Molecular Simulation Study of Montmorillonite in Contact with Variably

Subscriber access provided by UB + Fachbibliothek Chemie | (FU-Bibliothekssystem)

Article

Molecular Simulation Study of Montmorillonite in Contact with Variably Wet Supercritical Carbon Dioxide Ahmad Kadoura, Arun Kumar Narayanan Nair, and Shuyu Sun J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.7b01027 • Publication Date (Web): 07 Mar 2017 Downloaded from http://pubs.acs.org on March 9, 2017

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 free 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 accessible to all readers and 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.

The Journal of Physical Chemistry C 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

Molecular Simulation Study of Montmorillonite in Contact with Variably Wet Supercritical Carbon Dioxide Ahmad Kadoura, Arun Kumar Narayanan Nair∗, and Shuyu Sun Physical Science and Engineering Division (PSE), Computational Transport Phenomena Laboratory, King Abdullah University of Science and Technology (KAUST), Thuwal, 23955-6900, Saudi Arabia. March 3, 2017

∗

To whom correspondence should be addressed, email: [email protected]

1 ACS Paragon Plus Environment


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 We perform grand canonical Monte Carlo simulations to study the detailed molecular mechanism of intercalation behavior of CO2 in Na-, Ca-, and Mgmontmorillonite exposed to variably hydrated supercritical CO2 at 323.15 K and 90 bar. The simulations indicate that the intercalation of CO2 strongly depends on the relative humidity (RH). The intercalation of CO2 in the dehydrated interlayer is inhibited, followed by the swelling of the interlayer region due to uptake of water and CO2 as the RH increases. In all of the hydrated clay samples, the amount of the intercalated CO2 generally decreases as a function of increasing RH, which is attributed mainly to the weakening of the interaction between CO2 and clay. At low RH values, Ca- and Mg-montmorillonite are relatively more efficient in capturing CO2 . The amount of CO2 trapped in all clay samples shows similar values above RH of ≈ 60%. Molecular dynamics simulations show that the diffusion coefficient of each species generally increases with increasing RH due to the associated expansion of the interlayer distance of the clay. For all the hydrated samples, the diffusion coefficients of CO2 and water in the interlayers are mostly comparable due to the fact that CO2 molecules are well solvated. The diffusion of CO2 in each hydration state is mostly independent of the type of cation in accordance with the fact that CO2 molecules hardly migrate into the first hydration shell of the interlayer cations.


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


1

Introduction

Geologic sequestration of carbon dioxide has emerged as a strategy for reducing emissions of anthropogenic CO2 into the atmosphere. 1–14 Successful storage of supercritical CO2 (scCO2 ) depends on the changes in the permeability of overlying cap rock formations, owing to interaction with scCO2 . Cap rocks are typically composed of shale or mudstone rich in clay minerals, including swelling clays such as the smectite mineral montmorillonite. Smectite clay minerals are aluminosilicates consisting of negatively charged layers compensated usually by solvated cations in the interlayer space, such as Na+ , Ca2+ , or Mg2+ . In addition, studies have demonstrated the possibility to utilize scCO2 for the extraction of oil and gas from shale formations. 2,11 Consequently, in this paper we focus our study on competitive sorption and transport of water and CO2 in the interlayers of montmorillonite under geologic sequestration conditions. In montmorillonite, octahedral alumina sheets are sandwiched between tetrahedral silica sheets forming 2:1 or tetrahedral-octahedral-tetrahedral (TOT) layers. These initial steps toward molecular understanding of water and CO2 in montmorillonite interlayers are important for determining impacts from the long-term exposure of carbon dioxide to geological formations. 8,12,14 The basal d−spacing is in the range 9.6−10.4 ˚ A (0W) for dry montmorillonite and increases in the presence of water usually to the range 11.5−12.5 ˚ A forming a monolayer (1W) water arrangement. 15–24 Water can further form a bilayer (2W) with basal d−spacings ranging from 14.5 to 15.5 ˚ A or three layers (3W) with basal d−spacings ranging from 18.0−19.1 ˚ A. According to experimental investigations, residual water is required to intercalate pure CO2 in dry clay (≈ 0W) and the potential collapse or expansion of the interlayer volume depends on the initial hydration state of the clay and CO2 . 4–7 For instance, 1W clay is stable but 2W and higher hydration states loses water when exposed to dry scCO2 . 7 Romanov et al. also demonstrated that CO2 can be intercalated in the interlayers of clay minerals. 8,12 Interaction with wet scCO2 can lead to a stepwise increase in the basal d−spacing values as a function of increasing relative



