Nanomaterials 2019, 9, x; doi: FOR PEER REVIEW www.mdpi.com/journal/nanomaterials Article 1 Antibacterial Nanostructured Ti Coatings by 2 Magnetron Sputtering: from Laboratory Scales to 3 Industrial Reactors 4 Rafael Alvarez 1,2,*, Sandra Muñoz-Piña 3, María U. González 4, Isabel Izquierdo-Barba 5,6, Iván 5 Fernández-Martínez 3, Víctor Rico 1, Daniel Arcos 5,6, Aurelio García-Valenzuela 1, Alberto 6 Palmero 1, María Vallet-Regi 5,6, Agustín R. González-Elipe 1 and José M. García-Martín 4,* 7 1 Instituto de Ciencia de Materiales de Sevilla (CSIC-US), Américo Vespucio 49, 41092 Seville, Spain 8 2 Departamento de Física Aplicada I. Escuela Politécnica Superior. Universidad de Sevilla. c/ Virgen de África 9 7, 41011, Seville, Spain 10 3 Nano4Energy SLNE, C/ Jose Gutierrez Abascal 2, 28006, Madrid, Spain 11 4 Instituto de Micro y Nanotecnología, IMN-CNM, CSIC (CEI UAM+CSIC), Isaac Newton 8, 28760 Tres 12 Cantos, Madrid, Spain 13 5 Dpto de Química en Ciencias Farmacéuticas, Facultad de Farmacia, Universidad Complutense de Madrid, 14 Instituto de Investigación Sanitaria Hospital 12 de Octubre i+12, Plaza Ramón y Cajal s/n, 28040 Madrid, 15 Spain 16 6 CIBER de Bioingeniería Biomateriales y Nanomedicina (CIBER-BBN), Spain 17 * Correspondence: J.M.G.-M. josemiguel.garcia.martin@csic.es; R.A. rafael.alvarez@icmse.csic.es 18 Received: date; Accepted: date; Published: date 19 Abstract: Based on an already tested laboratory procedure, a new magnetron sputtering 20 methodology to simultaneously coat two-sides of large area implants (up to ~15 cm2) with Ti 21 nanocolumns in industrial reactors has been developed. By analyzing the required growth 22 conditions in a laboratory setup, a new geometry and methodology have been proposed and tested 23 in a semi-industrial scale reactor. A bone plate (Depuy Synthes) and a pseudo-rectangular bone 24 plate extracted from a patient have been coated following the new methodology, obtaining that 25 their osteoblast proliferation efficiency and antibacterial functionality were equivalent to the 26 coatings grown in the laboratory reactor on small areas. In particular, two kinds of experiments 27 have been performed: analysis of bacterial adhesion and biofilm formation, and 28 osteoblasts-bacteria competitive in vitro growth scenarios. In all these cases, the coatings show an 29 opposite behavior towards osteoblast and bacterial proliferation, demonstrating that the proposed 30 methodology represents a valid approach for industrial production and practical application of 31 nanostructured titanium coatings. 32 Keywords: Magnetron Sputtering; Oblique Angle Deposition; Nanostructured Titanium Thin 33 Films; Antibacterial Coatings; Osteoblast Proliferation; Industrial Scale 34 35 1. Introduction 36 Addressing the problem of infection from the very first stage, i.e. inhibiting the formation of the 37 bacterial biofilm, is a crucial n important step to prevent in preventing bone implant rejection. 38 Recent studies indicate that nanostructured surfaces can be a less aggressive alternative to 39 antibiotics to avoid infections [1, 2], with the additional advantage of improving the behavior of 40 osteoblasts, the cells that regenerate bone [3, 4]. In this regard, the fabrication of nanostructured 41 surfaces that may simultaneously favor the growth of osteoblast and hinder bacteria proliferation 42 represents a milestone in this research field with important implications, not only regarding 43 mailto:josemiguel.garcia.martin@csic.es mailto:rafael.alvarez@icmse.csic.es Nanomaterials 2019, 9, x FOR PEER REVIEW 2 of 14 improving the quality of life of patients but also and by promoting a new generation of orthopedic 44 implants. In the last few years, various alternatives have been proposed to induce such selective 45 behavior, either by using nanostructures that incorporate drugs or bactericides elements such as 46 silver [5-7], or by surface processing with a strong corrugation at the nanoscale [8, 9]. In our earlier 47 work in 2015, we manufactured nanostructured coatings made of titanium (Ti) nanocolumns by 48 oblique angle deposition (OAD) with magnetron sputtering onto the surface of Ti-6Al4V discs, one 49 of the alloys most commonly used in orthopedic implants. In vitro experiments showed that these 50 nanocolumnar Ti coatings exhibited an efficient antibacterial behavior against Staphylococcus aureus 51 (the bacterial adhesion decreased and the biofilm formation was prevented) without altering their 52 biocompatibility (the osteoblasts proliferated and retained their mitochondrial activity) [10]. 53 Moreover, in a more recent work, we have shown that these coatings also render exhibit similar 54 antibacterial properties functionality against Gram negative bacteria (Escherichia coli) and, what is 55 more important in order to have a direct impact in the field of medical implants, that these coatings 56 could be prepared on small areas (~ 1 cm2) either in a laboratory setup or in a semi-industrial scale 57 equipment [11]. 