Cl] reuse grade increases.48 This tendency may be a consequence of both dissolution and the regeneration efficiency step, involving the regeneration of more amorphous CRM when the dissolving efficacy decreases (increase of the reuse grade). The FTIR spectra of the CRM solids are displayed in Figure 5. Characteristic cellulose bands are present in all CRMs. Bands at 1335 and 1317 cm−1 (corresponding to the OH in plane deformation and CH2 rocking vibration) are affected by the crystallinity decrease and hydrogen bonds breakage during the dissolution process.49 The band at 1160 cm−1 (COC stretching asymmetry) shows the cellulose vibrational energy changes as a consequence of the wood dissolution process.13 Bands at 1110 cm−1 (ring asymmetric stretching) and 897 cm−1 (ring stretching) confirm the crystallinity decrease and amorphicity increase already exposed with the XRD results. Furthermore, an intense band at 1635 cm−1, characteristic of the OH of water absorbed from cellulose, is also observed in the CRM solids.13,50 There are three bands that are not present in MCC. The band at 1735 cm−1 corresponds to CO esters, characteristic of hemi- celluloses. The band at 1262 cm−1 is a CO stretching band Table 2. LRM Composition (mg/100 mg dry pine wood) Lignin Rich Material lignin glucan xylan galactan arabinan mannan pine wood 34.5 ± 1.5 42.2 ± 0.9 3.9 ± 0.1 1.7 ± 0.2 0.6 ± 0.0 13.3 ± 1.4 1st cycle 53.3 ± 1.2 16.0 ± 1.5 7.3 ± 0.3 5.1 ± 0.5 0.4 ± 0.1 15.3 ± 1.5 2nd cycle 43.1 ± 0.1 28.1 ± 0.4 6.4 ± 0.1 3.8 ± 0.1 0.5 ± 0.0 12.5 ± 0.7 3rd cycle 41.4 ± 2.0 27.8 ± 0.2 7.8 ± 0.0 3.9 ± 0.1 0.6 ± 0.1 12.9 ± 0.7 4th cycle 42.7 ± 0.8 28.5 ± 0.2 7.1 ± 0.4 3.3 ± 0.2 0.6 ± 0.0 13.3 ± 0.1 5th cycle 40.8 ± 0.2 28.4 ± 0.2 7.1 ± 0.6 3.3 ± 0.2 0.7 ± 0.1 13.3 ± 1.4 Figure 6. FTIR spectra of LRM. ACS Sustainable Chemistry & Engineering Research Article DOI: 10.1021/acssuschemeng.6b02723 ACS Sustainable Chem. Eng. 2017, 5, 2384−2392 2388 http://pubs.acs.org/doi/suppl/10.1021/acssuschemeng.6b02723/suppl_file/sc6b02723_si_001.pdf http://pubs.acs.org/doi/suppl/10.1021/acssuschemeng.6b02723/suppl_file/sc6b02723_si_001.pdf http://pubs.acs.org/doi/suppl/10.1021/acssuschemeng.6b02723/suppl_file/sc6b02723_si_001.pdf http://pubs.acs.org/doi/suppl/10.1021/acssuschemeng.6b02723/suppl_file/sc6b02723_si_001.pdf http://dx.doi.org/10.1021/acssuschemeng.6b02723 derived from arabino-4-O-methylglucuronoxylans, predominant in pine wood, and the band at 804 cm−1 is due to glucomannan.50,51 These bands confirm the presence of hemicelluloses in the CRM. No characteristic lignin bands have been found. Additionally, there is no presence of [Amim][Cl], indicating that the washing step has removed the ionic liquid in the CRM solid. Characterization of LRM. The composition of LRM is determined according to the NREL methodology and exhibited in Table 2. Glucan content decreases significantly, revealing the efficacy of [Amim][Cl] as cellulose solvent.16 Note that the glucan content increases considerably from the first to the second cycle. In the first cycle (process developed with fresh [Amim][Cl]), there is only 16.0 mg of glucan/100 mg dry pine wood, whereas in successive cycles, this quantity increase to approximately 28 mg of glucan/100 mg of pine wood. The largest difference produced between the first and the second cycle shows the loss of efficacy with the reused [Amim][Cl]. However, there is still a notable difference between this solid composition and the untreated pine wood (42.2 mg of glucan/ 100 mg dry pine wood). This behavior has not been observed before, as most of the IL recycling processes are focused on biomass pretreatment prior to enzymatic hydrolysis. Separation between soluble and not soluble biomass in the IL has not been deeply studied. Lignin content increases considerably indicating that although the cellulose is dissolved, lignin remains. There is a notable difference between the untreated pine wood lignin content, the first cycle lignin content, and the second cycle lignin content (34.5, 53.3, and 43.1 mg of lignin/100 mg of dry pine wood). Pentoses differences between untreated and treated wood are observed mainly in the xylan, probably due to a decrease in glucan content that causes an increase in other components. Xylan content varies between 7.3 and 6.4 mg of xylan/100 mg dry pine wood without an appreciable tendency. Galactan content increases from the untreated wood (1.7 mg of galactan/100 mg dry pine wood) to the first cycle treated wood (5.1 mg galactan/100 mg dry pine wood). After the first cycle the content decreses slightly, probably as a consequence of glucose total content increase. The mannan concentration in pine wood is also relevant (13.3 mg of mannan/100 mg dry pine wood). The use of fresh and recovered [Amim][Cl] in mannan has a similar tendency to galactan. FTIR spectra have been developed in all LRM samples (Figure 6), confirming lignin, cellulose, and hemicellulose content. A fast comparison shows that all bands of the untreated pine wood are present in the LRM FTIR spectra. Bands at 1597 cm−1, corresponding to the aromatic skeletal vibration with CO stretching, and at 1424 cm−1, character- istic of aromatic skeletal vibrations combined with CH in plane deformation, have approximately the same intensity in all the samples.52 The band at 1510 cm−1, characteristic of aromatic skeletal vibration guaiacyl > syringyl, seems to be more intense after the first cycle LRM sample and decreases in the subsequent LRM Table 3. Mass Balance (mg/1000 mg of [Amim][Cl]) stream 1 2 3 + 4 + 5 6 7 8 9 10 11 12 description initial DMSO methanol LRM 1st wash LRM 2nd wash LRM CRM wash CRM washing fractions recovered IL First Cycle IL 1000.0 0.0 0.0 85.4 17.7 0.0 830.9 0.0 933.9 880.2 DMSO 111.1 1375.0 0.0 98.8 33.8 0.0 1242.7 0.0 1375.3 1356.1 methanol 0.0 0.0 19800.0 1403.8 1666.1 0.0 13080.0 0.0 16150.0 0.0 carbohydrates 25.6 0.0 0.0 8.9 9.9 lignin 14.3 0.0 0.0 10.7 0.0 0.2 Second Cycle IL 1000.0 0.0 0.0 100.5 13.9 0.0 772.2 0.0 886.6 862.4 DMSO 111.1 1375.0 0.0 94.0 11.0 0.0 884.6 0.0 989.6 320.4 methanol 0.0 0.0 19800.0 1708.9 1730.5 0.0 13126.4 0.0 16565.7 0.0 carbohydrates 25.7 0.0 0.0 11.8 9.5 lignin 14.5 0.0 0.0 9.8 0.0 0.4 Third Cycle IL 1000.0 0.0 0.0 102.0 17.9 0.0 872.9 0.0 992.8 870.5 DMSO 111.1 1375.0 0.0 110.5 15.7 0.0 1075.3 0.0 1201.5 308.8 methanol 0.0 0.0 19800.0 1919.9 2075.4 0.0 15716.8 0.0 19712.2 0.0 carbohydrates 25.7 0.0 0.0 13.3 8.7 lignin 14.7 0.0 0.0 10.4 0.0 0.7 Fourth Cycle IL 1000.0 0.0 0.0 99.8 13.6 0.0 846.7 0.0 960.1 911.1 DMSO 111.1 1375.0 0.0 116.1 14.2 0.0 1015.4 0.0 1145.7 459.7 methanol 0.0 0.0 19800.0 2194.4 2368.9 0.0 16293.1 0.0 20856.4 0.0 carbohydrates 25.7 0.0 0.0 0.0 7.8 lignin 14.4 0.0 0.0 0.0 0.0 1.0 Fifth Cycle IL 1000.0 0.0 0.0 108.4 15.5 0.0 824.3 0.0 948.2 913.1 DMSO 111.1 1375.0 0.0 123.8 17.0 0.0 1087.8 0.0 1228.6 539.2 methanol 0.0 0.0 19800.0 2264.6 2607.4 0.0 16854.7 0.0 21726.7 0.0 carbohydrates 25.7 0.0 0.0 14.9 7.1 lignin 14.4 0.0 0.0 11.5 0.0 1.4 ACS Sustainable Chemistry & Engineering Research Article DOI: 10.1021/acssuschemeng.6b02723 ACS Sustainable Chem. Eng. 2017, 5, 2384−2392 2389 http://dx.doi.org/10.1021/acssuschemeng.6b02723 samples, becoming less intense in the untreated pine wood sample. This change verifies the overall lignin concentration obtained in each sample already explained in Table 2. Finally, the guaiacyl ring breathing with the carbonyl stretching band at 1269 cm−1 also increases in the LRM first cycle sample, decreasing in the following cycles as a consequence of the difficulty of dissolving guaiacyl units characteristics of softwoods.52,53 The band at 1735 cm−1 typical of hemicellulose remains almost unaltered in all spectra, and the band at 1262 cm−1 is not observed.50,51 However, the band at 804 cm−1 characteristic of glucomannan, one of the most common hemicelluloses in pine, is observed to slightly decrease.54 Regarding cellulose bands, the diminution of the intensity at 1160 cm−1 confirms the reduction of cellulose content in LRM. No presence of the ionic liquid characteristic bands was observed (e.g., at 1577 cm−1). Global Mass Balance. Mass balance of the main components used in the overall process is shown in Table 3. Streams numbers are indicated in Scheme. 1. Pine wood fractions have been divided into two main components: lignin and polysaccharides (formed by cellulose and hemicellulose). Table 3 summarizes the calculations developed in the whole process with chromatographic, gravimetric, and NREL analysis techniques. In the whole mass balance, the differences between the initial wood introduced and the biomass obtained after the treatment demonstrate the considerable loss of carbohydrates as well as lignin that have not been located into the process and that presumably are being accumulated in the IL as some analytical techniques have confirmed. In the first cycle, losses for carbohydrates are around 26% and decrease significantly in the subsequent cycles (loss of approximately 14−16%). The accumulation of woody fractions is gradual and will reach a point that the removal of impurities will be necessary to continue the fractionation process. The addition of an antisolvent after certain number of IL reuse cycles may regenerate the IL and increase again its fractionation efficiency. ■ CONCLUSIONS Fresh and recycled [Amim][Cl] used in the fractionation of Pinus radiata wood have been evaluated. The [Amim][Cl] employed along the process has been recovered, obtaining yields of up to 96%. From the first to the fifth [Amim][Cl] recovery cycle, a gradual decrease of the [Amim][Cl] capacity to dissolve pine wood has been observed. As a result, the LRM composition reveals differences, mainly produced between the solids obtained after the first (fresh IL) and the second IL recovery cycle, with 42.2 and 16.0 g of glucan/100 g of dry pine wood, respectively. CRM regeneration yields have also been reduced, while solids amorphicity is increased in successive IL recovery cycles. Note that gradual accumulation of lignin, and presumably carbohydrates are the causes of the decrease in pine wood fractionation yields. Despite of this reduction, the [Amim][Cl] chemical structure is not altered in the overall process. In future work, the addition of an antisolvent after a certain number of cycles to regenerate the ionic liquid, remove biomass impurities, and clean the recovered ionic liquid should be studied. ■ ASSOCIATED CONTENT *S Supporting Information The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acssusche- meng.6b02723. Figures S1 and 2 and Table S1 as mentioned in the text (PDF) ■ AUTHOR INFORMATION Corresponding Author *Tel.: (0034) 913948505. E-mail: vicrigua@ucm.es (V.R.). ORCID Victoria Rigual: 0000-0001-7216-1506 Notes The authors declare no competing financial interest. ■ ACKNOWLEDGMENTS This work was performed thanks to the financial support of the “Ministerio de Economıá y Competitividad” under the funded project CTQ2013-42006-R. ■ ABBREVIATIONS CRM, cellulose rich material; LRM, lignin rich material; IL, ionic liquid; [Amim][Cl], 1-allyl-3-methylimidazolium chloride ■ REFERENCES (1) Lynam, J. G.; Chow, G. I.; Hyland, P. L.; Coronella, C. J. Corn stover pretreatment by ionic liquid and glycerol mixtures with their density, viscosity, and thermogravimetric properties. ACS Sustainable Chem. Eng. 2016, 4 (7), 3786−3793. (2) Zhu, C.; Richardson, R. M.; Potter, K. D.; Koutsomitopoulou, A. F.; Van Duijneveldt, J. S.; Vincent, S. R.; Wanasekara, N. D.; Eichhorn, S. J.; Rahatekar, S. S. High modulus regenerated cellulose fibers spun from a low molecular weight microcrystalline cellulose solution. ACS Sustainable Chem. Eng. 2016, 4 (9), 4545−4553. (3) Du, X.; Lucia, L. A.; Ghiladi, R. A. Development of a highly efficient pretreatment sequence for the enzymatic saccharification of loblolly pine wood. ACS Sustainable Chem. Eng. 2016, 4 (7), 3669− 3678. (4) Tian, X.; Rehmann, L.; Xu, C. C.; Fang, Z. Pretreatment of Eastern White Pine (Pinus strobes L.) for Enzymatic Hydrolysis and Ethanol Production by Organic Electrolyte Solutions. ACS Sustainable Chem. Eng. 2016, 4 (5), 2822−2829. (5) Elgharbawy, A. A.; Alam, M. Z.; Moniruzzaman, M.; Goto, M. Ionic liquid pretreatment as emerging approaches for enhanced enzymatic hydrolysis of lignocellulosic biomass. Biochem. Eng. J. 2016, 109, 252−267. (6) Singh, R.; Krishna, B. B.; Kumar, J.; Bhaskar, T. Opportunities for utilization of non-conventional energy sources for biomass pretreat- ment. Bioresour. Technol. 2016, 199, 398−407. (7) Xu, J.; Zong, M.-H.; Fu, S.-Y.; Li, N. Correlation between Physicochemical Properties and Enzymatic Digestibility of Rice Straw Pretreated with Cholinium Ionic Liquids. ACS Sustainable Chem. Eng. 2016, 4 (8), 4340−4345. (8) Xu, J.; Xiong, P.; He, B. Advances in improving the performance of cellulase in ionic liquids for lignocellulose biorefinery. Bioresour. Technol. 2016, 200, 961−970. (9) Weerachanchai, P.; Lee, J. M. Effect of organic solvent in ionic liquid on biomass pretreatment. ACS Sustainable Chem. Eng. 2013, 1 (8), 894−902. (10) Hu, D.; Xiao, L.; Li, L.; Zhong, C.; Ju, X.; Yan, L.; Wu, T.; Qing, M.; Hu, Z. Effects of Ionic Liquid 1-Ethyl-3-Methylimidazolium Diethylphosphate on Cellulase Produced by Paenibacillus sp. LLZ1. ACS Sustainable Chem. Eng. 2016, 4 (9), 4922−4926. ACS Sustainable Chemistry & Engineering Research Article DOI: 10.1021/acssuschemeng.6b02723 ACS Sustainable Chem. Eng. 2017, 5, 2384−2392 2390 http://pubs.acs.org http://pubs.acs.org/doi/abs/10.1021/acssuschemeng.6b02723 http://pubs.acs.org/doi/abs/10.1021/acssuschemeng.6b02723 http://pubs.acs.org/doi/suppl/10.1021/acssuschemeng.6b02723/suppl_file/sc6b02723_si_001.pdf mailto:vicrigua@ucm.es http://orcid.org/0000-0001-7216-1506 http://dx.doi.org/10.1021/acssuschemeng.6b02723 (11) Xu, J. K.; Sun, Y. C.; Sun, R. C. Ionic liquid pretreatment of woody biomass to facilitate biorefinery: Structural elucidation of alkali- soluble hemicelluloses. ACS Sustainable Chem. Eng. 2014, 2 (4), 1035−1042. (12) George, A.; Brandt, A.; Tran, K.; Zahari, S. M. S. N. S.; Klein- Marcuschamer, D.; Sun, N.; Sathitsuksanoh, N.; Shi, J.; Stavila, V.; Parthasarathi, R.; et al. Design of low-cost ionic liquids for lignocellulosic biomass pretreatment. Green Chem. 2015, 17 (3), 1728−1734. (13) Casas, A.; Alonso, M. V.; Oliet, M.; Santos, T. M.; Rodriguez, F. Characterization of cellulose regenerated from solutions of pine and eucalyptus woods in 1-allyl-3-methilimidazolium chloride. Carbohydr. Polym. 2013, 92 (2), 1946−1952. (14) Ha, S. H.; Mai, N. L.; An, G.; Koo, Y.-M. Microwave-assisted pretreatment of cellulose in ionic liquid for accelerated enzymatic hydrolysis. Bioresour. Technol. 2011, 102 (2), 1214−1219. (15) Sun, N.; Rodríguez, H.; Rahman, M.; Rogers, R. D. Where are ionic liquid strategies most suited in the pursuit of chemicals and energy from lignocellulosic biomass? Chem. Commun. 2011, 47 (5), 1405−1421. (16) Zhang, H.; Wu, J.; Zhang, J.; He, J. 1-Allyl-3-methylimidazolium Chloride Room Temperature Ionic Liquid: A New and Powerful Nonderivatizing Solvent for Cellulose. Macromolecules 2005, 38 (20), 8272−8277. (17) Geng, X.; Henderson, W. A. Impact of non-solvents on the structural features and enzymatic digestibility of cellulose regenerated from an ionic liquid. RSC Adv. 2014, 4 (59), 31226−31229. (18) Weerachanchai, P.; Lim, K. H.; Lee, J.-M. Influence of organic solvent on the separation of an ionic liquid from a lignin−ionic liquid mixture. Bioresour. Technol. 2014, 156, 404−407. (19) Doherty, T. V.; Mora-Pale, M.; Foley, S. E.; Linhardt, R. J.; Dordick, J. S. Ionic liquid solvent properties as predictors of lignocellulose pretreatment efficacy. Green Chem. 2010, 12 (11), 1967−1975. (20) Fang, Z., Smith, R. L., Jr., Qi, X., Eds. Production of Biofuels and Chemicals with Ionic Liquids; Springer: Dordrecht, Heidelberg, New York, London, 2014. (21) Stenutz, R. Tables for Organic Chemistry. http://www.stenutz. eu/chem/ (accessed February 7, 2017). (22) Auxenfans, T.; Buchoux, S.; Husson, E.; Sarazin, C. Efficient enzymatic saccharification of Miscanthus: Energy-saving by combining dilute acid and ionic liquid pretreatments. Biomass Bioenergy 2014, 62, 82−92. (23) Auxenfans, T.; Buchoux, S.; Djellab, K.; Avondo, C.; Husson, E.; Sarazin, C. Mild pretreatment and enzymatic saccharification of cellulose with recycled ionic liquids towards one-batch process. Carbohydr. Polym. 2012, 90 (2), 805−813. (24) Mai, N. L.; Ahn, K.; Koo, Y.-M. Methods for recovery of ionic liquidsA review. Process Biochem. 2014, 49 (5), 872−881. (25) Lynam, J. G.; Chow, G. I.; Coronella, C. J.; Hiibel, S. R. Ionic liquid and water separation by membrane distillation. Chem. Eng. J. 2016, 288, 557−561. (26) Auxenfans, T.; Buchoux, S.; Larcher, D.; Husson, G.; Husson, E.; Sarazin, C. Enzymatic saccharification and structural properties of industrial wood sawdust: Recycled ionic liquids pretreatments. Energy Convers. Manage. 2014, 88, 1094−1103. (27) Clough, M. T.; Geyer, K.; Hunt, P. A.; Son, S.; Vagt, U.; Welton, T. Ionic liquids: not always innocent solvents for cellulose. Green Chem. 2015, 17 (1), 231−243. (28) Sluiter, A.; Hames, B.; Ruiz, R.; Scarlata, C.; Sluiter, J.; Templeton, D.; Crocker, D. Determination of Structural carbohydrates and lignin in biomass; NREL, 2011. (29) Kline, L. M.; Hayes, D. G.; Womac, A. R.; Labbe, N. Simplified determination of lignin content in hard and soft woods via UV- spectrophotometric analysis of biomass dissolved in ionic liquids. BioResources 2010, 5 (3), 1366−1383. (30) Weerachanchai, P.; Lee, J.-M. Recyclability of an ionic liquid for biomass pretreatment. Bioresour. Technol. 2014, 169, 336−343. (31) Ibañ́ez, A. B.; Bauer, S. Downscaled method using glass microfiber filters for the determination of Klason lignin and structural carbohydrates. Biomass Bioenergy 2014, 68, 75−81. (32) Lozano, P.; Bernal, B.; Recio, I.; Belleville, M.-P. A cyclic process for full enzymatic saccharification of pretreated cellulose with full recovery and reuse of the ionic liquid 1-butyl-3-methylimidazolium chloride. Green Chem. 2012, 14 (9), 2631−2637. (33) Badgujar, K. C.; Bhanage, B. M. Factors governing dissolution process of lignocellulosic biomass in ionic liquid: Current status, overview and challenges. Bioresour. Technol. 2015, 178, 2−18. (34) Kathirgamanathan, K.; Grigsby, W. J.; Al-Hakkak, J.; Edmonds, N. R. Two-Dimensional FTIR as a Tool to Study the Chemical Interactions within Cellulose-Ionic Liquid Solutions. Int. J. Polym. Sci. 2015, 2015, 1. (35) Liu, L.; Ju, M.; Li, W.; Jiang, Y. Cellulose extraction from Zoysia japonica pretreated by alumina-doped MgO in AMIMCl. Carbohydr. Polym. 2014, 113, 1−8. (36) FitzPatrick, M.; Champagne, P.; Cunningham, M. F. Quantitative determination of cellulose dissolved in 1-ethyl-3- methylimidazolium acetate using partial least squares regression on FTIR spectra. Carbohydr. Polym. 2012, 87 (2), 1124−1130. (37) Moumene, T.; Belarbi, E. H.; Haddad, B.; Villemin, D.; Abbas, O.; Khelifa, B.; Bresson, S. Study of imidazolium dicationic ionic liquids by Raman and FTIR spectroscopies: The effect of the nature of the anion. J. Mol. Struct. 2015, 1083, 179−186. (38) Li, W.-Z.; Ju, M.-T.; Wang, Y.-N.; Liu, L.; Jiang, Y. Separation and recovery of cellulose from Zoysia japonica by 1-allyl-3- methylimidazolium chloride. Carbohydr. Polym. 2013, 92 (1), 228− 235. (39) Huang, K.; Wu, R.; Cao, Y.; Li, H.; Wang, J. Recycling and Reuse of Ionic Liquid in Homogeneous Cellulose Acetylation. Chin. J. Chem. Eng. 2013, 21 (5), 577−584. (40) Prado, R.; Erdocia, X.; Labidi, J. Study of the influence of reutilization ionic liquid on lignin extraction. J. Cleaner Prod. 2016, 111 (Part A), 125−132. (41) Tan, S. S. Y.; MacFarlane, D. R.; Upfal, J.; Edye, L. A.; Doherty, W. O. S.; Patti, A. F.; Pringle, J. M.; Scott, J. L. Extraction of lignin from lignocellulose at atmospheric pressure using alkylbenzenesulfo- nate ionic liquid. Green Chem. 2009, 11 (3), 339−345. (42) Liu, Z.; Fatehi, P.; Jahan, M. S.; Ni, Y. Separation of lignocellulosic materials by combined processes of pre-hydrolysis and ethanol extraction. Bioresour. Technol. 2011, 102 (2), 1264−1269. (43) Zhang, H.; Wu, J.; Zhang, J.; He, J. 1-Allyl-3-methylimidazolium chloride room temperature ionic liquid: a new and powerful nonderivatizing solvent for cellulose. Macromolecules 2005, 38 (20), 8272−8277. (44) Hao, Y.; Peng, J.; Hu, S.; Li, J.; Zhai, M. Thermal decomposition of allyl-imidazolium-based ionic liquid studied by TGA−MS analysis and DFT calculations. Thermochim. Acta 2010, 501 (1), 78−83. (45) Kumar, R.; Mago, G.; Balan, V.; Wyman, C. E. Physical and chemical characterizations of corn stover and poplar solids resulting from leading pretreatment technologies. Bioresour. Technol. 2009, 100 (17), 3948−3962. (46) Blokhin, A. V.; Voitkevich, O. V.; Kabo, G. J.; Paulechka, Y. U.; Shishonok, M. V.; Kabo, A. G.; Simirsky, V. V. Thermodynamic properties of plant biomass components. Heat capacity, combustion energy, and gasification equilibria of cellulose. J. Chem. Eng. Data 2011, 56 (9), 3523−3531. (47) Lan, W.; Liu, C.-F.; Yue, F.-X.; Sun, R.-C.; Kennedy, J. F. Ultrasound-assisted dissolution of cellulose in ionic liquid. Carbohydr. Polym. 2011, 86 (2), 672−677. (48) Adel, A. M.; El-Wahab, Z. H. A.; Ibrahim, A. A.; Al-Shemy, M. T. Characterization of microcrystalline cellulose prepared from lignocellulosic materials. Part II: Physicochemical properties. Carbo- hydr. Polym. 2011, 83 (2), 676−687. (49) Colom, X.; Carrillo, F. Crystallinity changes in lyocell and viscose-type fibres by caustic treatment. Eur. Polym. J. 2002, 38 (11), 2225−2230. ACS Sustainable Chemistry & Engineering Research Article DOI: 10.1021/acssuschemeng.6b02723 ACS Sustainable Chem. Eng. 2017, 5, 2384−2392 2391 http://www.stenutz.eu/chem/ http://www.stenutz.eu/chem/ http://dx.doi.org/10.1021/acssuschemeng.6b02723 (50) Carrillo, F.; Colom, X.; Sunol, J. J.; Saurina, J. Structural FTIR analysis and thermal characterisation of lyocell and viscose-type fibres. Eur. Polym. J. 2004, 40 (9), 2229−2234. (51) Schwanninger, M.; Rodrigues, J. C.; Pereira, H.; Hinterstoisser, B. Effects of short-time vibratory ball milling on the shape of FT-IR spectra of wood and cellulose. Vib. Spectrosc. 2004, 36 (1), 23−40. (52) Casas, A.; Alonso, M. V.; Oliet, M.; Rojo, E.; Rodríguez, F. FTIR analysis of lignin regenerated from Pinus radiata and Eucalyptus globulus woods dissolved in imidazolium-based ionic liquids. J. Chem. Technol. Biotechnol. 2012, 87 (4), 472−480. (53) Casas, A.; Palomar, J.; Alonso, M. V.; Oliet, M.; Omar, S.; Rodriguez, F. Comparison of lignin and cellulose solubilities in ionic liquids by COSMO-RS analysis and experimental validation. Ind. Crops Prod. 2012, 37 (1), 155−163. (54) Gírio, F. M.; Fonseca, C.; Carvalheiro, F.; Duarte, L. C.; Marques, S.; Bogel-Łukasik, R. Hemicelluloses for fuel ethanol: a review. Bioresour. Technol. 2010, 101 (13), 4775−4800. ACS Sustainable Chemistry & Engineering Research Article DOI: 10.1021/acssuschemeng.6b02723 ACS Sustainable Chem. Eng. 2017, 5, 2384−2392 2392 http://dx.doi.org/10.1021/acssuschemeng.6b02723 S1 Supporting information Recovery and Reuse of 1-Allyl-3-methylimidazolium Chloride in the Fractionation of Pinus radiata Wood Victoria Rigual, Tamara M. Santos, Juan Carlos Domínguez, M. Virginia Alonso, Mercedes Oliet and Francisco Rodriguez Number of pages: 3 Number of Figures: 2 Number of Tables: 1 S2 Figure S1: 1 H-NMR spectra of fresh IL/DMSO mixture (a) and recovered ionic liquid in 1st to 5th cycle (b-f) Fig. S2: Photographs of a) recovered IL + H2SO4 and b) recovered IL + water a) b) S3 Table S1: TG parameters of recovered [Amim][Cl] Reuse grade IL/DMSO Tonset (°C) T50% (°C) Tmax (°C) Fresh IL/DMSO 255.5 261.7 266.9 1 st cycle 256.3 261.1 266.0 2 nd cycle 259.9 264.4 268.6 3 rd cycle 264.7 267.8 271.4 4 th cycle 268.2 270.7 272.2 5 th cycle 270.4 272.9 273.8 PUBLICACIÓN IV “Application of microscopy techniques for a better understanding of biomass pretreatment” Victoria Rigual, Juan C. Domínguez, Sandra Rivas, Antonio Ovejero-Pérez, M. Virginia Alonso, Mercedes Oliet y Francisco Rodriguez Enviado a Industrial Crops and Products 1 Application of microscopy techniques for a better understanding of biomass pretreatment Victoria Rigual*, Juan C. Domínguez, Sandra Rivas, Antonio Ovejero-Pérez, M. Virginia Alonso, Mercedes Oliet and Francisco Rodriguez Department of Chemical Engineering and Materials Science, Faculty of Chemistry, Complutense University of Madrid, Avda. Complutense s/n, 28040 Madrid, Spain. *E-mail: vicrigua@ucm.es ; Tel: (+34)913948505 ABSTRACT In this work, the application of digital image analysis to the results obtained by two microscopy techniques, confocal fluorescence microscopy and scanning electron microscopy, is developed to characterize two pretreated biomasses (Pinus radiata and Eucalyptus globulus). After images acquisition with SEM and confocal fluorescence microscopy, image processing methodologies are proposed to determine equivalent diameters distribution (using x100 SEM micrographs), fractal dimension and lacunarity (using x400 SEM micrographs), and surface composition (using confocal fluorescence microscopy). Particles sizes distributions, with median equivalent diameters in the range 7.9-15.4 μm, follows an exponential tendency and are strongly affected by the severity of the autohydrolysis pretreatment severity, as well as the biomass type. Biomass surface composition (in terms of holocellulose/lignin ratios) is not the same than the obtained with standard protocols (NREL) and let distinguish the lignin deposition in the surface, especially in severe autohydrolysis conditions with pine. Fractal dimension (in the range of 2.65- 2.74) and lacunarity (0.05-0.16) have been found to have a strong dependence on the solid accessibility (linear correlations of 0.93 and 0.70) for the two biomasses independently of the pretreatment employed. Keywords: Image processing, scanning electron microscopy, accessibility, fractal dimension, lacunarity, ImageJ. mailto:vicrigua@ucm.es 2 1. Introduction Lignocellulosic biomass is an abundant and cost-effective-renewable resource with a production of 15-17x1017 Mt annually (Chen et al., 2017a). In order to convert native biomass into a form effective to enzymatic hydrolysis, pretreatments are necessary (Agbor et al., 2011). The goal of any pretreatment technology is to alter or remove structural and compositional impediments to hydrolysis, in order to improve the rate of enzyme hydrolysis and increase fermentable sugars yields of cellulose or hemicellulose (Miyamoto et al., 2018). In this sense, some of the key factors identified that increase accessibility are the crystallinity of cellulose, hemicellulose disruption, accessible surface area (porosity), lignin protection, and association of hemicellulose to lignin (Putro et al., 2016). Another factor that strongly affects cellulose digestibility is the porous structure, as pores define pathways that enhance cellulases reach cellulose (Arantes and Saddler, 2010). In order to measure all these key aspects for biomass accessibility, scanning electron microscopy, light microscopy, infrared and Raman spectroscopy have been employed (Li et al., 2011, Rigual et al., 2018a, Singh et al., 2015). Two of the most employed techniques are confocal fluorescence microscopy and scanning electron microscopy (SEM). Confocal fluorescence microscopy enhances the understanding of biomass deconstruction and lignin changes through the visualization of the cell wall modification. Furthermore, this technique let visualize the spatial and temporal interactions between enzymes and substrate (Ding et al., 2012, Donaldson and Vaidya, 2017). From the other side, SEM is one of the most powerful tools employed in surface characterization, morphology, and analysis of microstructures (Moradi et al., 2013; Karimi and Taherzadeh, 2016). Despite of the potential of these microscopy techniques, in most of the cases they are only employed qualitatively (Gonçalves et al., 2015, Haykir et al., 2013, Karimi and Taherzadeh, 2016, Luterbacher et al., 2015, Moniz et al., 2013). The employment of digital image processing enhances the quantification of the visually observed parameters and extract additional information from micrographs (Chen et al., 2017b). Software tools, such as ImageJ or Matlab have been studied in a wide range of disciplines to extract quantitative parameters that describe micrographs (Hadzieva et al., 2015, Santacruz-Vazquez et al., 2015, Utrilla-Coello et al., 2013). Particle size distribution is one of the most common extracted information from micrographs, e.g. to determine its effect on biomass hydrolysis, or to observe the pseudo-lignin formation during certain pretreatments (Momoh and Ouki, 2018, Shinde et al., 2018). Another interesting approach is the employment of fractal parameters. Fractal models were introduced in the analysis of forests, and synthetic aperture radar imagery in 60s (Henebry and Kux, 1995, Mandelbrot and Pignoni, 1983, Plotnick et al., 1996). Techniques developed in fractal models enable the research 3 of rough details, which are based on mathematical treatments and allows obtaining several parameters from images. The concept of fractal dimension is applied to non-traditional patterns to define the inherent (scale independent) degree of complexity of a micrograph (Mandelbrot and Pignoni, 1983). However, this parameter cannot discriminate textures and natural surfaces (Cheng, 1997, Hanen et al., 2009). In complement to the fractal dimension, Mandelbrot and Pignoni (1983) introduced the concept of lacunarity, an index that indicates the homogeneity degree of the studied structure from a multiscale point of view (Mandelbrot and Pignoni, 1983). These two interdisciplinary concepts have been barely applied to pretreated biomasses, and their calculation has been based on porosimetry techniques instead of image analysis (Chi et al., 2017, Zhao and Chen, 2013). In this work, image-processing techniques are proposed for the quantification of SEM and confocal fluorescence microscopy micrographs of pretreated hardwood and softwood. Image acquisition at different magnifications will be developed and consequently treated following the methodology proposed. Particles sizes distributions, textural parameters (fractal dimension and laccunarity) and surface chemistry parameters will be related to biomass specie and type and severity of the biomass pretreatment. Finally, microscopy parameters will be correlated with enzymatic accessibility, to determine the relation between every parameter in the search of a global factor that measure the material digestibility. 