Enhanced CO2 Adsorption and Separation in Ionic Liquid

Subscriber access provided by UNIVERSITY OF TOLEDO LIBRARIES

C: Surfaces, Interfaces, Porous Materials, and Catalysis 2

Enhanced CO Adsorption and Separation in Ionic Liquid-Impregnated Mesoporous Silica MCM-41: A Molecular Simulation Study Kishant Kumar, and Amit Kumar J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.7b11529 • Publication Date (Web): 26 Mar 2018 Downloaded from http://pubs.acs.org on March 26, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Enhanced CO2 Adsorption and Separation in Ionic Liquid-Impregnated Mesoporous Silica MCM-41: A Molecular Simulation Study

Kishant Kumar, Amit Kumar* Department of Chemical Engineering, Indian Institute of Technology Guwahati Guwahati, Assam 781039, India.

*Corresponding author, Email: [email protected]

1 ACS Paragon Plus Environment

The Journal of Physical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Abstract: Separation of CO2 from gas mixtures is of importance in CO2 capture from flue gas and in natural gas sweetening. In this paper, we have conducted Grand canonical Monte Carlo (GCMC) simulations to study the adsorption of CO2, N2 and CH4, and separation of their binary mixtures in mesoporous silica MCM-41 modified by incorporation of the pyrrolidinium-based ionic liquid 1-methyl-1-butyl-pyrrolidinium (C4PYR+) bis(trifluoromethanesulfonyl)imide (TF2N–) at two different loadings: ~21% and 42% by weight, hereinafter referred to as MCM-41-IL1 and MCM41-IL2 respectively. Although MCM-41-IL1 showed significantly higher adsorption of pure CO2 than pristine MCM-41, the amounts of pure N2 and CH4 adsorbed on MCM-41-IL1 were only slightly higher. Molecular dynamics simulation of pure CO2 in IL-loaded MCM-41 models revealed that CO2 molecules prefer locations near the pore walls as well as in the pore interior around IL molecules. GCMC simulations of gas mixture adsorption (CO2/N2 and CO2/CH4) showed that CO2 adsorption is highest in MCM-41-IL1 and least in pure MCM-41. The CO2/N2 and CO2/CH4 selectivities at 298.15 K and 1 bar followed the trend MCM-41-IL2 > MCM-41-IL1 > MCM-41 with values for MCM-41IL2 almost twice that for pure MCM-41. Thus, modifying the mesopores of MCM-41 with IL can result in significant enhancement in CO2 adsorption and selectivity.


Page 2 of 36

Page 3 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Introduction In recent years, there has been considerable interest in porous materials modified through functionalization and physical impregnation for the efficient capture and separation of CO2. Separation of CO2 from gas mixtures is of interest in several operations, most important of which may be CO2 capture from flue gas, as the release of large amounts of CO2 into the atmosphere has been identified as contributing to global warming through greenhouse effect. Some of the industrially relevant gas separation operations involving CO2 are: a) CO2 capture from flue gas containing CO2, N2, CO, NOx, SO2, O2 etc., b) CO2 separation from blast furnace gas containing CO2, CO, CxHy, H2S, N2, O2, and NOx, c) production and separation of CO2 from landfill gas containing CO2 and CH4, d) separation of CO2/CH4 for natural gas upgradation and high purity methane1–6. Many studies on metal-organic frameworks (MOFs), consisting of microporous channels and cages, have been conducted as they are considered to be promising materials for efficient gas storage and separation applications owing to their large surface areas and pore volumes as well as tunable pore size7. Functionalized MOFs are also being actively explored to enhance the separation and capture of different gases of interest8–12. However, covalently grafted long chain functional groups inside MOFs require large pore diameters and cages, which are difficult to synthesize because of their low thermal stability13–15. On the other hand, mesoporous

materials

are

well

suited

for

pore-modification

through

covalent

functionalization or physical impregnation of larger groups/molecules because of their large pore diameter, pore volume and surface area16–21. MCM-41 belongs to a family of mesoporous silica materials and has been extensively studied for applications in drug delivery22, separation and capture of gases16, and catalysis23. Pore size of these materials can be tuned during their synthesis by using appropriate surfactant and increasing the pH level16. Regular hexagonal arrangement of one-dimensional cylindrical 3 ACS Paragon Plus Environment


pores in MCM-41 makes this material suitable for understanding gas adsorption at the molecular level. Although numerous studies involving metal doping, drug loading, capturing small organic volatile compounds, and surface functionalization (e.g. for enhanced CO2 selectivity towards other gases) have been conducted experimentally on MCM-4116,22–30, very few reports on incorporation of ionic liquids (IL) inside the mesopores of MCM-41 are available in literature31–35. In contrast, impregnation of IL molecules inside MOFs have been reported by several researchers and found to be effective in CO2 capture and separation36–39. Computational exploration of ionic liquid-assisted IRMOF-1 at different IL loadings was carried out by Chen et al.36 which revealed the composite adsorbent to be an efficient system for capturing CO2 over N2. Moreover, MOFs with IL molecules confined in their pores were found to be more promising in H2S removal from natural gas (mainly CH4) in the computational study by Li et al.37 The gas separation performance of ZIF-8 was significantly modified by Koyuturk et al.38 by incorporation on the ionic liquid [BMIM][BF4] with CO2/CH4 selectivity becoming almost twice of its value on pristine ZIF-8 at low pressures. Ban et al.39 have reported remarkable permeability and selectivity in mixed matrix membranes derived from ionic liquid-modified ZIF-8 for CO2/N2 and CO2/CH4 separation. Computational and experimental studies of CO2 + IL systems have also been carried out in recent years. Daly et al.40 conducted experimental as well as computer simulation investigation of CO2 vibrational frequencies in an IL environment. IL molecules can be incorporated into the pores of mesoporous materials by either chemical grafting or physical impregnation. Vangeli et al.31 prepared immobilized imidazolium-based ionic liquid inside the pores of MCM-41 and Vycor silica, and successfully grafted the cationic part of the IL in order to further investigate the physicochemical and thermodynamic properties of the grafted material. In addition, they found the selectivity of CO2 over CO to be higher for IL-grafted porous materials. Stefanopoulos et al.32 conducted small-angle neutron 4 ACS Paragon Plus Environment

Page 4 of 36

Page 5 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


scattering (SANS), contrast-matching SANS, and nitrogen adsorption on the IL [BMIM][PF6] confined inside the pores of MCM-41 and SBA-15. Nitrogen adsorption at 77K in both the IL-loaded materials was lowered significantly compared to the respective pure MCM-41 and SBA-15. Tripathi et al.34 synthesized MCM-41 with the IL 1-ethyl-3methyl imidazolium tetrafluoroborate [EMIM][BF4] confined in the mesopores, and studied the thermal, electrical and structural properties of synthesized material. They confirmed the interaction of oxygen atoms on pore wall of MCM-41 with the cationic part of IL from Fourier-transform infrared spectroscopy vibrational bands. Thus, it is evident that there is ample scope to investigate the capabilities of IL-loaded mesoporous materials such as MCM-41 for adsorption and separation of gases. Molecular simulation techniques can be employed to determine the extent of enhancement in CO2 adsorption and separation, and computationally identify the optimum IL loadings in the mesopore, to screen the more promising combinations for subsequent experimental investigation. At present, imidazolium- and pyrrolidinium-based ionic liquids are the most widely investigated class of ILs for CO2 solubility. Lei et al.41 provide a detailed review of gas solubility

in

various

ionic

liquids.