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

humidity (RH) and swelling of clays to a 3W hydration state. 6,7 Recently, a quantitative understanding of the partitioning of CO2 and H2 O between montmorillonite and variably wet scCO2 at T = 323.15 K and P = 90 bar was obtained by means of pressurized X-ray diffraction measurements, high-pressure infrared (IR) spectroscopy, and quartz crystal microbalance titration experiments. 10,13 These experiments demonstrated that the intercalation of CO2 in the dehydrated interlayer is inhibited, followed by the expansion of the interlayer space due to uptake of water and CO2 as the RH increases. Further, they reported that the intercalation of CO2 into clays decreases as a function of increasing RH. Recently, density functional theory based calculations by Lee et al. predicted that Ca−CO2 interactions has a stabilizing effect on CO2 intercalation. 25 The molecular simulations employing classical force fields have also shown the details of the structural and transport properties of clay mineral−water−CO2 systems at elevated temperatures and pressures relevant to geological carbon storage. 26–33 Botan et al. 27 reported that hydrated clay can intercalate CO2 and the thermodynamically stable configurations are identified with basal d−spacings corresponding to the 1W and 2W hydration states, in agreement with experiments. 4–7 The study by Makaremi et al. 29 showed that the CO2 concentrations in the interlayers of smectite minerals exceed the solubility of CO2 in bulk water and in agreement with experiments, 4,10 the maximum CO2 adsorption occurs at ≈ 1W hydration state. Rao and Leng carried out molecular simulations to investigate the interaction between Na-montmorillonite and wet scCO2 . 31 The sorbed H2 O and CO2 concentrations from those simulations compared reasonably well with the experimental data. 10 However, little is known about the influence of cation type on the adsorption process. Note that the interlayer cations are usually hydrated allowing for expansion or contraction of the interlayer region of the clay depending on the RH. 19,20,22 Recent experimental studies showed that the interlayer cations with different hydration energies significantly influence CO2 intercalation in the interlayer of smectite clays. 13,24 These experiments, for example, revealed the concept of CO2 -cation interactions as the driving force for the intercalation and retention of CO2 in smectite clays under low 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


water environments. Our studies showed that molecular simulations constitute a powerful tool to generally explore the chemical and surface interactions. 34–37 In this work, grand canonical Monte Carlo (GCMC) and molecular dynamics (MD) simulations are performed to provide molecular-level understanding of the intercalation behavior of CO2 in Na-, Ca-, and Mg-montmorillonite exposed to variably hydrated scCO2 at 323.15 K and 90 bar (conditions relevant to geological carbon sequestration). The clay was selected for comparison with the experimental quantification of sorbed H2 O and CO2 as a function of percent H2 O saturation in scCO2 . 10,13 In addition, we report a comprehensive study of the structural and transport properties of interlayer species following CO2 intercalation. A fundamental understanding of the intercalation behavior is desirable for linking or upscaling the descriptions from the molecular level to the macroscopic scale. 38,39 Our simulations indicate that Ca- and Mg-montmorillonite are relatively more efficient in capturing CO2 . In our simulations, CO2 molecules hardly migrate into the first hydration shell of the interlayer cations, contrary to the common notion that ion-CO2 interactions determine the intercalation behavior of CO2 . This result is further supported by recent experiments 10,13 and led us to the most interesting discovery that the diffusion of CO2 in each hydration state is independent of the type of cation.

2

Simulation details

Since the simulation model and method are the same as those used in our previous studies, 36,37 we briefly describe them here. The clay model is based on the pyrophyllite unit cell structure (Si8 Al4 O20 (OH)4 ) and the simulation supercell contains a total of 64 unit cells with 2560 atoms constituting the mineral portion of the clay phase (Fig. 1). In particular, we consider Wyoming-type montmorillonite 10,13,16,17 of unit cell formula: M0.75/n (Si7.75 Al0.25 )(Al3.5 Mg0.5 )O20 (OH)4 , where M represents a counterion (Na+ , Ca2+ , or Mg2+ ) and n represents the charge on the ion. Based on this 5 ACS Paragon Plus Environment


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

formula, each clay sheet contains 16 isomorphic substitutions of Al by Mg ion in the octahedral sheet, 8 isomorphic replacements of Si by Al ion in the tetrahedral sheets, and 24 (12) compensating monovalent (divalent) cations in the interlayer space. All these replacements obey the requirements of Loewenstein’s rule (i.e., substitution sites cannot be bridged by a single O atom). 40 Our substitution of Si by Al in both the tetrahedral sheets of a TOT layer is consistent with the previous simulation studies. 17,31 The clay particles are assumed to be rigid and periodic boundary conditions are used in all three spatial dimensions. All the atoms in the simulation interact with each other via the pairwise additive Lennard-Jones (LJ) 12-6 function. 41 All LJ cross interaction terms are generated from the conventional Lorentz-Berthelot combining rules. In addition, the unscreened Coulomb potential is used to describe interactions between the charged atoms. We employ the CLAYFF force field 42 which consists of nonbonded (electrostatic and van der Waals) terms and more realistically represents the local charge inhomogeneities formed around each specific substituted site in the clay. Each water molecule is represented by the simple point charge (SPC) model with flexible intramolecular interactions 42 and CO2 is modeled using the flexible force field developed by Cygan et al. 43 (GCMC simulations used SPC 44 and EPM2 45 models for water and CO2 , respectively). The LJ parameters of the mobile Na+ , Ca2+ , and Mg2+ ions are proposed by Smith and Whitley, 18 Koneshan et al., 46 and ˚ Aqvist, 47 respectively. The above force fields predicted structural and dynamic properties of clay systems in fair agreement with experiments. 36,37,42 GCMC algorithm is employed to simulate the adsorption of H2 O and CO2 by montmorillonites in the µH2 O µCO2 V T ensemble, where µH2 O and µCO2 are the chemical potentials of H2 O and CO2 , respectively. The pressure of CO2 is assumed to be the bulk pressure because of the CO2 dominated phase, and we computed the corresponding chemical potential from the N P T ensemble Monte Carlo simulations using the Widom’s insertion method. 41 Therefore, the imposed CO2 fugacity is about 64.5 bar at 323.15 K and 90 bar. The imposed water chemical potential corresponds to a pressure of RH×P0 , where P0 is the saturated vapor pressure of SPC water (≈ 0.12 6 ACS Paragon Plus Environment

