1 Removal of chlorinated organic volatile compounds by gas phase adsorption with activated carbon Jesus Lemus*, M. Martin-Martinez, Luisa Gomez-Sainero, Miguel Angel Gilarranz, Jose Palomar, Juan J. Rodriguez Sección de Ingeniería Química (Departamento de Química Física Aplicada), Universidad Autónoma de Madrid, Cantoblanco, 28049 Madrid, Spain. *Telephone number: 34 91 4976938. Fax number: 34 91 4973516. e-mail: jesus.lemus@uam.es Abstract This paper discusses the removal of chlorinated volatile organic compounds (Cl-VOCs) from gas streams by means of fixed-bed adsorption with a commercial activated carbon (AC). Column experiments were performed at different conditions (inlet concentration, temperature, pressure, total flow rate and amount of adsorbent). A simple two-parameter model was applied to predict the entire breakthrough curves for chloromethane adsorption. The regeneration of the exhausted AC was performed at mild conditions (atmospheric pressure and room temperature). In order to gain a better knowledge on the effect of the surface chemistry of AC on the adsorption of Cl-VOCs, the quantum- chemical COSMO-RS method was used to simulate the interactions between AC surface groups and different Cl-VOCs as chloromethane, dichloromethane and trichloromethane. This information can be useful for tailoring the ACs with the objective of improving their adsorption capacities by further functionalization. To confirm this, the commercial AC tested was modified by mean of different thermal and oxidative treatments (nitric acid and ammonium persulfate), being the surface chemistry and textural properties of the resulting materials characterized by different techniques. mailto:jesus.lemus@uam.es 2 The modified ACs were then tested in column adsorption experiment with different Cl- VOCs. The uptake of these compounds increased with the basic character of the AC surface. Keywords: Activated carbon, adsorption, chlorinated volatile organic compounds, COSMO-RS. 3 1. Introduction Chlorinated volatile organic compounds (Cl-VOCs) play an important role in the chemical and pharmatheutical industries, where they are used as solvents and reagents. They are also employed in aerosols, adhesives, dry cleaning, etc. [1] Cl-VOCs are mainly regarded as xenobiotics, resistant to biodegradation and, hence, persistent in the environment. Most of them are toxic or carcinogenic and present potential hazard under exposure [2]. Therefore, Cl-VOCs are classified nowadays as hazardous gas pollutants and were included in the list of the seventeen highly harmful chemicals targeted in the emissions reduction effort of the U.S. Environmental Protection Agency [3,4]. Emissions of these compounds to the atmosphere contribute to the destruction of the ozone layer, to the formation of photochemical smog and to global warming. Hence, they are restricted by strong legal regulations. This enforces the need of developing effective technologies for the treatment of residual streams contaminated with Cl- VOCs. This paper will focus in three of the most common chlorinated compounds in off gases: monochloromethane (MCM), dichloromethane (DCM) and chloroform (TCM), which are associated to a number of industrial processes. Nowadays, the main technique for the removal of these pollutants is incineration, but it may lead to more dangerous byproducts than the original contaminants such as phosgene, dioxins and furans [5,6]. On the other hand, at low Cl-VOCs concentration the use of catalysts is required for reducing the thermal needs [7,8]. Thus, some other solutions are being investigated [9-11]. Hydrodechlorination using different active catalyst, like Pt or Pd, is one of the most promising [12-14]. High conversions have been reported for the most reactive chloromethanes (TCM>DCM>MCM), but catalyst 4 deactivation is so far a main drawback [15-17]. Biological treatments using biofilters and biotrickling filters have been also studied [18-20]. Adsorption, absorption and condensation are the three most common non-destructive techniques for the removal/recovery of Cl-VOCs from gas streams [21-24]. Adsorption with activated carbon (AC) has been widely used for the removal of gaseous organic pollutants [25,26]. The application of different complex sorbent materials for Cl-VOCs adsorption has been reported in the literature [27,28]. Long et al. achieved high adsorption capacities for TCM by means of a hypercrosslinked polymer as adsorbent. ACs present unique chemical and physical properties that make them well-known useful adsorbents [29,30]. Although AC adsorption has been used for decades in industry [31], the