58 In this work paper we focus analyze on the practical use of these coatings and their fabrication 59 on larger scales, an aspect that is mandatory for the development of actual applications [12]. In 60 general, regarding the minimization of costs and other economic and throughput issues, turning 61 laboratory-size devices into operational market-ready products is a crucial engineering challenge 62 that demands scaling up laboratory procedures to large area and mass production [13]. This issue is 63 quite evident when using the magnetron sputtering (MS) method: by this technique, a plasma is 64 made to interact with a solid target in a vacuum reactor, producing the sputtering of atomic species 65 from a well-defined race-track region, and their deposition on a substrate located few centimeters 66 away [14]. In a classical MS configuration, the substrate is placed parallel to the target, producing the 67 growth of highly compact and dense coatings, in a process that has been easily scaled up to 68 mass-production methods by simply building larger versions of laboratory reactors [15]. Following 69 this methodology, the magnetron sputtering technique has demonstrated being of great utility for 70 the production market ready devices in microelectronics [16], optical coatings [17] or sensors [18], 71 among other devices and products [19-23]. Unlike the classical configuration, the OAD geometry 72 promotes the arrival of sputtered atoms at the substrate along an preferential oblique direction, 73 inducing surface shadowing mechanisms and the formation of nanocolumnar arrays, which has 74 been usually achieved by rotating tilting the substrate with respect to the target in laboratory scale 75 procedures. However, and due to the strongly non-lineal linear nature of these atomistic processes, 76 scaling up the OAD methodology from laboratory to mass production scales is not straightforward, 77 requiring the development of new approaches [24, 25] and reactor designs [26], issues that have 78 scarcely been addressed in the literature [27, 28]. 79 In this line, herein we develop a new engineering approach to coat with Ti nanocolumns 80 two-sides surfaces of bone plates with areas up to ~15 cm2 that are commonly used to immobilize 81 bone segments, and would be adequate for the development of this and other biomedical 82 applications. To set up this new methodology we have proceeded in the following way: we have first 83 analyzed the fundamental conditions leading to the formation of the nanocolumnar structures in a 84 laboratory reactor, in particular, the energy and angular distribution of sputtered particles ejected 85 from the magnetron target; then, based on these results, we have proposed a new geometry to 86 operate at oblique angles in semi-industrial reactors that reproduces these energy and momentum 87 distributions at much larger scales. To prove the feasibility of the proposed design, we have 88 homogeneously and simultaneously coated the two sides of relatively large substrates, and analyzed 89 whether the antibacterial functionalities were the same as those obtained on surfaces manufactured 90 in a laboratory MS reactor. In particular, two kind of experiments have been performed: bacterial 91 adhesion and biofilm formation, and osteoblasts-bacteria competitive in vitro essays, the latter also 92 named “Race for the Surface” competition [29]. 93 2. Experimental SetupMaterials and Methods 94 Nanomaterials 2019, 9, x FOR PEER REVIEW 3 of 14 The Ti coatings were first deposited grown in a MS laboratory setup described in detail in ref. 95 [30] that from now forth will be dubbed l-reactor (See Figure 1a). It has an AJA magnetron head with 96 a circular 5 cm diameter Ti target and a cylindrical 9 cm long metallic chimney that collimates the 97 ballistic flux of sputtered material and traps many of the thermalized atoms. The base pressure in the 98 chamber reactor is in the order of mid 10−7 Pa range and the distance between target and substrate is 99 22 cm. The parameters used to fabricate the columnar coatings in this reactor with Ar as sputter gas 100 in ref. [30] are: Pressure=0.15 Pa, power (DC discharge) = 300 W, and tilt angle of the substrate with 101 respect to the target . The semi-industrial scale reactor, which from now forth will be called 102 i-reactor, operates at the company Nano4 Energy (see Figure 1b). The target is rectangular and much 103 larger (20x7.5 cm2) and, as a result of its balanced magnetic configuration, exhibits a racetrack with 104 the shape of a rectangle (the long and short sides being 13.5 and 4.2 cm, respectively) with lines that 105 are about 3mm wide. 106 107 108 Figure 1. (a) Laboratory and (b) Semi-industrial reactors employed to grow the Ti nanocolumns. 109 As a first step to scale up the growth conditions from the l-reactor to the i-reactor, we have 110 employed as substrates fixation plates used in open trauma fractures that are known for their high 111 post-operatory infection rate (15% for patients with good health and more than 20% if they belong to 112 risk groups). We have coated two different fixation plates provided by Dr. Ricardo Larrainzar, Head 113 of the Orthopedic Surgery and Traumatology Department at the “Infanta Leonor” University 114 Hospital, Madrid. One of them was a new tubular plate from Depuy Synthes (made of Stainless Steel 115 with length 5.2 cm, width 0.9 cm and thickness 1mm, with convex and concave sides), while the 116 other was a pseudo-rectangular plate extracted from a patient and properly sterilized (with length 117 12 cm, width 1.3 cm and thickness 4 mm. For depositions in the i-reactor we have followed this 118 Nanomaterials 2019, 9, x FOR PEER REVIEW 4 of 14 methodology: in a first stage, the plate was immersed in the plasma for cleaning purposes (pulsed 119 DC voltage at 150KHz, -500 V bias voltage and a pressure of 1.2 Pa), after which the plate was left to 120 cool down for 30 minutes. In a second stage the Ti coating was deposited using the particular 121 geometrical configuration presented in the Results and Discussion section. The deposition 122 conditions are: Ar Pressure = 0.4 Pa, Power (DC discharge) = 325 W and time = 25 min. Under these 123 conditions, the deposition rate was ~12 nm/min and the thickness of the films about 300 nm. Finally, 124 for the competitive studies between bacteria and osteoblasts, medical grade Ti6Al4V disks were also 125 coated in the i-reactor and used for comparison with the large area sample results. 