Ionic

liquids

containing

the

anion

bis(trifluoromethanesulfonyl)imide ([TF2N–]) in particular have been reported to exhibit high CO2 solubilities42,43. Therefore, in the present work, computational exploration of different loadings

of

the

pyrrolidinium-based

IL

1-methyl-1-butyl-pyrrolidinium

bis(trifluoromethanesulfonyl)imide ([C4PYR]+[TF2N]–) confined in the pores of MCM-41 was carried out in order to study the capture and separation of CO2 from N2 and CH4. Pure component adsorption study for CO2, N2 and CH4 was carried out at three different temperatures on pristine and IL-loaded MCM-41. Three-dimensional spatial distribution maps and radial distribution functions (RDFs) of CO2 were calculated to identify the locations preferred by CO2 inside the adsorbent pores. Further, grand canonical Monte Carlo 5 ACS Paragon Plus Environment


Page 6 of 36

(GCMC) simulations of CO2/N2 and CO2/CH4 binary mixtures were performed to establish the enhancement of CO2 selectivities in IL-loaded MCM-41. 2 Model and Simulation Details 2.1 Structures of MCM-41 and IL Geometry

of

IL

pair

of

methyl-1-butyl-pyrrolidinium

[C4PYR]+

and

bis(trifluoromethanesulfonyl)imide [TF2N]– was optimized using density functional theory (DFT) at B3LYP/6-311++G** level of theory using GAUSSIAN 09 software48. The structure of the cation and anion of the IL is shown in Fig. 1. Molecular model of MCM-41 was constructed by carving cylindrical pores in an amorphous silica block, ensuring that the pore arrangement is hexagonal and keeping the pore diameter similar to that observed in realistic MCM-41 samples (described in detail by Kumar et al.44). The dangling bonds of silicon and oxygen atoms on the surface of the pores were saturated with hydroxyl groups and hydrogen atoms respectively. The constructed model was further subjected to geometry optimization using Forcite module of Materials Studio 7.0 software package45. The MCM-41 model thus obtained has unit cell dimensions of 42.8 Å × 42.8 Å × 64.2 Å and an average pore diameter of 32.8 Å, calculated from the average distance between the center of the pore and all the surface hydrogen atoms. The simulation box containing one central principal pore and four quarter-pores at the four corners of the box is shown in Fig. 2(a). The accessible surface area, computed using nitrogen probe molecule of radius 1.82 Å, was found to be 1225.71 m2/g which is in good agreement with the experimental values of BET surface area in the range 1090–1150 m2/g for MCM-4146,47. Two separate IL-impregnated MCM-41 models having different loadings of the IL were then prepared by inserting 16 and 32 cation-anion pairs of the IL respectively into the central pore of the MCM-41 model using the Packmol software package49. An equal number of IL molecules was also inserted in the four quarter-pores of the


Page 7 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


MCM-41 models using Packmol so that the total number of IL molecules (cation-anion pairs) present in the simulation box in the two models is 32 and 64 respectively. The MCM-41/IL system was equilibrated by conducting molecular dynamics (MD) simulations in the NVT ensemble. The equilibrated IL-impregnated MCM-41 models, referred to as MCM-41-IL1 and MCM-41-IL2 henceforth and shown in Fig. 2(b) and 2(c) respectively, were used for adsorption simulations using GCMC technique.

2.2 Force Fields Adsorbates In this work, the non-bonded van der Waals interactions were modeled using the 12-6 Lennard-Jones (LJ) potential for all the adsorbates. N2 was modeled as having two LJ sites at the two nitrogen atoms with N–N bond length of 1.1 Å and three partial point charges, one each on the two nitrogen atoms and a third at the center of mass of the nitrogen molecule, in order to effectively capture the effect of quadrupole moment. LJ parameters and partial charges for N2 molecule were taken from transferrable potentials for phase equilibria (TraPPE) force field50. Three LJ sites were used to represent the rigid CO2 linear molecule with partial charges on the respective atoms. The C=O bond length (of 1.16 Å), the LJ parameters and the partial charges on the CO2 molecule were taken from the EPM2 model51. A united atom approach was used to model CH4 where only one LJ site was considered. TraPPE united atom (TraPPE -UA) force field, proposed by Martin and Siepmann52, was used to model the CH4 molecule. Adsorbents



Several studies on ILs have used optimized potentials for liquid simulations (OPLS) force field53–56 and satisfactorily reproduced experimentally determined properties of ILs. In the present work, we have used OPLS force field for the studied IL. MCM-41 was modeled as a rigid, frozen framework and force field parameters were taken from the work of Jing et al.57 During MD equilibration of the IL-loaded MCM-41 structures, the IL cations and anions were modeled as flexible ions moving inside the rigid, frozen MCM-41 framework. However, during GCMC simulations of gas adsorption in the MCM-41-IL structures, both the MCM-41 framework and the IL molecules were maintained rigid and fixed. Only nonbonded interactions were considered between the framework atoms, IL ions and the adsorbed gas molecules. To ensure that considering the IL molecules immobile during the GCMC simulations did not have a significant effect on the adsorption behavior, we calculated the accessible surface areas and pore volumes of five independent configurations of the MCM-41-IL1 system, taken from the last 10 ns of the MD equilibration, spaced 2 ns apart. The properties listed in Table S1 in the supporting information vary by less than 5% between the different MCM-41-IL1 configurations indicating that variation in arrangement of IL molecules in the mesopore does not have a significant effect on the accessible surface area and pore volume. As these properties are important characteristics of an adsorbent, it is expected that different the MCM-41-IL1 configurations will show similar adsorption behavior. To further ensure that considering the ionic liquids fixed does not result in substantial errors, pure CO2 adsorption simulations were carried out on five snapshots each of MCM-41-IL1 and MCM-41-IL2 systems (snapshots were taken from molecular dynamics simulations of respective MCM-41+IL systems in which ILs were mobile and flexible); the resulting adsorption isotherms are shown in Figure S1. The CO2 adsorption isotherms for the different snapshots are very close to each other for both MCM-41-IL1 and MCM-41-IL2, suggesting that neglecting the mobility (and flexibility) of ILs does not have a strong effect


Page 8 of 36

Page 9 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


on adsorption behavior. Thus, considering only a single configuration of MCM-41-IL1 and assuming IL molecules to be immobile should not result in significant errors. The detailed LJ parameters and partial charges for the adsorbate gases as well as the adsorbent material (MCM-41 and IL) are listed in Table S2 in the supporting information.

2.3 Simulation Details MD equilibration of the ionic liquid confined inside the pores of MCM-41 was carried out using the large-scale atomic/molecular massively parallel simulator (LAMMPS) software package58. MD simulation of CO2 inside IL-impregnated MCM-41 was also carried out using LAMMPS to determine the spatial distribution of CO2 and RDFs between different atom types. The SHAKE algorithm59 was used to constrain the C=O bond length. The particleparticle particle-mesh (PPPM) solver implemented in LAMMPS was used to calculate the non-bonded Coulombic potential. A cutoff of 12.5 Å was used for both the LJ potential and the real part of the Coulombic interactions. The MCM-41/IL system was equilibrated for 40 ns using MD simulation in the constant NVT ensemble with a time step of 1 fs. The final configuration of the IL confined in MCM-41 obtained at the end of the MD equilibration was used to conduct the GCMC simulation of gas adsorption. All the GCMC simulations were carried out using RASPA software package developed by Dubbeldam et al.60 This package has been extensively tested for a large number of systems and validated against experimental data61–63. Insertion, deletion, translation and rotation moves were employed in all GCMC simulations. In the case of mixture adsorption simulations, identity exchange trial moves were also used. For pure gas adsorption, the four types of moves are assigned a probability of 0.25 each. For mixture adsorption, identity exchange move is assigned a probability of 0.05 by reducing the probabilities of translation 9 ACS Paragon Plus Environment