Page 6 of 36

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


bar). 48 Note that our previous work 36,37 employed SPC/E model 44 for the GCMC simulations. The SPC/E model vapor pressure value is, however, a factor of 3 lower than the experimental value, 48 an unacceptably large deviation for the present simulations. The Widom’s insertion method in the N P T ensemble using the above pressures for each component in the mixture produced similar results. Note that our simulated H2 O and CO2 chemical potentials (or fugacities) are identical to those reported by Rao and Leng. 31 The GCMC results are obtained from the average over three independent simulations, each consisting of 8 × 107 Monte Carlo steps to guarantee equilibration followed by a production run of 2 × 107 steps. The final configurations outputted by these GCMC simulations are used as the initial configurations in our MD simulations. All MD simulations are carried out with the LAMMPS code. 49 Equilibration runs of 1 ns are performed in the N V T ensemble, followed by 4 ns production runs in the N V E ensemble. Three independent trajectories each of length 5 ns per simulation are computed to achieve good statistical averages. To relate the simulation results to experimental data, we choose the basal spacings (see Table 1) based on the measurements provided by Loring et al. 10 and Schaef et al. 13 The stable basal d−spacing is about 10 ˚ A for dry clays. In the presence of variably wet scCO2 , the next stable states are typically in the 11.5-12.5 ˚ A and 14.5-15.5 ˚ A ranges. Therefore, for simplicity, we adopt basal spacings d = 10.0, 12.0, and 15.0 ˚ A in the simulations covering, for example, the swelling states ≈ 0W, ≈ 1W, and ≈ 2W, respectively. Teich-McGoldrick et al. 22 used a similar approach for studying the swelling properties of smectite clays as a function of RH. We also note that a more accurate prediction of the swelling states of our clay model can be achieved from the stability analysis based on swelling free energy. 18,27



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

3

Results and discussion

3.1

Competitive sorption of H2 O and CO2 in Na-, Ca-, and Mg-montmorillonite at 323.15 K and 90 bar

GCMC simulations are carried out to understand H2 O and CO2 partitioning between montmorillonites and variably wet scCO2 . The amounts of the intercalated H2 O and CO2 as a function of RH are shown in Fig. 2. The sorbed H2 O and CO2 amounts from our simulations compare well with the corresponding experimental data. 10,13,31 However, the smooth increase in sorbed H2 O obtained in experiments seems inconsistent with the corresponding stepwise increase in simulations. This discrepancy can be explained based on the hydration-heterogeneity model 10,13 which suggests that an expandable clay might exist as a distribution of integral hydrations states (i.e., 0W, 1W etc.) 50 at a given RH. The X-ray diffraction data are not sufficiently resolved to rigorously evaluate this possibility and Fig. 2 shows simulation data obtained using only basal d−spacing values at the peak maximum, and simulation of such complex processes is out of scope of this study. The experiments 10,13 reported that the interaction of clay samples with variably wet scCO2 above 100% RH can lead to swelling to the 3W hydration state. Simulations show that, for each clay sample at 100% RH, the amount of adsorbed water is higher by about a factor of 1.5 in the 3W hydration state (basal d−spacing of 18 ˚ A) than in the corresponding 2W hydration state. In addition, in all cases, the amounts of intercalated CO2 in the 2W and 3W hydration states are mostly similar. The swelling of clays above RH of 100% is not further considered in this study. The entry of CO2 in the interlayer space of 0W state is inhibited, followed by the swelling of the interlayer volume due to uptake of water and CO2 as the RH increases. In all of the hydrated clay samples, the amount of the trapped CO2 decreases with increasing RH, but for plateau-like regions at intermediate and high RH values. Our simulation results show that Ca- and Mg-montmorillonite are relatively more efficient