question of the interactions between the adsorptive molecules and the AC surface is still open for a better understanding. The case of Cl-VOCs entails a particular interest due to their significance as hazardous pollutants [31]. In the present work, column adsorption experiments of three Cl-VOCs (monochlorometane, dichloromethane and chloroform) at low concentration using a commercial AC supplied by Merck have been carried out. First, the influence of several variables (inlet concentration, temperature, pressure, gas flow rate and bed length) has been analyzed in order to optimize the operating condition for an effective adsorption of DCM. The breakthrough curves are described by a two-parameter theoretical model introduced by Yoon and Nelson [28,32] which was demonstrated that successfully predicts experimental adsorption data for a wide variety of gas solutes (including chlorinated compounds) onto AC at different operating concentrations and flow rates, allowing a better understanding of the specific factors influencing contaminant breakthrough [28,33]. The length of the mass transfer zone (HMTZ), which characterizes the wave front of the fixed-bed column and depends on the influent flow rate, the rate of 5 adsorption and adsorption equilibrium, is also estimated for design considerations. Subsequently, the quantum-chemical COSMO-RS (Conductor-like Screening Model for Real Solvents) model, developed by Klamt and co-workers [34], is used to analyze Cl- VOC-AC interactions. Previous works have showed that COSMO-RS method can successfully be applied to predict thermodynamic adsorption [35,36] and absorption [37-39] data. In this work, COSMO-RS predictions of Henry’s law constants, as thermodynamic parameter of reference for solute-adsorbent interactions of MCM, DCM and TCM in differently functionalized AC models (Figure 1) were performed, with the aim of designing tailor-made AC adsorbents with improved surface chemistry for selective adsorption of Cl-VOCs. Finally, the influence of the surface chemical composition of AC on the adsorbent performance has been experimentally evaluated by testing thermally and chemically treated ACs showing different surface polarity. These ACs were tested as adsorbents in fixed-bed experiments for MCM, DCM and TCM. The characterization of the modified ACs by 77K N2 adsorption-desorption and temperature programmed desorption (TPD) allows correlating the porous structure and surface chemistry of the adsorbents with their capacity for retaining Cl-VOC from gas streams. 6 Fig.1 – Molecular models of Cl-VOCs and the nine different functionalized ACs tested. 2. Materials and methods 2.1. Materials and methods A commercial AC supplied by Merck (AC-MkU) was used as adsorbent both virgin and after modification upon to oxidative treatments with nitric acid and ammonium persulfate. Nitric acid and ammonium persulfate were supplied, respectively, by Riedel de Haën and Sigma-Aldrich (both with purity >98%). Nitric acid treatment was carried out by boiling 1g of AC in 10 mL of 6 N nitric acid solution for 20 minutes as described elsewhere [40], giving rise to the treated AC-MkN. Oxidation with ammonium persulfate was performed by treating 1g of AC in 10 mL of 1 M solution at room temperature [41], leading to the treated AC-MkS. After these treatments the samples were washed with distilled water until neutrality and dried overnight at 100ºC. Finally, (A) MCM (B) DCM (C) CLF (1) None functionalization (AC) (2) Carboxylic anhydride (AC-COOCO) (3) Hydroxyl (AC-OH) (4) Carboxylic acid (AC-CCC) (5) Carbonyl (AC-CO) (6) Nitrile (AC-CN) (7) Amide (AC-CONH2) (8) Amine (AC-NH2) (9) Pyridine (AC-Pyr) Solutes: Adsorbents: 7 the ACs were grinded and the fraction between 0.1-0.5 mm particle size was separated by sieving and used in the experiments. Commercial mixtures of N2 with MCM, DCM and TCM (at 4000 ppmv in all cases) and bare N2 (purity ~99.999%) were supplied by Praxair. 