and rotation moves equally and maintaining the probabilities of insertion and deletion moves unchanged at 0.25. During each MC cycle, one MC trial move (of any type) is attempted on all the N adsorbate molecules in the simulation box if the number of adsorbed molecules, ≥ 20. However, if < 20, the number of trial moves attempted in each cycle is equal to 20. A total of 2.5 × 10 MC cycles were carried out for equilibration followed by another 2.5 × 10 MC cycles for production of adsorption data. The results obtained from the GCMC simulation are the absolute amounts adsorbed which were converted to excess adsorbed amounts using equation (1), = − × × 1 where is the excess number of gas molecules adsorbed, is the absolute number of gas molecules adsorbed, is the number density of the bulk gas phase, is the void fraction of the simulation box and is the total volume of the simulation box. Dual-site Langmuir (DSL) model was used to fit the isotherm data for pure CO2 and CH4 gas while Langmuir model was used to fit pure N2 isotherm. Adsorption isotherm modelling was done at three different temperatures for pure gases. The equations describing the Langmuir and DSL models are provided in the supporting information (Eq. S1 and S2 for Langmuir model and Eq. S3 and S4 for DSL model). Further, Eq. S5 is used to calculate Henry’s constant from the parameters of the DSL model and Eq. S6 is employed to calculate the isosteric heats of adsorption from GCMC simulations. The ideal adsorbed solution theory (IAST) proposed by the Myers and Prausnitz64 was used to predict the mixture adsorption isotherms using the pure component adsorption data. Pure gas isotherm model parameters, obtained by fitting Langmuir or dual-site Langmuir models to simulation data, were provided as input to the IAST calculation and the set of equations


Page 10 of 36

Page 11 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


S7–S10 provided in the supporting information were solved simultaneously to obtain mixture adsorption isotherm predictions.

3. Results and Discussion 3.1 Adsorption of Pure Gases: CO2, N2 and CH4 GCMC simulations were carried out to study the effect of impregnation of IL into the mesopores of MCM-41 on the adsorption of CO2, N2 and CH4 as well as on the separation of CO2 from gas mixtures. Fig. 3 shows a comparison between the simulated adsorption isotherm of pure CO2 in MCM-41 at 303.15 K with experimental data of dos Santos et al.65 The two isotherms compare well indicating that the models and interaction parameters employed are appropriate for simulating CO2 adsorption on MCM-41. The surface areas and accessible pore volume fractions of the three MCM-41 models, calculated using the Materials Studio software with two different probe sizes, are listed in Table 1. The accessible pore volume fraction decreases significantly with increase in the amount of IL incorporated, as expected. However, the accessible surface area increases at the lower IL loading in MCM-41 and decreases as the loading is increased. The adsorption isotherms of pure CO2, N2, and CH4 in MCM-41, MCM-41-IL1 and MCM41-IL2 along with respective DSL model fits are shown in Fig. 4. The saturation loading is lowest in MCM-41-IL2 for all three gases due to the lower accessible surface area and pore volume fraction. The accessible surface area includes the surface of MCM-41 pore walls as well as the surface around the IL molecules that is accessible to the adsorbent molecules. In contrast, MCM-41-IL1 showed the highest loadings for all the three adsorbates, viz. CO2, N2 and CH4, over the pressure range studied (0–14 bar) possibly due to the larger accessible surface area. Interestingly, CO2 loading was significantly higher in MCM-41-IL1 than in the 11 ACS Paragon Plus Environment


Page 12 of 36

other two models whereas CH4 and N2 loadings were only slightly higher in MCM-41-IL1 than in pure MCM-41. This can be attributed to the stronger electrostatic interactions between the ionic liquid molecules and CO2 due to its larger quadrupole moment. On the other hand, less polar N2 and nonpolar CH4 molecules do not interact as strongly with the ionic liquid molecules and hence, their loadings are only marginally higher in MCM-41-IL1 (and lower in MCM-41-IL2) than in pure MCM-41. A notable feature of Fig. 4(a) is that up to pressures of ~5 bar, the amount of CO2 adsorbed in MCM-41-IL2 is higher than that in pure MCM-41. Although the accessible surface area and void fraction of MCM-41-IL2 are lower than that of pure MCM-41, the strong CO2-ionic liquid interactions are responsible for higher loading at low pressures. At higher pressures (> 5 bar), when the number of adsorbed CO2 molecules becomes high, the smaller accessible surface area and pore volume of MCM-41-IL2 becomes the limiting factor and despite strong CO2-ionic liquid interactions, the loading in MCM-41IL2 drops below that in pure MCM-41. The CO2 loadings at 1 bar and 298.15 K in pure MCM-41. MCM-41-IL1 and MCM-41-IL2 are 0.9, 1.5 and 1.2 mmol/g. In comparison, CO2 capacity in pore-expanded MCM-41 (30 nm pore diameter) grafted with monoamine silane at 1 bar and 303.15 K is 1.2 mmol/g16. However,

Rao

et

al.66

have

reported

that

MCM-41

grafted

with

50%

3-

aminopropyltriethoxysilane and 50 wt% polyethylenimine show higher CO2 capacities of 2.41 mmol/g and 3.53 mmol/g respectively at 1 atm and 298.15 K. Similarly, activated amorphous carbon shows high capacity for CO2 namely ~16 mmol/g at 8 bar, 299.15 K67 (compared to ~5 mmol/g at 8 bar, 298.15 K for MCM-41-IL1). The enhancement in CO2 adsorption capacity for IL-modified MOFs36,38,68 is usually less pronounced or absent compared to that observed for MCM-41-IL1 here, due to the smaller pores of MOFs. Further, the CO2 capacity of the IL-modified MCM-41 adsorbents in the present work is several times larger than the CO2 capacity of pure [C4PYR+][TF2N–] (as reported by Lee et al.43). 12 ACS Paragon Plus Environment

Page 13 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Fig. 4 also shows the DSL model fits to the adsorption isotherms of pure CO2, N2 and CH4. The respective models are observed to fit the simulated adsorption data well. Adsorption simulations were also carried out at two other temperatures and model fitting was carried out on the adsorption isotherms at the three different temperatures. Adsorption isotherms and fitted parameters at all three temperatures in the three MCM-41 models are shown in Fig. S2– S4 and Table S3–S5 of the supplementary information respectively. The Henry’s constants for CO2, CH4 and N2 in the three adsorbents, obtained by fitting the DSL model to adsorption data, were calculated using equation S5 (see the supporting information). Fig. S5 shows the Henry’s constants of the gases in the adsorbents at different temperatures. The Henry’s constants of CO2 are higher than that of the other gases due to stronger interactions of CO2 with the adsorbent. Further, Henry’s constants of CO2 are higher in the IL-loaded adsorbents, especially at lower temperatures, due to more favorable interactions between CO2 and IL molecules. The isosteric heats of adsorption were calculated for the gases from simulation, based on fluctuations in the grand canonical ensemble, using equation S6 (see supplementary information). Fig. 5 shows the isosteric heats of the adsorbate gases in the three adsorbents at different pressures. For CO2 (see Fig. 5(a)), isosteric heats are in general higher in IL-loaded MCM-41, due to stronger interaction with the adsorption sites present whereas for CH4 and N2 (Fig. 5(b) and (c) respectively), isosteric heats are similar for the different adsorbents. 3.2 Distribution of CO2 in the Frameworks The distribution of CO2 molecules inside pure and IL-loaded MCM-41 was further studied using 3D density distribution maps of CO2 and RDFs, obtained from MD trajectories of CO2 and IL molecules inside the MCM-41 pores. The number of CO2 molecules considered was 32 and 64 in the case of MCM-IL-1 and MCM-IL-2 respectively (half in the principal pore