2. Materials and methods 2.1. Materials and reagents Biomass samples were obtained from untreated pine and eucalyptus wood provided by the Instituto Nacional de Investigación y Tecnología Agraria (CIFOR-INIA). These samples were autohydrolyzed (AH) and ionic liquid pretreated with 1-ethyl-3-methylimidazolium acetate (IL) as explained in a previous work (Rigual et al., 2018b). Pretreated biomass was saccharified at low solids loadings according to Table 1. Consequently, samples were sieved below 150 μm and stored at 40 °C. No milling was run to ensure keeping the sample morphology obtained during the pretreatment. 4 Table 1: Pretreated biomass conditions and digestibility Biomass Pretreatment Temperature Severity factor 72 h Digestibility AH150E Eucalyptus Autohydrolysis 150 °C 2.95 30.95±0.11 AH175E Eucalyptus Autohydrolysis 175 °C 3.69 53.34±0.14 AH200E Eucalyptus Autohydrolysis 200 °C 4.42 72.73±1.26 AH150P Pine Autohydrolysis 150 °C 2.95 29.61±0.62 AH175P Pine Autohydrolysis 175 °C 3.99 18.71±0.25 AH200P Pine Autohydrolysis 200 °C 4.90 10.42±0.42 IL80E Eucalyptus Ionic liquid 80 °C 55.18±0.00 IL120E Eucalyptus Ionic liquid 120 °C 67.64±1.27 IL80P Pine Ionic liquid 80 °C 72.16±1.31 IL120P Pine Ionic liquid 120 °C 78.37±0.85 2.2. Images acquisition Scanning electron microscopy was employed to acquire micrographs. Before analysis samples were coated with gold to impart conductivity with a Q150T Turbo-pumped sputter Coater. A Jeol JSM 6400 operating at 20 kV was used. Analyses were performed in the technical facilities of the Spanish National Centre for Electron Microscopy. X100 micrographs were acquired in fast mode to study particle size distribution, while x400 micrographs were acquired in medium mode to study textural parameters (fractal dimensions and lacunarity). Confocal fluorescence microscopy was employed to visualize holocellulose (cellulose+hemicellulose) and lignin (autofluorescent). Holocellulose was differentially dyed with an aqueous solution of Calcofluor White Stain (1% v/v) at a concentration of 20mg of biomass/mL of Calcofluor solution. The suspension was mixed (for 5 minutes manually) and centrifuged to remove the supernatant. Consequently, dyed biomass was washed twice with water at a concentration of 20mg of biomass/mL of water. Dyed biomass was dried overnight and in dark in crucibles, ensuring all the samples were dried for the same time and under the same conditions (Zinchuk et al., 2007). Dyed biomass was mounted using 1drop of Prolong Gold antifade mountant. A Leica SP-2 AOBS confocal laser microscope was employed for image acquisition (Pérez-Pimienta et al., 2017). Holocellulose was detected in the range of 428-480 nm, whereas lignin was detected in the range of 547-658 nm. A 405 nm laser was employed to excite the samples. (Rigual et al., 2018a). In order to compare images, saturation, detection and emission wavelength ranges, and pinhole were fixed. The Z-stack of images was acquired at a step size of 2 μm and a magnification of x63. 5 2.3. Images processing 2.3.1. Particle size distribution X100 micrographs were employed to obtain particles size distributions of each sample. Two different processing procedures were tested, selecting the most adequate in every sample. In the former, grayscale micrographs were binarize using the imbinarize built-in function in Matlab, employing an adaptative threshold with a sensitivity of 70. Particles counting and size determination was calculated using the most appropriate of either a Matlab function (bwpropfilt built-in function) or the equivalent option in ImageJ (Analyze Analyze particles). In the latter, a marker-controlled watershed segmentation to separate touching objects was employed. In order to avoid counting huge- merged particles (touching particles taken as a single particle) all the results were filtrated manually. The equivalent diameters distribution was determined in the range comprised between 10-250 pixels (to avoid including noise and merged particles). Finally, results were scaled to μm. 2.3.2. Surface morphology Micrographs (x400) were employed to study the texture of pretreated samples. Every sample was cropped to a size that ensured the whole capture is composed by a textural image of a biomass particle. Images sizes varied in the range of 50x50 to 360x360 pixels. The grayscale image was measured in terms of fractal dimension and lacunarity, using Fraclac plugin of ImageJ open source software. The calculation method selected was Differential Volume Variation, that explicitly defines 3d volumes, Vi,j,ε over a grayscale image, according to Eqs 1 and 2. In this method, grayscale images are measured including a new dimension based on the intensity of gray pixels (Karperien, 2013). 𝐼𝜀 = ∑[1 + (𝑀𝑎𝑥 𝑃𝑖𝑥𝑒𝑙 𝑖𝑛𝑡𝑒𝑛𝑠𝑖𝑡𝑦(𝑖.𝑗.𝜀) − 𝑀𝑖𝑛 𝑃𝑖𝑥𝑒𝑙 𝑖𝑛𝑡𝑒𝑛𝑠𝑖𝑡𝑦(𝑖.𝑗.𝜀))] (1) 𝑉𝑖.𝑗.𝜀 = ∑ 𝐼𝑖.𝑗.𝜀𝜀2 (2) where I is the intensity, i and j are the location positions of the box and ε is the box size employed. Fractal dimensions Db, Dm and lacunarity were calculated according to Eqs 3-5: 𝐷𝑏 = 3 − ( lim 𝜀→0 ( 𝑙𝑛(𝑉𝜀) 𝑙𝑛( 1 𝜀 ) ) 2 ) (3) 𝐷𝑚 = 3 − ( lim 𝜀→0 ( 𝑙𝑛(𝑉𝜀̅̅ ̅̅ ) 𝑙𝑛( 1 𝜀 ) ) 2 ) (4) Lacunarity = ( (𝜎𝜀.𝑔̅̅ ̅̅ ̅̅ ) 𝜇𝜀.𝑔̅̅ ̅̅ ̅̅ ) 2 (5) 6 where σ is the standard deviation of pixels per box and μ is the mean of pixels per box. Additionally, Dx was calculated using the same method than for Db but in an average covering over all grids and then calculated. Differential scans were employed with 12 grid positions. The box size per series was set to 0 to let the software determine the number of grid calibres to use at each grid position. The minimum size of the box was fixed to 0 and the largest size was fixed to 45 % of the image. 2.3.3. Surface composition Image processing methodology is shown in Figure 1. Figure 1: Image processing methodology for confocal fluorescence microscopy images Samples were acquired in the two coupled channels, obtaining a two colors Z-stack combination of images. Z-stack images were projected into the max. Z projection using ImageJ open source software. Projected channels were combined to obtain the exposed two-colors image where violet represent lignin channel and green represents holocellulose. The composed imaged was split into the separated two-channels images. Every grayscale image (1024x1024 pixels per image) was brightness and contrast adjusted, and thresholded using the Otsu function (Otsu, 1979). Finally, every binary image was analyzed using the histogram tool of ImageJ. 3. Results and discussion Results obtained related pretreatment efficiency and digestibility to the particles sizes distribution, textural parameters and surface composition. 3.1. Influence of the pretreatment and biomass on particle size distribution The most significant micrographs obtained are shown in Figure 2. 7 Figure 2: SEM microscopy images of: a) AH150P, b) AH200E, c) IL80E and d) IL120P. AH samples decrease their particles sizes when the severity of the pretreatment increases, in both woods. However, only by employing micrographs there is a lot of information that cannot be analyzed, as particles sizes distributions are difficult to determine and the range of equivalent diameters of every sample cannot be measured either. In contrast, micrographs enhance the observation of shapes that tends to be more spherical at stronger AH conditions for both woods. From the other side, visual observation of IL pretreated samples show little differentiation between biomasses and employed temperatures for IL pretreatments, as all the IL micrographs have very similar like-sponge particles. These characteristic structures are created by the partial dissolution and regeneration of wood fibers in a fusing structure (Pérez-Pimienta et al., 2017, Sun et al., 2011). Figure 3 shows the probability distribution plotted as a histogram with a bin width of 5 μm and a cumulative probability of the tails comprised between 70 and 200 μm. All the particles with an equivalent diameter lower than 9 μm were discarded to avoid counting the noise from the image. In this sense, histograms showed a skewed distribution to the right. 8 Figure 3. Histograms of the probability distribution of the equivalent diameter: a) AH of eucalyptus, c) AH of pine, e) IL pretreatment of eucalyptus and g) IL pretreatment of pine; and Box plot of the cumulative distributions (25 %, 50 %, i.e. median, 75 %, and standard deviation values (dash lines): b) AH of eucalyptus, d) AH of pine, f) IL pretreatment of eucalyptus, h) IL pretreatment of pine. 9 Histograms in all the cases showed a shape close to an exponential distribution. However, the high probability found for particles of low size (10-15 μm) limited a proper fitting to an exponential distribution. It was observed that histograms tails (particle size >70 μm) were significantly higher for the IL pretreated samples than for AH pretreated samples, especially in pine. The higher standard deviations exhibited in the box plots for IL pretreated samples (in comparison to AH box plot results) also denote the same behavior. The small differences observed comparing box plotting of softwood and hardwood IL results remarks the mechanisms are independent of the biomass type, and so it is the particle size. In contrast, a strong dependence with the biomass type and severity factor has been found for AH pretreated samples. A strong influence of AH operating conditions was observed on the particle size distribution. In both woods (Figures 3b and 3d for eucalyptus and pine respectively) the first and third quartiles (25 and 75 % cumulative probabilities) and the medians (50 % cumulative probability) decreased as the severity factor was increased. As softwood is more recalcitrant than hardwood, the decrease was steeper for eucalyptus (from 15.43 to 7.93 μm) than for pine (from 11.40 to 9.95 μm) (Nitsos et al., 2016). IL operating conditions showed a slight influence on the particle size distribution. The pretreatment effect was similar for both types of woods, showing similar distributions. The effect of increasing the pretreatment temperature have an opposite behavior in hardwood and softwood, in the medians and the tails. In the case of hardwood, the pretreatment at lower temperatures resulted in a higher percentage of big particles, whereas in the case of softwood the opposite behavior occurred (medians of 11.31 and 9.03 μm at 80 and 120 °C respectively). 3.2. Influence of the pretreatment on the image texture Scanning electron microscopy of some of the most different cropped micrographs employed for fractal calculations are shown in Figure 4. Visual examination already reveals differences between textures shadows and pixels intensities, but the chosen method corrects these differences normalizing the intensity in the interval of intensities of each image, according to Eq. 1. Figure 4: x400 SEM micrographs of a) AH150E, b) AH150E, c) IL120E and d) IL120P 10 Before analysis of fractal dimension, one of the main issues is to select the best estimation method, and set up the parameters (Hadzieva et al., 2015). Results confirmed that the fractal dimension exhibits different values depending on the method used (Alvarez et al., 2013). In this case, it was checked that binarization and application of the box counting method in a threshold image was not appropriate to estimate fractal parameters, whereas treating the image directly as a grayscale image without adjusting brightness and contrast gave better results. This processing treated the image as a volume, which reflects the reality of the material (Mandelbrot and Pignoni, 1983). Differential volume variation was checked as the best grayscale method to determine fractal parameters, whereas the other two (differential, and differential volume variation plus1) gave inconsistencies (too high fractal dimensions and a lot of error between replicates). The number of box sizes employed varied with the size of the micrograph in the range of 18-100. It was shown that the number of box sizes fixing the grid was independent of the fractal dimension, as reproducibility was checked. Too low number of box sizes (obtained increasing the grid size) returned unacceptable standard deviations in the calculation of fractal dimension (Liu and Ostadhassan, 2017). Results confirmed the veracity with the reality according to physical criteria (values between 2 and 3), and mathematical (Db, Dm and Dx are in the same range, and regression lines plotting log of box size and log of pixel count have r2>0.96) (Seuront, 2015). Textural characteristics of the biomass are shown in Figure 5. Among all the fractal dimensions calculated, Db was chosen as reference. According to Figure 5a, all the values were in the range 2.65-2.74, which although appears to be small, is enough to see differences between samples (Chi et al., 2017, Risović et al., 2008). High fractal dimensions values are related to more convolutions and porosity in a texture and rougher structures, as a consequence of the increment in porosity (Alvarez et al., 2013, Chi et al., 2017, Rigual et al., 2018a). Autohydrolysis of hardwood increases the fractal dimension with the increasing severity factor, due to the breakage of the lignocellulosic network (Rigual et al., 2018b). In contrast, autohydrolysis of softwood decreases the image complexity, as morphology appears to have a smoother surface for severe conditions in pine, due to the possible formation of condensed structures (Pielhop et al., 2015). In the case of IL samples, all the samples appear to have fractal dimensions higher than for autohydrolysis (excepting AH200E). Furthermore, values are in all the cases in the same range (values between 2.73-2.74), which indicates that nor biomass type (hardwood of softwood), neither IL temperature pretreatment influence the fractal dimension obtained in pretreated biomass. Lacunarity calculation is based on pixel masses instead of box counts (Alvarez et al., 2013). The higher the lacunarity is, the more heterogenous the material is (Alvarez et al., 2013). Figure 5b shows lacunarity values. They are in the range 0.05-0.16, which are low, as samples have low heterogeneity in the whole micrograph. In this sense, samples AH175P and AH200P represent the 11 most heterogeneous textures, whereas AH200E and IL samples were the most homogeneous ones (Mandelbrot and Pignoni, 1983, Utrilla-Coello et al., 2013). Figure 5: a) Db and b) lacunarity values of pine and eucalyptus 3.3. Influence of the pretreatment on the surface composition Figure 6 shows some representative max. Z-stack micrographs obtained through confocal fluorescence microscopy. Figure 6: Confocal fluorescence microscopy images of a) AH200E, b) AH200P, c) IL80E and d) IL80P 12 Variations between pine and eucalyptus samples are very evident, especially in severe autohydrolysis conditions. While in eucalyptus only holocellulose (in green) is visually observed, in pine most pixels are coloured by lignin (in violet), indicating autohydrolysis pretretament affect in a different way in hardwood and softwood (Donaldson and Vaidya, 2017). In the case of IL samples, not visual differences are observed between pine and eucalyptus pretreated samples, that appear to have a disordered structure with holocellulose and lignin inter-crossed (Pérez-Pimienta et al., 2017, Raghavi et al., 2016). However, visual analysis of micrographs do not let distinguish holocellulose and lignin when colocalisation exists, as observed in some parts of Figure 6c (Donaldson and Vaidya, 2017). In all the cases where discussion is not so evident, the need of further quantification using image processing is essential to find more relevant information. The pixels quantification of holocellulose and lignin, in terms of holocellulose/lignin ratios are shown in Figure 7. Figure 7: Holocellulose/lignin ratio distribution on particles surface Excepting AH200E, all the holocellulose/lignin ratios are between 0.21-0.96, indicating that more lignin than holocellulose is always detected in all the samples. These results contrast to the characterization results, that indicate that the overall composition is always higher in holocellulose than in lignin (Rigual et al., 2018b). Although chemical characterisation of the material indicated that AH200E sample was composed by a 62% of lignin, it is clearly observed that no autofluorescent lignin is detected (Rigual et al., 2018b). This may be attributed to two reasons: the former is that no lignin is accumulated in the particles surface, and the latter is that lignin has been modified and is no longer autofluorescent (Chabbert et al., 2018). In any of the cases, differences in lignin chemistry between hardwood (richer in syringyl units) and softwood (richer in guayacyl units) are evinced in this analysis, which can be a complement to other lignin-in-the cell-wall characterisation techniques such as 2D-NMR (Chabbert et al., 2018, Crestini et al., 2017, Dutta et al., 2018). 13 In contrast IL temperature shows a slight increase of holocellulose/lignin ratios when the temperature increases. However, no differences have been observed between hardwood and softwood. This is attributed to the pretreatment process, involving partial dissolution and regeneration without the separation of dissolved fractions from the undissolved ones (Mäki-Arvela et al., 2010, Torr et al., 2016). Higher temperatures lead to higher dissolution of lignin, so during the regeneration, more holocellulose is visualized (probably attributed to cellulose). Holocellulose/lignin ratios are in the range of mild autohydrolysis samples, and again they do not correspond to the ratios obtained using the NREL methodology for the determination of structural carbohydrates and lignin (Rigual et al., 2018b, Sluiter et al., 2008). 3.4. Image analysis parameters affecting biomass digestibility Figure 8 shows the overall results, plotting every parameter obtained from different micrographs with the others (Db, lacunarity, equivalent dimeter, holocellulose/lignin ratio) and with 72 h enzymatic digestibility. Figure 8. Digital image analysis matrix Besides from linear fits here exposed, other regression techniques, such as random forest tree decisions, were also tested in the search of a multivariable correlation with 72 h enzymatic 14 digestibility. However, only a clear dependence on the lacunarity and fractal dimension were found, and the influence of the rest of studied variables could not be properly identified. Linear tendencies were found in fractal dimension vs digestibility (r2=0.93), lacunarity vs digestibility (r2= 0.70) and fractal dimension vs lacunarity (r2=0.79), whereas in the rest of the cases representative linear tendencies were not observed. Apart from the correlation with 72 h enzymatic digestibility, the correlation with other digestibility times (12 h and 24 h) was checked, being lower in all the cases. Fractal dimension and lacunarity were inversely correlated, so complex pretreated biomass structures are homogeneous. The correlations observed between fractal dimension and digestibility indicates that the higher the complexity of the texture is, the more convolutions the particle has, and the higher the accessibility is. In contrast, the opposite behaviour obtained with lacunarity indicates that more homogeneous structures are more accessible to enzymatic saccharification. It is important to highlight that this relation between fractal dimension and lacunarity has been found independently of the biomass (hardwood and softwood) and the pretreatment employed. Tendencies found with these samples, open a path for further research in textural analysis. The employment of confocal fluorescence microscopy did not lead to a tendency with digestibility as found by Chabbert et al. (2018), which can be attributed to the differences developed in the experimental procedure. The lack of linear correlation between holocellulose /lignin ratios is not an inconsistence, as results can be explained depending on the pretreatment and the biomass type. Furthermore, lignin removal is usually insufficient to correlate with sugar yield because hemicelluloses are covalently linked to lignin (Ding et al., 2012, Meng et al., 2015). The same statement is observed plotting the equivalent diameter with 72 h enzymatic digestibility: while in AH of eucalyptus the most severe conditions lead to higher enzymatic digestibilities and smaller particles sizes, this parameter is not determinant in any of the rest of the samples. Kadić et al. (2014) found a similar behavior, showing a little effect of the particle size distribution for a pretreated spruce and giant reed biomass and its enzymatic digestibility. 4. Conclusion In this work, a deep microscopy characterization of pretreated biomass samples employing image processing tools has been proposed. Results have evinced the potential of image processing to further characterize micrographs obtained. Quantification with numerical parameters, such as equivalent diameter or holocellulose lignin ratios have shown a dependence with the pretreatment conditions and the biomass employed. Interestingly, lacunarity and mainly fractal dimension have been found to linearly correlate (0.70 and 0.93, respectively) with enzymatic digestibility. The correlation found is independent of the biomass and the pretreatment type, opening a path of further study in textural analysis. 15 Acknowledgements This work was performed thanks to the financial support of the “Ministerio de Ciencia, Innovación y Universidades” under the funded project CTQ2017-88623-R, the contract BES-2014- 067788, and the “Juan de la Cierva” postdoctoral grant FJCI-2015-23765. References Agbor, V.B., Cicek, N., Sparling, R., Berlin, A., Levin, D.B., 2011. Biomass pretreatment: fundamentals toward application. Biotechnol. Adv. 29, 675-685. Alvarez, A., Passé-Coutrin, N., Gaspard, S., 2013. Determination of the textural characteristics of carbon samples using scanning electronic microscopy images: comparison with mercury porosimetry data. Adsorption 19, 841-850. Arantes, V., Saddler, J.N., 2010. Access to cellulose limits the efficiency of enzymatic hydrolysis: the role of amorphogenesis. Biotechnol. Biofuels 3, 4. Chabbert, B., Terryn, C., Herbaut, M., Vaidya, A., Habrant, A., Paës, G., Donaldson, L., 2018. Fluorescence techniques can reveal cell wall organization and predict saccharification in pretreated wood biomass. Ind. Crop. Prod. 123, 84-92. Chen, S., Zhang, X., Ling, Z., Xu, F., 2017a. Characterization of the micromorphology and topochemistry of poplar wood during mild ionic liquid pretreatment for improving enzymatic saccharification. Molecules 22, 115. Chen, S., Zhang, X., Ling, Z., Xu, F., 2017b. Characterization of the Micromorphology and Topochemistry of Poplar Wood during Mild Ionic Liquid Pretreatment for Improving Enzymatic Saccharification. Molecules 22, Cheng, Q., 1997. Multifractal modeling and lacunarity analysis. Math. Geol. 29, 919-932. Chi, C., Hui, Z., Liu, M., Zhang, S., Gong, Y., 2017. Effect of Acetic Acid Pretreatment on Wood Pore Structure and Fractal Dimension. BioResources 12, 3905-3917. Crestini, C., Lange, H., Sette, M., Argyropoulos, D.S., 2017. On the structure of softwood kraft lignin. Green Chem. 19, 4104-4121. Ding, S.Y., Liu, Y.S., Zeng, Y., Himmel, M.E., Baker, J.O., Bayer, E.A., 2012. How does plant cell wall nanoscale architecture correlate with enzymatic digestibility? Science 338, Donaldson, L., Vaidya, A., 2017. Visualising recalcitrance by colocalisation of cellulase, lignin and cellulose in pretreated pine biomass using fluorescence microscopy. Scientific reports 7, 44386. Dutta, T., Papa, G., Wang, E., Sun, J., Isern, N.G., Cort, J.R., Simmons, B.A., Singh, S., 2018. Characterization of Lignin Streams during Bionic Liquid-Based Pretreatment from Grass, Hardwood, and Softwood. ACS Sustain. Chem. Eng. 6, 3079-3090. Gonçalves, F.A., Ruiz, H.A., dos Santos, E.S., Teixeira, J.A., de Macedo, G.R., 2015. Bioethanol production from coconuts and cactus pretreated by autohydrolysis. Ind. Crop. Prod. 77, 1-12. Hadzieva, E., Bogatinoska, D.C., Gjergjeska, L., Shuminoska, M., Petroski, R., 2015. Review of the Software Packages for Estimation of the Fractal Dimension. Hanen, A., Imen, B., Asma, B.A., Patrick, D., Hédi, B.M., 2009. Multifractal modelling and 3D lacunarity analysis. Phys. Lett. A 373, 3604-3609. 