126 The microstructure morphology of the coatings was studied using with two different 127 techniques: Scanning Electron Microscopy (SEM) with a FEI Verios 460 Field Emission using 128 secondary electrons detection and Atomic Force Microscopy (AFM) with a Bruker Dimension Icon 129 microscope that operates operating in a non-contact mode and commercial probes (Nanosensors, 130 type PPP-FM). 131 In order tTo checktest the antibacterial properties capabilities of the coatings, The Depuy 132 Synthes bone plate was introduced suspended in a bacterial solution of S. aureus bacterial strain (108 133 bacteria ml-1) and incubated for 24 h in a 66% TSB + 0.2% glucose medium environment to promote 134 robust biofilm formation (20 g l-1 of Difco Bacto tryptic soy broth, Becton Dickinson, Sparks, MD). 135 After 24 h, the plate was washed three times with sterile PBS, stained with 3 l of Syto-9/propidium 136 iodide mixture, incubated for 15 min and washed with PBS. To specifically determine the formation 137 of the biofilm formation, we used calcofluor, a fluorescent dye that has been used to stain the biofilm 138 extracellular matrix of biofilms. In this case, 1 ml of calcofluor solution (5 mg ml-1) was used 139 inoculated after the addition of the Syto-9/propidium iodide mixture and was incubated for 15 min 140 at room temperature. The formation of the biofilm Biofilm formation was examined using a LEICA 141 SP2 Confocal Laser Scanning Microscope. In this way, live and dead bacteria can be distinguished, 142 with green and red color, respectively, as well as the extracellular matrix of the biofilm with blue 143 color. Further details can be found in Ref. [10]. 144 To further evaluate As an additional evaluation of the antimicrobial activity of the 145 nanostructured coatings, we carried out osteoblasts-bacteria competitive in vitro studies using the 146 coated and uncoated regions of Ti6Al4V disks described above. For this purpose, co-cultures of 147 MC3T3-E1 preosteoblast-like cells from mouse (Sigma-Aldrich) [31] and S. aureus 15981 laboratory 148 strain (ATCC) [10] were co-cultured over uncoated surfaces and on surfaces coated with the Ti 149 nanocolumns. Two different scenarios were simulated: i) accidental infection (S. aureus 150 concentrations of 102 cfu/ml), and ii) osteomyelitis scenario (S. aureus concentrations of 106 cfu/ml). In 151 both cases, the S. aureus suspensions were mixed with 104 cellscfu/ml of MC3T3-E1 152 preosteoblast-like cells, suspended in Todd Hewitt Broth (THB) and complete Dulbecco's modified 153 Eagle's medium (DMEM) and simultaneously seeded on the samples. After 6 hours of culture, 154 confocal microscopy studies were done and lactate dehydrogenase (LDH) levels were measured as a 155 parameter of osteoblast destruction. In this regard, for confocal microscope the actin of preosteoblast 156 cytoskeleton was stained with Atto565-conjugated phalloidin (red) and both cell nuclei and bacteria 157 stained with DAPI (blue). Moreover, LDH level was determined in the culture medium, which is 158 directly related to the rupture of the plasmatic membrane (cell death) that, when broken, releases all 159 organelles and enzymes present in the cytoplasm. Measurements were performed by using a 160 commercial kit (Spinreact) having an absorbance at 340 nm with a UV–Visible spectrophotometer. Two 161 measurements of three independent experiments -have been carried out. All data are expressed as 162 means ± standard deviations of a representative of three independent experiments carried out in 163 triplicate. Statistical analysis was performed using the Statistical Package for the Social Sciences (SPSS) 164 version 19 software. Statistical comparisons were made by analysis of variance (ANOVA). Scheffé test 165 was used for post hoc evaluations of differences among groups. In all of the statistical evaluations, p 166 <0.05 was considered as statistically significant. The most representative confocal images are shown in 167 this study. 168 3. Results and Discussion 169 Nanomaterials 2019, 9, x FOR PEER REVIEW 5 of 14 3.1. From laboratory to industrial reactors 170 In order to scale up the deposition procedure developed in the l-reactor we have employed a 171 well-known model to analyze the conditions required to grow the nanocolumnar films. In Figure 2a 172 we show the calculated polar angle of incidence of sputtered Ti atoms on the tilted substrate in the 173 l-reactor in our experimental conditions, as obtained by the SIMTRA code [32, 33]. There, we can 174 appreciate that the most probable angle of incidence over the substrate in this configuration is ~80º, 175 which is above the calculated angular threshold of about ~70º required to promote the formation of 176 the nanocolumns [30, 34]. Moreover, as indicated in Figure 2a, there is a fraction of deposition 177 species that arrive with lower angles of incidence, which corresponds to those atoms that have 178 undergone experienced collisions in the plasma gas and that have altered their original steering [35]. 