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


for trapping of CO2 at low RH values, which is in agreement with other studies. 25 This is because at a given RH value the adsorption of CO2 is mostly in a higher hydration state here for the Ca and Mg samples than in the Na case (see also Table 1). The amount of CO2 intercalated in all of the investigated samples shows similar values above RH of ≈ 60% possibly due to the corresponding adsorption of the components in the same hydration state (2W). The approximate amounts of CO2 in the interlayers of Na-montmorillonite (1W: 2.8H2 O/0.6CO2 and 2W: 7.9H2 O/0.3CO2 per O20 (OH)4 at 60% and 100% RH, respectively) are close to the results of previous simulations at 348.15 K, 125 bar, and 100% RH (1W: 3.7H2 O/0.6CO2 ; 27 4.1H2 O/0.3CO2 29 and 2W: 8.6H2 O/0.4CO2 ; 27 8.9H2 O/0.2CO2 29 per O20 (OH)4 ). The GCMC simulation results of sorbed CO2 contents typically overestimate the corresponding experimental measurements. The hydration-heterogeneity model may also explain the discrepancy between simulation and experimental data of sorbed CO2 contents. The minor differences in the layer charge 32,51 and/or the charge distribution 29 between our clay model (layer charge of −0.75e per O20 (OH)4 ) and experimental samples (layer charge of −0.61e per O20 (OH)4 ) may also contribute to these discrepancies. Note that part of the discrepancy may also come from deviations in the basal d-spacings between simulations and experiments. Other possible reasons for this difference could include turbostratic stacking and registry motion of clay sheets 27,52,53 which are not considered in our study. The final configurations of these GCMC simulation study were used as the initial configurations in our N V T simulations, and the results are given below.

3.2

Atomic density profiles

In order to explore the distribution of the various species in the interlayer space of clays, their number density profiles are estimated at different RH values. Fig. 3 shows the average density profiles of CO2 (carbon atoms) molecules in montmorillonites computed along the z-axis (perpendicular to the clay surface). Fig. S1 in the Supporting Information displays the corresponding distributions of ions and water oxygens in the interlayers. These distribution profiles strongly depend on the RH and the type of 9 ACS Paragon Plus Environment


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

interlayer ions. Pronounced changes are observed in the profiles of cations and water with increasing RH, mostly due to swelling. The various cations are hydrated to different extents depending on the RH and the adsorption of cations onto montmorillonites mostly leads to both inner-sphere (close to the clay surfaces) and outer-sphere (away from the clay surfaces) surface complexes, which is consistent to previous investigations. 27,29 In both the 1W and 2W hydration states, Na+ ions have three or fewer 27,29 number of distinct density peaks possibly depending on the layer charge distribution (see also Fig. S2, Supporting Information). We find that swelling in the presence of CO2 from the 1W to the 2W hydration state reduces the probability of finding Na+ ions near the surfaces, in good agreement with previous estimates. 27,29 Previous simulation studies have demonstrated that increasing the CO2 content in the interlayer space promotes the migration of Na+ ions to the clay basal surfaces. 26 The fact that such a state is hardly reached is supported by our simulation data which, instead indicates that the amount of the trapped CO2 within the hydrated interlayer decreases with increasing RH (see Fig. 2). Furthermore, density distributions show that water and CO2 molecules form well-defined layered structures similar to those observed for pure hydration states in ambient conditions (e.g., 1W, 2W, etc.). The profiles of CO2 molecules exhibit good agreement to those reported at reservoir conditions. 26,27,29 A high peak is observed for the first-layer adsorption in all cases, which shows that the strong CO2 −clay interaction causes the molecules to pack much closer to the surface. With increasing RH, the distribution of CO2 starts to form relatively low peaks. This behavior could be possibly attributed to the CO2 -clay interaction, and the fact that the expansion of the clay layers leads to more interlayer space and the CO2 molecules are distributed between those layers. Our simulations demonstrate that polar H2 O has much more affinity to the hydrophilic montmorillonite framework than the CO2 molecule. Under identical conditions, therefore, water molecules exhibit higher adsorption amount and reside closer to the clay surface than CO2 molecules. In addition, the positions of the peak maxima of water and CO2 coincide, indicating that the molecules are basically placed in the same plane. 10 ACS Paragon Plus Environment

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


The distribution of interlayer ions is typically correlated with ion hydration energy. 18 In both the 1W and 2W hydration states, the Ca2+ and Mg2+ ions form mostly the outer-sphere surface complexes. For the same hydration states, the Na+ ions exhibit significant population in the inner-sphere configurations. The trend that increasing hydration energy precludes the formation of inner-sphere surface complexes is consistent with previous studies. 18,20 Furthermore, the density profiles in the 1W and 2W hydration states indicate that the type of counterion hardly influence the corresponding positions of the interlayer water and CO2 lying in the direction perpendicular to the clay layers.