16 Haykir, N.I., Bahcegul, E., Bicak, N., Bakir, U., 2013. Pretreatment of cotton stalk with ionic liquids including 2-hydroxy ethyl ammonium formate to enhance biomass digestibility. Ind. Crop. Prod. 41, 430-436. Henebry, G., Kux, H., 1995. Lacunarity as a texture measure for SAR imagery. Remote Sens. 16, 565-571. Kadić, A., Palmqvist, B., Lidén, G., 2014. Effects of agitation on particle-size distribution and enzymatic hydrolysis of pretreated spruce and giant reed. Biotechnol. Biofuels 7, 77. Karimi, K., Taherzadeh, M.J., 2016. A critical review of analytical methods in pretreatment of lignocelluloses: Composition, imaging, and crystallinity. Bioresour. Technol. 200, 1008-1018. Karperien, A., 2013. FracLac for ImageJ. Li, C., Cheng, G., Balan, V., Kent, M.S., Ong, M., Chundawat, S.P.S., Sousa, L., Melnichenko, Y.B., Dale, B.E., Simmons, B.A., 2011. Influence of physico-chemical changes on enzymatic digestibility of ionic liquid and AFEX pretreated corn stover. Bioresour. Technol. 102, Liu, K., Ostadhassan, M., 2017. Quantification of the microstructures of Bakken shale reservoirs using multi-fractal and lacunarity analysis. J. Nat. Gas Sci. Eng. 39, 62-71. Luterbacher, J.S., Moran‐Mirabal, J.M., Burkholder, E.W., Walker, L.P., 2015. Modeling enzymatic hydrolysis of lignocellulosic substrates using confocal fluorescence microscopy I: filter paper cellulose. Biotechnol. Bioeng. 112, 21-31. Mäki-Arvela, P., Anugwom, I., Virtanen, P., Sjöholm, R., Mikkola, J.P., 2010. Dissolution of lignocellulosic materials and its constituents using ionic liquids—A review. Ind. Crop. Prod. 32, 175-201. Mandelbrot, B.B., Pignoni, R., 1983. The fractal geometry of nature. WH freeman New York. Meng, X., Wells, T., Jr., Sun, Q., Huang, F., Ragauskas, A., 2015. Insights into the effect of dilute acid, hot water or alkaline pretreatment on the cellulose accessible surface area and the overall porosity of Populus. Green Chem. 17, 4239-4246. Miyamoto, T., Mihashi, A., Yamamura, M., Tobimatsu, Y., Suzuki, S., Takada, R., Kobayashi, Y., Umezawa, T., 2018. Comparative analysis of lignin chemical structures of sugarcane bagasse pretreated by alkaline, hydrothermal, and dilute sulfuric acid methods. Ind. Crop. Prod. 121, 124-131. Momoh, O., Ouki, S., 2018. Development of a novel fractal-like kinetic model for elucidating the effect of particle size on the mechanism of hydrolysis and biogas yield from ligno-cellulosic biomass. Renew. Energy 118, 71-83. Moniz, P., Pereira, H., Quilhó, T., Carvalheiro, F., 2013. Characterisation and hydrothermal processing of corn straw towards the selective fractionation of hemicelluloses. Ind. Crop. Prod. 50, 145-153. Moradi, F., Amiri, H., Soleimanian-Zad, S., Ehsani, M.R., Karimi, K., 2013. Improvement of acetone, butanol and ethanol production from rice straw by acid and alkaline pretreatments. Fuel 112, 8-13. Nitsos, C.K., Choli-Papadopoulou, T., Matis, K.A., Triantafyllidis, K.S., 2016. Optimization of Hydrothermal Pretreatment of Hardwood and Softwood Lignocellulosic Residues for Selective Hemicellulose Recovery and Improved Cellulose Enzymatic Hydrolysis. ACS Sustain. Chem. Eng. 4, 4529-4544. 17 Otsu, N., 1979. A threshold selection method from gray-level histograms. IEEE transactions on systems, man, and cybernetics 9, 62-66. Pérez-Pimienta, J.A., Vargas-Tah, A., López-Ortega, K.M., Medina-López, Y.N., Mendoza-Pérez, J.A., Avila, S., Singh, S., Simmons, B.A., Loaces, I., Martinez, A., 2017. Sequential enzymatic saccharification and fermentation of ionic liquid and organosolv pretreated agave bagasse for ethanol production. Bioresour. Technol. 225, 191-198. Pielhop, T., Larrazábal, G.O., Studer, M.H., Brethauer, S., Seidel, C.-M., von Rohr, P.R., 2015. Lignin repolymerisation in spruce autohydrolysis pretreatment increases cellulase deactivation. Green Chem. 17, 3521-3532. Plotnick, R.E., Gardner, R.H., Hargrove, W.W., Prestegaard, K., Perlmutter, M., 1996. Lacunarity analysis: a general technique for the analysis of spatial patterns. Phys. Rev. E 53, 5461. Putro, J.N., Soetaredjo, F.E., Lin, S.-Y., Ju, Y.-H., Ismadji, S., 2016. Pretreatment and conversion of lignocellulose biomass into valuable chemicals. RSC Adv. 6, 46834-46852. Raghavi, S., Sindhu, R., Binod, P., Gnansounou, E., Pandey, A., 2016. Development of a novel sequential pretreatment strategy for the production of bioethanol from sugarcane trash. Bioresour. Technol. 199, 202-210. Rigual, V., Santos, T.M., Domínguez, J.C., Alonso, M.V., Oliet, M., Rodriguez, F., 2018a. Combining autohydrolysis and ionic liquid microwave treatment to enhance enzymatic hydrolysis of Eucalyptus globulus wood. Bioresour. Technol. 251, 197-203. Rigual, V., Santos, T.M., Domínguez, J.C., Alonso, M.V., Oliet, M., Rodriguez, F., 2018b. Evaluation of hardwood and softwood fractionation using autohydrolysis and ionic liquid microwave pretreatment. Biomass Bioenerg. 117, 190-197. Risović, D., Poljaček, S.M., Furić, K., Gojo, M., 2008. Inferring fractal dimension of rough/porous surfaces—A comparison of SEM image analysis and electrochemical impedance spectroscopy methods. Appl. Surf. Sci. 255, 3063-3070. Santacruz-Vazquez, V., Santacruz-Vazquez, C., LAGUNA CORTÉS, J.O., 2015. Physical characterization of freeze-dried foam prepared from Aloe vera gel and guar gum. Vitae 22, 75- 86. Seuront, L., 2015. On uses, misuses and potential abuses of fractal analysis in zooplankton behavioral studies: a review, a critique and a few recommendations. Physica A 432, 410-434. Shinde, S.D., Meng, X., Kumar, R., Ragauskas, A.J., 2018. Recent advances in understanding the pseudo-lignin formation in a lignocellulosic biorefinery. Green Chem. 20, 2192-2205. Singh, S., Cheng, G., Sathitsuksanoh, N., Wu, D., Varanasi, P., George, A., Balan, V., Gao, X., Kumar, R., Dale, B.E., 2015. Comparison of different biomass pretreatment techniques and their impact on chemistry and structure. Frontiers in Energy Research 2, 62. Sluiter, A., Hames, B., Ruiz, R., Scarlata, C., Sluiter, J., Templeton, D., Crocker, D., 2008. Determination of structural carbohydrates and lignin in biomass. Laboratory analytical procedure 1617, 1-16. Sun, N., Rodriguez, H., Rahman, M., Rogers, R.D., 2011. Where are ionic liquid strategies most suited in the pursuit of chemicals and energy from lignocellulosic biomass? Chem. Commun. 47, 1405-1421. 18 Torr, K.M., Love, K.T., Simmons, B.A., Hill, S.J., 2016. Structural features affecting the enzymatic digestibility of pine wood pretreated with ionic liquids. Biotechnol. Bioeng. 113, 540-549. Utrilla-Coello, R.G., Bello-Pérez, L.A., Vernon-Carter, E.J., Rodriguez, E., Alvarez-Ramirez, J., 2013. Microstructure of retrograded starch: Quantification from lacunarity analysis of SEM micrographs. J. Food Eng. 116, 775-781. Zhao, J., Chen, H., 2013. Correlation of porous structure, mass transfer and enzymatic hydrolysis of steam exploded corn stover. Chem. Eng. Sci. 104, 1036-1044. Zinchuk, V., Zinchuk, O., Okada, T., 2007. Quantitative colocalization analysis of multicolor confocal immunofluorescence microscopy images: pushing pixels to explore biological phenomena. Acta Histochem. Cytochem. 40, 101-111. PUBLICACIÓN V “Combining autohydrolysis and ionic liquid microwave treatment to enhance enzymatic hydrolysis of Eucalyptus globulus wood” Victoria Rigual, Tamara M. Santos, Juan C. Domínguez, M. Virginia Alonso, Mercedes Oliet y Francisco Rodriguez Bioresource Technology, 2018, 251, 197-203 Contents lists available at ScienceDirect Bioresource Technology journal homepage: www.elsevier.com/locate/biortech Combining autohydrolysis and ionic liquid microwave treatment to enhance enzymatic hydrolysis of Eucalyptus globulus wood Victoria Rigual⁎, Tamara M. Santos, Juan Carlos Domínguez, M. Virginia Alonso, Mercedes Oliet, Francisco Rodriguez Department of Chemical Engineering, Faculty of Chemistry, Complutense University of Madrid, Avda. Complutense s/n, 28040 Madrid, Spain G R A P H I C A L A B S T R A C T A R T I C L E I N F O Keywords: Pretreatment combination Eucalyptus wood Microwave ionic liquid pretreatment Fractal dimension Lacunarity A B S T R A C T The combination of autohydrolysis and ionic liquid microwave treatments of eucalyptus wood have been studied to facilitate sugar production in a subsequent enzymatic hydrolysis step. Three autohydrolysis conditions (150 °C, 175 °C and 200 °C) in combination with two ionic liquid temperatures (80 °C and 120 °C) were compared in terms of chemical composition, enzymatic digestibility and sugar production. Morphology was measured (using SEM) and the biomass surface was visualized with confocal fluorescence microscopy. The synergistic cooperation of both treatments was demonstrated, enhancing cellulose accessibility. At intermediate auto- hydrolysis conditions (175 °C) and low ionic liquid temperature (80 °C), a glucan digestibility of 84.4% was obtained. Using SEM micrographs, fractal dimension (as a measure of biomass complexity) and lacunarity (as a measure of homogeneity) were calculated before and after pretreatment. High fractals dimensions and low lacunarities correspond to morphologically complex and homogeneous samples, that are better digested by enzyme cocktails. 1. Introduction Eucalyptus species are one of the most promising resources in biorefinery processes because of their fast growth and wide adaptability to soils and climates (Carrillo et al., 2017). In the Iberian Peninsula, eucalyptus accounts for a volume of 2.8× 106 tons per year (Silva- Fernandes et al., 2015). The plant cell wall of eucalyptus show linkages between cellulose (crystalline and amorphous), branched hemicellulose (formed by acetylated glucuronoxylan and minors amounts of arabinosyl units) and lignin, all intercrossed together conforming the recalcitrance structure of biomass (Sun et al., 2016; Isikgor and Becer, 2015). Due to the re- calcitrance of the cross-linked structure of lignocellulosic biomass, the biorefinery of these resources is still a challenge (Demartini et al., 2015). Pretreatment processes that deconstruct the linkages and disrupt the structure are recquired and involve up to 40% of the total proces- sing costs of lignocellulosic biomass conversion (Pérez-Pimienta et al., 2017; Raghavi et al., 2016). https://doi.org/10.1016/j.biortech.2017.12.034 Received 17 October 2017; Received in revised form 11 December 2017; Accepted 12 December 2017 ⁎ Corresponding author. E-mail address: vicrigua@ucm.es (V. Rigual). Bioresource Technology 251 (2018) 197–203 Available online 14 December 2017 0960-8524/ © 2017 Elsevier Ltd. All rights reserved. T http://www.sciencedirect.com/science/journal/09608524 https://www.elsevier.com/locate/biortech https://doi.org/10.1016/j.biortech.2017.12.034 https://doi.org/10.1016/j.biortech.2017.12.034 mailto:vicrigua@ucm.es https://doi.org/10.1016/j.biortech.2017.12.034 http://crossmark.crossref.org/dialog/?doi=10.1016/j.biortech.2017.12.034&domain=pdf Autohydrolysis (AH) is considered an inexpensive, environmentally friendly and easy-to-handle process that removes hemicelluloses (Liu et al., 2015; Silva-Fernandes et al., 2015). In the case of eucalyptus, the auto ionization of water into hydronium ions and acetyl group hydra- tion, autocatalyze the process (Perez-Pimienta et al., 2016). Due to hemicelluloses removal, the size, surface area, or pore size of the re- maining solids are altered (Rissanen et al., 2016). Thus, AH is an ade- quate pretreatment for the development of integrated biorefining stra- tegies aimed at the fractionation of hemicellulose and the obtention of lignin- and cellulose-rich materials that can be further treated (Silva- Fernandes et al., 2015). Although autohydrolysis pretreatments obtain high value added products, high energy input is recquired (Elgharbawy et al., 2016). Compared to other pretreatments such as low alkaline pretreatment, the high yields of hemicellulose extraction and the use of only water, are advantageous. However, its utilization as unique pre- treatment does not yield enough high glucan digestibilities to con- template it as a methodology for biomass fractionation. Ionic liquids (ILs) are very promising biomass solvents that reduce recalcitrance, enabling deconstruction of the lignocellulosic network. High chemical and thermal stability and negligible vapor pressure are some of the advantages against other organic solvents (Zhu et al., 2016). ILs also reduce cellulose crystallinity, increasing cellulose ac- cessibility and favoring high cellulose to glucose conversions (Singh et al., 2016; Xu et al., 2016). ILs pretreatments followed by enzymatic saccharification have been reported as a high glucose release meth- odologies that reach digestibilities of around 90% in different bio- masses at relatively mild temperature conditions (below 140 °C). However, the cost of the ionic liquid as reagent in comparison of other chemicals used in pretreatments (NaOH, H2SO4, ammonia), the diffi- culties to recover the IL, the high viscosity of most of the ionic liquids and the low solid loadings used are factors that are needed to be ful- filled before an industrial process is developed. Some imidazolium aprotic ILs such as [Emim][OAc], [Bmim][Cl] and [Emim][DEP], have been proved to dissolve cellulose and lignocellulosic biomass (Ha et al., 2011; Li et al., 2009). [Emim][OAc] has been found to report the best results in biomass pretreatment and is the most widely used IL for this purpose to date (Brandt et al., 2017; Elgharbawy et al., 2016). Ad- ditionally, ILs are considered good microwave absorbers. The use of this heating method speeds up the treatment (Fang et al., 2015). Sequential and combined treatments are being studied in an attempt to improve cellulose to glucose digestibility, extracting also hemi- cellulose and lignin. The combination of physical (steam explosion, AH, AFEX, etc.) with chemical pretreatments such as alkaline peroxide or ethanol extraction have been tested (Akhtar et al., 2015; Chen et al., 2008). Fewer literature is found combining ILs with other pretreat- ments. Silveira et al. (2015) combined [Bmim][OAc] with supercritical CO2, obtaining high glucose yields from sugarcane bagasse. Rice straw was treated by Nguyen et al. (2010), combining ammonia and [Emim] [OAc], obtaining high glucose yields and reducing enzyme dosage and incubation time; while Heggset et al. (2016) combined ILs and low temperature alkaline treatment. AH and IL treatment of wood was de- veloped by Hauru et al. (2013), focusing on biomass fractionation in- stead of a pretreatment (prior to enzymatic hydrolysis) approach. To evaluate biomass digestibility, microscopy analysis are com- monly used. SEM images are being widely utilized to visually support results obtained in biomass pretreatment (Demartini et al., 2015; Nitsos et al., 2016; Singh et al., 2009). However, digital image analysis (DIA) of these images has not been further investigated. The fractal dimension concept, first introduced by Mandelbrot (1977), is being calculated in many scientific research areas such as (porous) materials character- ization, botany, medicine or cellular biology (Liu and Ostadhassan, 2017; Neves et al., 2014; Smith et al., 1996; Utrilla-Coello et al., 2013). In complex objects and textures, the fractal dimension parameter is applied to non-traditional patterns or objects to define the inherent grade of complexity. Fractal dimension can be defined as a scaling rule that indicates how a pattern’s details change with the scale (Karperien, 2013). Lacunarity, another DIA-employed concept, is described as in- homogeneity, gapiness, visual texture or rotational invariance. This concept introduced by Mandelbrot and Pignoni (1983) has also been previously used to measure changes in the features of a digital image with the level of resolution (Karperien, 2013; Kilic and Abiyev, 2011; Liu and Ostadhassan, 2017; Utrilla-Coello et al., 2013). In this work, the combination of AH and IL microwave treatments, towards maximizing sugar extraction from Eucalyptus globulus wood, is studied. AH and IL microwave treatments are performed at different temperatures, evaluating glucan and xylan digestibilities in the sub- sequent enzymatic hydrolysis step. For the first time, the terms of fractal dimension and lacunarity are calculated from SEM images and related against enzymatic digestibility, as a measure of cellulose ac- cessibility. Furthermore, surface compositional changes are observed by confocal fluorescence microscopy. Finally, the global mass balance sums up solid recoveries as well as compositional analysis, showing the most favorable conditions to maximize sugars production. 2. Materials and methods 2.1. Materials and reagents The raw material Eucalyptus globulus sawdust was provided by the National Institute for Agronomic Research (CIFOR-INIA). The IL 1- ethyl-3-methylimidazolium acetate ([Emim][OAc], > 95%) was sup- plied by Iolitec GmbH. Lignin was extracted from the raw material Eucalyptus globulus through an organosolv process, using sulfuric acid as a precipitating agent, and was used as reference material for UV/VIS measurements (detailed below). Enzymatic hydrolysis was developed using the enzymatic cocktail Accellerase 1500®, kindly donated by DuPont Industrial Biosciences. 2.2. Experimental 2.2.1. Autohydrolysis (AH) treatment A 450mL stainless steel pressure reactor (Parr Instrument Company, model 4567) was used to perform AH treatments. The equipment op- erates with a four-blade turbine impeller. An external electrical mantle was used to heat the reactor vessel, and a silicone fluid oil (100 cSt) was employed in the cooling step. Parr PID controller model 4848 was used to control parameters. All the temperature ramps consist of a heating ramp from ambient temperature to the set temperature (150, 175 or 200 °C) for 50 min, a 30-min isotherm at the set temperature, and a cooling ramp until the reactor temperature decreased to 40 °C. A liquid- to-solid ratio of 8:1 (g of water: g of dry biomass) was employed ac- cording to the literature (García-Domínguez et al., 2013; Romaní et al., 2014). 2.2.2. Ionic liquid (IL) microwave treatment First, AH Eucalyptus globulus samples were milled and sieved to obtain particle sizes< 150 µm. Consequently, 0.8 g of wood samples were mixed with 20 g of [Emim][OAc]. Samples were heated at 80 °C or 120 °C under microwave radiation in a Berghof SpeedWave Four mi- crowave oven, using a two-step programme detailed in a previous work (Casas et al., 2013). Afterwards, 50mL of deionized water was added as an antisolvent agent at 40 °C. The solution was stirred for 10 min. Solid and liquid fractions were separated by vacuum filtration. To maximize [Emim][OAc] recovery yield and avoid IL residues in the wood samples in the washing step (that distort solid recoveries results), treated samples were washed 5 times with 70mL of deionized water per wash. Recovered treated samples were stored at 40 °C for further treatments and analysis. 2.2.3. Enzymatic hydrolysis 1% wt/wt of untreated/treated wood was suspended in 50mM ci- trate buffer (pH 5.0) containing 0.002% of sodium azide, in a working V. Rigual et al. Bioresource Technology 251 (2018) 197–203 198 volume of 8mL. 0.25mL/g glucan of Accellerase 1500 enzymatic cocktail was added to the preparation. The enzymatic hydrolysis was carried out in an IKA KS 4000i orbital incubator at 150 rpm and 50 °C. In parallel, blank assays of each substrate were prepared to control the process in absence of enzymatic cocktail. Hydrolysate aliquots of 150 µL were taken at 3, 6, 12, 24, 48 and 72 h, centrifuged twice consecutive at 9000 rpm for 10 min and analyzed by HPLC to obtain monosaccharides composition. 2.3. Analytical methods 2.3.1. Chemical characterization of untreated eucalyptus wood, treated wood, and enzymatic hydrolysis hydrolysates Untreated eucalyptus wood and treated solids compositions were determined according to the NREL/TP-510-42618 methodology, adapted to small quantities of samples (50mg) (Sluiter et al., 2011; Ibáñez and Bauer, 2014). A two-step acid hydrolysis was employed to obtain acid-insoluble lignin and ashes (gravimetrically quantified after being dried at 105 °C) and a liquid fraction formed by soluble lignin and sugars. Acid-soluble lignin was quantified by UV spectroscopy at 205 nm, using an absorptivity coefficient of 110 L·g−1·cm−1. Sugars in hydrolysates were determined using an Agilent 1260 HPLC system operating with a refractive index detector. A 300×7.8mm Carbosep-CHO 682 column with Micro-Guard cartridges (BioRad, Life Science Group Hercules, Ca) was used to measure sugars concentration in acidic samples. Before analysis, samples were neu- tralized with CaCO3 and filtered under 0.45 µm. The column was op- erating at 80 °C using water as mobile phase, and the flow rate was 0.4 mL/min. The anhydrous corrections to hexoses (162/180) and pentoses (132/150) were applied to each sugar concentration, respec- tively. Sugar recovery standards (SRS) corrections were applied fol- lowing the procedure (Sluiter et al., 2011). In the case of sugars determination in non-acidic samples (enzy- matic hydrolysis), the abovementioned CARBOSep CHO-682 Guard kit+ lead column was used under the above specified conditions. Glucan and xylan digestibilities were defined as the percentage of glucan or xylan converted into glucose or xylose during enzymatic hydrolysis and were calculated according to Eqs. (1) and (2). ⎜ ⎟= ⎛ ⎝ ⎞ ⎠ ×Glucan digestibility (%) hydrolyzed glucan (g) glucan in the enzymatic hydrolysis (g) 100 (1) ⎜ ⎟= ⎛ ⎝ ⎞ ⎠ ×Xylan digestibility (%) hydrolyzed xylan (g) xylan in the enzymatic hydrolysis (g) 100 (2) According to the NREL/TP-5100-63351 procedure, anhydro cor- rections of 0.9 and 0.88 were applied to determine glucan and xylan content in hydrolysates (Resch et al., 2015). 2.3.2. IL content in the washing fractions after microwave treatment Washing fractions were analyzed by HPLC equipped with a UV detector measuring at 235 nm. An Eclipse Plus C18 4.6×100 mm column and a mobile phase of 50/50% v/v acetonitrile/water at a flow rate of 1mL/min and at 30 °C was employed. To avoid exceeding the toleration levels of salts in the column, samples were diluted before injection (Hyvärinen et al., 2014). 2.3.3. Lignin content in the IL Lignin content in the recovered [Emim][OAc] and accumulated lignin were defined according to Eqs. (3) and (4). ⎜ ⎟= ⎛ ⎝ ⎞ ⎠ ×Lignin in [Emim][OAc] (wt%) Mass of lignin in IL (g) Total mass of IL introduced (g) 100 (3) ⎜ ⎟= ⎛ ⎝ ⎞ ⎠ ×Accumulated lignin (wt%) Mass of lignin in IL (g) Total mass lignin introduced (g) 100 (4) A Buchi R-114 rotavapor was employed to remove water from the washing fractions and recover the IL. The lignin content in the IL was analyzed by UV/VIS spectroscopy using a Varian Cary 50 scan UV/VIS spectrophotometer. Before measuring, water content in the recovered IL was determined using a thermobalance and mass was corrected. Samples were diluted 40-fold in 0.1 N NaOH and filtered, to measure the absorbance at 280 nm (Weerachanchai and Lee, 2014). The total dissolved lignin concentration was obtained from the reference curve of eucalyptus wood organosolv lignin samples prepared following the methodology of Weerachanchai and Lee (2014). To avoid an excess in the absorbance measured due to IL treatment, the absorbance mea- surement of each sample was subtracted from a blank sample prepared with the microwave heated IL at the working temperature. 2.3.4. Surface morphology and compositional distribution of processed solids Scanning electron microscope (SEM) images of untreated and treated samples were acquired using a Jeol JSM 6400 operating at 20 kV. A gold sputtering onto the sample surface was used to impart electrical conductivity. Analyses were developed in the technical fa- cilities of the Spanish National Centre for Electron Microscopy. Micrographs were processed using ImageJ software to measure fractal dimension and lacunarity. Five micrographs of each sample of 128× 128 pixels size were cropped. Based on previous assays, the local thresholding Bernsen method using a radius of 15 was found to be appropriate to convert the grayscale image into a binary image. The FracLac plugin of ImageJ was used to calculate fractal dimension (Db) and lacunarity (Λ) using the box counting method (Alvarez et al., 2013). The smallest grid used was 1 and the largest grid was 45%. A Leica SP-2 AOBS confocal laser microscope was used to observe variations in surface samples. A laser at 405 nm was used to observe sample fluorescence. Two wavelength detector sources (428–480 and 547–658 nm) were used to distinguish holocellulose and lignin (auto- fluorescent). Beta 1-4 polysaccharides linkages were dyed using 0.1% Calcofluor white stain (Pérez-Pimienta et al., 2017). Afterwards, sam- ples were washed twice with distilled water and dried in the dark overnight. Prolong Gold antifade mountant (1 drop) was used to fix the sample until analysis. Separate samples of dyed (with calcofluor) and non-dyed alkali lignin and Avicel PH-105 were taken as control. Images were acquired at a step size of 2 µm and were coupled, together with the images taken with transmitted light into a z-axis maximum pro- jection using ImageJ software. Images acquisition was developed in the technical facilities of the Centre for Cytometry and Fluorescence Mi- croscopy of the Complutense University of Madrid (UCM). 3. Results and discussion To nominate samples, the AH temperature of the treatment will be designated after “AH” letters (AH150, AH175 and AH200) and IL temperature will be designated afterwards, followed by the microwave treatment temperature (IL80 and IL120). 