and the rest equally divided in the four quarter pores). Fig. 6(a) and 6(b) show the 3D isosurface plots of the distribution of CO2 molecules in MCM-41-IL1 and MCM-41-IL2, obtained by considering the location of the CO2 molecules in the adsorbent framework over 1000 snapshots from the last 20 ns of the MD trajectory spaced 20 ps apart. The plots reveal that in addition to the surface of the pore walls of MCM-41, CO2 molecules also prefer sites around IL molecules away from the pore walls. The average distribution of CO2 molecules away from the pore walls is considerably more pronounced in the case of MCM-41-IL2 (see Fig. 6(b)) which contains larger loading of IL molecules within the MCM-41 mesopores. The RDF of CO2 molecules around the cations and anions of IL molecules provide more quantitative information about the spatial distribution of CO2. Fig. 7(a) and 7(b) show the RDFs of CO2 molecules (their carbon centers to be precise) with respect to the nitrogen atoms of the cations and anions of the IL molecules respectively (marked as N1 and N in Fig. 1). The RDFs show that there are almost no CO2 molecules up to a distance of ~4 Å from the nitrogen atom of the cation (N1 in Fig. 1) and up to a distance of ~3.5 Å from the nitrogen atom of the anion (N in Fig. 1), for both MCM-41-IL1 and MCM-41-IL2. The RDFs show a broad peak around ~6Å for both cationic and anionic nitrogen atoms, in both MCM-41-IL1 and MCM-41-IL2, indicating higher probability of finding CO2 molecules at this distance around the IL ions. Thus, CO2 prefers locations around both the cations and anions of IL molecules. The RDFs of the oxygen atoms of CO2 molecules around the hydrogen atoms of the pyrrolidinium group of the cations (marked as HA in Fig. 1) are plotted in Fig. 7(c). The RDFs for both MCM-41-IL1 and MCM-41-IL2 do not show any discernible peaks indicating the absence of any hydrogen bonding. Fig. 7(d) shows the RDFs of the nitrogen atom of anion (marked N in Fig. 1) around the nitrogen atom of the pyrrolidinium group of cations (marked N1 in Fig. 1). A strong peak can be seen around ~6 Å for both MCM-41-IL1 and


Page 14 of 36

Page 15 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


MCM-41-IL2 confirming that the cations and anions of the IL are closely associated and paired with each other inside the MCM-41 pores during the course of the MD simulation. 3.3 Adsorption of Binary Mixtures of CO2/N2 and CO2/CH4 GCMC simulations were carried out to study the adsorption and separation of binary mixtures of CO2/N2 and CO2/CH4, both in the molar ratio 15:85, at 298.15 K. Figure 8 shows the mixture adsorption isotherms of N2, CO2 and CH4 in the three adsorbent models. The amount of N2 adsorbed is similar for MCM-41 and MCM-41-IL1 models at pressures up to ~6 bar, as is evident from Fig. 8(a). Even at higher pressures, the amount of N2 adsorbed in MCM-41 is not significantly higher than that in MCM-41-IL1. However, in MCM-41-IL2, the amount of N2 adsorbed is considerably lower than that in the other two models, over the entire pressure range shown. The adsorption isotherms of CH4 in Fig. 8(b) show similar trend with the amounts adsorbed in MCM-41 and MCM-41-IL1 very similar over the entire pressure range and that in MCM-41-IL2 substantially lower than the other two models. Both N2 and CH4 are relatively nonpolar molecules, and hence do not show any noticeable preference for the charged ions of the IL over the surface atoms of MCM-41 pore wall. The major factor determining the adsorption behavior of these gases is thus primarily the available surface area and pore volume (important at high pressures) which are the lowest for MCM-41-IL2 leading to low adsorption. The adsorption behavior of CO2 on the three models is however much different. In both mixtures, CO2 adsorption is highest in MCM-41-IL1 followed by MCM-41-IL2 and pure MCM-41 in that order. The greater quadrupole moment of CO2, compared to N2 and CH4, results in stronger interactions with the IL ions leading to higher adsorption in both the IL-loaded MCM-41 models over the entire pressure range studied (0.1–14 bar), despite the larger accessible surface area of pure MCM-41 than MCM-41-IL2. This may appear to be somewhat in contrast with pure CO2 adsorption behavior (see Fig. 4(a)) where MCM-41 shows higher adsorption than MCM-41-IL2 above 5 15 ACS Paragon Plus Environment


bar due to the adsorption in MCM-41-IL2 approaching saturation. However, the amount of CO2 in the binary mixtures is only 15%, so that even at a total pressure of 14 bar, the fugacity of CO2 in the mixture would be significantly lower than that expected at saturation. CO2 adsorption in the models clearly does not approach saturation at total pressures of up to 14 bar and therefore, the favorable interactions of CO2 molecules with IL dominate over the higher accessible pore volume of pure MCM-41, leading to larger adsorption of CO2 in MCM-41-IL2. In the case of MCM-41-IL1, the combined effect of strong CO2-IL interactions and large accessible surface area leads to higher CO2 adsorption than the other two models. Fig. 9 shows snapshots of CO2/N2 and CO2/CH4 binary mixture adsorption on the three models at 298.15 K and 12 bar. In pristine MCM-41, the adsorption of the CO2 molecules is primarily on the pore walls whereas some N2 and CH4 molecules are present away from the pore walls as well. In IL-loaded MCM-41 models, CO2 molecules are not only present near the pore wall surface but are also distributed away from the walls, in the pore interior, near the IL molecules. Although the mole fraction of CO2 in the bulk gas mixture is only 0.15, the amount of CO2 adsorbed is higher than N2 and comparable to CH4 (especially for the ILloaded models), as can be seen from Fig. 8 and visually verified from Fig. 9. The selectivities of CO2 over N2 and CH4 as a function of pressure, obtained from GCMC simulations and IAST calculations, are shown in Fig. 10. At 1 bar and 298.15 K, the CO2/N2 selectivity in pristine MCM-41, obtained from GCMC simulations, is ~15 whereas selectivities in MCM-41-IL1 and MCM-41-IL2 are ~23 and ~28 respectively (see Fig. 10(a)). The CO2/N2 selectivity values decrease with pressure and become relatively constant at higher pressures. Above 5 bar, pristine MCM-41 shows a selectivity of ~10 whereas ILloaded MCM-41 show selectivities in the range 17–20. Thus, the CO2/N2 selectivity is 50– 100% higher in the presence of IL in the MCM-41 mesopores over the entire pressure range 16 ACS Paragon Plus Environment