179 The kinetic energy distribution function of these Ti atoms is shown in Figure 2b and it is 180 characterized by a long tail that extends up to energies above 10 eV, where it is clear the existence of 181 numerous deposition species with kinetic energy above surface binding energy of Ti (~5 eV), i.e. 182 with enough energy to induce kinetic energy-induced displacements processes of surface atoms 183 upon deposition [30]. Using these distributions, we have solved the model developed in ref. [30] to 184 account for the growth of Ti thin films by MS. The solution, presented in Figure 2c, shows a typical 185 nanocolumnar array very similar to those experimentally obtained in reference [30], supporting that 186 the necessary conditions for the growth of the Ti nanocolumns on a flat surface, as reported in ref. 187 [10], must also hold in the present case: i) the preferential angle of incidence of Ti sputtered species 188 onto the surface must be centered at about ~80º with respect to the substrate normal, and ii) the 189 kinetic energy distribution of deposition species must contain a significant fraction of atoms with 190 energies above the binding energy of Ti surface atoms, i.e. ~ 5 eV. 191 192 193 Figure 2. First row: (a) Polar angle distributions and (b) kinetic energy distributions of incident Ti 194 atoms with respect to the surface normal in the l- and i-reactors. Second row: solution of the model 195 for the conditions in (c) the l-reactor, and in (d) the i-reactor. 196 Based on the results outlined above, we have focused on reproducing both angular and kinetic 197 energy distribution functions when operating the i-reactor on larger surfaces. In this way, and given 198 the target and reactor geometry, we propose the geometrical arrangement shown in Figure 3. There, 199 we have placed the substrate perpendicular to the target, in such a way that atoms steaming from 200 the race-track may reach the substrate along an oblique angle of . Moreover, this particular 201 Nanomaterials 2019, 9, x FOR PEER REVIEW 6 of 14 configuration ensures that both sides of the substrate could be coated simultaneously. In Figure 2a 202 we show the calculated profile of the incident angle distribution of Ti species under this new 203 configuration, where we can notice the similarities with that obtained in the l-reactor. This similarity 204 extends to the kinetic energy distribution functions (see Figure 2b). In Figure 2d, we also show that 205 the calculated nanostructure of the films in the i-reactor is formed by a nanocolumnar array, very 206 similar to that obtained in the l-reactor (Figure 2c), suggesting the adequacy of the geometrical 207 approach presented in Figure 3. 208 209 Figure 3. Proposed geometry ((a) cross-sectional and (b) front views) to coat the implants on two 210 sides simultaneously with Ti nanocolumns. 211 3.2. Coating the Tubular Plate from Depuy Synthes 212 Following the geometrical configuration presented in Figure 3 and the conditions described in 213 the Experimental Setup Materials and Methods section, we have coated the Depuy Synthes plate in 214 the i-reactor. A mask protecting circa a quarter of the plate was employed to have an uncoated zone 215 for the sake of comparison when performing in vitro analyses (see Figure 4). Scanning Electron 216 Microscopy (SEM) and Atomic Force Microscopy (AFM) have been employed used to characterize 217 the morphology of the coating, although the latter technique could be only be applied on the convex 218 side, as the tip holder of the microscope crashed with the lateral edges of the plate when 219 approaching the concave surface. The uncoated zone presented a mirror-like brightness, indicative 220 of small roughness. In agreement with this visual observation both, SEM and AFM images of this 221 zone (not shown) indicate the absence of gaps or noticeable bumps on the surface, while the RMS 222 roughness measured with the latter technique was 4 nm. 223 224 Figure 4. Different views of the Depuy Synthes plate coated in the i-reactor. A mask protecting about 225 a quarter of the plate was used in order to have an uncoated zone to allow for comparison when 226 performing in vitro analyses. 227 Nanomaterials 2019, 9, x FOR PEER REVIEW 7 of 14 Figure 5a shows an AFM topographic map taken on the coated zone (convex side of the plate). 228 It is noteworthy the good homogeneity of the coating and the existence of a microstructure that 229 consists of regularly separated nanocolumns. Figures 5b-c shows SEM images of the coating that 230 were obtained on the convex and concave sides of the plate. The former shows a well distributed and 231 homogeneous Ti nanocolumnar array, very similar to those arrays obtained under laboratory 232 conditions in reference [30] and [10]. However, on the concave side, end even though the coating is 233 also homogeneous and consists of nanocolumns, these are now smaller both in diameter and length 234 and are well packed, resembling a film with a rather compact structure. This means that, due to the 235 curvature of the concave side of the substrate, sputtered atoms arrive at the surface with an angle of 236 incidence below 80º at some locations, obtaining structures similar to those found in the l-reactor for 237 lower angles of incidence [30]. This difference could be minimized by placing the substrate closer to 238 the target in figure 3b, thus promoting the arrival of sputtered species along higher polar angles of 239 incidencesome atoms from the target have not reached this side of the surface due to the concave 240 shape of the plate. 