3.1. Solid recovery and chemical composition of untreated/treated biomass Table 1 shows the solid yield recovery and chemical composition of treated and untreated solids. The effects of AH+ IL treatments incur in a gradual decrease in the solid recovery that should be taken into ac- count in the global mass balance (explained below). The biggest de- crease is produced in the most severe conditions (AH200IL120) to 55 g of treated sample/100 g of untreated eucalyptus wood, while 84 g of treated sample/100 g of untreated eucalyptus wood is obtained in the least treated sample (AH150IL80). Furthermore, it is observed that the V. Rigual et al. Bioresource Technology 251 (2018) 197–203 199 AH treatment affects more to the solid recovery than the IL treatment, where solid yield recoveries decreasing in the range of 1.7 and 3.0% are observed between experiments performed at 80 °C and 120 °C. The low solid recoveries variations between samples autohydrolyzed at the same conditions, followed by different IL treatment temperature, supports the statement that ILs pretreat biomass without degrading cellulose (Elgharbawy et al., 2016). The relative composition of biomass compared to the untreated sample, varies in most experiments, indicating that specific biomass fractions are selectively removed (Procentese et al., 2015). The AH step incurs a slight increase in lignin content, as a consequence of hemi- celluloses extraction in the case of the samples autohydrolyzed at 150 °C and 175 °C. Lignin content is notoriously higher in samples au- tohydrolyzed at 200 °C as consequence of hemicellulose and cellulose degradation. This is in agreement with the increasing sugar content of the liquid fraction up to a point where degradation products are gen- erated (Nitsos et al., 2016; Silva-Fernandes et al., 2015). The IL tem- perature variation in the microwave step does not incur in significant changes in lignin compositions. The only relevant variation is observed between AH200IL80 and AH200IL120 samples, with variations of 6% among them. These results are supported by the UV measurements developed below, where higher lignin accumulation is produced in samples autohydrolyzed at severe conditions. Overall, it can be ob- served that AH treatments lead to hemicellulose extraction up to a point at which degradation of cellulose and hemicellulose starts. Glucan content is maximized in samples autohydrolyzed at 175 °C, which will contribute to increase final glucose yields in the process. Severe AH conditions produce cellulose degradation, as a consequence of the for- mation of degradation products (Nitsos et al., 2016). 3.2. Ionic liquid recovery and lignin extraction Washing fractions obtained after IL microwave treatment were collected separately and analyzed in order to determine the minimum amount of water required in this treatment. Three washings (210 mL), plus the regeneration step (50 mL) are required to recover the IL. Despite of operational losses during the process, IL recovery yields are in all the cases higher than 96%. [Emim][OAc] accumulation in the solid, could lead into mistakes in the global mass balance (due to overweighting solid fractions) and would cause a decrease in the en- zymatic hydrolysis yields as a result of cellulose inactivation (Xu et al., 2016). The problematic of salts accumulation when measuring in the presence of ILs limited the detection of sugars by HPLC at the stablished IL dilution conditions (Hyvärinen et al., 2014). To reuse the IL in consecutive cycles, antisolvent must be removed. Using vacuum distillation, at least 90% of the initial IL was recovered. However, differences in the coloration denote the accumulation of impurities (lignin) after the treatment. The quantity of lignin present in the IL gradually increases from 0.050 to 0.348 g of lignin per 100 g of IL, which corresponds to 10–70%wt of lignin introduced. Although the IL temperature influences the lignin accumulation in the IL, the biggest differences are observed as an effect of AH conditions. A subsequent step with an additional antisolvent (e.g., iso-propanol or ethanol) that enables lignin precipitation should be assessed not only to obtain a by- product stream in the overall process but also to remove impurities in the IL, enabling the recyclability in the following cycles of use (Auxenfans et al., 2014; Isikgor and Becer, 2015; Weerachanchai et al., 2014). 3.3. Enzymatic hydrolysis digestibility Fig. 2a shows glucan digestibility during enzymatic hydrolysis of samples subjected to AH+ IL treatment. The lowest glucan digestibility is obtained for untreated eucalyptus wood (15 g hydrolyzed glucan/ 100 g glucan introduced). Differences between untreated and treated samples demonstrate the effectiveness of EmimOAc that has been ex- tensively commented in literature (Elgharbawy et al., 2016; Lynam and Coronella, 2014; Sun et al., 2016). Furthermore, it is also observed how curve aspect differs between samples. While in the untreated wood and AH150IL120 sample digestibilities increase only up to 12–24 h, di- gestibilities continue increasing up to 24–48 h in the other samples. The least severe treatment tested already yields 68 g of glucan/ 100 g glucan, while the highest digestibility is obtained with 175 °C AH treatment followed by an IL step at 120° C. As occurred with solid re- covery yields, the effect of AH temperature in the enzymatic digest- ibility obtained is greater than the effect of IL temperature. The only case in which IL treatment temperature appears to show a relevant difference is in samples autohydrolyzed at 150 °C. In this sense, treat- ment at lower IL temperature conditions for samples treated at inter- mediate or severe AH conditions (AH175IL80 and AH200IL80) would be justified, saving energy costs in future biorefinery processes. The observations between samples autohydrolized at different conditions verify that conditions that are too severe (as studied at AH samples at 200 °C) do not report higher digestibilities, suggesting that intermediate AH conditions in combination with IL microwave treatments may get considerable high glucose digestibilities. Xylan digestibility for untreated eucalyptus wood and treated samples (excepting samples autohydrolyzed at 200 °C) was also calcu- lated and is shown in Fig. 1b. As occurred in glucan, the differences between untreated and treated samples are remarkable. Xylan digest- ibility varies between 5.7 and 74.3 g xylan/100 g initial glucan in- troduced for untreated eucalyptus wood and AH175IL80 samples, re- spectively. Again, the AH conditions are the main effect that cause variation in xylan digestibilities variations. Differences between sam- ples autohydrolyzed under the same conditions but with different IL temperature treatments varies in the range 6–12%. Although digest- ibilities are shown to be higher in the case of samples autohydrolyzed at 175 °C, the low content of xylan in these samples before hydrolysis (due to hemicellulose extraction under the fixed AH conditions), results in low xylan global yields. Table 1 Solid recovery and compositional analysis. ± values denote standard deviations. Solid Yield Composition (g/100 g ODW treated biomass) g/100 g ODW* untreated wood Lignin Glucan Xylan Galactan Arabinan Mannan Acetate % % % % % % % % Eucalyptus wood 100 27.21 ± 0.20 51.08 ± 1.32 18.14 ± 0.35 0.62 ± 0.10 0.36 ± 0.18 0.83 ± 0.41 4.12 ± 0.11 AH150IL80 84.35 28.86 ± 1.31 46.54 ± 0.61 12.45 ± 0.04 0.63 ± 0.05 0 0.77 ± 0.77 9.00 ± 0.71 AH150IL120 81.82 29.08 ± 0.97 53.83 ± 0.08 12.86 ± 0.32 0.60 ± 0.15 0 1.74 ± 0.87 4.22 ± 0.78 AH175IL80 63.11 30.51 ± 1.43 66.82 ± 1.82 1.92 ± 0.02 0 0 0 1.16 ± 0.26 AH175IL120 62.03 31.21 ± 0.79 61.22 ± 0.85 2.05 ± 0.01 0 0 0 3.33 ± 0.34 AH200IL80 56.33 55.86 ± 1.52 46.05 ± 0.98 0 0 0 0 0 AH200IL120 55.19 52.54 ± 1.71 45.32 ± 0.98 0 0 0 0 0 * ODW: Oven dry weight. V. Rigual et al. Bioresource Technology 251 (2018) 197–203 200 3.4. Biomass accessibility SEM and confocal fluorescence microscopy images may help to elucidate the results obtained after enzymatic hydrolysis treatment. 3.4.1. Surface texture: Fractal dimension and lacunarity The differences between untreated eucalyptus and treated samples are notorious. The overall surface of untreated eucalyptus wood (image not shown) is smoothed with no visible holes in the structure. Treated samples have a quite homogenous macrostructure with more porosity, possibly due to partial dissolution and regeneration of wood fibers in a fusing structure (Sun et al., 2011). The biggest difference visually ob- served in SEM micrographs is for AH200 samples. In these cases, the surfaces are rougher and agglomeration of smaller particles are ob- served. The effect of IL temperature on the solids morphology does not produce clear visual changes, excepting between AH150IL80 and AH150IL120 samples. To get additional information of the textural images obtained, the terms of fractal dimension and lacunarity using the box counting method were calculated for SEM micrographs and results are shown in Table 2. The fractal dimensions values obtained vary in the range 1.5202 (for untreated eucalyptus wood) to 1.7112 for AH175IL80. Although no previous references of this parameter calculation have been found in treated biomass samples, in other kinds of materials, values are in the range of 1.80–1.85 in the case of carbon samples or 1.72–1.75 in ret- rograded starch (Alvarez et al., 2013; Utrilla-Coello et al., 2013). The bigger a fractal dimension is, the more convolutions a porous system has (i.e., twists and turns) and the higher the complexity is (Alvarez et al., 2013; Karperien, 2013). It is curious to observe that fractal di- mensions obtained for AH175IL80, AH175IL120, AH200IL80 and AH200IL120 have very similar values (approximately 1.71), in the same way that 72-h glucan digestibility values are similar. Lacunarity values are in the range of 0.6955 (for untreated sample) to 0.3410 for AH200IL120. These values are in the range of those ob- tained with different samples such as rocks (0.05–0.20), or carbon samples (0.15–0.28) (Alvarez et al., 2013; Liu and Ostadhassan, 2017). In this case, the lower the lacunarity is, the more homogenous textures are obtained, as has been visually confirmed. Again, lacunarity values of AH150IL120, AH175IL80, AH175IL120, AH200IL80 and AH200IL120 are quite similar among themselves, and their reproduci- bility shows low standard deviations between replicates, while in the case of untreated eucalyptus wood and AH150IL80 samples, the results are more heterogeneous, and the standard deviation is greater. As a consequence, lacunarity and fractal dimension of each sample are also related between themselves. The higher the lacunarity is, the lower fractal dimensions are obtained. In this way, fractal dimension and la- cunarity could be correlated against glucan digestibility and become a parameter to take into account in the future as a measurement of cel- lulose accessibility for enzymatic hydrolysis. However, further studies should be developed in order to verify this tendency in other treatment technologies. 3.4.2. Surface chemical distribution Confocal fluorescence microscopy expose lignin and holocellulose surface distribution. Untreated eucalyptus wood surfaces show an or- dered and crosslinked lignin and holocellulose distribution, and both components, corresponding to both wavelengths ranges detections are not visually observed to be mixed. As ocurred with other analyses, the effect of the AH treatment results in more visual differences than the effect of the IL treatment temperature. Samples autohydrolyzed at mild conditions, show an altered structure (already observed in SEM images) where lignin and holocellulose have been rearranged. Hemicelluloses are still present in the sample which may explain the superficial holo- celluloses observed. At intermediate AH conditions, samples are mainly composed of lignin in the surface. Lignin has been noted as an in- hibitory agent of cellulases (Rahikainen et al., 2013). However, di- gestibilities observed under these conditions demonstrate that eu- calyptus wood digestibility is not decreased due to the presence of lignin in the surface. Samples autohydrolyzed under severe conditions show particles agglomeration. The transmitted coupled channel shows that not the whole particle is observed under fluorescence. This may be attributed to the opacity of the sample that limits laser penetration, Fig. 1. a) Glucan digestibility and b) xylan digestibility. Percentages are expressed in grams of glucan/xylan per 100 g of initial glucan/ xylan introduced in the enzymatic hydrolysis step. Table 2 Fractal dimension (Db), lacunarity (Λ) and 72 h digestibility of untreated and treated samples. The Db and Λ results were calculated using the Fraclac plugin from ImageJ software and the box counting method. ± values denotes standard deviations. Sample Fractal dimension (Db) Lacunarity (Λ) 72 h Enzymatic digestibility Untreated eucalyptus wood 1.5202 ± 0.0859 0.6955 ± 0.1200 15.05 ± 0.00 AH150IL80 1.6258 ± 0.0447 0.4976 ± 0.1357 68.35 ± 1.14 AH150IL120 1.6724 ± 0.0163 0.3933 ± 0.0354 74.23 ± 1.43 AH175IL80 1.7112 ± 0.0304 0.3470 ± 0.0351 84.41 ± 1.02 AH175IL120 1.7001 ± 0.0201 0.3735 ± 0.0340 86.46 ± 0.57 AH200IL80 1.7042 ± 0.0230 0.3824 ± 0.0554 81.81 ± 1.05 AH200IL120 1.7080 ± 0.0225 0.3410 ± 0.0287 83.11 ± 1.47 V. Rigual et al. Bioresource Technology 251 (2018) 197–203 201 probably due to the high content of lignin (dark). Despite of this opa- city, superficial cellulose is also observed, which indicates that the pretreatment combination have altered the internal part of the cell wall and part of the recalcitrance have been reduced. 3.5. Process mass balance Fig. 2 summarizes the calculations developed in the whole process with chromatographic, gravimetric, and NREL analyses and spectro- scopy techniques. Eucalyptus wood has been divided into three main components: cellulose, hemicellulose and lignin. Minimum, and max- imum values of each component have been included as reference. The results underlined corresponds to AH175IL80, which had the highest glucose yield (69.7 g glucose in eucalyptus/100 g glucose in the global process). The highest xylose yield was obtained in the AH150IL120 sample as a consequence of the low hemicellulose removal at mild AH conditions. Due to pseudo lignin formation in some samples, lignin extraction was not included in AH200IL80 and AH200IL120 samples. The global mass balance shows the importance of calculating global yield throughout the process. In this sense, the sample with maximum glucan digestibility (AH175IL120) and maximum glucose yield (AH175IL80) is not the same. From a biorefinery perspective, the combination of autohydrolysis and ionic liquid pretreatment can be a model for low environmental impact pretreatment combination. Despite of the high cost of these processes, these two pretreatments fulfil the disadvantages of some of the other current alternatives (acid hydrolysis, alkaline pretreatment, or steam explosion): the low corrosion the materials to be used and the low generation of degradation products that lower yields and worsen purification processes (Elgharbawy et al., 2016; Heggset et al., 2016). The overall mass balance has shown that lignin has been over- estimated in the case of AH200 samples, probably due to “pseudo- lignin” that cannot be distinguished from Klason lignin (Huijgen et al., 2012). This overestimation was observed in the solid characterization analysis and in the lignin quantification in the IL. 4. Conclusions In this work, combinations of autohydrolysis and microwave ionic liquid treatments combination, towards sugars production maximiza- tion from eucalyptus wood, were tested. The highest glucan digest- ibilities were obtained for AH175IL80 and AH175IL120 samples (84.4 and 85.5%). The AH175IL80 sample yielded the highest glucan pro- duction (39.6 g of glucose/100 g of untreated wood). Fractal dimension and lacunarity were found to be promising parameters to understand biomass accessibility. Confocal fluorescence microscopy images showed that superficial lignin does not limit glucan digestibility. Mass balance evinced the importance of quantifying solid recoveries along the pro- cess towards a realistic glucose quantification in the global process. Acknowledgement This work was performed thanks to the financial support of the “Ministerio de Economía y Competitividad” under the funded project CTQ2013-42006-R and the contract BES-2014-067788. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.biortech.2017.12.034. References Akhtar, N., Gupta, K., Goyal, D., Goyal, A., 2015. Recent advances in pretreatment technologies for efficient hydrolysis of lignocellulosic biomass. Environ. Prog. Sustain. Energy 35 (2), 489–511. Alvarez, A., Passé-Coutrin, N., Gaspard, S., 2013. Determination of the textural char- acteristics of carbon samples using scanning electronic microscopy images: compar- ison with mercury porosimetry data. Adsorption 19 (2–4), 841–850. Auxenfans, T., Buchoux, S., Husson, E., Sarazin, C., 2014. Efficient enzymatic sacchar- ification of Miscanthus: energy-saving by combining dilute acid and ionic liquid pretreatments. Biomass Bioenergy 62, 82–92. Brandt, A., Gschwend, F., Fennell, P., Lammens, T., Tan, B., Weale, J., Hallett, J., 2017. An economically viable ionic liquid for the pretreatment of lignocellulosic biomass. Green Chem. 19, 3078–3102. Carrillo, I., Vidal, C., Elissetche, J.P., Mendonça, R.T., 2017. Wood anatomical and che- mical properties related to the pulpability of Eucalyptus globulus: a review. South. For. 1–8. Casas, A., Alonso, M.V., Oliet, M., Santos, T.M., Rodriguez, F., 2013. Characterization of cellulose regenerated from solutions of pine and eucalyptus woods in 1-allyl-3-me- thilimidazolium chloride. Carbohydr. Polym. 92 (2), 1946–1952. Chen, H., Han, Y., Xu, J., 2008. Simultaneous saccharification and fermentation of steam exploded wheat straw pretreated with alkaline peroxide. Process Biochem. 43 (12), 1462–1466. Demartini, J.D., Foston, M., Meng, X., Jung, S., Kumar, R., Ragauskas, A.J., Wyman, C.E., 2015. How chip size impacts steam pretreatment effectiveness for biological con- version of poplar wood into fermentable sugars. Biotechnol. Biofuels 8 (1). Elgharbawy, A.A., Alam, M.Z., Moniruzzaman, M., Goto, M., 2016. Ionic liquid pre- treatment as emerging approaches for enhanced enzymatic hydrolysis of lig- nocellulosic biomass. Biochem. Eng. J. 109, 252–267. Fang, Z., Smith, R.L., Qi, X. (Eds.), 2015. Production of Biofuels and Chemicals with Ionic Liquids. Springer, New York. García-Domínguez, M.T., García-Domínguez, J.C., Feria, M.J., Gómez-Lozano, D.M., López, F., Díaz, M.J., 2013. Furfural production from Eucalyptus globulus: optimizing by using neural fuzzy models. Chem. Eng. J. 221, 185–192. Ha, S.H., Mai, N.L., An, G., Koo, Y.-M., 2011. Microwave-assisted pretreatment of cellu- lose in ionic liquid for accelerated enzymatic hydrolysis. Bioresour. Technol. 102 (2), 1214–1219. Hauru, L.K.J., Ma, Y., Hummel, M., Alekhina, M., King, A.W.T., Kilpelainen, I., Penttila, P.A., Serimaa, R., Sixta, H., 2013. Enhancement of ionic liquid-aided fractionation of Fig. 2. Mass balance per 100 g of untreated eucalyptus wood during sequential autohydrolysis and IL microwave treatment. Underlined values represent results for the highest glucose productions (AH175IL80). V. Rigual et al. Bioresource Technology 251 (2018) 197–203 202 http://dx.doi.org/10.1016/j.biortech.2017.12.034 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0005 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0005 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0005 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0010 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0010 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0010 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0015 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0015 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0015 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0020 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0020 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0020 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0025 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0025 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0025 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0030 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0030 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0030 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0035 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0035 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0035 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0040 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0040 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0040 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0045 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0045 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0045 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0050 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0050 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0055 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0055 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0055 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0060 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0060 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0060 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0065 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0065 birchwood. Part 1: autohydrolysis pretreatment. RSC Adv. 3 (37), 16365–16373. Heggset, E.B., Syverud, K., Øyaas, K., 2016. Novel pretreatment pathways for dissolution of lignocellulosic biomass based on ionic liquid and low temperature alkaline treat- ment. Biomass Bioenergy 93, 194–200. Huijgen, W., Smit, A., De Wild, P., Den Uil, H., 2012. Fractionation of wheat straw by prehydrolysis, organosolv delignification and enzymatic hydrolysis for production of sugars and lignin. Bioresour. Technol. 114, 389–398. Hyvärinen, S., Mikkola, J.-P., Murzin, D.Y., Vaher, M., Kaljurand, M., Koel, M., 2014. Sugars and sugar derivatives in ionic liquid media obtained from lignocellulosic biomass: comparison of capillary electrophoresis and chromatographic analysis. Catal. Today 223, 18–24. Ibáñez, A.B., Bauer, S., 2014. Downscaled method using glass microfiber filters for the determination of Klason lignin and structural carbohydrates. Biomass Bioenergy 68, 75–81. Isikgor, F.H., Becer, C.R., 2015. Lignocellulosic biomass: a sustainable platform for the production of bio-based chemicals and polymers. Polym. Chem. 6 (25), 4497–4559. Karperien, A. 2013. FracLac for Image J. vol. 2017, FracLac. https://imagej.nih.gov/ij/ plugins/fraclac/FLHelp/Introduction.htm. Kilic, K.I., Abiyev, R.H., 2011. Exploiting the synergy between fractal dimension and lacunarity for improved texture recognition. Signal Process. 91 (10), 2332–2344. Li, Q., He, Y.-C., Xian, M., Jun, G., Xu, X., Yang, J.-M., Li, L.-Z., 2009. Improving enzy- matic hydrolysis of wheat straw using ionic liquid 1-ethyl-3-methyl imidazolium diethyl phosphate pretreatment. Bioresour. Technol. 100 (14), 3570–3575. Liu, J., Li, M., Luo, X., Chen, L., Huang, L., 2015. Effect of hot-water extraction (HWE) severity on bleached pulp based biorefinery performance of eucalyptus during the HWE–Kraft–ECF bleaching process. Bioresour. Technol. 181, 183–190. Liu, K., Ostadhassan, M., 2017. Quantification of the microstructures of Bakken shale reservoirs using multi-fractal and lacunarity analysis. J. Nat. Gas Sci. Eng. 39, 62–71. Lynam, J.G., Coronella, C.J., 2014. Glycerol as an ionic liquid co-solvent for pretreatment of rice hulls to enhance glucose and xylose yield. Bioresour. Technol. 166, 471–478. Mandelbrot, B. 1977. Fractals, Form, Chance and Dimension WH Freeman and Co. Mandelbrot, B.B., Pignoni, R. 1983. The fractal geometry of nature. WH freeman New York. Neves, L.A., Nascimento, M.Z., Oliveira, D.L.L., Martins, A.S., Godoy, M.F., Arruda, P.F.F., de Santi Neto, D., Machado, J.M., 2014. Multi-scale lacunarity as an alternative to quantify and diagnose the behavior of prostate cancer. Expert Syst. Appl. 41 (11), 5017–5029. Nguyen, T.-A.D., Kim, K.-R., Han, S.J., Cho, H.Y., Kim, J.W., Park, S.M., Park, J.C., Sim, S.J., 2010. Pretreatment of rice straw with ammonia and ionic liquid for lig- nocellulose conversion to fermentable sugars. Bioresour. Technol. 101 (19), 7432–7438. Nitsos, C.K., Choli-Papadopoulou, T., Matis, K.A., Triantafyllidis, K.S., 2016. Optimization of hydrothermal pretreatment of hardwood and softwood lignocellulosic residues for selective hemicellulose recovery and improved cellulose enzymatic hydrolysis. ACS Sustain. Chem. Eng. 4 (9), 4529–4544. Perez-Pimienta, J.A., Flores-Gómez, C.A., Ruiz, H.A., Sathitsuksanoh, N., Balan, V., da Costa Sousa, L., Dale, B.E., Singh, S., Simmons, B.A., 2016. Evaluation of agave ba- gasse recalcitrance using AFEX™, autohydrolysis, and ionic liquid pretreatments. Bioresour. Technol. 211, 216–223. Pérez-Pimienta, J.A., Vargas-Tah, A., López-Ortega, K.M., Medina-López, Y.N., Mendoza- Pérez, J.A., Avila, S., Singh, S., Simmons, B.A., Loaces, I., Martinez, A., 2017. Sequential enzymatic saccharification and fermentation of ionic liquid and organo- solv pretreated agave bagasse for ethanol production. Bioresour. Technol. 225, 191–198. Procentese, A., Johnson, E., Orr, V., Garruto Campanile, A., Wood, J.A., Marzocchella, A., Rehmann, L., 2015. Deep eutectic solvent pretreatment and subsequent sacchar- ification of corncob. Bioresour. Technol. 192, 31–36. Raghavi, S., Sindhu, R., Binod, P., Gnansounou, E., Pandey, A., 2016. Development of a novel sequential pretreatment strategy for the production of bioethanol from su- garcane trash. Bioresour. Technol. 199, 202–210. Rahikainen, J.L., Martin-Sampedro, R., Heikkinen, H., Rovio, S., Marjamaa, K., Tamminen, T., Rojas, O.J., Kruus, K., 2013. Inhibitory effect of lignin during cellulose bioconversion: the effect of lignin chemistry on non-productive enzyme adsorption. Bioresour. Technol. 133, 270–278. Resch, M., Baker, J., Decker, S. 2015. Low solids enzymatic saccharification of lig- nocellulosic biomass. NREL Laboratory Analytical Procedure. Rissanen, J.V., Murzin, D.Y., Salmi, T., Grénman, H., 2016. Aqueous extraction of hemicelluloses from spruce – from hot to warm. Bioresour. Technol. 199, 279–282. Romaní, A., Ruiz, H.A., Pereira, F.B., Teixeira, J.A., Domingues, L., 2014. Integrated approach for effective bioethanol production using whole slurry from autohydrolyzed Eucalyptus globulus wood at high-solid loadings. Fuel 135, 482–491. Silva-Fernandes, T., Duarte, L.C., Carvalheiro, F., Marques, S., Loureiro-Dias, M.C., Fonseca, C., Gírio, F., 2015. Biorefining strategy for maximal monosaccharide re- covery from three different feedstocks: Eucalyptus residues, wheat straw and olive tree pruning. Bioresour. Technol. 183, 203–212. Silveira, M.H.L., Vanelli, B.A., Corazza, M.L., Ramos, L.P., 2015. Supercritical carbon dioxide combined with 1-butyl-3-methylimidazolium acetate and ethanol for the pretreatment and enzymatic hydrolysis of sugarcane bagasse. Bioresour. Technol. 