2.2. Experimental The adsorption of Cl-VOCs was evaluated in a continuous flow system (depicted in Scheme 1), which consists of a quartz fixed-bed column (1/4 inch diameter), coupled to a gas chromatograph with a flame ionization detector. Experiments were conducted under different conditions of temperature, pressure, total flow rate, inlet chloromethane concentration and bed length. Desorption experiments were performed after exhaustion of the adsorbents using the same experimental set-up, passing a N2 flow of 100 mL·min-1 through the column. The inlet gas, with a chloromethane concentration from 200 to 4000 ppmv, was prepared by mixing adequate proportions of the starting chloromethane /N2 commercial mixture and N2. The values of initial concentration (C0) reported are those registered by gas chromatograph. The porous structure of the starting and modified AC was characterized from 77 K N2 adsorption–desorption using a Micromeritics apparatus (Tristar II 3020 model). The samples were previously outgassed at 423 K for 8 h to a residual pressure of 10-5 Torr. The BET equation was used to obtain the apparent surface area (ABET) and the Dubinin– Radushkevich equation for micropore volume estimation. The difference between the 8 volume of N2 adsorbed (as liquid) at 0.95 relative pressure and the micropore volume was taken as mesopore volume [42]. The amount of oxygen surface groups of the ACs was determined by temperature programmed desorption (TPD), heating 0.1 g of the AC sample up to 1100ºC in a vertical quartz tube under continuous N2 flow of 1 NL·min-1 at a heating rate of 10ºC·min-1. The evolved amounts of CO and CO2 were analyzed by means of a non- dispersive infrared absorption analyzer (Siemens, model Ultramat 22). The CO and CO2 TPD profiles were deconvoluted using PeakFit 4.12, selecting a multiple Gaussian function to fit each deconvolution peak [43]. Scheme 1. - Diagram of the gas adsorption system. 2.3. Computational details The molecular geometry of all compounds (MCM, DCM, TCM and AC models) was optimized at the B3LYP/6-31++G** computational level in the ideal gas phase using the quantum chemical Gaussian03 package [44]. Vibrational frequency calculations C l-V O C / N 2 Oven Reactor GC (FID) CV CV Gas N 2 FIC2 FIC1FT1 TIC TIC PIC FT2 9 were performed in each case to confirm the presence of an energy minimum. Once the molecular models were optimized, Gaussian03 was used to compute the COSMO files. The ideal screening charges on the molecular surface for each species were calculated by the continuum solvation COSMO model using BVP86/TZVP/DGA1 level of theory. Subsequently, COSMO files were used as an input in COSMOthermX [45] code to calculate the thermodynamic properties (Henry’s law constants of Cl-VOCs for different AC models). According to our chosen quantum method, the functional and the basis set, we used the corresponding parameterization (BP_TZVP_C21_0108) for COSMO-RS calculations in COSMOtherm code. 3. Results 3.1. Analysis of operating conditions Figures 2 to 6 depict the breakthrough curves obtained for the adsorption of DCM onto the commercial AC-MkU at the different operating conditions. For the sake of comparison, we have obtained the values of DCM adsorption capacity at saturation (qs) of AC-MkU from the breakthrough curves for each experiment, calculated by: 𝑞𝑞𝑠𝑠 = 𝑄𝑄 𝑚𝑚 ∫ 𝐶𝐶0 − 𝐶𝐶 𝑑𝑑𝑑𝑑𝑡𝑡𝑠𝑠 0 (1) where Q is the gas flow rate (N⋅L·min-1); m is the mass of adsorbent in the column (mg); t the time flow at which C=C0 and C0 and C the inlet and outlet DCM concentrations (mg·L-1). Also, the theoretical model developed by Yoon and Nelson [32,46-48] was applied to describe the obtained breakthrough curves, which were fitted to the expression : 10 𝑡𝑡 = 𝑡𝑡0.5 + 1 𝑘𝑘′ ln 𝐶𝐶 𝐶𝐶0−𝐶𝐶 (2) where t is the operation time (min), C0 and C are the inlet and outlet solute concentration (mg·L-1), respectively, t0.5 is the time at which the outlet concentration is one half of the inlet and k’ is a proportionality constant with dimension of reciprocal time (min-1). In addition, the length of mass transfer zone (HMTZ) has been estimated from the breakthrough curves using the expression: 𝐻𝐻𝑀𝑀𝑀𝑀𝑀𝑀 = 𝐻𝐻 ∙ (𝑡𝑡0.95−𝑡𝑡0.05) 𝑡𝑡0.95 (3) where H is length of the entire AC bed in the fixed-bed column, t0.9 and t0.1 are the times at which the outlet DCM concentrations are 5 and 95% of the inlet one, respectively. Table 1 summarizes the results of the adsorption experiments carried out with DCM and the original activated carbon AC-MkU at the different operating conditions tested, including qs values, the obtained adjustable parameters (t0.5 and k’) and correlation coefficient of Yoon and Nelson model and the HMTZ data estimated from the breakthrough curves of Figures 2-6. As can be seen, the experimental breakthrough curves are fairly well described by Yoon and Nelson model. As expected, the saturation capacity (qs) is neither affected by the gas flow-rate (Q) nor by the amount of AC (m) provided the inlet DCM concentration is maintained (practically) constant (Figures 2-3). On the other hand, that saturation capacity increases with the operating pressure due to the higher relative pressure of DCM. Looking at the length of the mass-transfer zone (HMTZ) it consistently increases with the gas flow-rate. Increasing the pressure at the same normal flow-rate (four last experiments of Table 1) also decreases HMTZ due to the lower values of superficial velocity. 