3.3

Radial distribution functions

Simulated radial distribution functions (RDFs) describing CO2 carbon (Cc )−clay oxygen atom (Os ) spatial correlations in the interlayer space of clays are compared in Fig. 4. Fig. S3 in the Supporting Information provides corresponding distributions of CO2 with oxygen atom in water (Ow ), and intermolecular oxygen (Oc ) and carbon atoms in CO2 . Additionally, the resulting nearest-neighbor coordination numbers are collected in Table 2. The number of CO2 molecules around each carbon atom is found to be about 1.8 in the 1W state of Na-montmorillonite at 30% RH, which increases to about 2.5 in the cases of Ca- and Mg-montmorillonite at 5% RH. Interestingly, these numbers are well below the coordination number of 12 determined for scCO2 at high densities and the crystal. 54 The Cc −Oc peaks centered at about 4.2 ˚ A include a shoulder at about 3.2 ˚ A representing most likely a T-shaped molecular configuration. 54–57 As the RH increases, the relative intensity of the peak at ≈ 3.2 ˚ A generally increases in all studied cases. In the 1W hydration state, pairs of CO2 molecules likely adopt the slipped parallel geometry at shorter distances (see Fig. S4). However, the closest neighbors in the 2W state are largely oriented in a distorted T-shaped geometry and pairs at relatively large distances likely take the slipped parallel arrangement. This later behavior is similar to that of pure scCO2 and scCO2 /H2 O mixtures. 54–56 The RDFs of Cc −Os indicate the weakening of the interaction between CO2 and clay with 11 ACS Paragon Plus Environment


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

increasing RH, which is pronounced in the case of expansion of 1W to 2W hydration state in all investigated systems. This is consistent with the observation that intercalation of CO2 decreases as a function of increasing RH. 13,31 The Cc −Ow RDFs display a weak shoulder near 3.2 ˚ A which is more pronounced in the 2W state than in the 1W state, consistent with larger number of water molecules surrounding CO2 in the former case. 27,29 However, CO2 molecules are surrounded by approximately the same number of total oxygen atoms for all hydration states. For instance, Na-montmorillonite has about 4.1 water and 13.0 clay oxygen atoms around each carbon atom in the 1W state (60% RH), and 9.1 water and 5.2 clay oxygen atoms in the 2W state (80% RH), in reasonable agreement with the results of Botan et al. 27 (1W: 4.9Ow /11Os and 2W: 11Ow /5.7Os ). Simulated RDFs detailing cation−CO2 oxygen spatial correlations in the interlayer space of clays are compared in Fig. 5. Fig. S5 in the Supporting Information provides corresponding distributions of cations with oxygen atom in water and oxygen atom of clay sheet. The resulting nearest-neighbor coordination numbers are collected in Table 3. According to these distributions, cation-Ow spatial correlations display sharp peaks in g(r) at about 2.0-2.5 ˚ A and 4.0-5.0 ˚ A, evidently because of solvation effects of Na+ , Ca+2 , and Mg+2 ions. 57 Those peaks are relatively strong in the Mg-montmorillonite, suggesting that solvation effects are crucial in organizing water and ions in the interlayer region of smectite clay minerals. Ions are able to hydrate in the presence of water which enhances the probability of finding Na+ ions near the surface oxygen atoms in comparison to Ca+2 and Mg+2 ions. Our simulations show that Na-montmorillonite has about 3.6 water and 1.5 clay oxygen atoms, and 5.0 water and 0.6 clay oxygen atoms around each Na+ ion in the 1W (60% RH) and 2W (80% RH) hydration states, respectively, consistent with the results of Makaremi et al. 29 (1W: 4Ow /1.2Os and 2W: 5.5Ow /0Os ). The RDFs show no evidence of direct interaction of the cations with CO2 in the first hydration shells of strongly hydrated ions (e.g., Ca2+ and Mg2+ ), and negligible coordination of Na+ to CO2 in the first hydration shell of a Na+ ion with < 0.1. The attenuated total reflection IR data did not isolate coordination number of ∼ 12 ACS Paragon Plus Environment

Page 12 of 36

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


any features attributable to a direct cation−CO2 interaction, 10,13 which is supported by our simulation results. This result is also consistent with a prior analysis of the RDFs for CO2 in hydrated Na- and Ca-montmorillonite. 31,33 Previous simulation studies have shown that arbitrarily increasing the CO2 content in the interlayer (e.g., 3H2 O/2CO2 molecules per O20 (OH)4 in the 1W state) promotes the migration of the CO2 molecule into the first hydration shell of a Ca2+ ion. 25 Similarly, 3H2 O/1CO2 molecules per O20 (OH)4 in the 1W state created activation barriers of ≈ 2.8 and 0.5 kcal/mol for a H2 O and a CO2 molecule, respectively, to move out of the first hydration shell of a Na+ ion. 30 Myshakin et al. reported a blue shift in the asymmetric stretch frequency of the CO2 molecule which is primarily coordinated to a Na+ ion due to insufficient number of water molecules (1H2 O/0.5CO2 molecules per O20 (OH)4 in the 1W state). 26 The later hydration state with negligible water contents is not observed even at low RH values (see Fig. 2).