192, 389–396. Singh, R., Krishna, B.B., Kumar, J., Bhaskar, T., 2016. Opportunities for utilization of non- conventional energy sources for biomass pretreatment. Bioresour. Technol. 199, 398–407. Singh, S., Simmons, B.A., Vogel, K.P., 2009. Visualization of biomass solubilization and cellulose regeneration during ionic liquid pretreatment of switchgrass. Biotechnol Bioeng 104 (1), 68–75. Sluiter, A., Hames, B., Ruiz, R., Scarlata, C., Sluitr, J., Templeton, D., Crocker, D., 2011. Determination of structural carbohydrates and lignin in biomass. NREL. Smith Jr, T.G., Lange, G.D., Marks, W.B., 1996. Fractal methods and results in cellular morphology— dimensions, lacunarity and multifractals. J. Neurosci. Methods 69 (2), 123–136. Sun, N., Rodríguez, H., Rahman, M., Rogers, R.D., 2011. Where are ionic liquid strategies most suited in the pursuit of chemicals and energy from lignocellulosic biomass? Chem. Commun. 47 (5), 1405–1421. Sun, S., Sun, S., Cao, X., Sun, R., 2016. The role of pretreatment in improving the en- zymatic hydrolysis of lignocellulosic materials. Bioresour. Technol. 199, 49–58. Utrilla-Coello, R.G., Bello-Pérez, L.A., Vernon-Carter, E.J., Rodriguez, E., Alvarez- Ramirez, J., 2013. Microstructure of retrograded starch: quantification from lacu- narity analysis of SEM micrographs. J. Food Eng. 116 (4), 775–781. Weerachanchai, P., Lee, J.-M., 2014. Recyclability of an ionic liquid for biomass pre- treatment. Bioresour. Technol. 169, 336–343. Weerachanchai, P., Lim, K.H., Lee, J.-M., 2014. Influence of organic solvent on the se- paration of an ionic liquid from a lignin–ionic liquid mixture. Bioresour. Technol. 156, 404–407. Xu, J., Xiong, P., He, B., 2016. Advances in improving the performance of cellulase in ionic liquids for lignocellulose biorefinery. Bioresour. Technol. 200, 961–970. Zhu, C., Richardson, R.M., Potter, K.D., Koutsomitopoulou, A.F., Van Duijneveldt, J.S., Vincent, S.R., Wanasekara, N.D., Eichhorn, S.J., Rahatekar, S.S., 2016. High modulus regenerated cellulose fibers spun from a low molecular weight microcrystalline cel- lulose solution. ACS Sustain. Chem. Eng. 4 (9), 4545–4553. V. Rigual et al. Bioresource Technology 251 (2018) 197–203 203 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0065 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0070 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0070 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0070 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0075 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0075 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0075 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0080 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0080 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0080 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0080 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0085 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0085 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0085 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0090 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0090 https://imagej.nih.gov/ij/plugins/fraclac/FLHelp/Introduction.htm https://imagej.nih.gov/ij/plugins/fraclac/FLHelp/Introduction.htm http://refhub.elsevier.com/S0960-8524(17)32167-3/h0100 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0100 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0105 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0105 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0105 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0110 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0110 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0110 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0115 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0115 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0120 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0120 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0135 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0135 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0135 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0135 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0140 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0140 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0140 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0140 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0145 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0145 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0145 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0145 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0150 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0150 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0150 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0150 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0155 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0155 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0155 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0155 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0155 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0160 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0160 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0160 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0165 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0165 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0165 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0170 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0170 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0170 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0170 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0180 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0180 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0185 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0185 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0185 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0190 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0190 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0190 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0190 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0195 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0195 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0195 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0195 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0200 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0200 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0200 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0205 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0205 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0205 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0210 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0210 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0215 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0215 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0215 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0220 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0220 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0220 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0225 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0225 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0230 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0230 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0230 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0235 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0235 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0240 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0240 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0240 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0245 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0245 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0250 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0250 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0250 http://refhub.elsevier.com/S0960-8524(17)32167-3/h0250 S1 Supplementary data Title: Combining autohydrolysis and ionic liquid microwave treatment of Eucalyptus globulus wood Authors: Victoria Rigual, Tamara M. Santos, Juan C. Domínguez, M. Virginia Alonso, Mercedes Oliet and Francisco Rodriguez Affiliation and address: Department of Chemical Engineering, Faculty of Chemistry, Complutense University of Madrid, Avda. Complutense s/n, 28040 Madrid, Spain. Corresponding author: Victoria Rigual E-mail: vicrigua@ucm.es ; Tel: +34913948505 Figure S1: IL recovered in the washing step mailto:vicrigua@ucm.es S2 Figure S2: Lignin accumulation in the IL Figure S3: Recovered IL in samples from left to right AH150IL80, AH150IL120, AH175IL80, AH175IL120, AH200IL80 and AH200IL120 S3 Figure S4: a) SEM and b) Confocal fluorescence microscopy images of untreated eucalyptus wood a) b) PUBLICACIÓN VI “Autohydrolysis and microwave ionic liquid pretreatment of Pinus radiata: imaging visualization and analysis to understand enzymatic digestibility” Victoria Rigual, Juan C. Domínguez, Tamara M. Santos, Sandra Rivas, M. Virginia Alonso, Mercedes Oliet y Francisco Rodriguez Enviado a Industrial Crops and Products 1 Autohydrolysis and microwave ionic liquid pretreatment of Pinus radiata: imaging visualization and analysis to understand enzymatic digestibility Victoria Rigual*, Juan C. Domínguez, Tamara M. Santos, Sandra Rivas, M. Virginia Alonso, Mercedes Oliet and Francisco Rodriguez Department of Chemical Engineering and Materials Science, Faculty of Chemistry, Complutense University of Madrid, Avda. Complutense s/n, 28040 Madrid, Spain. *E-mail: vicrigua@ucm.es ; Tel: +34913948505 GRAPHICAL ABSTRACT ABSTRACT Sequential pretreatments combine advantages of different pretreatment towards a complete biomass fractionation. In this work, autohydrolysis and IL microwave pretreatments are combined and their effect to enhance enzymatic hydrolysis is studied. Mild and severe autohydrolysis are combined with four IL temperatures (50, 80, 120 and 150 °C). Pretreated solids are enzymatically hydrolyzed and compared in terms of chemical composition and morphology. Digital image analysis is employed to numerically determine the heterogeneity of the solids using fractal dimension and laccunarity parameters. In this study, the negative effect of severe autohydrolysis over the subsequent IL pretreatment is demonstrated. Mild autohydrolysis and high IL conditions (AH150IL120) results in digestibilities of 78.8 g of glucan/100 g of glucan introduced. High fractal values (in the range of 2.5461-2.7124) and low laccunarities (0.0818-0.2563) enhance the enzymatic accessibility of pine wood. Furthermore, the negative effect of softwood lignin accumulation in the surface is observed using confocal fluorescence microscopy. Keywords: Softwood, autohydrolysis, ionic liquid, fractal dimension, enzymatic hydrolysis Abbreviations: autohydrolysis (AH), ionic liquid (IL); fractal dimensión (Db), Digital image analysis (DIA) mailto:vicrigua@ucm.es 2 1. Introduction The biorefinery concept aims at the selective separation of its components and its versatile utilization to produce high-value added biomass-derived compounds, using clean processes (Ståhl et al., 2018). Apart from converting carbohydrates into biofuels, the implementation of other coproducts provides a promising opportunity to improve the economic viability, sustainability and overall carbon conversion (Wang et al., 2017). For this purpose, the selection of the most adequate technology to process each type of feedstock is essential. Variations in the nature of every feedstock, originate pretreatments behave differently (Li et al., 2013b). Woody biomass is one of the candidates to become in raw material in a biorefinery, due to its properties, availability and versatility (Nitsos et al., 2016) (Huang et al., 2018). The most abundant wood in the Northern Hemisphere is softwood, and pine in particular (loblolly pine, lodgepole pine, scots pine, insignis pine, and maritime pine) (Wang et al., 2017) (Kandhola et al., 2017; Li et al., 2013a). A lot of countries count with well-established systems for the sustainable production of these forests, and use wood as a raw material in the pulp, paper, and furniture-industry, leaving a considerable portion of waste behind from harvesting that can be benefitted (Kandhola et al., 2017). Despite of the availability of pine, its resistance to biocatalitic conversion processes (sugar extraction through enzymatic hydrolysis) is lower than hardwood or herbaceous, due to its recalcitrance (Hendriks and Zeeman, 2009; Zhang et al., 2016). To overcome the hardness of woody biomass many pretreatments have been proposed, such as acid, alkali, steam explosion, organosolv or autohydrolysis (Trinh et al., 2015). Autohydrolysis pretreatment only uses hot water as solvent at temperatures below 230 °C and low autogenous pressures (Nitsos et al., 2016). Water autoionization generates acidic and basic hydronium, and hydroxide ions (H3O + and OH-) that enhances hemicellulose extraction and partial lignin recovery (Moniz et al., 2013; Ståhl et al., 2018). The main benefits of autohydrolysis are the use of inexpensive and non-hazardous chemicals, corrosion minimization, cost effectivity, and environmentally friendliness (Nitsos et al., 2016). During the last decades, ionic liquids (ILs) were considered as promising solvents in numerous research areas. The low vapour pressure, non- flammability, and switchability are some of the strongest points of ILs (Trinh et al., 2015). The employment of ILs has been widely employed in biomass to obtain fermentable sugars, furans, and lignin. In the utilization of ILs for wood dissolution, factors such as wood type, particle size, solvent system, water content, and temperature must be taken into account (Kilpeläinen et al., 2007; Mäki- Arvela et al., 2010; van Osch et al., 2017). Apart from conventional heating, other alternatives such as microwave pulses or ultrasound radiation accelerate the dissolution process with ionic ILs (Liu et al., 2014; Sun et al., 2009b; Sun et al., 2009a) (Sun et al., 2009a). In the conversion of biomass, the IL 1-ethyl-3-methylimidazolium acetate ([Emim]OAc) is one of the most used IL because it 3 reduces crystallinity and biomass recalcitrance (Yoo et al., 2017). However, high cost of this IL and lack of effective IL recycling strategies are challenges to address (Kandhola et al., 2017). In the last years, biomass fractionation to obtain hemicellulose, lignin and accessible-to- enzymes cellulose is being studied using sequential or combined pretreatments. The combination of physical (ammonia fiber explosion, autohydrolysis or steam explosion) and chemical (alkaline peroxide) pretreatments have been tested in various biomasses (Akhtar et al., 2015; Chen et al., 2008). Some studies have also combined ILs pretreatments with supercritical CO2, ammonia or low temperature alkaline pretreatment (Hauru et al., 2013; Heggset et al., 2016; Nguyen et al., 2010; Silveira et al., 2015). Autohydrolysis pretreatment alters lignin structure, but does not decrystallize cellulose or increase the surface area. On the contrary, ILs enhance accessibility towards enzymatic saccharification, decristallizing cellulose. The combination of these two pretreatments may improve some of the factors that the ideal pretreatment must achieve (Kandhola et al., 2017). Focusing on saccharification, a deep understanding of the mechanisms that favors enzyme accessibility is crucial. The accessibility can be affected by cellulose crystallinity, content and distribution of lignin and hemicellulose, and surface morphology (Sun et al., 2016).To study these factors, SEM, TEM, or AFM are techniques very relevant for a better understanding and insights into the structure that cannot be obtained from other analysis (Karimi and Taherzadeh, 2016). However, lignocelluloses have a multi-scale complexity that limits the conclusions extracted from microscopy images (Karimi and Taherzadeh, 2016). In a previous work, the application of digital image analysis (DIA) of SEM micrographs to convert these up-to-date qualitative techniques into numeric values (Rigual et al., 2018). The calculation of independent-of-the-multi-scale parameters, such as fractal dimension was initially developed, but more data of different biomasses and pretreatments need to be calculated for a better correlation and validation (Chen et al., 2017; Rigual et al., 2018). In this work, sequential combination of autohydrolysis and IL microwave pretreatment of Pinus radiata is studied. The effect of severity factor (temperature and time) is evaluated in terms of wood digestibility and most appropriate conditions towards saccharification. The washing step to remove the IL from the solid is optimized to minimize water consumption. Digestibility results are related to chemical composition and also morphological parameters. Digital image anlaysis is employed to measure fractal dimension, laccunarity, and particle surface composition in relation to the accessibility of the material. 4 2. Materials and methods 2.1. Materials and reagents 1-ethyl-3-methylimidazoliumacetate ([Emim][OAc], > 98 %, Iolitec GmbH) was employed for wood dissolution. Previously, [Emim][OAc] was dried in a vacuum oven for 24 h at 60 ˚C. Pinus radiata sawdust was provided by the Instituto Nacional de Investigación y Tecnología Agraria y Alimentaria (CIFOR-INIA). Accellerase 1500® enzymes cocktail was kindly provided by DuPont Industrial Biosciences. 2.2. Sequential autohydrolysis + microwave IL pretreatments Autohydrolysis pretreatments were performed using a 450 mL stainless steel Parr pressure reactor, model 4567, fitted with a four-blade turbine impeller. The reactor was heated by an external mantle and cooled by an internal stainless steel loop using a100 cSt silicone fluid oil. A 4848 Parr PID controller was used to control the temperature. Pinus radiata wood sawdust was mixed with deionized water in a liquid-to-solid ratio of 10:1 (g water: g dry biomass). Two different conditions were used for autohydrolysis pretreatment. Mild and severe autohydrolysis were carried out for 30 minutes at 150 °C, and for 90 minutes at 200 °C, respectively. A 4 % (w/w) mixtures of autohydrolyzed Pinus radiata (particle size < 150 µm) and [Emim][OAc] were introduced in teflon vials of a Berghof SpeedWave Four microwave oven in a two-steps programme up to the operation temperature, in a total time of 20+30 min (total time=50 min) (Casas et al., 2013). Operation temperatures were 50, 80, 120, and 150 °C. Consequently, deionized water was added to precipitate the pretreated wood. To determine the minimum water washing consumption required, treated samples were washed 5 times with 70 mL of deionized water, separating every fraction for analysis (Rigual et al., 2018). Recovered treated samples were stored at 40 °C. Experiments nomenclature was designed according to the temperature of the AH pretreatment (AH150 or AH200) and the IL pretreatment temperature (IL50, IL80, IL120, IL150). 2.3. Enzymatic hydrolysis Enzymatic saccharification at a solid loading of 1% (w/w) was carried out in a total volume of 8 mL in an orbital incubator at 150 rpm and 50 °C according to NREL/TP-5100- 63351 procedure (Resch et al., 2015). The enzyme loading was normalized in order to homogenize the glucan content according to the NREL/TP-510-42618 protocol (A. Sluiter, 2011). The enzymatic cocktail Accellerase 1500® was added in a dosage of 0.25 mL of enzymatic 5 cocktail/g glucan in the untreated or pretreated wood. Untreated/pretreated wood was suspended in 50mM citrate buffer (pH 5.0) containing 0.002 % of sodium azide, in a volume of 8 mL. Aliquots of 150 µL were periodically taken at 3, 6, 12, 24, 48, and 72 h and centrifuged twice at 9000 rpm for 10 minutes. Glucose was measured using the above- mentioned method for sugars analysis employing Carbosep CHO-682 column. Glucan digestibility was calculated according to Eq. 3. Digestibility (%)= hydrolysed glucan (g) glucan introduced (g) × 100 (3) Anhydro correction of 0.9 was applied to glucose concentration obtained, to determine glucan content. 2.4. Analytical procedure 2.4.1. Characterization of IL liquid by-stream To determine the minimum amount of water needed to remove IL from the pretreated solid, washing fractions were filtrated and analyzed by HPLC equipped with a UV detector measuring the [Emim][OAc] absorbance at 235 nm and an Eclipse Plus C18 column. The column was operated at 30 °C using a flow rate of 1 mL/min of a mixture of acetonitrile/ water 50/50 (v/v). Lignin presence in the IL by-stream was analyzed using a Varian Cary 50 scan UV/VIS spectrophotometer at 280 nm. Organosolv lignin from Pinus radiata wood was used as reference material, obtained with a mixture of 60 % (w/w) ethanol/water at 200 ˚C for 50 minutes. Sulfuric acid was used to precipitate the lignin from the mixture. For sample preparation, vacuum distillation was employed to remove water and recover the IL. Final water content in the recovered IL was calculated using a thermobalance. IL by-stream samples were diluted in 0.1 N NaOH. The total dissolved lignin concentration was obtained from the reference curve of pine wood organosolv lignin samples prepared. (Weerachanchai and Lee, 2014) A blank was employed, prepared at the IL operating conditions in the pretreatment. The lignin content in the recovered [Emim][OAc] and the lignin extracted were calculated by Eqs. 1 and 2. Lignin in [Emim][OAc] (wt %)= Lignin in IL (g) IL (g) x100 (1) Cumulative lignin (wt %)= Lignin in IL (g) Lignin introduced (g) x100 (2) 6 2.4.2. Pretreated solids compositional analysis Untreated/pretreated solid compositions were determined according to the National Renewable Energy Laboratory methodology (NREL/TP-510-42618) adapted to smaller quantities of samples (500 mg of biomass employed for the characterization) (Ibáñez and Bauer, 2014; Sluiter et al., 2008). Solids were treated in a two-step acid hydrolysis resulting in an insoluble residue compound by acid-insoluble lignin and ashes, and a liquid fraction containing soluble lignin and sugars. The acid insoluble lignin was gravimetrically measured after drying the residue for 12 h at 105 ˚C. The amount of acid-soluble lignin was determined from the UV absorbance of the hydrolysate at 240 nm and an absorptivity of 12 L·g-1·cm-1. Sugars in the hydrolysate were determined by HPLC using a RI detector equipped with a Carbosep CHO-682 column with Micro-Guard cartridges (BioRad, Life cience Group Hercules, Ca), using a flow rate of 0.4 mL/min at 80 °C. Acetic acid was measured in a Rezex ROA-Organic Acid H+ (8 %) 300x7.8 mm column at 60 °C with a mobile phase (0.005 M H2SO4) eluted at 0.6 mL/min. The anhydrous correction to hexoses (162/180) and pentoses (132/150) were applied to each sugar concentration, respectively. Sugar recovery standards (SRS) correction was also applied following the NREL procedure. 2.4.3. Surface morphology Untreated/pretreated samples were coated with gold utilizing a Q150T Turbo- pumped sputter Coater. The morphological structure was observed by SEM in a Jeol JSM 6400 operating at 20 kV. Analyses were developed in the technical facilities of the Spanish National Centre for Electron Microscopy. Every x1000 micrograph was cropped to select different 300x300 pixels images, ensuring that every image was completely composed by a texture of the particle. Grayscale images were brightness and contrast adjusted manually. Fractal dimension (Db) and laccunarity (Ʌ) were calculated using the box counting method and grayscale micrographs setting the diferencial volume variation scans. This option defines 3d volumes over a grayscale image, and pixels’ intensity (from 0 to 255) are included to count the whole pattern as a volume (Karperien, 2013) (Alvarez et al., 2013; Rigual et al., 2018). The fractal dimension calculated was obtained from the log-log regression line of the sum of all intensities. The grid design had 100 grid positions. The smallest grid was 1 and the largest 45 % of the image. ImageJ open source software was employed to convert micrographs and calculate Db and Ʌ. 2.4.4. Surface composition Confocal fluorescence microscopy was employed to distinguish holocellulose and lignin employing a Leica SP-2 AOBS in the surface as previously reported (Rigual et al., 2018). Samples were dyed with Calcofluor white stain (0.1 %) and fixed in dark with prolong 7 Gold antifade mountant (Pérez-Pimienta et al., 2017). Emission wavelength ranges are 428- 480 nm for hollocellulose and 547-658 nm for lignin, while a 405 nm laser was employed to excite samples. It must be highlighted that, in order to compare properly images saturation, detection and emission wavelength ranges and pinhole were fixed. Furthermore, all the samples were dyed under the same conditions and stored the same period of time before the analysis (Zinchuk et al., 2007). The Z-stack images were projected into the max Z projection using ImageJ software. Projected channels were combined to obtain the merged image. Afterwards, RGB channels were separated. Blue channel was omitted, while red and green channels were assigned to lignin and hollocellulose (cellulose and hemicellulose), respectively. Both channels images (1024x1024 pixels per image) were brightness and contrast adjusted and thresholded individually using the Otsu method (Otsu, 1979). Finally, number of pixels in every binary image were counted separately. 3. Results and discussion 3.1. IL removal in the pretreated solid The recovered IL after each washing as well as the cumulative recovered IL (%) is shown in Fig. A1. IL recovery is mainly produced during the solid precipitation step. However, it is necessary to wash with water at least three-fold, which is equivalent to 10.5 mL H2O/g [Emim][OAc], to ensure the removal of the IL from solid fractions. [Emim][OAc] accumulation in the solid can lead into mistakes in the global mass balances (due to overweighting solid fractions) and cause a decrease in the enzymatic hydrolysis yields as a result of cellulose inactivation (Xu et al., 2016). 3.2. Lignin accumulation in the IL The accumulated lignin expressed in g lignin/100 g IL (eq. 1) and g lignin/100 g lignin (eq. 2) is shown in Fig. 1. Lignin extracted in the liquid fraction is very low concentrated and reach up to 0.18 g lignin/100 g ionic liquid introduced under the most severe conditions tested (AH200IL150). In both autohydrolysis conditions tested, the higher the temperature of the IL pretreatment is, the higher the amount of lignin is extracted. 