241 242 Figure 5. Microscopy images of the Depuy Synthes plate after deposition of Ti nanocolumns: (a) 243 AFM topographic map on the convex side of the plate; (b) SEM image of the nanocolumnar 244 structures in the convex and (c) concave side of the plate. 245 3.3. Coating of pseudo-rectangular plate extracted from a patient 246 As an additional test of the geometrical arrangement presented in Figure 3, we analyze the 247 microstructure and morphology of a pseudo-rectangular plate extracted from a patient, as described 248 in the Materials and MethodsExperimental Setup section. This plate was so large that it did not fit 249 the entrance gate of the observation chamber of the SEM equipment and could only be analyzed by 250 AFM. The initial gloss of the plate, which is rather matt instead of mirror-like, indicates that its 251 roughness is high [36] (see Figure 6). To ascertain this, we performed a study of its morphology 252 before the coating process. Figure 7a shows representative AFM images obtained in areas with 253 different scale sizes (left and right of the figure, respectively) before deposition. On the higher 254 magnification scale, the plate has a RMS roughness of 7 nm. However, in the image obtained in the 255 same area but over a wider field of view (8 micron side), it can be seen that these flat areas are 256 separated by deep cracks, with depths above one micron. This implies that, after the deposition 257 process, most cracks will remain uncovered because their walls cast a shadowed region that avoids 258 the arrival of most atoms inside, preventing the formation of nanocolumns. Consequently, the 259 coatings will be inhomogeneous and there will be a large part of the implant surface (i.e., smooth 260 areas of the initial surface) exhibiting well-formed nanocolumns, while a small percentage of it (deep 261 cracks) will remain uncoated. 262 Nanomaterials 2019, 9, x FOR PEER REVIEW 8 of 14 263 Figure 6. Photographs of the pseudo-rectangular plate extracted from a patient, before and after 264 deposition of Ti nanocolumns. 265 Following the sputtering process, the surface of the plate darkens considerably (c.f., middle and 266 bottom panel in Figure 6), which indicates that a nanostructured coating has been successfully 267 formed on both sides [37]. Figure 7b-c contain representative AFM images of the obtained coatings 268 on both sides of the implant, upper and lower, respectively. They are composed of titanium 269 nanocolumns, with a non-uniform distribution that depends on the morphology of the plate in each 270 specific region: the columns grown on flat areas do have the same height, but those grown on the 271 walls of the holes have lower height, as the initial surface was deeper. For example, Figure 7c shows 272 an area with a very deep crack (depth about 1 micron) where it can be appreciated that the height of 273 the columns is maximum at the top and gradually decreases when moving into the crack, until no 274 columns are formed at the bottom. Overall, the columnar morphology of the coating is remarkably 275 similar to that obtained on small substrates in the l-reactor in ref. [30] and [10]. 276 277 Nanomaterials 2019, 9, x FOR PEER REVIEW 9 of 14 Figure 7. AFM images of the pseudo-rectangular plate extracted from a patient, obtained in areas 278 with different size (left and right of the figure, respectively): (a) before deposition; (b) top side after 279 deposition; (c) bottom side after deposition. 280 3.4. Bacterial adhesion and biofilm formation 281 Once we have checked the nanocolumnar topography of the coatings produced in the i-reactor, 282 we analyze whether these maintain same functionality as those produced in the l-reactor, i.e. if they 283 are biocompatible and possess antibacterial capability. Following the bacterial growth procedure 284 described in the Materials and MethodsExperimental Setup section, live and dead bacteria could be 285 distinguished, with green and red color, respectively, as well as the extracellular matrix of the 286 biofilm with blue color. Results appear in Figure 8 where we can clearly notice the bacteria 287 proliferation on the uncoated region of the plate, which contains numerous living and dead bacteria, 288 along with numerous blue staining, typical of extracellular matrix covering on the bacterial colonies. 289 However, this blue stain does not appear in the coated zone in Figure 8, indicating the absence of 290 bacterial biofilm in this case. 291 292 Figure 8. Antimicrobial activity in an osteoblasts-bacteria competitive in vitro scenario. Green color 293 corresponds to live bacteria, red to dead bacteria and blue color corresponds to the extracellular 294 matrix of the bacterial biofilm. 295 In order to further evaluate the antimicrobial activity of the nanostructured coatings, 296 osteoblasts-bacteria competitive in vitro studies, already described in the Materials and 297 MethodsExperimental Setup section, were also carried out in two different scenarios using the 298 coated and uncoated regions of Ti6Al4V disks. 