11 Table 1. Operating conditions and results of the adsorption experiments of DCM with commercial AC-MkU. Q m C0 T P qs HZTM k R2 (N mL·min-1) (g) (ppmv) (ºC) (atm) (mg·g-1) (cm) (h-1) 100 0.25 1015 35 1.5 111.1 1.4 21.8 0.995 60 0.25 1003 35 1.5 115.7 0.9 19.6 0.950 40 0.25 983 35 1.5 115.9 0.6 18.7 0.977 100 0.42 1000 35 1.5 111.7 1.3 20.9 0.957 100 0.25 1015 35 1.5 111.1 1.4 21.8 0.995 100 0.16 1000 35 1.5 111.0 1.2 19.8 0.952 100 0.25 3998 35 1.5 200.8 1.3 48.4 0.937 100 0.25 2711 35 1.5 165.8 1.0 47.1 0.982 100 0.25 2001 35 1.5 187.1 1.1 30.4 0.982 100 0.25 1431 35 1.5 150.5 1.3 24.1 0.997 100 0.25 1015 35 1.5 111.1 1.4 21.8 0.995 100 0.25 808 35 1.5 108.0 1.2 19.1 0.979 100 0.25 500 35 1.5 107.5 0.8 17.0 0.952 100 0.25 207 35 1.5 61.7 1.0 10.3 0.992 100 0.25 996 100 1.5 29.4 2.7 33.9 0.950 100 0.25 1047 75 1.5 44.8 2.0 32.8 0.989 100 0.25 1024 50 1.5 81.7 1.3 29.9 0.958 100 0.25 1015 35 1.5 111.1 1.4 21.8 0.995 100 0.25 1066 35 4 156.1 0.8 27.5 0.979 100 0.25 1012 35 3 143.4 0.9 27.2 0.954 100 0.25 1001 35 2 121.7 1.1 24.5 0.999 100 0.25 1015 35 1.5 111.1 1.4 21.8 0.995 Operating conditions 12 Fig.2- Experimental (dots) and predicted (lines) breakthrough curves of DCM adsorption with AC-MkU at different gas flow rates. Experiments carried out with 250 mg of AC-MkU, ~1000 ppmv of inlet DCM concentration, 35ºC and 1.5 atm. Figure 2 shows the breakthrough curves obtained varying the amount of adsorbent (AC- MkU) in the column, thus giving rise to different bed lengths and, consequently, contact times. No significant effects on the thermodynamics and the kinetics of adsorption process were observed, as can be seen from the values of qs, k’ and HMTZ values of Table 1. 0.0 0.2 0.4 0.6 0.8 1.0 0 50 100 150 200 250 300 C /C 0 Time (min) 100 NmL/min 60 NmL/min 40 NmL/min 13 Fig.3- Experimental (dots) and predicted (lines) breakthrough curves of DCM adsorption with AC-MkU using different bed lengths. Experiments carried out at 1000 ppmv of inlet DCM concentration, 35ºC, 1 atm and 100 N mL·min-1 of gas flow rate. Figure 4A shows the breakthrough curves for DCM adsorption with AC-MkU at 35ºC and different DCM inlet concentrations. The values of adsorption capacity at saturation (qs) and the rate constant (k) are collected in Table 1. The breakthrough curve is steeper at higher inlet concentration, since the concentration gradient (driving force) increases giving rise to higher values of the kinetic constant k. Obviously saturation of the adsorbent bed occurs in less time due to the higher DCM mass flow. As expected, the time required for 50% DCM breakthrough (t0.5) decreases when increases the inlet DCM concentration. From these breakthrough curves, the adsorption isotherm of DCM can be obtained, being depicted in Figure 4B. The experimental data were fitted to the well-known Langmuir equation: qe = qmBPe 1+BPe (3) 0.0 0.2 0.4 0.6 0.8 1.0 0 50 100 150 200 C /C 0 Time (min) 4cm 7cm 11cm 14 where qm refers to the monolayer adsorption capacity, Pe is the partial pressure of DCM at equilibrium and B is the Langmuir constant. The equilibrium data of Figure 4B reveal a fairly high adsorption capacity of AC-MkU for DCM (~200 mg·g-1). This value compares very well with the reported by Long et al. [27] using a complex hydrophobic hypercrosslinked polymer, recently proposed as an effective adsorbent for removal of Cl-VOCs. Fig.4- (A) Experimental (dots) and predicted (lines) breakthrough curves at different DCM inlet concentrations; and (B) adsorption isotherm obtained from the breakthrough curves at 35ºC. Experiments carried out with 250 mg of AC-MkU, 35ºC, 1 atm and 100 N mL·min-1 gas flow rate. Another important operating condition is temperature. Figure 5 shows the breakthrough curves obtained at 35, 50, 75 and 100ºC for DCM adsorption with AC-MkU at 1000ppmv inlet concentration. The adsorption capacity is significantly higher at lower temperature, indicative of physical adsorption. The favorable effect of temperature on the adsorption rate is confirmed by the k values (Table 1). Increasing the temperature leads to a higher volumetric gas flow giving rise to higher HMTZ values, in spite of higher adsorption rate. 0 50 100 150 200 0.00 0.10 0.20 0.30 0.40 A ds or pt io n ca pa ci ty (m g g-1 ) DCM Partial Pressure (KPa) (B)(A) 0.0 0.2 0.4 0.6 0.8 1.0 0 50 100 150 200 250 300 C /C 0 Time (min) 4000 ppm 3000 ppm 2000 ppm 1500 ppm 1000 ppm 800 ppm 500 ppm 200 ppm 15 Fig.5- Experimental (dots) and predicted (lines) breakthrough curves of DCM adsorption at different temperatures. Experiments carried out with 250 mg of AC-MkU, 1000 ppmv of initial DCM concentration, at 1 atm and 100 N mL·min-1 gas flow rate. Figure 6 shows the breakthrough curves of DCM adsorption with AC-MkU at different total pressures. Increasing this pressure means a higher DCM relative pressure thus determining higher values of DCM uptake at equilibrium as can be seen from the qs values of Table 1. The volumetric gas flow decreases as the operating pressure increases leading to higher contact time and lower HMTZ values. Simultaneously the higher driving force improves the adsorptive rate (see the corresponding k values in Table 1). 