8 Figure 1. Lignin content in the IL. Results are expressed in g lignin/100g IL (Eq. 1) and in g lignin/100 g AH lignin (Eq. 2). Severe autohydrolysis leads to higher quantities of lignin extracted (up to 0.18 g of lignin/100 g of IL). However, the relative quantity of lignin extracted at mild autohydrolysis and IL pretreatment at high temperatures (AH150IL150) is 10.9 g/100 g of lignin and is superior in relation to the same case for severe autohydrolysis conditions (9.4 g/100 g of lignin). This behavior can be a result of the partial lignin recondensation in the autohydrolysis step (Shi et al., 2016). The quantity of lignin accumulated in the IL is much lower than the obtained in the pretreatment of eucalyptus, where 10 – 70 % (w/w) of lignin introduced in a combination of AH+IL pretreatment were extracted (Rigual et al., 2018). Differences between hardwood and softwood lignin solubility evince that lignin reactivity to [Emim][OAc] is different, making more difficult the extraction of softwood lignin. The overall quantity of lignin that is accumulated into the IL is up to 10 % of the initial lignin introduced. 3.3. Solid fractions obtained after AH+IL pretreatments The composition of untreated pine and solid recovered after pretreatment expressed in g/100 g of untreated/pretreated biomass are shown in Table 1. Note that the solid yields decrease considerably after pretreatments combination, which in general terms involve a decrease of each component in the global balance. The effect of autohydrolysis pretreatment leads to a decrease of the solid yield of at least 15 % in mild AH conditions and 34 % in severe AH conditions as a consequence of hemicellulose removal (Herrera et al., 2016). The effect of IL pretreatment is stronger at mild than at severe AH conditions. In samples autohydrolyzed under the same conditions, there is a loss of solid yields of around 14 % for IL treatment temperatures between 50 and 150 °C; while the decrease is of 10 % in samples autohydrolyzed under severe conditions. As mentioned before, lignin recondensation may be the cause of the difficult processing of biomass at severe 9 conditions (Miyamoto et al., 2018; Shi et al., 2016). Lignin relative content for untreated pine wood was 34.09 % and the hemicellulose content was 21.72 % being mannose the main monosaccharide present in the hemicellulose fraction (Mou et al., 2013). Lignin relative content barely varies in the mild AH samples (AH150IL50, AH150IL80, AH150IL120 and AH150IL150), neither in almost all severe AH samples (AH200IL50, AH200IL80 and AH200IL120). Only in the most severe conditions (AH200IL150), lignin content (48 %) is differentially superior to the other samples (43-45 %). This behavior is attributed appart from the hemicellulose removal during autohydrolyzed samples, to the partial degradation of cellulose and lignin due to excessive temperature (Clough et al., 2015; Nitsos et al., 2016). Comparing the effect of microwave IL pretreatment temperature, the lignin removal increases when the IL temperature increases, as a consequence of lignin solubilization in presence of [Emim][OAc] (Dutta et al., 2017). At mild AH, the maximum lignin removal is 25.3 % for AH150IL150 sample, which corresponds to the maximum accumulated lignin in the IL. However, at these conditions cellulose and hemicellulosic fractions are also degraded. At severe AH, less relative lignin is removed ̴15.69-21.06 %. As pine lignin is mainly composed by guayacil units, the β-aryl ether unit is the largest interunit linkage, being very recalcitrant, difficulting the breakage and favoring recondensation (Dutta et al., 2018). Furthermore, the lack of methoxyl groups at the ortho position to the phenolic hydroxyl groups enhance the production of condensed and branched structures (Wang et al., 2017). Global content of glucan decreases as a consequence of degradation, while relative glucan composition increases at severe AH as a consequence of hemicellulose removal. Glucan losses are more evident in samples AH200IL50, AH200IL80, AH200IL120 and AH200IL150. Hemicellulose removal is conditioned to the autohydrolysis pretreatment. Under mild conditions most of the mannan (as part of hemicellulose) is still present in the solid and the effect of the IL temperature from 50 to 120 °C does not involve a decrease of this polysaccharide. An excessive IL temperature (150 °C) confirms hemicelluloses degradation (Kandhola et al., 2017). In the AH200IL150 sample, the relative composition of acetyl groups have increased. This is due to the IL decomposition (Clough et al., 2015; Rashid et al., 2016). 1 0 T a b le 1 : S o lid reco v ery a n d co m p o sitio n a l a n a ly sis. ± v a lu es d en o te sta n d a rd d ev ia tio n C O M P O S IT IO N (g /1 0 0 g O D W b io m ass) S o lid Y ield (% ) L ig n in (% ) G lu can (% ) X y lan (% ) G alactan (% ) A rab in an (% ) M an n an (% ) A cetate (% ) U n treated p in e 1 0 0 .0 0 3 4 .0 9  0 .7 9 4 2 .8 9  0 .2 2 4 .3 4  0 .2 6 2 .1 7  0 .3 1 0 .8 5  0 .1 6 1 1 .7 5  0 .0 5 2 .6 1  0 .4 1 A H 1 5 0 IL 5 0 8 4 .7 4 3 6 .1 2 ± 0 .7 2 4 6 .7 8 ± 1 .0 9 1 .9 0 ± 0 .4 3 1 .4 6 ± 0 .2 3 0 .2 2 ± 0 .0 2 9 .5 0 ± 0 .8 5 4 .5 6 ± 0 .7 7 A H 1 5 0 IL 8 0 8 1 .5 7 3 5 .2 8 ± 1 .1 0 4 8 .3 4 ± 0 .3 9 5 .2 0 ± 0 .4 4 2 .7 4 ± 0 .4 3 0 .0 0 ± 0 .0 0 9 .3 7 ± 0 .2 0 1 .6 0 ± 0 .1 7 A H 1 5 0 IL 1 2 0 7 9 .7 9 3 6 .7 4 ± 2 .0 5 4 7 .7 5 ± 0 .9 6 3 .5 4 ± 0 .3 1 0 .5 3 ± 0 .0 1 0 .6 3 ± 0 .0 1 9 .3 2 ± 0 .0 5 2 .1 4 ± 0 .2 2 A H 1 5 0 IL 1 5 0 7 0 .4 4 3 6 .4 0 ± 0 .9 1 4 8 .1 4 ± 0 .3 4 1 .5 9 ± 0 .6 5 1 .2 8 ± 0 .3 0 0 .4 4 ± 0 .1 0 3 .6 5 ± 0 .2 0 4 .4 0 ± 0 .3 1 A H 2 0 0 IL 5 0 6 6 .3 7 4 3 .5 5 ± 0 .4 7 4 9 .8 9 ± 0 .8 1 0 .0 0 ± 0 .0 0 0 .0 0 ± 0 .0 0 0 .4 2 ± 0 .0 1 0 .0 0 ± 0 .0 0 0 .5 5 ± 0 .3 2 A H 2 0 0 IL 8 0 6 5 .0 5 4 5 .1 4 ± 0 .7 9 4 9 .6 9 ± 1 .0 5 0 .2 3 ± 0 .0 0 0 .0 0 ± 0 .0 0 0 .0 0 ± 0 .0 0 0 .0 0 ± 0 .0 0 1 .4 3 ± 0 .1 8 A H 2 0 0 IL 1 2 0 6 4 .2 8 4 3 .7 9 ± 1 .7 5 5 0 .5 0 ± 0 .7 6 0 .4 3 ± 0 .0 0 0 .3 7 ± 0 .0 2 0 .0 0 ± 0 .0 0 0 .3 4 ± 0 .3 4 0 .3 2 ± 0 .1 1 A H 2 0 0 IL 1 5 0 5 6 .9 4 4 8 .0 6 ± 1 .9 2 4 9 .9 5 ± 0 .4 4 0 .0 0 ± 0 .0 0 0 .0 0 ± 0 .0 0 0 .3 8 ± 0 .0 2 0 .0 0 ± 0 .0 0 3 .3 6 ± 0 .0 6 11 3.4. Enzymatic digestibility The enzymatic digestibility of untreated and pretreated pine wood is shown in Figure 2. It is well-known that the effect of dissolving and regenerating cellulose in untreated wood enhances the enzymatic digestibility (Geng and Henderson, 2014). AH150IL50, AH150IL80, AH150IL120 and AH150IL150 samples exhibit digestibilities higher than the analogous sample autohydrolyzed at severe conditions (AH200IL50, AH200IL80, AH200IL120 and AH200IL150). Using conventional heating, Torr et al. demonstrated that the increment in the temperature from 80 to 120 °C increases hydrolysis yields (Torr et al., 2016). Even increasing the temperature up to 130 °C, glucose digestibility increases (Trinh et al., 2015). In this work, the same tendency is also followed in mild autohydrolyzed samples. However, digestibility is decreased in AH150IL150 sample. Degradation products that inactivate enzymes, structural modifications of softwood lignins, or alteration of the morphological structure (further commented) can be some of the reasons of this result (Lai et al., 2014; Torr, 2012). Severe autohydrolyzed samples tendencies show more drastic differences. In this case, AH200IL50 digestibilities are very low (maximum digestibility obtained 19.2 %) even below the untreated wood. Although microwave can enhance the pretreatment, these results evince that at least 80 °C should be reached for IL pretreatment to improve accessibility (Liu et al., 2014). Furthermore, severe autohydrolysis difficult the subsequent pretreatment and enzymatic digestibility as, in all the cases, digestibilities are higher in mild than in severe autohydrolyzed samples (Nitsos et al., 2016). This effect was opposite to the observed in our previous work using hardwood (Eucalyptus globulus), so the chemical composition of softwood, specially the different nature of softwood lignin may be affecting digestibility as previously commented (Rigual et al., 2018; Wang et al., 2018). IL temperatures above 80 °C worsen the enzymatic digestibility, probably because there are less active sites of the pretreated biomass that can be cleaveaged by the enzyme, as will be discussed in the next section (Sun et al., 2014). In the case of untreated pine wood, digestibility is increased in the first 12 hours up to 20.2 % of glucan. After 12 h the increment is very low, reaching only 26.3 % of glucan at 72 h. This behavior (main increment is produced up to 12 hours) happens in AH150IL50 and AH200IL50. Samples AH150IL80, AH150IL120 need up to 24 h to reach almost the maximum digestibility. Rates are lower imn AH150IL150, and AH200IL80, where longer times of up to 48 h should be assessed (Sun et al., 2014). In the case of AH200IL120 and AH200IL150, the increment of digestibility is even produced up to 72 h. The increment of the time employed to complete the hydrolysis may be produced as a consequence of the antisolvent used in the IL microwave pretreatment (water). Other antisolvents, such as alcohols may increase hydrolysis rates and reduce the time, although other factors like cost and environmental impact should be considered (Asakawa et al., 2016; Geng and Henderson, 2014). 12 AH150IL120 is the most appropriate combination to obtain an easily hydrolysed-by-enzymes substrate, obtaining a maximum digestibility of ̴ 79 %. However, the removal of hemicellulose in a prior step (in the context of a global utilization of every fraction) is not obtained at mild autohydrolysisis. The AH200IL80 conditions fulfil also good digestibility’s of a free of hemicellulose solid 72.6 %. Nevertheless, the win-win scenario obtained in hardwood is not achieved with softwoods and a compromise solution must be found (Rigual et al., 2018). Figure 2. Glucan digestibility of Pine wood: a) untreated and mild autohydrolyzed+IL pretreatment; b) untreated and severe autohydrolyzed+IL pretreatment. Results are expressed based on Eq 3 3.5. Factors affecting enzymatic digestibility: surface morphology and composition 3.5.1. Surface morphology: Fractal dimension and lacunarity Untreated pine wood (Figure 3a) presents a compact surface, while AH150IL80 pretreated wood (Figure 3b) exhibits the alteration of the morphology, forming a fusing structure with a lot of pores revealing the network disruption(Sun et al., 2011; Trinh et al., 2015). On the contrary, AH150IL150 (Figure 3c) depicts a very closed structure with no visible holes available. In the case of samples autohydrolyzed at severe conditions (Figures 3d and 3e), images are much similar to the AH150IL150 sample. Particles have very few visible holes and a much-smoothed surface. This morphology contrast with the obtained in samples only pretreated with IL, where holes are visible (Karimi and Taherzadeh, 2016). 13 Figure 3: SEM microscopy images of a) untreated pine wood, b) AH150IL120, c) AH150IL150, d) AH200IL80, and e) AH200IL150 Using the box counting method, Db (Fig A2) and Ʌ for each sample were calculated. Values of Db and Ʌ were plotted against 72 hours enzymatic digestibilitiy (Figures 3a and 3b). The Db values obtained in this work were between 2.5461 and 2.7124. Although this interval may look narrow, Db values are in a very small interval in other biocompounds (carbon samples, retrograded starch) previously measured and are of the same order of magnitude (1.0-1.9) (Alvarez et al., 2013; Rigual et al., 2018; Smith Jr et al., 1996; Utrilla-Coello et al., 2013). This parameter is related to the complexity of a texture or a silhouette (twists and turns), increasing in value when the complexity and convolutions are more notorious, which is confirmed with the micrographs obtained. Thus, the higher Db indicates more accessibility. According to results obtained (Figure 5a), the lowest Db value is attributed to untreated pine (2.5461), followed by AH200IL50 (2.6606) and AH150IL50 (2.6705). Remark that the relation stablished between Db and enzymatic digestibility (bigger fractal dimensions are related to higher enzymatic digestibilties) is not followed for the sample AH200IL50, indicating that other factors are also affecting the pretreated biomass digestibility as is discussed below. Laccunarity as a measurement of homogeneity and visual gappiness ranged between 0.0818 (AH150IL80) and 0.2563 (untreated pine), which is also in accordance with values calculated for other kind of materials and shapes (starch, or carbon) (Liu and Ostadhassan, 2017; Smith Jr et al., 1996; Utrilla-Coello et al., 2013). Homogeneous morphologies resulted in low laccunarity values that are also related to high digestibilities. Therefore, the sample AH200IL50 does not follow again the same tendency than others as occurred with its fractal dimension. The particular behaviour of this sample is explained below as a consequence of surface composition. Additionally, both parameters (Db and ʌ) are related between themselves, as higher Db corresponds to lower Ʌ laccunarities (Rigual et al., 2018). 3.5.2. Surface composition The use of CFM showed relevant differences in lignin surface distribution. Some of the most representative images of the maximum Z projection of the merged image (formed by the combination of holocellulose and lignin channel, in green and red respectively) are shown in Fig. 14 4. The effect of the IL cause remarkable changes of the swelling of pine network in comparison to the initial material (Singh et al., 2009). Furthermore, distribution of lignin and holocellulose in the surface is different. Note that mild autohydrolyzed samples still have a considerable content of hemicellulose, while severe autohydrolyzed samples are almost free of hemicelluloses. In this sense, it is observed that AH150IL120 presents more hollocellulose in the surface. Thus, this fact joined to the high fractal dimension obtained, results in high accessibilities. However, the AH150IL150 sample has a high concentration of lignin in the surface. In the case of severe autohydrolyzed samples, hollocellulose is only attributed to cellulose as hemicellulose was previously removed in the autohydrolysis step. In the AH200IL80 sample more cellulose in the surface is observed than in AH200IL150 sample. In order to further quantify the differences observed, the negative effect of lignin deposition on the surface was hypothesized. In this sense, 72 h enzymatic digestibility was plotted against holocellulose/lignin ratio in the surface (Fig. 4c). This ratio was determined after counting the number of pixels of every channel. Note that lignin surface concentration is related to the enzymatic hydrolysis digestibility, increasing the enzymatic digestibility when the lignin content in the surface is decreased. The hemicellulose present in untreated and hamper the tendency observed for the rest of the samples. There is not a relation between holocellulose/ lignin ratio calculated in terms of composition and holocellulose/ lignin ratio in the surface (Figure 3d). This lack of relation remarks differences of composition in the particles surface and inside the particles, and also evinces that compositional analysis requires additional techniques to better understand factors affecting enzymatic digestibility. Pine wood structures after pretreatment are much more compact and lignin is more recalcitrant due to the nature of softwood lignin (Shi et al., 2016). Additionally, the negative effect of softwood lignin in enzymatic hydrolysis has been widely demonstrated (Lai et al., 2014; Li et al., 2013b). In summary CFM may be a powerful technique, to be employed as a complement, as it adds information about chemical composition and the distribution of holocellulose and mainly lignin in the surface. 15 Figure 4. Representation of different DIA parameters vs enzymatic digestibility: a) Db vs enzymatic digestibility; b) ratio holocellulose/lignin in the surface vs glucan digestibility; c) ratio holocellulose/lignin vs ratio holocellulose/lignin in the surface Figure 5: Confocal fluorescence microscopy images of a) untreated pine wood, b) AH150IL120, c) AH150IL150, d) AH200IL80, and e) AH200IL150 3.6. Total yields Glucan and glucose results obtained after pretreatment and after saccharification for every condition are shown in Figure 6. The decrease in glucose yield with respect to mild autohydrolyzed samples is notorious, due to the high degradation of solids in severe autohydrolyzed samples. In the case of samples with low sugars degradation, such as AH150IL50 16 and AH150IL80 (with 90.4 and 89.9 of the initial glucan still present in the solid), digestibilities are lower, of 48.6 and 73.1 %, respectively (total glucose yield of 43.9 and 65.7, respectively). In a biorefinery perspective, it is difficult to find a compromise selection that enhances hemicellulose extraction and high glucose yields. Although mild autohydrolysis pretreatments limit the hemicelluloe removal, severe autohydrolysis limit glucose production. Thus, a compromise must be reached in function of the goal to achieve. In this way, best results were obtained for AH150IL120 sample, which yields highest glucose productions 68.4 %. Figure 6: Overall glucan and glucose yield a) after sequential pretreatments and, b) after enzymatic saccharification The enzymatically hydrolyzed solid is mainly composed by lignin (86.15% of the initial lignin is still in that solid) and this solid is free of hemicelluloses, which may help in the application of this lignin rich material (i.e. lignin depolimerization or thermal conversion) (Schutyser et al., 2018). In the scenario of severe autohydrolysed samples, a free of hemicellulose solid has been obtained. In the case of the AH200IL80, while 72.6 % of glucan was produced after 72 h this value was reduced up to 54.2 g of glucose/100 g of initial glucose due to the losses produced during pretreatment (only 74.7 % of the initial glucan introduced in the process was recovered after pretreatment). 4. Conclusions In the present work, a sequential pretreatment of autohydrolysis + microwave IL pretreatment towards the enhancement of the digestibility of the solid has been studied. Lignin extraction is higher using mild AH pretreatments. Mild AH requires high IL temperatures (AH150IL120), while severe AH should be operated at moderate IL temeperatures (AH200IL80) to maximize glucan digestibility. Morphology and surface composition can be quantified, in conjunction with compositional analysis to have an overall outlook of different factors affecting digestibility. In softwood, high fractal dimension 17 (Db=2.6962), low laccunarity (ʌ=0.0865) and high holocellulose/lignin ratios (0.70422) on the surface may considerably decrease the digestibility of the material. Acknowledgements This work was performed thanks to the financial support of the “Ministerio de Economía y Competitividad” under the funded project CTQ2013-42006-R and Victoria Rigual thanks he contract BES-2014-067788. Sandra Rivas thanks her “Juan de la Cierva” postdoctoral grant. Authors thank DuPont Industrial Bioscience for the donation of enzymatic cocktails. References A. Sluiter, B.H., R.Ruiz, C.Scarlata, J.Sluitr, D.Templeton and D. Crocker, 2011. Determination of Structural carbohydrates and lignin in biomass. NREL. Akhtar, N., Gupta, K., Goyal, D., Goyal, A., 2015. Recent advances in pretreatment technologies for efficient hydrolysis of lignocellulosic biomass. Environ. Prog. Sustain. Energy 35, 489-511. Alvarez, A., Passé-Coutrin, N., Gaspard, S., 2013. Determination of the textural characteristics of carbon samples using scanning electronic microscopy images: comparison with mercury porosimetry data. Adsorption 19, 841-850. Asakawa, A., Oka, T., Sasaki, C., Asada, C., Nakamura, Y., 2016. Cholinium ionic liquid/cosolvent pretreatment for enhancing enzymatic saccharification of sugarcane bagasse. Industrial Crops and Products 86, 113-119. Casas, A., Alonso, M.V., Oliet, M., Santos, T.M., Rodriguez, F., 2013. Characterization of cellulose regenerated from solutions of pine and eucalyptus woods in 1-allyl-3- methilimidazolium chloride. Carbohydr. Polym. 92, 1946-1952. Clough, M.T., Geyer, K., Hunt, P.A., Son, S., Vagt, U., Welton, T., 2015. Ionic liquids: not always innocent solvents for cellulose. Green Chem. 17, 231-243. Chen, H., Han, Y., Xu, J., 2008. Simultaneous saccharification and fermentation of steam exploded wheat straw pretreated with alkaline peroxide. Process Biochem. 43, 1462-1466. Chen, Y., Wang, J., Feng, J., 2017. Understanding the fractal dimensions of urban forms through spatial entropy. Entropy 19, 600. Dutta, T., Isern, N.G., Sun, J., Wang, E., Hull, S., Cort, J.R., Simmons, B.A., Singh, S., 2017. Survey of Lignin-Structure Changes and Depolymerization during Ionic Liquid Pretreatment. ACS Sustainable Chemistry and Engineering 5, 10116-10127. Dutta, T., Papa, G., Wang, E., Sun, J., Isern, N.G., Cort, J.R., Simmons, B.A., Singh, S., 2018. Characterization of lignin streams during bionic liquid-based pretreatment from grass, hardwood and softwood. ACS Sustainable Chemistry & Engineering. Geng, X., Henderson, W.A., 2014. Impact of non-solvents on the structural features and enzymatic digestibility of cellulose regenerated from an ionic liquid. RSC Adv. 4, 31226-31229. 18 Hauru, L.K.J., Ma, Y., Hummel, M., Alekhina, M., King, A.W.T., Kilpelainen, I., Penttila, P.A., Serimaa, R., Sixta, H., 2013. Enhancement of ionic liquid-aided fractionation of birchwood. Part 1: autohydrolysis pretreatment. RSC Adv. 3, 16365-16373. Heggset, E.B., Syverud, K., Øyaas, K., 2016. Novel pretreatment pathways for dissolution of lignocellulosic biomass based on ionic liquid and low temperature alkaline treatment. Biomass and Bioen. 93, 194-200. Hendriks, A.T.W.M., Zeeman, G., 2009. Pretreatments to enhance the digestibility of lignocellulosic biomass. Bioresource Technology 100, 10-18. Herrera, R., da Silva, D.T., Llano-Ponte, R., Labidi, J., 2016. Characterization of pine wood liquid and solid residues generated during industrial hydrothermal treatment. Biomass and Bioenergy 95, 174-181. Huang, X., Kocaefe, D., Kocaefe, Y., Pichette, A., 2018. Combined effect of acetylation and heat treatment on the physical, mechanical and biological behavior of jack pine (Pinus banksiana) wood. European Journal of Wood and Wood Products 76, 525-540. Ibáñez, A.B., Bauer, S., 2014. Downscaled method using glass microfiber filters for the determination of Klason lignin and structural carbohydrates. Biomass Bioenerg. 68, 75-81. Kandhola, G., Djioleu, A., Carrier, D.J., Kim, J.-W., 2017. Pretreatments for Enhanced Enzymatic Hydrolysis of Pinewood: a Review. Bioenerg. Res. 10, 1138-1154. Karimi, K., Taherzadeh, M.J., 2016. A critical review of analytical methods in pretreatment of lignocelluloses: Composition, imaging, and crystallinity. Bioresource Technology 200, 1008- 1018. Karperien, A., 2013. FracLac for ImageJ. FracLac, https://imagej.nih.gov/ij/plugins/fraclac/FLHelp/Introduction.htm. Kilpeläinen, I., Xie, H., King, A., Granstrom, M., Heikkinen, S., Argyropoulos, D.S., 2007. Dissolution of Wood in Ionic Liquids. Journal of Agricultural and Food Chemistry 55, 9142- 9148. Lai, C., Tu, M., Shi, Z., Zheng, K., Olmos, L.G., Yu, S., 2014. Contrasting effects of hardwood and softwood organosolv lignins on enzymatic hydrolysis of lignocellulose. Bioresource technology 163, 320-327. Li, C., Sun, L., Simmons, B., Singh, S., 2013a. Comparing the recalcitrance of eucalyptus, pine, and switchgrass using ionic liquid and dilute acid pretreatments. Bioenerg. Res. 6. Li, C., Sun, L., Simmons, B.A., Singh, S., 2013b. Comparing the Recalcitrance of Eucalyptus, Pine, and Switchgrass Using Ionic Liquid and Dilute Acid Pretreatments. Bioenerg. Res. 6, 14- 23. Liu, J.-F., Cao, Y., Yang, M.-H., Wang, X.-J., Li, H.-Q., Xing, J.-M., 2014. Enhanced saccharification of lignocellulosic biomass with 1-allyl-3-methylimidazolium chloride (AmimCl) pretreatment. Chinese Chemical Letters. Liu, K., Ostadhassan, M., 2017. Quantification of the microstructures of Bakken shale reservoirs using multi-fractal and lacunarity analysis. J. Nat. Gas Sci. Eng. 39, 62-71. Mäki-Arvela, P., Anugwom, I., Virtanen, P., Sjöholm, R., Mikkola, J.P., 2010. Dissolution of lignocellulosic materials and its constituents using ionic liquids—A review. Industrial Crops and Products 32, 175-201. https://imagej.nih.gov/ij/plugins/fraclac/FLHelp/Introduction.htm 19 Miyamoto, T., Mihashi, A., Yamamura, M., Tobimatsu, Y., Suzuki, S., Takada, R., Kobayashi, Y., Umezawa, T., 2018. Comparative analysis of lignin chemical structures of sugarcane bagasse pretreated by alkaline, hydrothermal, and dilute sulfuric acid methods. Industrial Crops and Products 121, 124-131. Moniz, P., Pereira, H., Quilhó, T., Carvalheiro, F., 2013. Characterisation and hydrothermal processing of corn straw towards the selective fractionation of hemicelluloses. Industrial Crops and Products 50, 145-153. Mou, H.-Y., Orblin, E., Kruus, K., Fardim, P., 2013. Topochemical pretreatment of wood biomass to enhance enzymatic hydrolysis of polysaccharides to sugars. Bioresource Technology 142, 540- 545. Nguyen, T.-A.D., Kim, K.-R., Han, S.J., Cho, H.Y., Kim, J.W., Park, S.M., Park, J.C., Sim, S.J., 2010. Pretreatment of rice straw with ammonia and ionic liquid for lignocellulose conversion to fermentable sugars. Bioresource Technology 101, 7432-7438. Nitsos, C.K., Choli-Papadopoulou, T., Matis, K.A., Triantafyllidis, K.S., 2016. Optimization of Hydrothermal Pretreatment of Hardwood and Softwood Lignocellulosic Residues for Selective Hemicellulose Recovery and Improved Cellulose Enzymatic Hydrolysis. ACS Sustain. Chem. Eng. 4, 4529-4544. Otsu, N., 1979. A threshold selection method from gray-level histograms. IEEE transactions on systems, man, and cybernetics 9, 62-66. Pérez-Pimienta, J.A., Vargas-Tah, A., López-Ortega, K.M., Medina-López, Y.N., Mendoza- Pérez, J.A., Avila, S., Singh, S., Simmons, B.A., Loaces, I., Martinez, A., 2017. Sequential enzymatic saccharification and fermentation of ionic liquid and organosolv pretreated agave bagasse for ethanol production. Bioresour. Technol. 225, 191-198. Rashid, T., Kait, C.F., Regupathi, I., Murugesan, T., 2016. Dissolution of kraft lignin using Protic Ionic Liquids and characterization. Industrial Crops and Products 84, 284-293. Resch, M., Baker, J., Decker, S., 2015. Low solids enzymatic saccharification of lignocellulosic biomass. NREL Laboratory Analytical Procedure. Rigual, V., Santos, T.M., Domínguez, J.C., Alonso, M.V., Oliet, M., Rodriguez, F., 2018. Combining autohydrolysis and ionic liquid microwave treatment to enhance enzymatic hydrolysis of Eucalyptus globulus wood. Bioresource technology 251, 197-203. Schutyser, W., Renders, T., Van den Bosch, S., Koelewijn, S.-F., Beckham, G., Sels, B., 2018. Chemicals from lignin: an interplay of lignocellulose fractionation, depolymerisation, and upgrading. Chemical Society Reviews. Shi, J., Pattathil, S., Parthasarathi, R., Anderson, N.A., Im Kim, J., Venketachalam, S., Hahn, M.G., Chapple, C., Simmons, B.A., Singh, S., 2016. Impact of engineered lignin composition on biomass recalcitrance and ionic liquid pretreatment efficiency. Green Chemistry 18, 4884-4895. Silveira, M.H.L., Vanelli, B.A., Corazza, M.L., Ramos, L.P., 2015. Supercritical carbon dioxide combined with 1-butyl-3-methylimidazolium acetate and ethanol for the pretreatment and enzymatic hydrolysis of sugarcane bagasse. Bioresour. Technol. 