3.4

Preferential adsorption sites

To identify the preferential adsorption sites of the different species on the montmorillonite substrate, we estimated the in-plane (xy) density distributions. The computed distributions are shown in Figs. S6-S14, Supporting Information. Note that in the case of interlayer ions, these calculations are given for each ion found along the zaxis within either the first adsorption layers (monolayers) located nearest to the clay surfaces or away from those surfaces. The in-plane distribution of interlayer species depends on factors such as the locations of the isomorphic substitution in the clay, 29 the charge and size of the cations, 52 and the ditrigonal ring locations. 27,29,37 We find that the cations bound to the inner-sphere surface complexes are generally associated only with the sites of tetrahedral substitution. Typically, cations coordinated to the outersphere surface complexes are loosely associated with only the octahedral substitutions in the 1W hydration state for all RH. Furthermore, those cations are associated with both the tetrahedral and octahedral substitutions in the 2W hydration state. Water molecules near the surface, in all cases, display correlation with the sites of tetrahedral 13 ACS Paragon Plus Environment


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

and, to a lesser extent, octahedral substitutions. In addition, there is a tendency of water to spread along the montmorillonite surface and develop a periodic pattern reminiscent of the underlying ditrigonal rings, as the RH is increased. Recently, we have shown that such a feature persisted, albeit to a much lesser extent, also away from the clay surfaces. 37 A similar ordered structure of water also exists near an external Na-montmorillonite surface. 23 In agreement with previous simulations, 29,31 our results show that CO2 and water molecules near the surface are positioned in almost mutually exclusive regions, and the distribution of water coincides with the cation region. This can be explained by considering that the solvation energies for these cations are larger in water than in carbon dioxide. 57 Note that, Rao and Leng obtained mutually exclusive regions for the adsorbates with clays containing Al substitutions in the adjacent tetrahedral sheets located far away from each other. 31 In our work, the Al substitutions in the adjacent tetrahedral sheets are located close to each other (see Figs. S6-S14) consistent with earlier studies. 17 This can lead to significantly larger negative charge around those sites which further enhances the formation of mutually exclusive regions for the adsorbates. There is also a tendency of CO2 to spread along the montmorillonite surface with swelling, consistent with the weakening of the interaction between CO2 and clay. To improve the statistics, the different species are brought back to the unit cell and the computed in-plane distributions are shown in Figs. S15-S23, Supporting Information. Interestingly, CO2 molecules display a tendency to occupy the ditrigonal cavities on the basal surface, in agreement with previous studies. 27 The specific distribution patterns are, however, determined by properties such as turbostratic stacking 53 and registry motion of clay sheets, 52 which are not considered here. A qualitatively similar preferential adsorption behavior was obtained under near-surface geological conditions. 16,37 Inspection of the in-plane density maps reveals that the trajectories of the divalent counterions and their accompanying coordinated water molecules are denser than the case of Na+ . Similarly, comparing with Na-montmorillonite, the much lower hydration energy in Ca-montmorillonite resulted in a much more stable hydration structure of 14 ACS Paragon Plus Environment

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


Ca2+ under ambient conditions. 21 Furthermore, the in-plane density profiles in the 1W and 2W hydration states show that the type of counterion hardly influence the corresponding position of the interlayer CO2 lying parallel to the clay layers. Note that these results are also consistent with the above analysis of the ion-water and ion-CO2 RDFs.

3.5

Orientaions of water and CO2

In the 1W state, the water dipole orients parallel to the closest clay platelet, while in the 2W state, the preferred angle of the water dipole with the normal vector of the closest clay platelet is ≈ 135◦ (see Figs. S24-S26, Supporting Information). Similar results were obtained at ambient conditions. 20 Notably, the type of cation only has a small influence on the orientation of the H−H vector with respect to the axis perpendicular to the clay surface in the 1W state. For instance, a parallel orientation for the H−H vector along the surface is absent in the case of Na-montmorillonite. Botan et al. proposed a parallel orientation, based on the in-plane density distributions, for CO2 molecules along the surface. 27 This observation is consistent with our results which show a normal distribution centered around 90◦ indicating that the CO2 molecules are preferentially oriented parallel to the surface, in all studied systems. Due to swelling, the CO2 molecules explore a wider range of angles in good agreement with earlier studies. 26

3.6

Dynamical properties

Examples of the two-dimensional mean square displacements (MSDs) of the different molecules in the clay interlayer as a function of simulation time are provided in Figure S27, Supporting Information. The self-diffusion coefficients have been calculated from the linear slope of the MSDs of the molecules as a function of time, and the results are given in Table S1, Supporting Information. All these coefficients are normalized by the corresponding bulk values at about 323.15 K and presented in Fig. 6. The