0.0 0.2 0.4 0.6 0.8 1.0 0 25 50 75 100 125 C /C 0 Time (min) 35ºC 50ºC 75ºC 100ºC 16 Fig.6- Experimental (dots) and predicted (lines) breakthrough curves of DCM adsorption with AC-MkU at different total pressures. Experiments carried out with 250 mg of AC-MkU, 1000 ppmv of inlet DCM concentration, 35ºC and 100 N mL·min-1 gas flow rate. Regeneration of the adsorbent is an important issue regarding the potential application of adsorption. Figure 7 shows examples of DCM adsorption/desorption curves obtained with AC-MkU. Desorption was performed under continuous N2 flow (100 NmL·min-1) at the same temperature and pressure than adsorption (35ºC and 1 atm in this case). Complete desorption of DCM was achieved and the adsorption capacity after four successive regeneration cycles remained virtually constant. 0.0 0.2 0.4 0.6 0.8 1.0 0 25 50 75 100 125 150 C /C 0 Time (min) 1.5atm 2atm 3atm 4atm 17 Fig.7- Experimental adsorption (■) desorption (□) curves of DCM. Experiments carried out with fixed 250 mg of AC-MkU, at 35ºC and 1 atm, 1000 ppmv DCM/N2 and 100 N mL·min-1 of total gas flow. 3.2. Adsorption with chemically modified ACs. It is well-known that both the porous structure and the surface chemistry of activated carbons are important issues determining the adsorption of different solutes [49]. Increasing the surface area of sorbent materials will generally lead to improved adsorption, but the effect of surface chemical functional groups of ACs on the solute adsorption is closely related to the chemical nature of the adsorptive. Therefore, we have analyzed the possible interactions between different functional groups in the surface of AC and Cl-VOCs. For this purpose, molecular models of AC with a wide variety of functional groups (Figure 1) were optimized by quantum-chemical calculations and used in COSMO-RS to calculate the Henry constants (partition coefficient between gas phase and AC) for the three Cl-VOCs (MCM, DCM and TCM). Those Henry constants were used for predicting the adsorption capacities. The results 0.0 0.2 0.4 0.6 0.8 1.0 0 200 400 600 800 1000 1200 C /C 0 Time (min) 18 are collected in Table 2. A lower value of the Henry constant means a higher uptake of Cl-VOC by AC, thus, according with the values of Table 2 the adsorption capacities follow the order TCM > DCM > MCM for all the AC models studied. Therefore, the predicted uptake of Cl-VOCs by ACs increases with the number of chlorine atoms of the molecule, consistently with the lower volatility of the solute. In addition, the results of Table 2 indicate that the adsorption capacity is affected by the nature of the functional groups on the AC surface. The presence of hydrogen-bond acceptor groups, as carbonyl or carboxylic anhydride, enhances solute-AC interactions, and, consequently, increases the adsorption uptake in some extent, being this effect more significant as the acidic character of Cl-VOCs increases in the order TCM>DCM>MCM. On the other hand, the contribution of nitrile and pyridine groups is also expected to have some relevance, although those groups are less frequent in ACs. Table 2. Henry constants (bar) of MCM, DCM and TCM in the AC models showed in Figure 1 predicted by COSMO-RS at 298K. Functionalization MCM DCM TCM (1) AC none 3.93 0.46 0.21 (2) AC-COOCO carboxylic anhydride 3.52 0.30 0.15 (3) AC-OH hydroxyl 3.57 0.45 0.24 (4) AC-CCC carboxylic acid 4.53 0.47 0.22 (5) AC-CO carbonyl 3.79 0.20 0.08 (6) AC-CN nitrile 3.84 0.34 0.16 (7) AC-CONH2 amide 4.92 0.43 0.20 (8) AC-NH2 amine 3.64 0.43 0.25 (9) AC-Pyr pyridine 3.82 0.23 0.07 AC model Figure 1 19 In order to check experimentally the influence of the surface chemistry of AC on Cl- VOCs adsorption, the starting commercial AC (AC-MkU) was subjected to thermal treatment in N2 atmosphere at 900ºC(AC-Mk900) and to oxidative treatment with nitric acid (AC-MkN) and ammonium persulfate (AC-MkS). Figure 8 shows the TPD profiles of these activated carbons. Both oxidized carbons, AC-MkN