192, 389-396. Singh, S., Simmons, B.A., Vogel, K.P., 2009. Visualization of biomass solubilization and cellulose regeneration during ionic liquid pretreatment of switchgrass. Biotechnol Bioeng 104, 68-75. 20 Sluiter, A., Hames, B., Ruiz, R., Scarlata, C., Sluiter, J., Templeton, D., 2008. Determination of Structural Carbohydrates and Lignin in Biomass. Laboratory Analytical Procedure (LAP). National Renewable Energy Laboratory (NREL), Golden, CO, USA. Smith Jr, T.G., Lange, G.D., Marks, W.B., 1996. Fractal methods and results in cellular morphology — dimensions, lacunarity and multifractals. J. Neurosci. Methods 69, 123-136. Ståhl, M., Nieminen, K., Sixta, H., 2018. Hydrothermolysis of pine wood. Biomass and Bioenergy 109, 100-113. Sun, N., Parthasarathi, R., Socha, A.M., Shi, J., Zhang, S., Stavila, V., Sale, K.L., Simmons, B.A., Singh, S., 2014. Understanding pretreatment efficacy of four cholinium and imidazolium ionic liquids by chemistry and computation. Green Chemistry 16, 2546-2557. Sun, N., Rahman, M., Qin, Y., Maxim, M.L., Rodriguez, H., Rogers, R.D., 2009a. Complete dissolution and partial delignification of wood in the ionic liquid 1-ethyl-3-methylimidazolium acetate. Green Chemistry 11, 646-655. Sun, N., Rahman, M., Qin, Y., Maxim, M.L., Rodriguez, H., Rogers, R.D., 2009b. Complete dissolution and partial delignification of wood in the ionic liquid 1-ethyl-3-methylimidazolium acetate. Green Chem 11. Sun, N., Rodríguez, H., Rahman, M., Rogers, R.D., 2011. Where are ionic liquid strategies most suited in the pursuit of chemicals and energy from lignocellulosic biomass? Chem. Commun. 47, 1405-1421. Sun, S., Sun, S., Cao, X., Sun, R., 2016. The role of pretreatment in improving the enzymatic hydrolysis of lignocellulosic materials. Bioresour. Technol. 199, 49-58. Torr, K.M., 2012. The impact of ionic liquid pretreatment on the chemistry and enzymatic digestibility of Pinus radiata compression wood. Green Chem. 14, 778-787. Torr, K.M., Love, K.T., Simmons, B.A., Hill, S.J., 2016. Structural features affecting the enzymatic digestibility of pine wood pretreated with ionic liquids. Biotechnology and Bioengineering 113, 540-549. Trinh, L.T.P., Lee, Y.J., Lee, J.-W., Lee, H.-J., 2015. Characterization of ionic liquid pretreatment and the bioconversion of pretreated mixed softwood biomass. Biomass and Bioenergy 81, 1-8. Utrilla-Coello, R.G., Bello-Pérez, L.A., Vernon-Carter, E.J., Rodriguez, E., Alvarez-Ramirez, J., 2013. Microstructure of retrograded starch: Quantification from lacunarity analysis of SEM micrographs. Journal of Food Engineering 116, 775-781. van Osch, D.J., Kollau, L.J., van den Bruinhorst, A., Asikainen, S., Rocha, M.A., Kroon, M.C., 2017. Ionic liquids and deep eutectic solvents for lignocellulosic biomass fractionation. Physical Chemistry Chemical Physics 19, 2636-2665. Wang, H., Ben, H., Ruan, H., Zhang, L., Pu, Y., Feng, M., Ragauskas, A.J., Yang, B., 2017. Effects of Lignin Structure on Hydrodeoxygenation Reactivity of Pine Wood Lignin to Valuable Chemicals. ACS Sustainable Chemistry & Engineering 5, 1824-1830. Wang, W., Tan, X., Yu, Q., Wang, Q., Qi, W., Zhuang, X., Wang, Z., Yuan, Z., 2018. Effect of stepwise lignin removal on the enzymatic hydrolysis and cellulase adsorption. Industrial Crops and Products 122, 16-22. Weerachanchai, P., Lee, J.-M., 2014. Recyclability of an ionic liquid for biomass pretreatment. Bioresour. Technol. 169, 336-343. 21 Xu, J., Xiong, P., He, B., 2016. Advances in improving the performance of cellulase in ionic liquids for lignocellulose biorefinery. Bioresource Technology 200, 961-970. Yoo, C.G., Pu, Y., Ragauskas, A.J., 2017. Ionic liquids: Promising green solvents for lignocellulosic biomass utilization. Current Opinion in Green and Sustainable Chemistry 5, 5-11. Zhang, L., Pu, Y., Cort, J.R., Ragauskas, A.J., Yang, B., 2016. Revealing the Molecular Structural Transformation of Hardwood and Softwood in Dilute Acid Flowthrough Pretreatment. ACS Sustainable Chemistry & Engineering 4, 6618-6628. Zinchuk, V., Zinchuk, O., Okada, T., 2007. Quantitative colocalization analysis of multicolor confocal immunofluorescence microscopy images: pushing pixels to explore biological phenomena. Acta histochemica et cytochemica 40, 101-111. S1 Autohydrolysis and microwave ionic liquid pretreatment of Pinus radiata: imaging visualization and analysis to understand enzymatic digestibility Victoria Rigual*, Juan C. Domínguez, Tamara M. Santos, Sandra Rivas, M. Virginia Alonso, Mercedes Oliet and Francisco Rodriguez Department of Chemical Engineering and Materials, Faculty of Chemistry, Complutense University of Madrid, Avda. Complutense s/n, 28040 Madrid, Spain. *E-mail: vicrigua@ucm.es ; Tel: +34913948505 Figure A1: Recovered IL AH150 IL50 AH150 IL80 AH150 IL120 AH150 IL150 AH200 IL50 AH200 IL80 AH200 IL120 AH200 IL150 0 20 40 60 80 100 Regeneration step 1st Wash 2nd Wash 3rd Wash 4th Wash 5th Wash R e c o v e re d I L ( % ) mailto:vicrigua@ucm.es S2 Figure A2: Calculation of fractal dimension by the box counting method using a color scale PUBLICACIÓN VII “One-pot conversion of woody biomass feedstocks using protic ionic liquids” Victoria Rigual, Gabriella Papa, Alberto Rodriguez, Maren Wehrs, Kwang Ho Kim, M. Virginia Alonso, Mercedes Oliet, John Gladden, Aindrila Mukhopadhyay, Blake A. Simmons y Seema Singh Enviado a Journal of Cleaner Production 1 One-pot conversion of woody biomass feedstocks using protic ionic liquids Victoria Riguala, Gabriella Papab,c, Alberto Rodriguezb,c, Maren Wehrsb,c.d, Kwang Ho Kime, M. Virginia Alonsoa, Mercedes Olieta, John Gladdenc,f, Aindrila Mukhopadhyayb,c, Blake A. Simmonsb,c, Seema Singhc,f* aDepartment of Chemical Engineering, Faculty of Chemistry, Complutense University of Madrid, Avda. Complutense s/n, 28040 Madrid, Spain b Biological Systems and Engineering Division, Lawrence Berkeley National Laboratory, 1 Cyclotron Road, Berkeley CA, 94720, USA c Joint BioEnergy Institute, 5885 Hollis Street, Emeryville, CA, USA d Institut für Genetik, Technische Universität Braunschweig, Germany e Clean Energy Research Center, Korea Institute of Science and Technology, Seoul 02792, Republic of Korea f Biological and Engineering Sciences Center, Sandia National Laboratories, 7011 East Avenue, Livermore, California 94551, USA *E-mail: seesing@sandia.gov mailto:seesing@sandia.gov 2 ABSTRACT One-pot ionic liquid processes are a promising approach to biomass conversion. This study evaluated the one-pot configuration for the deconstruction and conversion of two different woody feedstocks: eucalyptus and pine. The ionic liquids studied were 2-hydroxyethylammonium acetate ([2-HEA][OAc]), bis-2-hydroxyethyl ammonium acetate ([B2-HEA][OAc]), triethylammonium acetate ([TEA][OAc]) and cholinium lysinate ([Ch][Lys]). Biomass digestibility and the impact of ILs at 11% w/w on the saccharification efficiency of the two feedstocks were evaluated. We observed that protic ILs were less toxic to fungal hosts than [Ch][Lys] in media at concentrations ≤10 % w/w, and sugar digestibility of up to 75 % was observed for eucalyptus pretreated with [2- HEA][OAc]. The residual lignin-rich solids obtained with [2-HEA][OAc] contained more guaiacyl units than the initial biomass lignin. The fungal strain Rhodosporidium toruloides was used for bisabolene production on the hydrolysates. This work evaluate challenges of a woody-based One- pot ionic liquid process using protic ionic liquids. Keywords: One-pot, ionic liquids, eucalyptus, pine, Rhodosporidium toruloides, enzymatic hydrolysis 1. Introduction Lignocellulose is the primary building block of plant cell walls, giving rise to 2x1011 tonnes of biomass produced per year (van Osch et al., 2017). Among all woody biomass types, pine and eucalyptus have been considered as potential bioenergy crops (Simmons et al., 2008). Eucalyptus is a hardwood considered one of the most promising biorefinery feedstocks due to its fast growth and wide adaptability . Softwoods are also a major resource that has a significant importance in Europe, where two thirds (67.9 %) of the total level of removals under bark were coniferous (Hämäläinen et al., 2011; Muñoz, 2011). Pretreatment technologies may comprise up to 40 % of the total cost of biomass processing (Corona et al., 2018; Elgharbawy et al., 2016). In last decade, certain ionic liquids (ILs) have emerged as an effective alternative due to low volatilities, non-flammability, and biomass solubilisation (Fan et al., 2019; Singh et al., 2016; Xu, J. et al., 2016). The top performing IL used to-date in biomass processing is 1-ethyl-3-methylimidazolium acetate ([C2C1Im][OAc]), but is one of the most expensive and is toxic to hydrolytic enzymes and biofuels-producing hosts used downstream (George et al., 2014; Kahani et al., 2017; Khan et al., 2018; Ouellet et al., 2011; Prado et al., 2016). As a result, processes involving the use of [C2C1Im][OAc] need to include multiple washes with water to remove the IL, and/or extensive enzyme and host engineering to improve their IL tolerance (Frederix et al., 2016). 3 One-pot processes are based on biocompatible ILs, since they combine pretreatment and saccharification or pretreatment, saccharification and biofuel production in a single reactor (Auxenfans et al., 2017). This new approach has emerged as a compelling alternative to conventional ILs Protic ionic liquids (PILs) are easily synthesized by the equimolar neutralization of an organic amine and a mineral base (Maciel et al., 2019). Certain PILs have been demonstrated to be relatively inexpensive to synthesis and their effectivity have been confirmed (Weigand et al., 2017). For example, triethylammonium hydrogen sulfate was successfully used in conventional lignocellulosic biomass pretreatment (George et al., 2015) and 2- hydroxyethylammonium acetate ([2-HEA][OAc]]) was proved to convert switchgrass using a one-pot configuration (Sun et al., 2017). Despite this attractive background, the viability of the PIL based one-pot process on the conversion of woody biomass feedstocks has not been reported yet. To successfully implement a fermentation/conversion stage, conceive either as a separate vessel or in the same, information about toxicity is required (Frederix et al., 2016). Saccharomyces cerevisiae is the most widely employed yeast for cellulosic ethanol, and may be conceived as a starting point for comparison and yeast selection in bioprocesses. Additionally, the ability to produce genetic engineering toolkits and employ non- conventional yeasts have increased during last years (Yaguchi et al., 2018). Recently, the oleaginous yeast Rhodosporidium toruloides has been identified as an attractive biocatalyst for substrate present in complex hydrolysates, able to metabolize a wide range of sugars and lignin derived aromatic compounds (Qadeer et al., 2017; Yaegashi et al., 2017). In this work, the feasibility of one-pot pretratment and saccharification processes of eucalyptus and pine using PILs were evaluated. Three PILs (2-hydroxylethylammonium acetate, bis-2-hydroxyethyl ammonium acetate and triethylammonium acetate) were tested and compared with [Ch][Lys]. In parallel, the effects of these PILs on the growth of two fungal hosts (S. cerevisae and R. toruloides) were studied. In a bigger batch, process was repeated with selected system and finer particle size. Solid fractions were characterized, while hydrolysates containing IL are evaluated to study selected fungal host growth (and the influence of pH adjustment), sugars and lignin consumption, and advance biofuel production 2. Materials and methods 2.1. Materials and reagents Commercial chemicals were of reagent grade and purchased from Sigma-Aldrich (St. Louis, MO). Pinus radiata and Eucalyptus globulus were kindly provided by the laboratory of Dr. G. L. Gresham (Idaho National Laboratory). The values for glucan, xylan, galactan, mannan, acid soluble lignin, Klason lignin and extractives were of 38.25±0.58, 6.88±0.37, 2.82±0.14, 10.86±0.08, 4 4.77±1.40, 29.32±0.5 and 1.52± 0.01, for pine; and 37.92±1.50, 12.46±1.24, 3.02± 0.34, 0± 0, 2.86± 0.12, 26.23± 1.67, 3.08± 0.0 for eucalyptus respectively. Additionally, Eucalyptus globulus fine powder was provided by the Instituto Nacional de Investigación y Tecnología Agraria y Alimentaria of Spain (INIA). A commercial cellulase (Cellic® CTec3) and hemicellulase (Cellic® HTec3) enzymatic cocktail were provided by Novozymes, North America (Franklinton, NC). Strains used in this study (Rhodosporidium toruloides BIS3, JBEI registry ID: JBx 065244, and Saccharomyces cerevisiae BY4742) were obtained from the Joint Bioenergy Institute strain registry (https://registry.jbei.org/login). 2.2. Ionic liquid synthesis The IL [Ch][Lys] was prepared as reported previously (Sun et al., 2014b). The PILs [2- HEA][OAc], [B2-HEA][OAc] and [TEA][OAc] were synthetized in an equimolar ratio (1:1) of the Brønsted acid (glacial acetic acid) and the Brønsted base (2-hydroxyethylamine, bis(2- hydroxyethyl)amine or trimethylamine, respectively) . The synthesis was carried out in a 1 L round- bottom flask which was immersed in an ice bath and kept open to atmosphere. Glacial acetic acid (150 g, 2.498 mol) was added dropwise to a stirring solution of ethanolamine, diethanolamine or triethylamine respectively. After the complete addition (2-4 hours), 5 mL of pure ethanol was added for rinsing. The resultant solution was stirred for another 6 hours. Figure S1 shows the 1H-NMR spectra of the synthesized PILs. 2.3. One-pot pretreatment and saccharification screening Pine or eucalyptus wood chips were milled with a Thomas-Wiley Mini Mill fitted with a 425-500 µm screen (Model 3383-L10 Arthur H. Thomas Co., Philadelphia, Pa, USA), and stored in cold room at 4 °C. 425-500 μm pine or eucalyptus were mixed (100 mg) with [Ch][Lys], [2- HEA][OAc] or [B2-HEA][OAc] at a 10% (w/w) biomass loading in a 15 mL capped glass pressure tube and pretreated for 6 hours in an oil bath heated at 150 °C and stirred. 10% (w/w) biomass soaking in [TEA][OAc], at lower temperature (30 °C) for 7 days reaction times has been used to prevent degradation (Td=48 °C), as previously suggested (Shmukler et al., 2016). After pretreatment, samples were cooled to room temperature. The IL-treated biomass was diluted to achieve final IL concentrations of 11.02±0.2 % w/v (i.e. final solid loading of 1 %w/w) and pH was adjusted to 5 if needed with glacial acetic acid or the corresponding base (triethylamine). Enzymatic saccharification was run at 50 °C, for 72 hours on an Enviro Genie SI-1200 rotator platform (Scientific Industries, Inc., Bohemia, NY). The enzyme complexes Cellic CTEC3 + Cellic HTEC3 (9:1 v/v) were used at a loading of 25 mg protein/g glucan. A 0.08 mL aliquot of the hydrolysate was taken after 0.5, 12, 24, 48 and 72 hours. The samples obtained at different times of hydrolysis were analyzed by HPLC (see Analytical Method for details) for monosaccharides detection. Enzymatic digestibility was defined as the percentage of glucan released from https://registry.jbei.org/login 5 saccharification divided by the maximum glucan available in untreated biomass as obtained by compositional analysis. In the calculation of cellulose conversion to glucose, it was considered a cellulose: glucose ratio of 1: 1.11 . 2.4. Screening of yeast growth in presence of ionic liquids. S. cerevisiae BY4742 and R. toruloides BIS3 were maintained on agar plates at 4 °C for a maximum of 4 weeks. A single colony was used to inoculate an overnight culture of 5 mL YPD medium (10 g/L yeast extract, 20 g/L peptone, 20 g/L dextrose) and incubated in an orbital shaker at a frequency of 200 rpm at 30 °C for 12 hours. The overnight culture was resuspended in fresh YPD media for 4 h. The culture was then used to inoculate YPD media containing different concentrations of IL. The initial OD600 for S. cerevisiae was 0.15, and for R. toruloides 0.20, in a total volume of 5 mL. To prepare the IL/YPD media, solutions of 16.67, 11.11, 5.55, 2.78, 1.11 and 0.56 % (w/w) of IL/water (for each IL) were prepared and mixed in a 9:1 ratio with 10x YPD (100 g/L yeast extract, 200 g/L peptone, 200 g/L dextrose) to create YPD media with 15 %, 10 %, 5 %, 2.5 %, 1 % and 0.5 % IL. R. toruloides growth was conducted with and without pH adjustment of the solution, as typical growth media for this yeast is at pH 7.5 (Yaegashi et al., 2017). However, growth tests for S. cerevisiae were run without pH adjustment, as they were previously demonstrated to effectively grow in the pHs of the solutions tested (Sun et al., 2016). In this sense pH of solutions were R. toruloides was going to be tested, were adjusted using 2 M NaOH until a final pH of 7.5. Afterwards, microbial growth was monitored in 96-well plates containing 200 μL cell suspension by measuring OD600 with a Tecan F200 Microplate (Mennëdorf, Switzerland) reader with high orbital speed preheated to 30 °C. Experiments were run for 72 h in three biological replicates for each condition. The whole procedure was developed under sterile conditions. 2.4. One pot pretreatment and saccharification, and R. toruloides growth in lignocellulosic hydrolysates A larger batch of the described one-pot pretreatment and saccharification process was conducted by using the most favourable conditions (eucalyptus and [2-HEA][OAc]), with and without pH adjustment. 400 mg of eucalyptus powder (<150 µm) were used. Pretreatment and saccharification conditions were the same as in the screening (150 °C, for 6 h, for pretreatment and 50 °C, for 72 h, for saccharification). After saccharification, the solid and liquid fractions were separated by centrifugation. The solid fraction was washed and lyophilized for further analysis. The liquid fraction was used to test 6 R. toruloides growth, sugars and lignin consumption, and bisabolene production. Hydrolysates were separated into two fractions: pH adjusted (to 7.5) and non-pH adjusted. Growth was tested in hydrolysates containing 10%, 7.5%, 5%, 2.5% and 1.25 % (w/w) of IL. Synthetic defined (SD) medium was prepared mixing yeast nitrogen base without amino acids (BD Difco, Sparks, MD) and complete supplement mixture (Sunrise Science, San Diego, CA) following the manufacturers’ instructions and added to the hydrolysates. The inoculation and growth conditions were the same as those used previously. The growth was monitored with a Beckman Coulter DTX800 (Brea, California, United States) multimode detector, at 30°C and selecting high orbital speed mode. R. toruloides experiments in test tubes containing 10 mL batch medium were conducted to assess sugar and lignin monomers consumption and bisabolene production. R. toruloides was cultivated in one-pot hydrolysates supplemented with the SD medium. Ionic liquid in hydrolysates was at 10%, 7.5% and 5% (w/w), for both, pH-adjusted and non-pH adjusted samples. Each sample contained 2 mL of dodecane that formed a top-layer with the medium and is used to capture the bisabolene produced (Sundstrom et al., 2018). Conversion process was run at 30 °C and 200 rpm for 96 h, and samples were collected after 24, 48, 72, and 96 h. 2.5. Analytical methods NMR spectra of the ILs prepared were acquired at 298 K using a Bruker Avance-600 MHz instrument in DMSO-d6 and calibrated with the corresponding DMSO peak (δH=2.50 ppm for 1H). Monomeric sugars collected from different experiments of this study were quantified through high-performance liquid chromatography (HPLC) using Agilent 1260 Infinity system (Santa Clara, California, United States) equipped with a BioRad 300 × 7.8 mm Aminex 87 H column, a BioRad cation H guard column, and a refractive index detector heated at 35 °C. The samples were run according to the NREL/TP-5100-63351 procedure (Resch et al., 2015). The solids obtained after one-pot pretreatment and saccharification were characterized in terms of solid recovery, compositional analysis, lignin units and morphology. Compositional analysis of untreated and one-pot pretreated and saccharified biomass (eucalyptus) was performed according to the NREL/TP-510-42618 methodology, adapted to small quantities of sample (50 mg) (A. Sluiter, 2011; Ibáñez and Bauer, 2014). Acid-soluble lignin was quantified by UV spectroscopy at 205 nm, using an absorptivity coefficient of 110 L/g·cm. Changes in lignin monomers composition of one-pot pretreated and saccharified solids were revealed using a pyrolysis-GC–MS methodology reported previously (Papa et al., 2012; Shi et al., 2016). A Pyroprobe 5200 from CDS Analytical, Inc. (Chemical Data Systems, Oxford, PA, USA) 7 was used to pyrolize approximately 1 mg of sample that was loaded in a quartz tube sealed with glass wool. Surface morphology of untreated and pretreated samples was assessed through scanning electron microscopy (SEM). Images were acquired using a Jeol JSM 6400 (Akishima, Tokyo, Japan) operating at 20 kV. Analyses were developed at the technical facilities of the Spanish National Centre for Electron Microscopy. Phenolic compounds were detected at the beginning and at the end of the cultivation step using HPLC. The standards employed were 4-hydroxybenzoic acid, vanillic acid, 4-coumaric acid, ferulic acid, 4-hydroxybenzaldehyde, vanillin and benzoic acid. The method employed was previously reported (Eudes et al., 2014; Rodriguez et al., 2017). To detect bisabolene, samples from the dodecane overlay were diluted into ethyl acetate and analysed by gas chromatography-mass spectrometry (GC-MS), using an Agilent Technologies 6890N system, equipped with a 5973 mass selective detector and a DB-5ms column (30 m x 250 µm x 0.25 µm, Agilent Technologies, USA) (Özaydın et al., 2013; Yaegashi et al., 2017). 3. Results and discussion 3.1. Influence of the ionic liquid and biomass on the enzymatic digestibility In Figure 1, pine and eucalyptus glucan digestibilities after pretreatment and saccharification are presented. Results show an evident difference between hardwood (eucalyptus) and softwood (pine). Untreated pine and eucalyptus reported the smallest digestibilities (8.4 and 7.4 %, respectively). The most effective IL for both feedstocks was [Ch][Lys], with a final 72 h digestibility of 22% for pine, and 44 % for eucalyptus respectively. This is equivalent to three times untreated pine digestibility and six times untreated eucalyptus digestibility, respectively. Eucalyptus pretreated with [2-HEA][OAc] exhibited a digestibility of 42 % within 72 h. In contrast, pine pretreated with [2-HEA][OAc] reached only a digestibility of 17.4 %. Results corroborates the higher ability of [Ch][Lys] to more efficiently pretreat eucalyptus, compared to pine (Dutta et al., 2018). Regarding PILs, previous literature reported higher cellulose digestibility after pretreatment, although the work was conducted on herbaceous biomass and the conditions employed were different (Sun et al., 2017). The enzymatic kinetic curves shows similar aspect independently of the IL employed. After 24 h, no relevant increments of digestibility were observed in any of the samples, probably due to the absence of accessible-to enzymes cellulose. Differences between IL-pretreated (with [Ch][Lys] and PILs) hardwood and softwood digestibilities may be attributed to the kind of lignin present in each feedstock. In the case of 8 softwood (pine), pretreatment may lead to lignin recondensation, limiting the enzymatic hydrolysis (Shi et al., 2016). However, the cleavage effect of PILs and [Ch][Lys] over hardwood lignin appears to not be affected by the same mechanism. PILs have been reported to induce the acid rupture of the lignocellulosic network, which may worsen enzymatic hydrolysis, due to lignin recondensation (De Gregorio et al., 2016; Shuai and Saha, 2017). Figure 1: Enzymatic digestibility of a) pine and b) eucalyptus after one-pot pretreatment and saccharification. In all the cases pH was adjusted after pretreatment and prior to saccharification. Although [Ch][Lys] enhances considerably the enzymatic digestibility of the solid, the pH of solutions containing 11 % w/w [Ch][Lys] is 10.6 – 10.7. A pH adjustment step, prior to enzymatic hydrolysis is essential (Xu, F. et al., 2016). To reach a pH of 5 we had to add 2.2-2.6 moles of glacial acetic acid per mole of [Ch][Lys]. In contrast, the pH of [2-HEA][OAc] is close to the optimum for enzymatic cocktails (pH 5) and nor or negligible glacial acetic acid (0.02 moles of glacial acetic acid per mole of [2-HEA][OAc]) is needed to add to reach a pH of 5. In the case of [B2-HEA][OAc], pH at 11 % w/w was 6,4-7, which would lead to a considerable decrease of efficiency. 0.26-0.29 moles of acetic acid per mole of [B2-HEA]][OAc] were employed to adjust the pH to 5. Finally, the slightly low pH of [TEA][OAc] (pH equal to 4.7 in solutions containing 11 % w/w) required to be adjusted using 0.25-0.28 moles of triethylamine per mole of [TEA][OAc]. 3.2. Yeast growth in YPD media containing PILs The effect of IL toxicity toward a subsequent fermentation step was determined by comparing the growth curves of the two yeasts (S. cerevisiae and R. toruloides) in presence of [Ch][Lys], [2-HEA][OAc], [B2-HEA][OAc] and [TEA][OAc]. Growth curves were firstly 9 compared in terms of minimal biocidal concentration (MBC), defined as the minimal concentration tested where growth was observed, and results are shown in Figures 2a, 2b, and 2c. The MBC was of ~ 5 % IL in almost all the samples. In previous literature, the PILs [2-HEA][OAc] and [B2- HEA][OAc], as well as [Ch][Lys] had already been tested with S. cerevisiae and appreciable yeast growth had been observed even at 10% of [Ch][Lys] (Sun et al., 2016). Differences in this study (Figure 2c) in comparison to literature may be attributed to the different experimental conditions employed, such as using 96-well plates instead of culture tubes. Figure 2: Yeasts growth in presence of IL a): Rhodosporidium toruloides without pH adjustment at MBC; b): Rhodosporidium toruloides with pH adjustment at MBC; c): Saccharomyces cerevisiae with pH adjustment at MBC; d) Rhodosporidium toruloides with pH adjustment at 2.5, 5, 10 and 15 % [2-HEA][OAc]; e) Rhodosporidium toruloides without pH adjustment at 2.5, 5, 10 and 15 % [B2-HEA][OAc]; f) Rhodosporidium toruloides with pH adjustment at 2.5, 5, 10 and 15 % [B2-HEA][OAc] The same tendencies are followed in the subgraphs 2a and 2c (growth in presence of [B2- HEA][OAc] > [2-HEA][OAc] > [TEA][OAc] > [Ch][Lys]), while in Figure 2b the negative effect of the low pH of [TEA][OAc] is compensated adjusting the pH, showing higher growths. Comparing 10 the growth in R. toruloides, in Figures 2a and 2b, the growth is higher when the pH adjustment step is included, excepting for [B2-HEA][OAc], in which the maximal OD600 was reached in non-pH adjusted curves, being superior to the fresh media without IL. This may indicate that under these conditions, R. toruloides is using [B2-HEA][OAc] as substrate. The initial pH of that sample is close to neutral, (6.4 – 7) and the addition of NaOH to reach the required pH (7.5) has worsen the growth. Previous literature indicated that toxicity of ILs appears to be directly proportional to the alkyl groups substituted in the cation (Yu et al., 2016). However, in this case [B2-HEA][OAc] clearly appears to be the least toxic among the ILs tested. 