Page 16 of 36

Page 17 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


explored. The higher affinity of CO2 toward the IL molecules than N2 results in the significantly higher selectivities observed in MCM-41-IL1 and MCM-41-IL2. The IAST predictions of CO2/N2 selectivity in Fig. 10(a) show similar trend to the GCMC results, although the numerical values are different, especially for adsorbents containing IL. The observed CO2/N2 selectivity in unmodified MCM-41 is comparable to values reported in literature for other unfunctionalized porous materials such as ordered mesoporous carbon. Yuan et al.70 reported CO2/N2 selectivity of ~8 at 318 K, 100 kPa for synthesized ordered mesoporous carbon having a surface area of 2255 m2/g. Some microporous MOFs such as Cu-BTC exhibit higher CO2/N2 selectivity (~20 at 298 K and 100 kPa for CuBTC71). ILmodified MOFs have also been reported to show enhanced selectivity compared to pristine MOF. Chen et al.36 reported CO2/N2 selectivity of ~5, ~11 and ~20 for unmodified IRMOF, and IRMOF with 40 wt% and 86 wt% loading of the ionic liquid [BMIM][PF6] respectively. CO2/N2 selectivity on ZIF-8 showed enhancement from ~8 to ~11.5 on incorporation of 28 wt% [BMIM][BF4]38 and to ~13 on incorporation of 24 wt% [BMIM][PF6]68. Thus, the enhanced selectivities observed for our IL-loaded MCM-41 systems are in general higher than those reported for IL-loaded MOFs in literature, but lower than the selectivity in pure IL ([C4PYR+][TF2N–]) reported by Tomé et al.72 to be 32 (ratio of solubilities of CO2 and N2 in the IL). CO2/CH4 selectivities, shown in Fig. 10(b), also follow the trend MCM-41-IL2 > MCM-41-IL1 > MCM-41, with the values being significantly higher in the IL-loaded models. The CO2/CH4 selectivities obtained from GCMC simulations in pristine MCM-41, MCM-41-IL1 and MCM-41-IL2 at 1 bar are ~3.7, ~5.4 and ~6.8 respectively, whereas at pressures above 5 bar, the selectivity values are ~3, ~4.6 and ~5.5–6 respectively. In comparison, CO2/CH4 selectivity on ZIF-8 shows enhancement from ~2.5 to ~3.5 on incorporation of 28 wt% [BMIM][BF4] (as reported by Boyuturk et al.38) or 24 wt% 17 ACS Paragon Plus Environment


[BMIM][PF6] (as reported by Kinik et al.68). The ratio of solubilities of CO2 and CH4 in the ionic liquid [C4PYR+][TF2N–] is ~14 (Tomé et al.72). Thus, although the IL-loaded MCM-41 materials studied here exhibit greater enhancement of CO2/CH4 selectivity than IL-loaded MOFs reported in literature, the selectivities are smaller than in pure IL. The IAST predictions for CO2/CH4 selectivities display the same overall trend as GCMC results with MCM-41-IL2 showing highest selectivities. Moreover, above 2 bar pressure, the IAST results show good quantitative match with values obtained from GCMC simulations. MCM41-IL2 is the most desirable of the three adsorbents in terms of the selectivity of CO2 over CH4 and N2, over the entire pressure range studied (up to 14 bar). 4. Conclusions We have studied the effect of impregnating the pores of mesoporous silica MCM-41 with the IL C4PYR-TF2N on the adsorption behavior of CO2, N2, CH4 and the separation of binary mixtures CO2/N2 and CO2/CH4. Two models were generated by packing IL molecules in the mesopores of MCM-41 at different loadings (~21% and 42% by weight) and equilibrating the composite structure using MD simulations. The accessible surface area of the models showed the trend MCM-41-IL1 > MCM-41 > MCM-41-IL2, whereas the void fraction was observed to follow MCM-41 > MCM-41-IL1 > MCM-41-IL2. The adsorption of pure CO2 in the model with 21% IL content, i.e. MCM-41-IL1, was observed to be significantly higher than that in pristine MCM-41. However, the CO2 capacity at 1 bar and 298.15 K was still lower than that reported in literature for amine-grafted MCM-41 and activated carbon. In the model with 42% IL content, i.e. MCM-41-IL2, pure CO2 adsorption was slightly higher than that in pristine MCM-41 at pressures below 5 bar and lower above that. In the case of pure N2 and CO2 adsorption, although MCM-41-IL1 showed the highest adsorbed amounts, the enhancement with respect to pristine MCM-41 was small. The high adsorption loading of CO2 in MCM-41-IL1 can be attributed to the larger accessible area and stronger interactions 18 ACS Paragon Plus Environment

Page 18 of 36

Page 19 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


with the IL molecules. For the less polar N2 and CH4 (essentially nonpolar), the slight enhancement in loading in MCM-41-IL1 is primarily due to increased accessible area. In the case of MCM-41-IL2, favorable interaction between CO2 and IL molecules results in higher CO2 adsorption than pure MCM-41 at pressures below 5 bar. However, at higher pressures, the effect of interactions is offset by the lower accessible area and void fraction due to which saturation sets in leading to lower CO2 adsorption than in pure MCM-41. The RDFs and spatial distribution maps of CO2 show that in addition to the MCM-41 pore walls, CO2 molecules are also distributed in the neighborhood of IL molecules. The simulation of binary CO2/N2 and CO2/CH4 mixtures revealed that the amount of CO2 adsorbed is higher in ILloaded MCM-41 models whereas the amounts of N2 and CH4 adsorbed were higher in the pristine MCM-41. Selectivity of CO2 over N2 and CO2 over CH4 was 50–100% higher in the IL-loaded MCM-41 models than in pure MCM-41. Although the amount of CO2 adsorbed is higher in MCM-41-IL1, the selectivities are higher in MCM-41-IL2. The selectivities obtained with IL-loaded MCM-41 models were typically higher than those reported in literature for IL-modified MOFs. Thus, modification of MCM-41 mesopores by impregnation of IL molecules can potentially improve the CO2 separation behavior of mesoporous silica. The relative improvements in the amount of CO2 adsorbed and CO2 selectivity may be tuned by adjusting the loading of IL in the mesopores. Supporting Information Brief description of Langmuir model, dual-site Langmuir model, ideal adsorbed solution theory, calculation of isosteric heat of adsorption; Textural properties of different configurations of MCM-41-IL1 model; LJ potential parameter and charges in adsorbent and adsorbate molecule; Parameters of the DSL model for CO2 adsorption on the three MCM-41 models; Parameters of the DSL model for CH4 adsorption on the three MCM-41 models; Parameters of the DSL model for N2 adsorption on the three MCM-41 models; CO2 19 ACS Paragon Plus Environment


adsorption isotherms for five different configurations of MCM-41-IL1 and MCM-41-IL2; Simulated adsorption isotherms and corresponding DSL model fits for CO2 adsorption at three temperatures on pure MCM-41, MCM-41-IL1 and MCM-41-IL2; Simulated adsorption isotherms and corresponding DSL model fits for CH4 adsorption at three temperatures on pure MCM-41, MCM-41-IL1 and MCM-41-IL2; Simulated adsorption isotherms and corresponding DSL model fits for N2 adsorption at three temperatures on pure MCM-41, MCM-41-IL1 and MCM-41-IL2; Henry’s constant for CO2, CH4 and N2 adsorption in the three adsorbent models.

Acknowledgment The authors wish to acknowledge the Department of Science and Technology, Government of India for FIST program support under project SR/FST/ET11-047/2010, which provided funding for the high-performance computing cluster used to carry out the simulations reported in this work. References (1)

Li, J.; Kuppler, R.; Zhou, H. Selective Gas Adsorption and Separation in MetalOrganic Frameworks. Chem. Soc. Rev. 2009, 38, 1477–1504.

(2)

Yuan, B.; Wu, X.; Chen, Y.; Huang, J.; Luo, H.; Deng, S. Adsorption of CO2, CH4, and N2 on Ordered Mesoporous Carbon: Approach for Greenhouse Gases Capture and Biogas Upgrading. Environ. Sci. Technol. 2013, 47, 5474–5480.

(3)

Liu, J.; Tian, J.; Thallapally, P. K.; Mcgrail, B. P. Selective CO2 Capture from Flue Gas Using Metal-Organic Frameworks―A Fixed Bed Study. J. Phys. Chem. C 2012, 116, 9575–9581.