and AC-MkS, present a substantially increased amount of surface oxygen groups (SOG) with respect to the starting activated carbon, evidenced by the significantly higher amounts of CO and CO2 evolved upon TPD, whereas the thermal treatment in inert atmosphere provokes a loose of SOG. Particularly significant was the oxidation with nitric acid, where the amount of oxygen surface groups increased by threefold (Table 3). It is also remarkable that the oxygen groups with hydrogen bond acceptor character (carbonyls, anhydrides) are majority (>75 %) in the adsorbent surface of all ACs (Table 3). A small although monotonically decrease of the BET surface area can be observed as the amount of SOG increases. Fig.8- CO2 (A) and CO (B) TPD profiles for the starting and modified AC. 0.00 0.20 0.40 0.60 0.80 1.00 0 200 400 600 800 1000 1200 C O e vo lv ed (µ m ol ·g -1 ) T (ºC) 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0 200 400 600 800 1000 1200 C O 2 ev ol ve d ( µ m ol ·g -1 ) T (ºC) MkU Mk900 MkS MkN (A) (B) 20 Table 3. Textural characteristic and assessment of oxygen surface groups from deconvolution of TPD profiles for the starting and modified ACs. Figure 9 compares the results of column tests of TCM, DCM and MCM adsorption with the starting and modified ACs. Table 4 summarizes the operating conditions and the values obtained for qs, k and HMTZ. As can be seen, the adsorption capacities significantly increased with the number of chlorine atoms of the Cl-VOC molecule, in agreement with the COSMO-RS predictions. With regard to the effect of the surface chemistry of the ACs, it can be seen that the adsorption capacity increases in the order AC-MkN > AC-MkS > AC-MkU > AC-Mk900, i.e. with the amount of SOG of the AC, despite the small decrease of surface area. Therefore, the experimental evidences confirm the COSMO-RS predictions respect to the positive affect of hydrogen-bond acceptor groups on the adsorption of chloromethanes. Fig.9- Experimental (dots) and predicted (lines) breakthrough curves of Cl-VOCs adsorption with the starting and modified ACs (250 mg AC-MkU, 35ºC, 100 N mL·min-1 total gas flow and 1000 ppmv of inlet Cl-VOC concentration) ABET Vmicrop. Vmesop. (m2·g-1) Carboxylic acids Anhydrides Lactones Pirones Total Anhydrides Phenols Carbonyl Chromenes Total AC-MkU 927 0.36 0.22 63.1 19.9 39.1 60.7 182.8 50.4 73.9 315.1 51.0 490.4 AC-Mk900 906 0.36 0.23 37.9 27.5 18.6 27.7 111.7 47.5 57.0 69.2 148.5 322.2 AC-MkS 872 0.34 0.26 198.5 16.3 53.5 83.4 351.7 196.2 158.6 532.0 165.7 1052.5 AC-MkN 856 0.33 0.24 209.3 182.8 53.5 128.1 573.8 437.5 351.8 874.9 214.6 1878.8 Groups evolved as CO2 (µmol·g-1) Groups evolved as CO (µmol·g-1) V (cm3·g-1) 0 20 40 60 80 100 120 140 Time (min) 0 10 20 30 40 50 Time (min) 0.0 0.2 0.4 0.6 0.8 1.0 0 20 40 60 80 100 120 140 160 C /C 0 Time (min) Mk900 MkU MkS MkN (B) DCM (C) MCM (A) TCM MkU Mk900 MkS MkN 21 Table 4. Results of the column tests of Cl-VOCs adsorption with the starting and modified ACs. 4. Conclusions Efficient and fast adsorption of dichloromethane from gas stream onto commercial activated carbon was shown by fixed-bed experiments at a wide range of operating conditions (total flow rate, amount of adsorbent, inlet concentration, temperature and pressure). The effects of these operating conditions on the equilibrium capacity and the rate of adsorption have been analyzed. Desorption experiments proved that complete regeneration of the exhausted adsorbent can be successfully achieved under mild conditions (atmospheric pressure and room temperature). The quantum-chemical COSMO-RS method allowed predicting what the relative order of adsorption of chloromethanes (TCM > DCM > MCM) and the favorable effect of introducing surface oxygen groups on the activated carbon. Those theoretical findings were experimentally confirmed indicating that the COSMO-RS method provides an interesting tool as a Solute Adsorbent Q m C0 T P qs HZTM k R2 (N mL·min-1) (g) (ppmv) (ºC) (atm) (mg·g-1) (cm) (h-1) TCM AC-Mk900 100 0.25 846 35 1.5 197.3 1.1 18.0 0.989 TCM AC-MkU 100 0.25 854 35 1.5 203.1 0.7 27.5 0.979 TCM AC-MkS 100 0.25 849 35 1.5 205.0 0.8 25.0 0.993 TCM AC-MkN 100 0.25 856 35 1.5 213.4 0.9 21.0 0.999 DCM AC-Mk900 100 0.25 971 35 1.5 103.0 1.4 21.2 0.998 DCM AC-MkU 100 0.25 1015 35 1.5 111.1 1.4 21.8 0.995 DCM AC-MkS 100 0.25 971 35 1.5 116.4 1.1 23.2 0.956 DCM AC-MkN 100 0.25 995 35 1.5 123.9 1.0 22.6 0.987 MCM AC-Mk900 100 0.25 1068 35 1.5 8.4 5.1 34.0 0.986 MCM AC-MkU 100 0.25 1135 35 1.5 14.1 5.1 21.9 0.935 MCM AC-MkS 100 0.25 1046 35 1.5 16.2 5.3 16.2 0.983 MCM AC-MkN 100 0.25 1192 35 1.5 22.2 3.3 25.6 0.995 Operating