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

bulk diffusion coefficients of Na+ , Ca2+ , Mg2+ , H2 O, and CO2 are about 2.3, 1.2, 1.0, 4.0, and 3.6 ×10−9 m2 /s, respectively. 58–61 Generally, the diffusion coefficients of all species increase with increasing RH due to the associated expansion of the interlayer space. The diffusion coefficient of each species in the interlayer is smaller than its corresponding bulk value for all RH. Our results show that the diffusion coefficients of these species decrease by about one to three orders of magnitude under the extreme confinement, as compared to the bulk values. As expected, the decrease is more pronounced for the cations because of the strong electrostatic interactions between those ions and the charged clay surfaces. For all studied systems, the diffusion of CO2 increases by about an order of magnitude as clays swell from the 1W to the 2W hydration state, in agreement with previous simulations. 27,29 Moreover, the diffusion of CO2 in each of these hydration states is mostly independent of the type of cation in line with the RDFs which show no evidence of direct interaction of the interlayer cations with CO2 . Consistent with the results of the analysis of solvation energies, 57 the mobility of water in the hydration states of Na-montmorillonite is larger than those computed for the corresponding states of Ca- and Mg-montmorillonite. Our previous simulations have shown that the diffusion of CO2 in the clay interlayer is much larger than that of water because of the less hydrated environment. 37 For all cases, the diffusion coefficients of CO2 and water in the interlayer are mostly comparable in this study, which is consistent with other simulations. 27,29 This is because each CO2 molecule in the interlayer is well solvated by clay oxygens and water molecules even at low RH (see also Table 2).

3.7

Conclusions

Molecular simulations show that the molecular partitioning between montmorillonites and the variably wet scCO2 under sequestration conditions is greatly influenced by the CO2 −clay interactions. Our analysis of the adsorption process and the interlayer structure indicates that while the CO2 −clay interaction has a stabilizing effect, CO2 molecules hardly migrate into the first hydration shell of cations. These results 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


are supported by simulations 27,29,31,33 and experiments. 10,13 Artificially increasing the CO2 content in the interlayer encourages the movement of CO2 molecules into the first hydration shell of cations. 25,26,30 The fact that such a state is hardly reached is strengthened by our simulation results which, instead shows that the amount of the trapped CO2 within the hydrated interlayer generally decreases with increasing RH. Recent computational studies of very low H2 O systems reported that CO2 can solvate interlayer cations and cause a blue-shift in the asymmetric stretch vibration of CO2 . 26 Our previous investigations have shown that the diffusion of CO2 in the clay interlayer is much larger than that of water for such systems. 37 The diffusion coefficients of CO2 and water in the clay interlayer are mostly comparable in the current study because each CO2 molecule is well solvated even at low RH values. The diffusion of CO2 in each hydration state is not much dependent on the type of cation in line with the fact that the solvation of CO2 is sufficient to form a solvation shell and preclude a direct interaction of CO2 with the interlayer cations. Furthermore, the diffusion coefficients of all species generally increase with increasing RH due to the associated expansion of the interlayer distance of the clay. These findings imply that clays rich in highly charged cations are more likely to capture and mobilize CO2 in conditions close to those existing in subsurface geological reservoirs. Molecular simulations of the interactions of CO2 with clays provide key information about the possible mechanisms that leads to intercalation and diffusion of CO2 within smectite minerals. These findings should provide a basis for future studies on the effects of the long-term exposure of scCO2 to geological formations. 8,12,14 Regardless, the processes detailed here will be relevant to geological carbon storage and CO2 enhanced shale gas production.

Acknowledgments The research reported in this publication was supported by funding from King Abdullah University of Science and Technology (KAUST), Kingdom of Saudi Arabia. A. K. and A. K. N. N. gratefully acknowledge computational facilities and the MedeA



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

environment provided at KAUST. A.K. also acknowledges Prof. Abbas Firoozabadi and Dr. Zhehui Jin at Yale, and IFP-EN and Laboratory of Chemical Physics, CNRSUniversité Paris Sud (MedeA) for helpful comments on their Monte Carlo simulation codes.

Supporting Information Additional details of simulation analysis are provided in the Supporting Information.


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


References (1) Benson, S.M.; Cole, D.R. CO2 sequestration in deep sedimentary formations. Elements 2008, 4, 325-331. (2) Busch, A.; Alles, S.; Gensterblum, Y.; Prinz, D.; Dewhurst, D. N.; Raven, M. D.; Stanjek, H.; Krooss, B. M. Carbon dioxide storage potential of shales. Int. J. Greenhouse Gas Control 2008, 2, 297-308. (3) Gaus, I. Role and impact of CO2 -rock interactions during CO2 storage in sedimentary rocks. Int. J. Greenhouse Gas Control 2010, 4, 73-89. (4) Giesting, P.; Guggenheim, S.; Koster van Groos, A. F.; Busch, A. Interaction of carbon dioxide with Na-exchanged montmorillonite at pressures to 640 bar: Implications for CO2 sequestration. Int. J. Greenhouse Gas Control 2012, 8, 73-81. (5) Giesting, P.; Guggenheim, S.; Koster van Groos, A. F.; Busch, A. X-ray diffraction study of K-and Ca-exchanged montmorillonites in CO2 atmospheres. Environ. Sci. Technol. 2012, 46, 5623-5630. (6) Ilton, E. S.; Schaef, H. T.; Qafoku, O.; Rosso, K. M.; Felmy, A. R. In situ X-ray diffraction study of Na+ saturated montmorillonite exposed to variably wet super critical CO2 . Environ. Sci. Technol. 2012, 46, 4241-4248. (7) Schaef, H. T.; Ilton, E. S.; Qafoku, O.; Martin, P. F.; Felmy, A. R.; Rosso, K. M. In situ XRD study of Ca2+ saturated montmorillonite (STX-1) exposed to anhydrous and wet supercritical carbon dioxide. Int. J. Greenhouse Gas Control 2012, 6, 220 229. (8) Romanov, V. N. Evidence of irreversible CO2 intercalation in montmorillonite. Int. J. Greenhouse Gas Control 2013, 81, 220-226.