(4)

Yu, K.; Kiesling, K.; Schmidt, J. R. Trace Flue Gas Contaminants Poison Coordinatively Unsaturated Metal-Organic Frameworks: Implications for CO2 Adsorption and Separation. J. Phys. Chem. C 2012, 116, 20480–20488.

(5)

Zhang, C.; Sun, Z.; Chen, S.; Wang, B. Enriching Blast Furnace Gas by Removing Carbon Dioxide. J. Environ. Sci. (China) 2013, 25, S196–S200.

(6)

Islamoglu, T.; Kim, T.; Kahveci, Z.; El-Kadri, O. M.; El-Kaderi, H. M. Systematic Postsynthetic Modification of Nanoporous Organic Frameworks for Enhanced CO2 20 ACS Paragon Plus Environment

Page 20 of 36

Page 21 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Capture from Flue Gas and Landfill Gas. J. Phys. Chem. C 2016, 120, 2592–2599. (7)

Yu, J.; Xie, L.-H.; Li, J.-R.; Ma, Y.; Seminario, J. M.; Balbuena, P. B. CO2 Capture and Separations Using MOFs: Computational and Experimental Studies. Chem. Rev. 2017, 117, 9674–9754.

(8)

Liu, Z.; Wu, Y.; Liu, B.; Oh, S. C.; Fan, W.; Qian, Y.; Xi, H. Tuning the Adsorption and Separation Properties of Noble Gases and N2 in CuBTC by Ligand Functionalization. RSC Adv. 2016, 6, 91093–91101.

(9)

Arstad, B.; Fjellvåg, H.; Kongshaug, K. O.; Swang, O.; Blom, R. Amine Functionalised Metal Organic Frameworks (MOFs) as Adsorbents for Carbon Dioxide. Adsorption 2008, 14, 755–762.

(10)

Lin, Y.; Kong, C.; Chen, L. Amine-Functionalized Metal–organic Frameworks: Structure, Synthesis and Applications. RSC Adv. 2016, 6, 32598–32614.

(11)

Hu, J.; Liu, J.; Liu, Y.; Yang, X. Improving Carbon Dioxide Storage Capacity of Metal Organic Frameworks by Lithium Alkoxide Functionalization: A Molecular Simulation Study. J. Phys. Chem. C 2016, 120, 10311–10319.

(12)

Abid, H. R.; Shang, J.; Ang, H.-M.; Wang, S. Amino-Functionalized Zr-MOF Nanoparticles for Adsorption of CO2 and CH4. Int. J. Smart Nano Mater. 2013, 4, 72– 82.

(13)

Makal, T. A.; Wang, X.; Zhou, H. C. Tuning the Moisture and Thermal Stability of Metal-Organic Frameworks through Incorporation of Pendant Hydrophobic Groups. Cryst. Growth Des. 2013, 13, 4760–4768.

(14)

Howarth, A. J.; Liu, Y.; Li, P.; Li, Z.; Wang, T. C.; Hupp, J. T.; Farha, O. K. Chemical, Thermal and Mechanical Stabilities of Metal–organic Frameworks. Nat. Rev. Mater. 2016, 1, 15018.

(15)

Chen, D. M.; Tian, J. Y.; Chen, M.; Liu, C. Sen; Du, M. Moisture-Stable Zn(II) MetalOrganic Framework as a Multifunctional Platform for Highly Efficient CO2 Capture and Nitro Pollutant Vapor Detection. ACS Appl. Mater. Interfaces 2016, 8, 18043– 18050.

(16)

Loganathan, S.; Tikmani, M.; Ghoshal, A. K. Novel Pore-Expanded MCM-41 for CO2 Capture: Synthesis and Characterization. Langmuir 2013, 29, 3491–3499.

(17)

Pajzderska, A.; Gonzalez, M. A.; Mielcarek, J.; Wasicki, J. Water Behaviour in MCM41 as a Function of Pore Filling and Temperature Studied by NMR and Molecular Dynamics Simulations. J. Phys. Chem. C 2014, 118, 23701–23710.

(18)

He, Y.; Seaton, N. A. Heats of Adsorption and Adsorption Heterogeneity for Methane, Ethane, and Carbon Dioxide in MCM-41. Langmuir 2006, 22, 1150–1155.

(19)

Kruk, M.; Jaroniec, M.; Ko, C. H.; Ryoo, R. Characterization of the Porous Structure of SBA-15. Chem. Mater. 2000, 12, 1961–1968.

(20)

Wu, S.-H.; Mou, C.-Y.; Lin, H.-P. Synthesis of Mesoporous Silica Nanoparticles. Chem. Soc. Rev. 2013, 42, 3862–3875.

(21)

Heikkilä, T.; Salonen, J.; Tuura, J.; Hamdy, M. S.; Mul, G.; Kumar, N.; Salmi, T.; Murzin, D. Y.; Laitinen, L.; Kaukonen, A. M.; et al. Mesoporous Silica Material TUD21 ACS Paragon Plus Environment


1 as a Drug Delivery System. Int. J. Pharm. 2007, 331, 133–138. (22)

Vallet-Regi, M.; Rámila, A.; del Real, R. P.; Pérez-Pariente, J. A New Property of MCM-41:Drug Delivery System. Chem. Mater. 2001, 13, 308–311.

(23)

Mondal, J.; Nandi, M.; Modak, A.; Bhaumik, A. Functionalized Mesoporous Materials as Efficient Organocatalysts for the Syntheses of Xanthenes. J. Mol. Catal. A Chem. 2012, 363–364, 254–264.

(24)

Vallet-Regí, M.; Balas, F.; Arcos, D. Mesoporous Materials for Drug Delivery. Angew. Chemie - Int. Ed. 2007, 46, 7548–7558.

(25)

Hung, C.; Bai, H.; Karthik, M. Ordered Mesoporous Silica Particles and Si-MCM-41 for the Adsorption of Acetone: A Comparative Study. Sep. Purif. Technol. 2009, 64, 265–272.

(26)

Jiao, J.; Cao, J.; Xia, Y.; Zhao, L. Improvement of Adsorbent Materials for CO2 Capture by Amine Functionalized Mesoporous Silica with Worm-Hole Framework Structure. Chem. Eng. J. 2016, 306, 9–16.

(27)

Wang, G.; Otuonye, A. N.; Blair, E. A.; Denton, K.; Tao, Z.; Asefa, T. Functionalized Mesoporous Materials for Adsorption and Release of Different Drug Molecules: A Comparative Study. J. Solid State Chem. 2009, 182, 1649–1660.

(28)

Karpov, S. I.; Roessner, F.; Selemenev, V. F. Studies on Functionalized Mesoporous Materials - Part I: Characterization of Silylized Mesoporous Material of Type MCM41. J. Porous Mater. 2014, 21, 449–457.

(29)

Santos, H. A.; Salonen, J.; Bimbo, L. M.; Lehto, V.-P.; Peltonen, L.; Hirvonen, J. Mesoporous Materials as Controlled Drug Delivery Formulations. J. Drug Deliv. Sci. Technol. 2011, 21, 139–155.

(30)

Wang, S. Ordered Mesoporous Materials for Drug Delivery. Microporous Mesoporous Mater. 2009, 117, 1–9.

(31)

Vangeli, O. C.; Romanos, G. E.; Beltsios, K. G.; Fokas, D.; Kouvelos, E. P.; Stefanopoulos, K. L.; Kanellopoulos, N. K. Grafting of Imidazolium Based Ionic Liquids on the Pore Surface of Nanoporous Materials-Study of Physiochemical and Thermodynamic Properties. J. Phys. Chem. B 2010, 114, 6480–6491.