299 3.4.1. Accidental infection scenario 300 In this first case scenario, the MC3T3-E1/S. aureus ratio seeded was 100:1. Good osteoblast 301 adhesion was observed in the uncoated and coated surfaces (Figures 9a and 9b). However, several 302 lacunae can be observed in the case of the uncoated surface (Figure 9a), were colonies of S. aureus are 303 present. On the contrary, the nanocolumnar surface appears almost fully coated by a MC3T3-E1 304 preosteoblast-like cells monolayer that reaches about 90% coverage, as can be seen in Figure 10a 305 (right). 306 Nanomaterials 2019, 9, x FOR PEER REVIEW 10 of 14 307 Figure 9. Competitive co-culture MC3T3-E1/S. aureus: (a) 100:1 ratio (accidental infection scenario), 308 uncoated region after 6 hours; (b) 100:1 ratio (accidental infection scenario), coated region after 6 309 hours; (c) ratio 1:100 (osteomyelitis scenario), uncoated region after 6 hours; (d) ratio 1:100 310 (osteomyelitis scenario), coated region after 6 hours. 311 LDH levels were measured as a parameter of cells destruction, illustrated in Figure 10b (right). 312 There, it is evidenced that preosteoblast cells destruction is much higher on the uncoated surface 313 than on the nanocolumnar coating under accidental infection scenarios. 314 315 Figure 10. (a) Fraction of surface covered by preosteoblasts after 6 hours under osteomylelitis (left) 316 and accidental infection (right) scenarios; (b) LDH levels after 6 hours under osteomyelitis (left) and 317 accidental infection (right) scenarios. 318 3.4.2. Osteomyelitis scenario 319 In this case the MC3T3-E1/S. aureus ratio seeded was 1:100. After 6 hours of culture, Ti6Al4V is 320 covered by an important amount of bacteria that have colonized most of the implant surface (Figures 321 9c-d). The number of osteoblast cells is significantly reduced and the cells exhibit rounded 322 morphology with low spreading degree. On the contrary, the nanocolumnar surface shows a higher 323 degree of osteoblast proliferation and spread, thus occupying an important amount of surface 324 Nanomaterials 2019, 9, x FOR PEER REVIEW 11 of 14 (around 50 % as observed in Figure 10a, left). It must be highlighted the very low presence of S. 325 aureus in this sample, compared with Ti6Al4V. The LDH measurements also evidenced much higher 326 preosteoblast destruction in the case of Ti6Al4V (see Figure 10b, left). 327 As a final comment, it is important to underline that the existence of the cracks on the fixation 328 plates reported above implies that the coating is not fully homogeneous and therefore, based on the 329 results presented in [10], its efficiency as antibacterial coatings can be affected. This issue could be 330 minimized by making use of a rather standard industrial technique, by which the substrate rotates 331 around certain axis to enhance the film homogeneity. In this manuscript we have not attempted this 332 approach, as we aim at scaling up an already reported laboratory technique that operates on static 333 substrates. Yet, it is likely that the existence of cracks is minimized when the substrate rotates 334 around an axis parallel to the target (parallel to the substrate holder line in figure 3b), so sputtered 335 species may arrive at the film following a constant polar angle of incidence, but different azimuthal 336 angles. 337 338 4. Conclusions 339 We have developed a methodology based on a new geometry to coat two-side surfaces with 340 areas up to ~15 cm2 with Ti nanocolumns by magnetron sputtering at oblique angles, and 341 demonstrated its feasibility using a semi-industrial scale reactor. This method was developed by 342 calculating the necessary conditions for the growth of these structures in a laboratory-size reactor 343 and reproducing them in a different geometry, suitable to coat larger areas in an industrial-scale 344 reactor. These conditions are defined to control the incident angle distribution function of Ti atoms 345 in the gaseous phase in such a way that they arrive at the surface along an oblique direction of about 346 ~80º, and they possess a kinetic energy distribution function with a relevant proportion of deposition 347 atoms with energies above the surface binding energy of Ti on the film surface. 348 After checking the homogeneity and features of the nanocolumnar structures deposited on 349 different fixation plates on both sides, we have analyzed the antibacterial functionality of the coating 350 and demonstrated its equivalence to those produced in a laboratory reactor. In particular, two kind 351 of experiments have been performed: analysis of bacterial adhesion and biofilm formation, and 352 osteoblasts-bacteria competitive in vitro scenarios, the latter also named “Race for the Surface” 353 competition. In all these cases, we showed the opposite behavior of these surfaces towards osteoblast 354 and bacterial proliferation, and demonstrated that the proposed method represents a valid approach 355 to coat large surfaces on both sides in industrial reactors, maintaining same properties as 356 laboratory-produced coatings on much smaller surfaces. 