conditions 22 guide for activated carbon tunning addressed to the adsorption of specific compounds from gas stream. Acknowledgements The authors are grateful to the Spanish “Ministerio de Ciencia e Innovación (MICINN)” and “Comunidad de Madrid” for financial support (projects CTQ2011- 26758, CTQ2009-09983 and S2009/PPQ-1545). We are very grateful to “Centro de Computación Científica de la Universidad Autónoma de Madrid” for computational facilities. References [1] de Pedro ZM, Gomez-Sainero LM, Gonzalez-Serrano E, Rodriguez JJ. Gas-phase hydrodechlorination of dichloromethane at low concentrations with palladium/carbon catalysts. Ind Eng Chem Res 2006 NOV 8;45(23):7760-6. [2] Iranpour R, Coxa H, Deshusses M, Schroeder E. Literature review of air pollution control biofilters and biotrickling filters for odor and volatile organic compound removal. Environ.Prog. 2005 OCT;24(3):254-67. [3] Dobrzynska E, Posniak M, Szewczynska M, Buszewski B. Chlorinated Volatile Organic Compounds-Old, However, Actual Analytical and Toxicological Problem RID A-3187-2009. Crit.Rev.Anal.Chem. 2010;40(1):41-57. [4] W.J. Hayes ERL. Handbook of Pesticide Toxicology. San Diego: Academic Press; 1991. [5] Verhulst D, Buekens A, Spencer P, Eriksson G. Thermodynamic behavior of metal chlorides and sulfates under the conditions of incineration furnaces. Environ.Sci.Technol. 1996 JAN;30(1):50-6. [6] Buekens A, Huang H. Comparative evaluation of techniques for controlling the formation and emission of chlorinated dioxins/furans in municipal waste incineration. J.Hazard.Mater. 1998 SEP 11;62(1):1-33. [7] Ballikaya M, Atalay S, Alpay H, Atalay F. Catalytic combustion of methylenechloride. Combustion Sci.Technol. 1996;120(1-6):169-84. [8] Atalay S, Alpay H, Atalay F. Catalyst preparation and testing for catalytic combustion of chloromethanes. ; 2002. 23 [9] Li C, Lee W, Chen C, Wang Y. CH2Cl2 decomposition by using a radio-frequency plasma system. Journal of Chemical Technology and Biotechnology 1996 AUG;66(4):382-8. [10] Everaert K, Baeyens J. Catalytic combustion of volatile organic compounds. J.Hazard.Mater. 2004 JUN 18;109(1-3):113-39. [11] Wang J, Chen J. Removal of dichloromethane from waste gases with a bio-contact oxidation reactor. Chem.Eng.J. 2006 OCT 15;123(3):103-7. [12] Gomez-Sainero L, Seoane X, Arcoya A. Hydrodechlorination of carbon tetrachloride in the liquid phase on a Pd/carbon catalyst: kinetic and mechanistic studies. Applied Catalysis B-Environmental 2004 OCT 22;53(2):101-10. [13] Gomez-Sainero L, Seoane X, Fierro J, Arcoya A. Liquid-phase hydrodechlorination of CCI4 to CHCl3 on Pd/carbon catalysts: Nature and role of Pd active species. Journal of Catalysis 2002 JUL 25;209(2):279-88. [14] Gomez-Sainero L, Seoane X, Tijero E, Arcoya A. Hydrodechlorination of carbon tetrachloride to chloroform in the liquid phase with a Pd/carbon catalyst. Study of the mass transfer steps. Chemical Engineering Science 2002 SEP;57(17):3565-74. [15] Alvarez-Montero MA, Gomez-Sainero LM, Juan-Juan J, Linares-Solano A, Rodriguez JJ. Gas-phase hydrodechlorination of dichloromethane with activated carbon-supported metallic catalysts. Chem.Eng.J. 2010 AUG 15;162(2):599-608. [16] Alvarez-Montero MA, Gomez-Sainero LM, Martin-Martinez M, Heras F, Rodriguez JJ. Hydrodechlorination of chloromethanes with Pd on activated carbon catalysts for the treatment of residual gas streams. Applied Catalysis B-Environmental 2010 APR 26;96(1-2):148-56. [17] Alvarez-Montero MA, Gomez-Sainero LM, Mayoral A, Diaz I, Baker RT, Rodriguez JJ. Hydrodechlorination of chloromethanes with a highly stable Pt on activated carbon catalyst. Journal of Catalysis 2011 APR 25;279(2):389-96. [18] Zuber L, Dunn I, Deshusses M. Comparative scale-up and cost estimation of a biological trickling filter and a three-phase airlift bioreactor for the removal of methylene chloride from polluted air. J.Air Waste Manage.Assoc. 1997 SEP;47(9):969- 75. [19] Cox H, Deshusses M, Converse B, Schroeder E, Iranpour R. Odor and volatile organic compound treatment by biotrickling filters: Pilot-scale studies at hyperion treatment plant. Water Environ.Res. 2002 NOV-DEC;74(6):557-63. [20] Yu Jian-ming, Chen Jian-meng, Wang Jia-de. Removal of dichloromethane from waste gases by a biotrickling filter. Journal of Environmental Sciences-China 2006;18(6):1073-6. [21] Anfruns A, Martin MJ, Montes-Moran MA. Removal of odourous VOCs using sludge-based adsorbents. Chem.Eng.J. 2011 FEB 1;166(3):1022-31. 