(32)

Romanos, G. E.; Stefanopoulos, K. L.; Vangeli, O. C.; Mergia, K.; Beltsios, K. G.; Kanellopoulos, N. K.; Lairez, D. Investigation of Physically and Chemically Ionic Liquid Confinement in Nanoporous Materials by a Combination of SANS, ContrastMatching SANS, XRD and Nitrogen Adsorption. J. Phys.: Conf. Ser. 2012, 340, 012087.

(33)

Sanaeishoar, H.; Sabbaghan, M.; Mohave, F. Synthesis and Characterization of MicroMesoporous MCM-41 Using Various Ionic Liquids as Co-Templates. Microporous Mesoporous Mater. 2015, 217, 219–224.

(34)

Tripathi, A. K.; Verma, Y. L.; Singh, R. K. Thermal, Electrical and Structural Studies on Ionic Liquid Confined in Ordered Mesoporous MCM-41. J. Mater. Chem. A 2015, 3, 23809–23820.

(35)

Cheng, J.; Li, Y.; Hu, L.; Zhou, J.; Cen, K. CO2 Adsorption Performance of Ionic Liquid [P66614][2-Op] Loaded onto Molecular Sieve MCM-41 Compared to Pure Ionic 22 ACS Paragon Plus Environment

Page 22 of 36

Page 23 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Liquid in Biohythane/Pure CO2 Atmospheres. Energy & Fuels 2016, 30, 3251–3256. (36)

Chen, Y.; Hu, Z.; Gupta, K. M.; Jiang, J. Ionic Liquid/Metal Organic Framework Composite for CO2 Capture: A Computational Investigation. J. Phys. Chem. C 2011, 115, 21736–21742.

(37)

Li, Z.; Xiao, Y.; Xue, W.; Yang, Q.; Zhong, C. Ionic Liquid/metal-Organic Framework Composites for H2S Removal from Natural Gas: A Computational Exploration. J. Phys. Chem. C 2015, 119, 3674–3683.

(38)

Koyuturk, B.; Altintas, C.; Kinik, F. P.; Keskin,S.; Uzun, A. Improving Gas Separation Performance of ZIF-8 by [BMIM][BF4] Incorporation: Interactions and Their Consequences on Performance. J. Phys. Chem. C 2017, 121,10370–10381.

(39)

Ban, Y.; Li, Z.; Li, Y.; Peng, Y.; Jin, H.; Jiao, W.; Guo, A.; Wang, P.; Yang, Q.; Zhong, C. et al. Confinement of Ionic Liquids in Nanocages: Tailoring the Molecular Sieving Properties of ZIF-8 for Membrane-Based CO2 Capture. Angew. Chemie - Int. Ed. 2015, 54, 15483–15487.

(40)

Daly, C. A.; Berquist, E. J.; Brinzer, T.; Garrett-Roe, S.; Lambrecht, D. S.; Corcelli, S. A. Modeling Carbon Dioxide Vibrational Frequencies in Ionic Liquids: II. Spectroscopic Map. J. Phys. Chem. B 2016, 120, 12633–12642.

(41)

Lei, Z; Dai, C.; Chen, B. Gas Solubility in Ionic Liquids. Chem. Rev. 2014, 114, 1289– 1326.

(42)

Yim, J.-H., Song, H. N.; Yoo, K.-P.; Lim, J. S. Measurement of CO2 Solubility in Ionic Liquids: [BMP][Tf2N] and [BMP][MeSO4] by Measuring Bubble-Point Pressure. J. Chem. Eng. Data 2011, 56, 1197–1203.

(43)

Lee, B.-C.; Nam, S. G. High-pressure Solubility of Carbon Dioxide in Pyrrolidiniumbased Ionic Liquids: [bmpyr][dca] and [bmpyr][Tf2N]. Korean J. Chem. Eng. 2015, 32, 521–533.

(44)

Kumar, K.; Kumar, A. Adsorptive Separation of Carbon Dioxide from Flue Gas using Mesoporous MCM-41: A Molecular Simulation Study. Korean J. Chem. Eng. 2018, 35, 535–547.

(45)

Accelrys Inc., Materials Studio, 7.0, San Diego: Accelrys Inc., 2013.

(46)

Northcott, K. A.; Miyakawa, K.; Oshima, S.; Komatsu, Y.; Perera, J. M.; Stevens, G. W. The Adsorption of Divalent Metal Cations on Mesoporous Silicate MCM-41. Chem. Eng. J. 2010, 157, 25–28.

(47)

Oshima, S.; Perera, J. M.; Northcott, K. A.; Kokusen, H.; Stevens, G. W.; Komatsu, Y. Adsorption Behavior of Cadmium(II) and Lead(II) on Mesoporous Silicate MCM-41. Sep. Sci. Technol. 2006, 41, 1635–1643.

(48)

Gaussian 09, Revision B.01, Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Petersson, G. A.; Nakatsuji, H. et al. Gaussian, Inc., Wallingford CT, 2010.

(49)

Martínez, L.; Andrade, R.; Birgin, E. G.; Martínez, J. M. Packmol: A Package for Building Initial Configurations for Molecular Dynamics Simulations. J. Comput. 23 ACS Paragon Plus Environment


Chem., 2009, 30, 2157–2164. (50)

Potoff, J. J.; Sandler, S. I. Vapor-Liquid Equilibria of Mixtures Containing Alkanes, Carbon Dioxide, and Nitrogen. AIChE J. 2001, 47, 1676–1682.

(51)

Harris, J. G.; Yung, K. H. Carbon Dioxide’s Liquid-Vapor Coexistence Curve And Critical Properties as Predicted by a Simple Molecular Model. J. Phys. Chem. 1995, 99, 12021–12024.

(52)

Martin, M. G.; Siepmann, J. I. Transferable Potentials for Phase Equilibria. 1. UnitedAtom Description of N -Alkanes. J. Phys. Chem. B 1998, 102, 2569–2577.

(53)

Jorgensen, W. L.; Maxwell, D. S.; Tirado-Rives, J. Development and Testing of the OLPS All-Atom Force Field on Conformational Energetics and Properties of Organic Liquids. J. Am. Chem. Soc. 1996, 118, 11225–11236.

(54)

Jorgensen, W. L.; Mcdonald, N. A. Development of an All-Atom Force Field for Heterocycles. Properties of Liquid Pyridine and Diazenes . J. Mol. Struct.: THEOCHEM 1998, 424, 145–155.

(55)

McDonald, N. A.; Jorgensen, W. L. Development of an All-Atom Force Field for Heterocycles. Properties of Liquid Pyrrole, Furan, Diazoles, and Oxazoles. J. Phys. Chem. B 1998, 102, 8049–8059.

(56)

Rizzo, R. C.; Jorgensen, W. L. OPLS All-Atom Model for Amines: Resolution of the Amine Hydration Problem. J. Am. Chem. Soc. 1999, 121, 4827–4836.

(57)

Jing, Y.; Wei, L.; Wang, Y.; Yu, Y. Molecular Simulation of MCM-41: Structural Properties and Adsorption of CO2, N2 and Flue Gas. Chem. Eng. J. 2013, 220, 264– 275.

(58)

Plimpton, S. J. Fast Parallel Algorithms for Short-Range Molecular Dynamics. J. Comput. Phys. 1995, 117, 1−19.

(59)

Ryckaert, J.-P.; Ciccotti, G.; Berendsen, H. J. C. Numerical Integration of the Cartesian Equations of Motion of a System with Constraints: Molecular Dynamics of n-Alkanes. J. Comput. Phys. 1977, 23, 327–341.