357 Author Contributions: Formal analysis, Rafael Alvarez, Isabel Izquierdo-Barba and Alberto 358 Palmero; Funding acquisition, Daniel Arcos, María Vallet-Regi, Agustín R. González-Elipe and José 359 M. García-Martín; Investigation, Rafael Alvarez, Sandra Muñoz-Piña, Maria U. González, Isabel 360 Izquierdo-Barba, Ivan Fernández-Martínez, Victor Rico, Daniel Arcos, Aurelio García-Valenzuela, 361 Alberto Palmero and José M. García-Martín; Methodology, Rafael Alvarez, Sandra Muñoz-Piña, 362 Maria U. González, Isabel Izquierdo-Barba, Ivan Fernández-Martínez, Victor Rico, Daniel Arcos, 363 Aurelio García-Valenzuela, Alberto Palmero, María Vallet-Regi, Agustín R. González-Elipe and José 364 M. García-Martín; Project administration, José M. García-Martín; Resources, Maria U. González, 365 Ivan Fernández-Martínez, Victor Rico, Alberto Palmero and José M. García-Martín; Supervision, 366 Alberto Palmero and José M. García-Martín; Visualization, Rafael Alvarez, Sandra Muñoz-Piña, 367 Maria U. González, Isabel Izquierdo-Barba, Alberto Palmero and José M. García-Martín; Writing – 368 original draft, Rafael Alvarez, Isabel Izquierdo-Barba, Alberto Palmero and José M. García-Martín; 369 Writing – review & editing, Rafael Alvarez, Sandra Muñoz-Piña, Maria U. González, Ivan 370 Fernández-Martínez, Daniel Arcos, Alberto Palmero, María Vallet-Regi, Agustín R. González-Elipe 371 and José M. García-Martín. 372 Funding: The authors thank acknowledge the financial support from the European Regional 373 Development Funds program (EU-FEDER) and the MINECO-AEI (201560E055, 374 Nanomaterials 2019, 9, x FOR PEER REVIEW 12 of 14 MAT2014-59772-C2-1-P, MAT2016-75611-R and MAT2016-79866-R and network 375 MAT2015-69035-REDC) for financial support. The authors also thank acknowledge the financial 376 support of the University of Seville (V and VI PPIT-US) and Fundación Domingo Martínez for 377 financial support. M.V.-R. also thanks acknowledges funding from the European Research Council 378 (Advanced Grant VERDI; ERC-2015-AdG Proposal 694160). The authors also acknowledge thank the 379 service from the MiNa Laboratory at IMN funded by CM (S2018/NMT-4291 TEC2SPACE), MINECO 380 (CSIC13-4E-1794) and the EU (FEDER, FSE) for the support. 381 Acknowledgments: The authors acknowledge helpful discussions with Dr. Ricardo Larrainzar, 382 Head of the Orthopedic Surgery and Traumatology Department at the “Infanta Leonor” University 383 Hospital, Madrid. The National Center for Accelerators (Seville, Spain) is also acknowledged. 384 Conflicts of Interest: “The authors declare no conflict of interest.” 385 References 386 1. Puckett, S.D.; Taylor, E.; Raimondo, T.; Webster, T.J. The relationship between the nanostructure of 387 titanium surfaces and bacterial attachment. Biomaterials 2010, 31, 706-713. 388 2. Jahed, Z.; Lin, P.; Seo, B.B.; Verma, M.S.; Gu, F.X.; Tsui, T.Y.; Mofrad, M.R.K. Responses of Staphylococcus 389 aureus bacterial cells to nanocrystalline nickel nanostructures. Biomaterials 2014, 35, 4249-4254. 390 3. Webster, T.J.; Ejiofor, J.U. Increased osteoblast adhesion on nanophase metals: Ti, Ti6Al4V, and CoCrMo. 391 Biomaterials 2004, 25, 4731. 392 4. de Oliveira, P.T.; Zalzal, S.F.; Beloti, M.M.; Rosa, A.L.; Nanci, A. Enhancement of in vitro osteogenesis on 393 titanium by chemically produced nanotopography. J. Biomed. Mater. Res. A. 2007, 80, 554-64. 394 5. Kazemzadeh-Narbat, M.; Lai, B.F.L.; Ding, C.; Kizhakkedathu, J.N.; Hancock, R.E.W.; Wang, R. 395 Multilayered coating on titanium for controlled release of antimicrobial peptides for the prevention of 396 implant-associated infections. Biomaterials 2013, 34, 5969-5977. 397 6. Mei, S.; Wang, H.; Wang, W.; Tong, L.; Pan, H.; Ruan, C.; Ma, Q.; Liu, M.; Yang, H.; Zhang, L.; Cheng, Y.; 398 Zhang, Y.; Zhao, L.; Chu, P.K. Antibacterial effects and biocompatibility of titanium surfaces with graded 399 silver incorporation in titania nanotubes. Biomaterials 2014, 35, 4255-4265. 400 7. Cheng, H.; Xiong, W.; Fang, Z.; Guan, H.; Wu, W.; Li, Y.; Zhang, Y.; Alvarez, M.M.; Gao, B.; Huo, K.; Xu, J.; 401 Xu, N.; Zhang, C.; Fu, J.; Khademhosseini, A.; Li, F. Strontium (Sr) and silver (Ag) loaded nanotubular 402 structures with combined osteoinductive and antimicrobial activities. Acta Biomaterialia 2016, 31, 388-400. 403 8. Bhadra, C.M.; Truong, V.K.; Pham, V.T.H.; Al Kobaisi, M.; Seniutinas, G.; Wang, J.Y.; Juodkazis, S.; 404 Crawford, R.J.; Ivanova, E.P. Antibacterial titanium nano-patterned arrays inspired by dragonfly wings 405 Sci.Rep. 2015, 5, 16817. 406 9. Bagherifard, S.; Hickey, D.J.; de Luca, A.C.; Malheiro, V.N.; Markaki A.E.; Guagliano M.; Webster T.J. The 407 influence of nanostructured features on bacterial adhesion and bone cell functions on severely shot peened 408 316L stainless steel. Biomaterials 2015, 73, 185-197. 409 10. Izquierdo-Barba, I.; García-Martín, J.M.; Alvarez, R.; Palmero, A.; Esteban, J.; Pérez-Jorge, C.; Arcos, D.; 410 Vallet-Regí, M. Nanocolumnar coatings with selective behavior towards osteoblast and Staphylococcus 411 aureus proliferation. Acta Biomaterialia 2015, 15, 20-28. 412 11. Medina-Cruz, D.; González, M.U.; Tien-Street, W.; Fernández-Castro, M.; Vernet-Crua, A.; 413 Fernández-Martínez, I.; Martínez, L.; Huttel, Y.; Webster, T.J.; García-Martín, J.M. Synergic antibacterial 414 coatings combining titanium nanocolumns and tellurium nanorods. Nanomed-Nanotechnol 2019, 17, 415 36-46. 416 12. Anders, A. Plasma and ion sources in large area coating: A review. Surf. Coat. Tech. 2005, 200, 1893. 417 13. Walter, C.; Sigumonrong, D.P.; El-Raghy, T.; Schneider, J.M. Towards large area deposition of Cr2AlC on 418 steel. Thin Solid Films 2006, 515, 389. 419 14. Bräuer, G.; Szyszka, B.; Vergöhl, M.; Bandorf, R. Magnetron sputtering – Milestones of 30 years. Vacuum 420 2010, 84, 1354. 421 15. Betz, U.; Olsson, M.K.; Marthy, J.; Escola, M.F.; Atamny, F. Thin films engineering of indium tin oxide: 422 Large area flat panel displays application. Surf. Coat. Tech. 2006, 200, 5751-5759. 423 16. Choi, J.H.; Kim, Y.M.; Park, Y.W.; Park, T.H.; Jeong, J.W.; Choi, H.J.; Song, E.H.; Lee, J.W.; Kim, C.H.; Ju, 424 B.K. Highly conformal SiO2/Al2O3 nanolaminate gas-diffusion barriers for large-area flexible electronics 425 applications. Nanotechnology 2010, 21, 475203. 