24 [22] Gallego E, Roca FJ, Perales JF, Guardino X. Comparative study of the adsorption performance of a multi-sorbent bed (Carbotrap, Carbopack X, Carboxen 569) and a Tenax TA adsorbent tube for the analysis of volatile organic compounds (VOCs). Talanta 2010 MAY 15;81(3):916-24. [23] Ren X, Chen C, Nagatsu M, Wang X. Carbon nanotubes as adsorbents in environmental pollution management: A review. Chem.Eng.J. 2011 JUN 1;170(2- 3):395-410. [24] Ghoshal A, Manjare S. Selection of appropriate adsorption technique for recovery of VOCs: an analysis. J Loss Prev Process Ind 2002 NOV;15(6):413-21. [25] Navarri P, Marchal D, Ginestet A. Activated carbon fibre materials for VOC removal. Filtration Sep. 2001 JAN-FEB;38(1):34-40. [26] Zhang X, Zhao X, Hu J, Wei C, Bi HT. Adsorption dynamics of trichlorofluoromethane in activated carbon fiber beds. J.Hazard.Mater. 2011 FEB 28;186(2-3):1816-22. [27] Long C, Liu P, Li Y, Li A, Zhang Q. Characterization of Hydrophobic Hypercrosslinked Polymer as an Adsorbent for Removal of Chlorinated Volatile Organic Compounds. Environ.Sci.Technol. 2011 MAY 15;45(10):4506-12. [28] Tsai W, Chang C, Ho C, Chen L. Adsorption properties and breakthrough model of 1,1-dichforo-1-fluoroethane on granular activated carbon and activated carbon fiber. Sep.Sci.Technol. 2000;35(10):1635-50. [29] Xiang Z, Lu Y, Gong X, Luo G. Absorption and desorption of gaseous toluene by an absorbent microcapsules column. J.Hazard.Mater. 2010 JAN 15;173(1-3):243-8. [30] Ahmaruzzaman M. Adsorption of phenolic compounds on low-cost adsorbents: A review. Adv.Colloid Interface Sci. 2008 NOV 4;143(1-2):48-67. [31] Khan F, Ghoshal A. Removal of Volatile Organic Compounds from polluted air. J Loss Prev Process Ind 2000 NOV;13(6):527-45. [32] Yoon YH, Nelson JH. Application of Gas-Adsorption Kinetics .1. a Theoretical- Model for Respirator Cartridge Service Life. Am.Ind.Hyg.Assoc.J. 1984 1984;45(8):509-16. [33] Aksu Z, Gonen F. Biosorption of phenol by immobilized activated sludge in a continuous packed bed: prediction of breakthrough curves. Process Biochemistry 2004 JAN 30;39(5):599-613. [34] Klamt A, Eckert F. COSMO-RS: a novel and efficient method for the a priori prediction of thermophysical data of liquids (vol 172, pg 43, 2000). Fluid Phase Equilib. 2003;205(2):357-. [35] Palomar J, Lemus J, Gilarranz MA, Rodriguez JJ. Adsorption of ionic liquids from aqueous effluents by activated carbon. Carbon 2009 6;47(7):1846-56. 25 [36] Mehler C, Klamt A, Peukert W. Use of COSMO-RS for the prediction of adsorption equilibria. AIChE journal 2002;48:1093. [37] Palomar J, Gonzalez-Miquel M, Bedia J, Rodriguez F, Rodriguez JJ. Task-specific ionic liquids for efficient ammonia absorption. Separation and Purification Technology 2011 10/27;82(0):43-52. [38] Palomar J, Gonzalez Miquel M, Polo A, Rodriguez F. Understanding the Physical Absorption of CO2 in Ionic Liquids Using the COSMO-RS Method. Industrial engineering chemistry research 2011;50(6):3452-63. [39] Gonzalez-Miquel M, Palomar J, Omar S, Rodriguez F. CO(2)/N(2) Selectivity Prediction in Supported Ionic Liquid Membranes (SILMs) by COSMO-RS. Ind Eng Chem Res 2011 MAY 4;50(9):5739-48. [40] PRADOBURGUETE C, LINARESSOLANO A, RODRIGUEZREINOSO F, DELECEA C. The Effect of Oxygen-Surface Groups of the Support on Platinum Dispersion in Pt/carbon Catalysts. Journal of Catalysis 1989 JAN;115(1):98-106. [41] Moreno-Castilla C, Carrasco-Marin F, Mueden A. The creation of acid carbon surfaces by treatment with (NH4)(2)S2O8. Carbon 1997;35(10-11):1619-26. [42] Sing K. Adsorption methods for the characterization of porous materials. Advances in colloid and interface science 1998;76:3-11. [43] Rey A, Faraldos M, Casas JA, Zazo JA, Bahamonde A, Rodriguez JJ. Catalytic wet peroxide oxidation of phenol over Fe/AC catalysts: Influence of iron precursor and activated carbon surface. Applied Catalysis B-Environmental 2009 FEB 2;86(1-2):69- 77. [44] Frisch MJ, Trucks GW, Schlegel HB, Scuseria GE, Robb MA, Cheeseman JR, et al. Gaussian03, revision B.05; 2004;. [45] COSMOtherm C2.1, release 01.06. Leverkusen, Germany: GmbH & Co. KG; 2003. [46] Yoon YH, Nelson JH. Application of Gas-Adsorption Kinetics .2. a Theoretical- Model for Respirator Cartridge Service Life and its Practical Applications. Am.Ind.Hyg.Assoc.J. 1984 1984;45(8):517-24. [47] Lodewyckx P, Wood G, Ryu S. The Wheeler-Jonas equation: a versatile tool for the prediction of carbon bed breakthrough times. Carbon 2004;42(7):1351-5. [48] Burkert CAV, Barbosa GNO, Mazutti MA, Maugeri F. Mathematical modeling and experimental breakthrough curves of cephalosporin C adsorption in a fixed-bed column. Process Biochemistry 2011 JUN;46(6):1270-7. [49] Marsh H, Rodriguez-Reinoso F. Activated carbon. Great Britain: Elsevier; 2006. Abstract Acknowledgements