(60)

Dubbeldam, D.; Calero, S.; Ellis, D. E.; Snurr, R. Q. RASPA: Molecular Simulation Software for Adsorption and Diffusion in Flexible Nanoporous Materials. Mol. Simul. 2015, 7022, 1–21.

(61)

Chen, G.; Zhang, J.; Lu, S.; Pervukhina, M.; Liu, K.; Xue, Q.; Tian, H.; Tian, S.; Li, J.; Clennell, M. B.; et al. Adsorption Behavior of Hydrocarbon on Illite. Energy and Fuels 2016, 30, 9114–9121.

(62)

Gutiérrez-Sevillano, J. J.; Caro-Pérez, A.; Dubbeldam, D.; Calero, S. Molecular Simulation Investigation into the Performance of Cu-BTC Metal-Organic Frameworks for Carbon Dioxide-Methane Separations. Phys. Chem. Chem. Phys. 2011, 13, 20453.

(63)

Poursaeidesfahani, A.; Torres-Knoop, A.; Rigutto, M.; Nair, N.; Dubbeldam, D.; Vlugt, T. J. H. Computation of the Heat and Entropy of Adsorption in Proximity of Inflection Points. J. Phys. Chem. C 2016, 120, 1727–1738.

(64)

Myers, A. L.; Prausnitz, J. M. Thermodynamics of Mixed-Gas Adsorption. Am. Inst. Chem. Eng. J. 1965, 11, 121–127. 24 ACS Paragon Plus Environment

Page 24 of 36

Page 25 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(65)

Dos Santos, T. C.; Bourrelly, S.; Llewellyn, P. L.; de M. Carneiro, J. W.; Ronconi, C. M. Adsorption of CO2 on Amine-Functionalised MCM-41: Experimental and Theoretical Studies. Phys. Chem. Chem. Phys. 2015, 17, 11095–11102.

(66)

Rao, N.; Wang, M.; Shang, Z.; Hou, Y.; Fan, G.; Li, J. CO2 Adsorption by AmineFunctionalized MCM-41: A Comparison between Impregnation and Grafting Modification Methods. Energy Fuels 2018, 32, 670–677.

(67)

Jiménez, V.; Ramírez-Lucas, A.; Díaz, J. A.; Sánchez, P.; Romero, A. CO2 Capture in Different Carbon Materials. Environ. Sci. Technol. 2012, 46, 7407–7414.

(68)

Kinik, F. P.; Altintas, C.; Balci, V.; Koyuturk, B.; Uzun, A.; Keskin, S. [BMIM][PF6] Incorporation Doubles CO2 Selectivity of ZIF-8: Elucidation of Interactions and Their Consequences on Performance. ACS Appl. Mater. Interfaces 2016, 8, 30992–31005.

(69)

B. Yuan, B.; Wu, X.; Chen, Y.; Huang, J.; Luo, H.; Deng, S. Adsorption of CO2, CH4, and N2 on Ordered Mesoporous Carbon: Approach for Greenhouse Gases Capture and Biogas Upgrading. Environ. Sci. Technol. 2013, 47, 5474–5480.

(70)

Yang, Q.; Xue, C.;Zhong, C.; Chen, J.-F. Molecular simulation of separation of CO2 from flue gases in Cu-BTC metal-organic framework. AIChE J., 2007, 53, 2832–2840.

(71)

Tomé, L. C.; Mecerreyes, D.; Freire, C. S. R.; Rebelo, L. P. N.; Marrucho, I. M. Pyrrolidinium-based polymeric ionic liquid materials: New perspectives for CO2 separation membranes. J. Membr. Sci. 2013, 428, 260–266.



Page 26 of 36

Table 1: Accessible pore volume and surface area of different MCM-41 models considered in this work. Adsorbent Name

Accessible pore volume fraction (probe radius 1.4 Å)

Accessible pore volume fraction (probe radius 1.82 Å)

MCM-41

0.4962

0.4576

1061.22

951.30

MCM-41-IL1

0.2812

0.2178

1492.77

1217.57

MCM-41-IL2

0.1446

0.0951

1050.70

724.14

Surface area in m2/g (probe radius 1.4 Å)


Surface area in m2/g (probe radius 1.82 Å)

Page 27 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 1: Molecular structures of the cation (C4PYR+) and the anion (TF2N–) of the IL.

(a)

(b)

(c)

Figure 2: Molecular models of (a) pure MCM-41, (b) MCM-41-IL1, and (c) MCM-41-IL2.



Figure 3: Comparison of simulated adsorption isotherm of pure CO2 on MCM-41 with experimental data of dos Santos et al.65 at 303.15 K.


Page 28 of 36

Page 29 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(a)

(b)

(c)

Figure 4: Adsorption isotherms of pure (a) CO2 (b) N2 and (c) CH4 at 298.15 K in the pure MCM-41 and IL-loaded MCM-41 models. Lines represent dual-site Langmuir model fits.



(a)

Page 30 of 36

(b)

(c)

Figure 5: Isosteric heats of adsorption of pure (a) CO2, (b) CH4 and (c) N2 at 298.15 K in the pure MCM-41 and IL-loaded MCM-41 models.


Page 31 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(a)

(b)

Figure 6: 3D spatial distribution maps of CO2 in (a) MCM-41-IL1, and (b) MCM-41-IL2, obtained from MD simulations at 298.15 K. The framework atoms (MCM-41 and IL) have not been shown for clarity.



Page 32 of 36

(a)

(b)

(c)

(d)

Figure 7: RDFs between (a) C4PYR cation and CO2, (b) TF2N anion and CO2, (c) Hydrogen atoms on pyrrolidinium ring of cation (marked as HA in Fig. 1) and oxygen atoms of CO2, and (d) the cation and the anion of the IL. In (a), (b) and (d), the reference atom for the cation is the nitrogen atom of the pyrrolidinium ring (marked as N1 in Fig. 1), that for the anion is the central (and only) nitrogen atom (marked as N in Fig. 1) and that for CO2 is the carbon atom center of the CO2 molecules.


Page 33 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(b)

(a)

Figure 8: Simulated adsorption isotherms of (a) N2 and CO2 for 15:85 (by mole) binary CO2/N2 mixture, and (b) CH4 and CO2 for 15:85 (by mole) binary CO2/CH4 mixture. All isotherms are at 298.15K (lines have been shown only as guide to the eye).



Page 34 of 36

(a)

(b)

(c)

(d)

(e)

(f)

Figure 9: Snapshots of adsorption of CO2/N2 mixture (15:85 molar ratio) in (a) pure MCM41, (b) MCM-41-IL1 and (c) MCM-41-IL2, and of CO2/CH4 mixture (15:85 molar ratio) in (d) pure MCM-41, (e) MCM-41-IL1 and (f) MCM-41-IL2 at 298.15 K and 12 bar. The adsorbent (MCM-41 and IL) is shown in wireframe representation whereas the atoms in the adsorbate gas molecules are represented as spheres. CO2 is shown with carbon atoms in cyan and oxygen atoms in red; N2 is shown with nitrogen atoms in blue; CH4 is shown as single effective sphere in green.


Page 35 of 36 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(a) CO2/N2

(b) CO2/CH4

Figure 10: Adsorptive selectivity of CO2 over (a) N2 and (b) CH4 in the three adsorbent models, obtained from GCMC simulations and IAST calculations. 35 ACS Paragon Plus Environment


TOC Graphic


Page 36 of 36

Enhanced CO2 Adsorption and Separation in Ionic Liquid

Recommend Documents