426 Nanomaterials 2019, 9, x FOR PEER REVIEW 13 of 14 17. Oliva-Ramirez, M.; Gonzalez-Garcia, L.; Parra-Barranco, J.; Yubero, F.; Barranco, A.; Gonzalez-Elipe, A.R. 427 Liquids analysis with optofluidic bragg microcavities. ACS Appl. Mater. Interfaces 2013, 5, 6743. 428 18. Martin, M.; Salazar, P.; Alvarez, R.; Palmero, A.; Lopez-Santos, C.; Gonzalez-Mora, J.L.; Gonzalez-Elipe, 429 A.R. Cholesterol biosensing with a polydopamine-modified nanostructured platinum electrode prepared 430 by oblique angle physical vacuum deposition. Sens. Actuator B-Chem 2017, 240, 37-45. 431 19. Ollitrault, J.; Martin, N.; Rauch, J.Y.; Sanchez, J.B.; Berger, F. Improvement of ozone detection with GLAD 432 WO3 films. Mater. Lett. 2015, 155, 1. 433 20. Sengstock, C.; Lopian, M.; Motemani, Y.; Borgmann, A.; Khare, C.; Buenconsejo, P.J S.; Schildhauer, T.A.; 434 Ludwig, A.; Koller, M. Structure-related antibacterial activity of a titanium nanostructured surface 435 fabricated by glancing angle sputter deposition. Nanotechnology 2014, 25, 195101. 436 21. Yoo, Y.J.; Lim, J.H.; Lee, G.J.; Jang, K.I.; Song, Y.M. Ultra-thin films with highly absorbent porous media 437 fine-tunable for coloration and enhanced color purity. Nanoscale 2017, 9, 2986. 438 22. Yang, L.; Zhao, Y.; Feng, Y.; Shan, J.S.; Liang, X.Y.; Huang, J.; Mi, J.H.; Wang, L.J.; Shi, W.M. The influence 439 of incident angle on physical properties of a novel back contact prepared by oblique angle deposition. 440 Appl. Surf. Sci. 2016, 363, 252. 441 23. Polat, D.B.; Keles, O. The effect of copper coating on nanocolumnar silicon anodes for lithium ion batteries. 442 Thin Solid Films 2015, 589, 543. 443 24. Godinho, V.; Alvarez, R.; Caballero-Hernandez, J.; Schierholz, R.; Bera, B.; Demarche, J.; Palmero, A.; 444 Fernandez, A.; Lucas, S. On the formation of the porous structure in nanostructured a-Si coatings 445 deposited by dc magnetron sputtering at oblique angles. Nanotechnology 2014, 25, 355705. 446 25. Schierholz, R.; Lacroix, B.; Godinho, V.; Caballero-Hernández, J.; Duchamp, M.; Fernández, A. STEM–447 EELS analysis reveals stable high-density He in nanopores of amorphous silicon coatings deposited by 448 magnetron sputtering. Nanotechnology 2015, 26, 075703. 449 26. Garcia-Valenzuela, A.; Alvarez, R.; Rico, V.; Cotrino, J.; Gonzalez-Elipe, A.R.; Palmero, A. Growth of 450 nanocolumnar porous TiO2 thin films by magnetron sputtering using particle collimators. Surf. Coat. 451 Tech. 2018, 343, 172-177. 452 27. Krause, K.M.; Taschuk, M.T.; Brett, M.J. Glancing angle deposition on a roll: Towards high-throughput 453 nanostructured thin films. J. Vac. Sci. Technol. A 2013, 31, 031507-1. 454 28. Ziegler, N.; Sengstock, C.; Mai, V.; Schildhauer, T.A.; Köller, M.; Ludwig, A. Glancing-Angle Deposition of 455 Nanostructures on an Implant Material Surface. Nanomaterials 2019, 9, 60. 456 29. Pham, V.T.H.; Truong, V.K.; Orlowska, A.; Ghanaati, S.; Barbeck, M.; Booms, P.; Fulcher, A.J.; Bhadra, 457 C.M.; Buividas, R.; Baulin, V.; Kirkpatrick, C.J.; Doran, P.; Mainwaring, D.E.; Juodkazis, S.; Crawford, R.J.; 458 Ivanova, E.P. “Race for the Surface”: Eukaryotic Cells Can Win. ACS Appl. Mater. Interfaces 2016, 834, 459 22025-22031. 460 30. Alvarez, R.; Garcia-Martin, J.M.; Garcia-Valenzuela, A.; Macias-Montero, M.; Ferrer, F.J.; Santiso, J.; Rico, 461 V.; Cotrino, J.; Gonzalez-Elipe A.R.; Palmero, A. Nanostructured Ti thin films by magnetron sputtering at 462 oblique angles. J. Phys. D: Appl. Phys. 2016, 49, 045303. 463 31. Izquierdo-Barba, I.; Santos-Ruiz, L.; Becerra, J.; Feito, M.J.; Fernández-Villa, D.; Serrano, M.C.; 464 Díaz-Güemes, I.; Fernández-Tomé, B.; Enciso, S.; Sánchez-Margallo, F.M.; Monopoli, D.; Afonso, H.; 465 Portolés, M.T.; Arcos, D.; Vallet-Regí, M. Synergistic effect of Si-hydroxyapatite coating and VEGF 466 adsorption on Ti6Al4V-ELI scaffolds for bone regeneration in an osteoporotic bone environment. Acta 467 Biomaterialia 2019, 83, 456-466. 468 32. van Aeken, K. SIMTRA. Available online: www.draft.ugent.be (Accessed on 29 July 2019). 469 33. van Aeken, K.; Mahieu, S.; Depla, D. The metal flux from a rotating cylindrical magnetron: a Monte Carlo 470 simulation. J. Phys. D. Appl. Phys. 2008, 41, 20530. 471 34. Dervaux, J.; Cormier, P.A.; Moskovkin, P.; Douheret, O.; Konstantinidis, S.; Lazzaroni, R.; Lucas, S.; 472 Snyders, R. Synthesis of nanostructured Ti thin films by combining glancing angle deposition and 473 magnetron sputtering: A joint experimental and modeling study. Thin Solid Films 2017, 636, 644-657. 474 35. Alvarez, R.; Romero‐Gomez, P.; Gil‐Rostra, J.; Cotrino, J.; Yubero, F.; Gonzalez‐Elipe, A.R.; Palmero, A. 475 Growth of SiO2 and TiO2 thin films deposited by reactive magnetron sputtering and PECVD by the 476 incorporation of non‐directional deposition fluxes. Phys. Status Solidi (a) 2013, 210, 796-801. 477 36. Stover, J.C. Optical Scattering – Measurement and Analysis. 1990. SPIE Optical Engineering Press. DOI: 478 10.1117/3.975276. 479 Nanomaterials 2019, 9, x FOR PEER REVIEW 14 of 14 34.37. Vitrey, A.; Alvarez, R.; Palmero, A.; González, M.U.; García-Martín, J.M. Fabrication of black-gold 480 coatings by glancing angle deposition with sputtering. Beilstein J. Nanotechnol. 2017, 8, 434–439. 481 © 2019 by the authors. Submitted for possible open access publication under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/). 482