3.3. One-pot pretreatment and saccharification of eucalyptus with [2-HEA][OAc] A fully integrated one-pot process must offer good digestibilities and acceptable host tolerances to the hydrolysates containing IL. Additionally, the minimization or removal of pH adjustment steps should be pursued. [2-HEA][OAc] + Eucalyptus globulus system offers the highest digestibilities, requiring nor or negligible pH adjustment step. However, Figure 2d shows that the growth, although exists, is being decreased by the presence of [2-HEA][OAc]. From the other side, [B2-HEA][OAc] would be a better alternative in terms of biocompatibility, as reflected in Figures 4e and 4f. However, the low digestibilities and the need to adjust the pH before saccharification strongly would limit this system from the pretreatement. A compromise solution needs to be searched, without limiting digestibilities but ensuring the growth of yeasts. In this study, we decided to choose the system [2-HEA][OAc] + Eucalyptus globulus that although limits the growth of R. toruloides would offer a good base that ensures high glucose digestibilities after saccharification, for a further development of conversion processes. In order to maximize sugar yields, initial particle size had to be reduced from 425 – 500 µm to <150 µm. The pretreatment + saccharification using eucalyptus particle size <150 µm lead to digestibilities of around 75 g/100 g of glucan (Figure 3). The enzymatic kinetic curves of glucose produced, observed with eucalyptus particle sizes <150 µm in [2-HEA][OAc] increased after 24- 48 hours. Curves obtained are similar to the observed with this material and the conventional IL [C2C1Im][OAc] in literature, suggesting that PILs effectivity highly depends on the particle size, increasing their efficiency when the particle size is smaller avoiding diffusional limitations (Li et al., 2013). Appart from glucose, a small amount of xylose is also released (41 %) and the production of this pentose continued after 48-72 h. Hydrolysates with 11 % [2-HEA][OAc], contains 4.55 g/L of glucose and 0.71 g/L of xylose. Additionally, vanillic acid, 4p-coumaric acid , vanillin and benzoic acid in quantities below 5 mg/g of untreated eucalyptus were found. No presence of lignin monomers derived from S structures were detected in the hydrolysate, while the presence of benzoic acid and p-coumaric acid can be a consequence of lignin depolimerisation and rupture (Sun et al., 2004). 11 Lignin-rich solid obtained after saccharification was recovered and characterized. Table 1a shows the compositional analysis of the recovered solid. Solid recovery was decreased because of carbohydrates hydrolysis into the liquid fraction and partial lignin depolymerisation of the liquid phase. The overall mass balance closure was 70.8%, due to losses during the analytical washing step and possible lignin depolymerisation to lignin monomers in the liquid phase. Figure 3: One-pot pretreatment and saccharification Eucalyptus in [2-HEA][OAc] The solid is mainly composed of lignin (69%), and glucan (23 %). Partial lignin extraction into the liquid fraction is produced, as only 68.4 g of the initial lignin remains in the solid. In the same way, only 11 % of the initial glucan remains in the solid, indicating that one-pot pretreatment and saccharification can convert most of the sugars. Solid contains also small amounts of xylan (1.6 %) and acetate (1.7 %). These values correspond to a total xylan extraction of 98 % and a total acetate extraction of 92 % from the untreated eucalyptus wood, that corroborates the liquid phase characterization, so almost complete extraction of most of the hemicelluloses (acetylated glucuronoxylan) has been produced (Isikgor and Becer, 2015). Table 1b shows the influence of syringyl (S)/ guaiacyl (G) ratio moieties (i.e. S/G ratio) as one of the features that impact recalcitrance of plant biomass (Studer et al., 2011). Listed in Table 1b, the S/G ratio detected by py-GC-MS decreased after pretreatment and saccharification, indicating that guaiacyl units content increased. As a consequence, syringyl units were extracted to the liquid phase during the pretreatment. For untreated samples the major components identified were 2-methoxy-4-methyl phenol, syringil acetone, 2-allyl-3,6-dimethoxyphenylmethanol and 4’- hydroxy-3’,5’-dimethoxyacetophenone). Key components identified in the pretreated + saccharified solid included 3-methoxy-catechol, 1-(3-hydroxy-4-methoxyophenylethanone), and 4-((1E)3- hydroxy-1-propenyl-2-methoxyphenol). 12 Table 1a. Recovered solid and chemical composition of untreated and one-pot pretreated and saccharified solid. Table 1b. Lignin composition of untreated and enzymatic in situ saccharified remaining solid. Scanning electron microscopy images (SEM) before and after pretreatment + saccharification differs considerably. While untreated eucalyptus particles are uniform and compact with no visible holes, remaining solids after saccharification are formed by an agglomeration of smaller particles, evidencing the material rupture. This structure has been altered by the enzymatic hydrolysis step, but also may be a cause of the PIL attack of the material. 3.4. R. toruloides growth in eucalyptus one-pot hydrolysates containing [2-HEA][OAc] The influence of pH adjustment on the growth of R. toruloides in [2-HEA][OAc] hydrolysates from eucalyptus was investigated and the results are shown in Figure 4. In the tests using 200 L microtiter well plates, we have not observed growth in hydrolysates containing 5 % w/w [2-HEA][OAc] without adjusting the pH (Figure 4a). However, the same growth was observed in Figure 4b (hydrolysate containing 5 % w/w of [2-HEA][OAc]) than in the equivalent synthetic sample (Figures 2b and 2d). This result certify that there is no presence of other compounds in the hydrolysates, apart from the IL, that negatively affect the growth of R. toruloides. Experiments run in test tubes (Figure 4c) confirmed that R. toruloides exhibit good growth rates in hydrolysates at pH 7.5 containing 5 % of [2-HEA][OAc], evincing the importance of vessel geometry and oxygen availability for an optimization of the conversion step (Bommareddy et al., 2015). Solid Yield COMPOSITION (g/100g ODW treated biomass) g/100 g ODW* untreated wood Lignin Glucan Xylan Galactan Arabinan Mannan Acetate % % % % % % % % Eucalyptus wood 100±0.00 27.21±0.20 51.08±1.32 18.14±0.35 0.62±0.10 0.18±0.18 0.82±0.41 4.12±0.11 OP-IL 25.53±0.66 69.75±3.05 20.61±0.68 1.43±0.17 0.50±0.02 0±0.0 5.78±0.81 1.28±0.04 Lignin characterisation Soluble lignin Insoluble lignin S/G ratio % % % Eucalyptus wood 17.2 82.8 3.94±0.03 OP-IL and saccharification 9.9 90.1 2.59±0.13 13 Analysing growth curves, at least two factors seem to be limiting the growth of R. toruloides in this media: the low sugars concentration at low IL concentration (below 5 g/L of glucose), and the presence of IL in the hydrolysate. At low [2-HEA][OAc] concentrations (1.25 % of [2- HEA][OAc]), differences between hydrolysates pH-adjusted and non-pH-adjusted were negligible, indicating that R. toruloides was able to grow despite low pH values in presence of low amounts of ILs (Yaegashi et al., 2017). Figure 4: Growth of Rhodosporidium toruloides in hydrolysates: a) without pH adjustment in 200 µL well-plates; b) with pH adjustment in 200 µL well-plates; and c) with and without pH adjustment in test-tubes. Percentages indicate the quantity (%w/w) of [2-HEA][OAc] present in hydrolysates 14 3.4.1. Carbon source consumption and production Complete sugar consumption (2.6 g/L) by the yeast was observed after 24 h, for experiments conducted in presence of 5% [2-HEA][OAc] and pH adjusted (Figure 5). However, without adjusting the pH, only 30 % of the initial sugars present in the hydrolysate were consumed. The limiting growth observed in that sample (max OD600 of ̴ 0.4 at 48 h, according to Figure 4c) verifies that in that case inhibition is produced due to the low pH of the hydrolysates (equal to 5) and consequently limits sugars consumption. In the sample containing 7.5 % w/w [2-HEA][OAc], 3.9 g/L of sugars were available. Only 20 % of sugars were consumed, confirming again that toxicity of hydrolysates due to the presence of [2-HEA][OAc] (max OD600 of ̴ 0.4 at 48 h, according to Figure 4c) is limiting sugars consumption. In the other conditions tested, the scarce growth of R. toruloides also limited its sugars consumption. Figure 5: Effect of pH adjustment on glucose and xylose consumption over the course of fermentations using Rhodosporidium toruloides in the presence of IL [2-HEA][OAc]. The fermentation was carried out at pH 7.5 (in pH adjusted samples) and pH 5 for non-pH adjusted samples, 30 °C, with initial cell density corresponded to 0.2 OD 600 nm in 10 mL hydrolyzate with total 2.28-4.55 g/L of sugar concentration R. toruloides was able to consume some lignin monomers under tested conditions. In the growth test performed at 5% IL without adjusting the pH, analysis via HPLC revealed that small quantities (<0.01 g/L) of the lignin monomers 4p-coumaric acid, benzoic acid, and vanillin were consumed. Additionally, in the sample containing 5 % [2-HEA][OAc], adjusting the pH, the vanillic acid present in hydrolysates was also consumed. Chromatograms confirmed that lignin-derived compounds consumption was higher in the pH-adjusted sample at 5 % w/w [2-HEA][OAc] where 15 total lignin monomers are consumed, than in the non-pH adjusted sample at 5 % w/w [2- HEA][OAc]. These results are in agreement with Yaegashi et al (2017), indicating that some lignin monomers are not toxic to R. toruloides, being even able to consumed them. As a result of the strict conditions employed (low substrate present in the media and the toxicity of [2-HEA][OAc]), 120 µg/L of bisabolene, were obtained in the sample pH-adjusted at 5 % w/w [2-HEA][OAc]. None bisabolene production was observed in the rest of the samples, confirming previously commented limiting factors. In future research, the employment of high- gravity configurations as suggested is a must, in order to increase sugar concentration and bisabolene production. 4. Conclusions The viability of a one-pot pretreatment+ saccharification process employing woody biomass has been confirmed. Protic ILs have proven to be more effective pretreating eucalyptus than pine, without the need of a pH adjustment step prior to saccharification. One-pot pretreatment and saccharification process using [2-HEA][OAc] and eucalyptus, yielded a glucan digestibility of 75 %. The solid obtained after saccharification evinced syringyl units had been selectively removed. R. toruloides cultivation in hydrolysates determined that host growth, sugar and lignin-derived compounds consumption, and bisabolene conversion is limited when pH of hydrolysates is not adjusted to 7.5 and at [2-HEA][OAc] concentrations above 5 %. E-supplementary data of this work can be found in online version of the paper Acknowledgements This work was performed with financial support from the “Ministerio de Economía y Competitividad” under the funded project CTQ2013-42006-R, the contract BES-2014-067788 and the grant EEBB-I-17-12049. This work was also supported by the U.S. Department of Energy, Office of Science, Office of Biological and Environmental Research, through the Joint BioEnergy Institute under contract no. DE-AC02-05CH11231 with the Lawrence Berkeley National Laboratory. Authors thank the laboratory of Dr. G. L. Gresham (Idaho National Laboratory) and Instituto Nacional de Investigación y Tecnología Agraria y Alimentaria of Spain (INIA) for providing biomass, as well as the Spanish National Centre for Electron Microscopy for SEM analysis. Finally, authors thank Jian Sun for his knowledge about the process provided, and the IL synthesis. 16 References A. Sluiter, B.H., R.Ruiz, C.Scarlata, J.Sluitr, D.Templeton and D. Crocker, 2011. Determination of Structural carbohydrates and lignin in biomass. NREL. Auxenfans, T., Husson, E., Sarazin, C., 2017. Simultaneous pretreatment and enzymatic saccharification of (ligno) celluloses in aqueous-ionic liquid media: A compromise. Biochem. Eng. J. 117, Part A, 77-86. http://dx.doi.org/10.1016/j.bej.2016.10.004. Bommareddy, R.R., Sabra, W., Maheshwari, G., Zeng, A.-P., 2015. Metabolic network analysis and experimental study of lipid production in Rhodosporidium toruloides grown on single and mixed substrates. Microb. Cell Fact. 14(1), 1. https://doi.org/10.1186/s12934-015-0217-5. Corona, A., Parajuli, R., Ambye-Jensen, M., Hauschild, M.Z., Birkved, M., 2018. Environmental screening of potential biomass for green biorefinery conversion. J. Clean. Prod. 189, 344-357. https://doi.org/10.1016/j.jclepro.2018.03.316. De Gregorio, G.F., Weber, C.C., Gräsvik, J., Welton, T., Brandt, A., Hallett, J.P., 2016. Mechanistic insights into lignin depolymerisation in acidic ionic liquids. Green Chem. 18(20), 5456-5465. https://doi.org/ 10.1039/C6GC01295G. Elgharbawy, A.A., Alam, M.Z., Moniruzzaman, M., Goto, M., 2016. Ionic liquid pretreatment as emerging approaches for enhanced enzymatic hydrolysis of lignocellulosic biomass. Biochem. Eng. J. 109, 252-267. https://doi.org/10.1039/C6GC01295G. Eudes, A., Juminaga, D., Baidoo, E.E.K., Collins, F.W., Keasling, J.D., Loqué, D., 2014. Correction: Production of hydroxycinnamoyl anthranilates from glucose in Escherichia coli. Microb. Cell Fact. 13, 8-8. https://doi.org/10.1186/1475-2859-13-8. Fan, C., Dong, H., Liang, Y., Yang, J., Tang, G., Zhang, W., Cao, Y., 2019. Sustainable synthesis of HKUST-1 and its composite by biocompatible ionic liquid for enhancing visible-light photocatalytic performance. J. Clean. Prod. 208, 353-362. https://doi.org/10.1016/j.jclepro.2018.10.141. Frederix, M., Mingardon, F., Hu, M., Sun, N., Pray, T., Singh, S., Simmons, B.A., Keasling, J.D., Mukhopadhyay, A., 2016. Development of an E. coli strain for one-pot biofuel production from ionic liquid pretreated cellulose and switchgrass. Green Chem. https://doi.org/ 10.1039/C6GC00642F. George, A., Brandt, A., Tran, K., Zahari, S.M.S.N.S., Klein-Marcuschamer, D., Sun, N., Sathitsuksanoh, N., Shi, J., Stavila, V., Parthasarathi, R., Singh, S., Holmes, B.M., Welton, T., Simmons, B.A., Hallett, J.P., 2015. Design of low-cost ionic liquids for lignocellulosic biomass pretreatment. Green Chem. 17(3), 1728-1734. https://doi.org/10.1039/C4GC01208A. George, A., Brandt, A., Zahari, S., Klein-Marcuschamer, D., Parthasarathi, R., Sun, N., Sathitsuksanoh, N., Shi, J., Stavila, V., Tran, K., Singh, S., Holmes, B.M., Welton, T., Simmons, B., Hallett, J., 2014. Design of low-cost ionic liquids for lignocellulosic biomass pretreatment. Green Chem. https://doi.org/ 10.1039/C4GC01208A. Hämäläinen, S., Näyhä, A., Pesonen, H.-L., 2011. Forest biorefineries–A business opportunity for the Finnish forest cluster. J. Clean. Prod. 19(16), 1884-1891. https://doi.org/10.1016/j.jclepro.2011.01.011. Ibáñez, A.B., Bauer, S., 2014. Downscaled method using glass microfiber filters for the determination of Klason lignin and structural carbohydrates. Biomass Bioenerg. 68, 75-81. http://dx.doi.org/10.1016/j.biombioe.2014.06.013. Isikgor, F.H., Becer, C.R., 2015. Lignocellulosic biomass: a sustainable platform for the production of bio-based chemicals and polymers. Polym. Chem. 6(25), 4497-4559. https://doi.org/10.1039/C5PY00263J. https://doi.org/10.1186/s12934-015-0217-5 https://doi.org/ https://doi.org/ https://doi.org/ https://doi.org/ https://doi.org/ 17 Kahani, S., Shafiei, M., Abdolmaleki, A., Karimi, K., 2017. Enhancement of ethanol production by novel morpholinium ionic liquids. J. Clean. Prod. 168, 952-962. https://doi.org/10.1016/j.jclepro.2017.09.008. Khan, A.S., Man, Z., Bustam, M.A., Kait, C.F., Nasrullah, A., Ullah, Z., Sarwono, A., Ahamd, P., Muhammad, N., 2018. Dicationic ionic liquids as sustainable approach for direct conversion of cellulose to levulinic acid. J. Clean. Prod. 170, 591-600. https://doi.org/10.1016/j.jclepro.2017.09.103. Li, C., Tanjore, D., He, W., Wong, J., Gardner, J.L., Sale, K.L., Simmons, B.A., Singh, S., 2013. Scale-up and evaluation of high solid ionic liquid pretreatment and enzymatic hydrolysis of switchgrass. Biotechnol. Biofuels 6(1). https://doi.org/10.1186/1754-6834-6-154. Maciel, V.G., Wales, D.J., Seferin, M., Lie Ugaya, C.M., Sans, V., 2019. State-of-the-art and limitations in the life cycle assessment of ionic liquids. J. Clean. Prod. https://doi.org/10.1016/j.jclepro.2019.01.133. Muñoz, P.D., 2011. Forestry in the EU and the world. A statistical portrait, in: Eurostat (Ed.). Belgium. Ouellet, M., Datta, S., Dibble, D.C., Tamrakar, P.R., Benke, P.I., Li, C., Singh, S., Sale, K.L., Adams, P.D., Keasling, J.D., 2011. Impact of ionic liquid pretreated plant biomass on Saccharomyces cerevisiae growth and biofuel production. Green Chem. 13(10), 2743-2749. https://doi.org/10.1039/C1GC15327G. Özaydın, B., Burd, H., Lee, T.S., Keasling, J.D., 2013. Carotenoid-based phenotypic screen of the yeast deletion collection reveals new genes with roles in isoprenoid production. Metab. Eng. 15, 174-183. https://doi.org/10.1016/j.ymben.2012.07.010. Papa, G., Varanasi, P., Sun, L., Cheng, G., Stavila, V., Holmes, B., Simmons, B.A., Adani, F., Singh, S., 2012. Exploring the effect of different plant lignin content and composition on ionic liquid pretreatment efficiency and enzymatic saccharification of Eucalyptus globulus L. mutants. Bioresour. Technol. 117, 352-359. https://doi.org/10.1016/j.biortech.2012.04.065. Prado, R., Erdocia, X., Labidi, J., 2016. Study of the influence of reutilization ionic liquid on lignin extraction. J. Clean. Prod. 111, Part A, 125-132. http://dx.doi.org/10.1016/j.jclepro.2015.04.003. Qadeer, S., Khalid, A., Mahmood, S., Anjum, M., Ahmad, Z., 2017. Utilizing oleaginous bacteria and fungi for cleaner energy production. J. Clean. Prod. 168, 917-928. https://doi.org/10.1016/j.jclepro.2017.09.093 Resch, M., Baker, J., Decker, S., 2015. Low solids enzymatic saccharification of lignocellulosic biomass. NREL Laboratory Analytical Procedure. Rodriguez, A., Salvachúa, D., Katahira, R., Black, B.A., Cleveland, N.S., Reed, M., Smith, H., Baidoo, E.E.K., Keasling, J.D., Simmons, B.A., Beckham, G.T., Gladden, J.M., 2017. Base- Catalyzed Depolymerization of Solid Lignin-Rich Streams Enables Microbial Conversion. ACS Sustain. Chem. Eng. 5(9), 8171-8180. https://doi.org/10.1021/acssuschemeng.7b01818. Shi, J., Pattathil, S., Parthasarathi, R., Anderson, N.A., Im Kim, J., Venketachalam, S., Hahn, M.G., Chapple, C., Simmons, B.A., Singh, S., 2016. Impact of engineered lignin composition on biomass recalcitrance and ionic liquid pretreatment efficiency. Green Chem. 18(18), 4884-4895. https://doi.org/10.1039/C6GC01193D Shuai, L., Saha, B., 2017. Towards high-yield lignin monomer production. Green Chem. 19(16), 3752-3758. https://doi.org/ 10.1039/C7GC01676J Simmons, B.A., Loque, D., Blanch, H.W., 2008. Next-generation biomass feedstocks for biofuel production. Genome Biol 9, 242. https://doi.org/ 10.1186/gb-2008-9-12-242. https://doi.org/10.1016/j.biortech.2012.04.065 18 Singh, R., Krishna, B.B., Kumar, J., Bhaskar, T., 2016. Opportunities for utilization of non- conventional energy sources for biomass pretreatment. Bioresour. Technol. 199, 398-407. http://dx.doi.org/10.1016/j.biortech.2015.08.117 Studer, M.H., DeMartini, J.D., Davis, M.F., Sykes, R.W., Davison, B., Keller, M., Tuskan, G.A., Wyman, C.E., 2011. Lignin content in natural Populus variants affects sugar release. Proceedings of the National Academy of Sciences 108(15), 6300-6305. https://doi.org/10.1039/C6GC01193D Sun, J., Konda, N.V.S.N.M., Shi, J., Parthasarathi, R., Dutta, T., Xu, F., Scown, C.D., Simmons, B.A., Singh, S., 2016. CO2 enabled process integration for the production of cellulosic ethanol using bionic liquids. Energy & Environmental Science 9(9), 2822-2834. https://doi.org/10.1039/C6EE00913A Sun, J., Konda, S., Ramakrishnan, P., Dutta, T., Valiev, M., Xu, F., Simmons, B., Singh, S., 2017. One-pot integrated biofuel production using low-cost biocompatible protic ionic liquids. Green Chem. 19, 3152-3163. https://doi.org/ 10.1039/C7GC01179B Sun, X.F., Sun, R.C., Tomkinson, J., Baird, M.S., 2004. Degradation of wheat straw lignin and hemicellulosic polymers by a totally chlorine-free method. Polym. Degrad. Stab. 83(1), 47-57. https://doi.org/10.1016/S0141-3910(03)00205-2. Sundstrom, E., Yaegashi, J., Yan, J., Masson, F., Papa, G., Rodriguez, A., Mirsiaghi, M., Liang, L., He, Q., Tanjore, D., 2018. Demonstrating a separation-free process coupling ionic liquid pretreatment, saccharification, and fermentation with Rhodosporidium toruloides to produce advanced biofuels. Green Chem. https://doi.org/10.1039/C8GC00518D van Osch, D.J., Kollau, L.J., van den Bruinhorst, A., Asikainen, S., Rocha, M.A., Kroon, M.C., 2017. Ionic liquids and deep eutectic solvents for lignocellulosic biomass fractionation. PCCP 19(4), 2636-2665. https://doi.org/ 10.1039/C6CP07499E. Weigand, L., Mostame, S., Brandt, A., Welton, T., Hallett, J., 2017. Feedstock considerations in lignocellulosic biomass fractionation using low-cost ionic liquids. Faraday Discuss. 202, 331-349. https://doi.org/ 10.1039/C7FD00059F. Xu, F., Sun, J., Konda, N.M., Shi, J., Dutta, T., Scown, C.D., Simmons, B.A., Singh, S., 2016. Transforming biomass conversion with ionic liquids: process intensification and the development of a high-gravity, one-pot process for the production of cellulosic ethanol. Energy Environ. Sci. 9(3), 1042-1049. https://doi.org/10.1039/C5EE02940F. Xu, J., Xiong, P., He, B., 2016. Advances in improving the performance of cellulase in ionic liquids for lignocellulose biorefinery. Bioresour. Technol. 200, 961-970. https://doi.org/10.1016/j.biortech.2015.10.031. Yaegashi, J., Kirby, J., Ito, M., Sun, J., Dutta, T., Mirsiaghi, M., Sundstrom, E.R., Rodriguez, A., Baidoo, E., Tanjore, D., 2017. Rhodosporidium toruloides: a new platform organism for conversion of lignocellulose into terpene biofuels and bioproducts. Biotechnol. biofuels 10(1), 241. https://doi.org/10.1186/s13068-017-0927-5. Yaguchi, A., Spagnuolo, M., Blenner, M., 2018. Engineering yeast for utilization of alternative feedstocks. Curr. Opin. Biotech. 53, 122-129. https://doi.org/10.1016/j.copbio.2017.12.003 Yu, Q., Zhuang, X., Wang, W., Qi, W., Wang, Q., Tan, X., Kong, X., Yuan, Z., 2016. Hemicellulose and lignin removal to improve the enzymatic digestibility and ethanol production. Biomass Bioenerg. 94, 105-109. https://doi.org/10.1016/j.biombioe.2016.08.005. https://doi.org/10.1016/S0141-3910(03)00205-2 https://doi.org/10.1016/j.biortech.2015.10.031 https://doi.org/10.1186/s13068-017-0927-5 S1 One-pot conversion of woody biomass feedstocks using protic ionic liquids Victoria Riguala, Gabriella Papab,c, Alberto Rodriguezb,c, Maren Wehrsb,c.d, Kwang Ho Kime, M. Virginia Alonsoa, Mercedes Olieta, John Gladdenc,f, Aindrila Mukhopadhyayb,c, Blake A. Simmonsb,c, Seema Singhc,f* aDepartment of Chemical Engineering, Faculty of Chemistry, Complutense University of Madrid, Avda. Complutense s/n, 28040 Madrid, Spain b Biological Systems and Engineering Division, Lawrence Berkeley National Laboratory, 1 Cyclotron Road, Berkeley CA, 94720, USA c Joint BioEnergy Institute, 5885 Hollis Street, Emeryville, CA, USA d Institut für Genetik, Technische Universität Braunschweig, Germany e Clean Energy Research Center, Korea Institute of Science and Technology, Seoul 02792, Republic of Korea f Biological and Engineering Sciences Center, Sandia National Laboratories, 7011 East Avenue, Livermore, California 94551, USA *E-mail: seesing@sandia.gov Figure S1: 1H-NMR spectra of the synthetized PILs 10 9 8 7 6 5 4 3 2 1 0 10 9 8 7 6 5 4 3 2 1 0 10 9 8 7 6 5 4 3 2 1 0 4 3 2 ppm NH3 + OH O O – CH3 1 23 4 1 [2-HEA][OAc] [B2-HEA][OAc] 4 3 2 1 ppm NH2 + OH OH 2 3 4 O O – CH3 1 [TEA][OAc] 1 (ppm) CH3CH3 NH + CH3 O O – CH3 1 2 3 2 3 mailto:seesing@sandia.gov S2 Table S1: Lignin derived pyrolysis products and relative retention times in order of elution found in untreated eucalyptus and [2-HEA][OAc] pretreated sample Lignin monomer Retention time (min) Unit G/S Sample 2-methoxyphenol 5.65 G Untreated, Pretreated Creosol 6.85 G Untreated, Pretreated 2-methoxy-4-methyl phenol 9.46 G Untreated 3-methoxy-catechol 9.54 G Pretreated 4-vinyl guaiacol 11.69 G Untreated, Pretreated 2,6-dimethoxy-phenol 12.82 S Untreated, Pretreated Vanillic acid methylesther 14.14 G Untreated, Pretreated 2-methoxy-5-propenyl-E-phenol 14.26 G Untreated, Pretreated 2-methoxy-4-vinylphenol 15.05 G Untreated, Pretreated 2-methoxy-4-propenylphenol (trans) 15.17 G Untreated, Pretreated α-ethyl-p-methoxy benzylalcohol 15.33 G Untreated, Pretreated 1-(3-hydroxy-4- methoxyphenylethanone) 15.90 G Pretreated 4-3-hydroxy-1-propenyl-2- methoxyphenol 16.58 G Untreated, Pretreated 4-vinyl-2,6-dimethoxy-phenol 17.13 S Untreated, Pretreated 4-allyl-2,6-dimethoxy-phenol 17.64 S Untreated, Pretreated 4-propenyl-trans 2,6-dimethoxy-phenol 18.35 S Untreated, Pretreated Syringaldehide 18.55 S Untreated, Pretreated 2-allyl-1,4-dimethoxy-3-methylbenzene 18.77 S Untreated, Pretreated 4-2-propenyl-2,6-dimethoxy-phenol 19.04 S Untreated, Pretreated Syringol-4-propenyl (trans) 19.42 S Untreated, Pretreated 4-((1E)3-hydroxy-1-propenyl-2- methoxyphenol 19.6 S Pretreated Syringil acetone 19.85 S Untreated 2-allyl-3,6-dimethoxyphenylmethanol 20.19 S Untreated 4’-hydroxy-3’,5’- dimethoxyacetophenone 20.51 S Untreated S3 Figure S2: SEM micrographs of untreated eucalyptus and remaining solid after enzymatic in situ saccharification Figure S3: Lignin monomers chromatogram a) without pH adjustment and b) with pH adjustment a) 7 8 9 10 11 12 13 7 8 9 10 11 12 13 Time (min) t=72 h 254nm t=0 h 254 nm t=72 h 280 nm t=0 h 280 nm t=72 h 310 nm t=0 h 310nm 7 8 9 10 11 12 13 7 8 9 10 11 12 13 Time (min) t=72 h 254 nm t= 0 h 254 nm t=72 h 280 nm t= 0 h 280 nm t=72h 310 nm t= 0h 310 nmb) Tesis Victoria de los Ángeles Rigual Hernández PORTADA ÍNDICE ÍNDICE DE FIGURAS ÍNDICE DE TABLAS RESUMEN / ABSTRACT RESUMEN ABSTRACT CAPÍTULO 1: LÍQUIDOS IÓNICOS EN EL PROCESADO DE BIOMASA CAPÍTULO 2: OBJETIVOS Y PLANTEAMIENTO DE LA TESIS DOCTORAL CAPÍTULO 3: PRETRATAMIENTOS EN UNA ETAPA CAPÍTULO 4: PRETRATAMIENTOS MULTIETAPA:AUTOHIDRÓLISIS + LÍQUIDOS IÓNICOS CAPÍTULO 5: PROCESO “ONE-POT”:PRETRATAMIENTO + HIDRÓLISISENZIMÁTICA CAPÍTULO 6: CONCLUSIONES / CONCLUSIONS CAPÍTULO 7: NOMENCLATURA CAPÍTULO 8: BIBLIOGRAFÍA ANEXO I: LISTADO DE LÍQUIDOS IÓNICOS YESTRUCTURA QUÍMICA ANEXO II: PUBLICACIONES