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

(9) Jeon, P. R.; Choi, J.; Yun, T. S.; Lee, C. H. Sorption equilibrium and kinetics of CO2 on clay minerals from subcritical to supercritical conditions: CO2 sequestration at nanoscale interfaces. Chem. Eng. J. 2014, 255, 705-715. (10) Loring, J.S.; Ilton, E.S.; Chen, J.; Thompson, C.J.; Martin, P.F.; Bénézeth, P.; Rosso, K.M.; Felmy, A.R.; Schaef, H.T. In situ study of CO2 and H2 O partitioning between NaMontmorillonite and variably wet supercritical carbon dioxide. Langmuir 2014, 30 (21), 6120-6128. (11) Edwards, R. W.; Celia, M. A.; Bandilla, K. W.; Doster, F.; Kanno, C. M. A model to estimate carbon dioxide injectivity and storage capacity for geological sequestration in shale gas wells. Environ. Sci. Technol. 2015, 49, 9222-9229. (12) Romanov, V.; Soong, Y.; Carney, C.; Rush, G.E.; Nielsen, B.; O’Connor, W. Mineralization of carbon dioxide: A literature review. ChemBioEng Rev. 2015, 2, 231-256. (13) Schaef, H. T.; Loring, J. S.; Glezakou, V. A.; Miller, Q. R.; Chen, J.; Owen, A. T.; Lee, M. S.; Ilton, E. S.; Felmy, A. R.; McGrail, B. P.; Thompson, C. J. Competitive sorption of CO2 and H2 O in 2:1 layer phyllosilicates. Geochim. Cosmochim. Acta 2015, 161, 248-257. (14) Busch, A.; Bertier, P.; Gensterblum, Y.; Rother, G.; Spiers, C. J.; Zhang, M.; Wentinck, H. M. On sorption and swelling of CO2 in clays. Geomech. Geophys. Geo-energ. 2016, 2, 111-130. (15) Fu, M. H.; Zhang, Z. Z.; Low, P. F. Changes in the properties of a montmorillonite-water system during the adsorption and desorption of water: Hysteresis. Clays Clay Miner. 1990, 38, 485 492. (16) Boek, E. S.; Coveney, P. V.; Skipper, N. T. Monte Carlo molecular modeling studies of hydrated Li-, Na-, and K-smectites: Understanding the role of potassium as a clay swelling inhibitor. JACS 1995, 117 (50), 12608-12617. 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


(17) Chávez-Páez, M.; Van Workum, K.; De Pablo, L.; de Pablo, J. J. Monte Carlo simulations of Wyoming sodium montmorillonite hydrates. J. Chem. Phys. 2001, 114 (3), 1405-1413. (18) Whitley, H. D.; Smith, D. E. Free energy, energy, and entropy of swelling in Cs, Na, and Srmontmorillonite clays. J. Chem. Phys. 2004, 120 (11), 5387-5395. (19) Ferrage, E.; Lanson, B.; Sakharov, B. A.; Drits, V. A. Investigation of smectite hydration properties by modeling experimental X-ray diffraction patterns: part I. montmorillonite hydration properties. Am. Mineral. 2005, 90, 1358-1374. (20) Tambach, T. J.; Bolhuis, P. G.; Hensen, E. J.; Smit, B. Hysteresis in clay swelling induced by hydrogen bonding: accurate prediction of swelling states. Langmuir 2006, 22 (3), 1223-1234. (21) Zhang, L.; Lu, X.; Liu, X.; Zhou, J.; Zhou, H. Hydration and mobility of interlayer ions of (Nax , Cay )-montmorillonite: a molecular dynamics study. J. Phys. Chem. C 2014, 118 (51), 29811-29821. (22) Teich-McGoldrick, S.L.; Greathouse, J.A.; Jove-Colon, C.F.; Cygan, R.T. Swelling properties of montmorillonite and beidellite clay minerals from molecular simulation: comparison of temperature, interlayer cation, and charge location effects. J. Phys. Chem. C 2015, 119 (36), 20880-20891. (23) Greathouse, J.A.; Hart, D.B.; Bowers, G.M.; Kirkpatrick, R.J.; Cygan, R.T. Molecular simulation of structure and diffusion at smectite-water interfaces: using expanded clay interlayers as model nanopores. J. Phys. Chem. C 2015, 119 (30), 17126-17136. (24) Michels, L.; Fossum, J.O.; Rozynek, Z.; Hemmen, H.; Rustenberg, K.; Sobas, P.A.; Kalantzopoulos, G.N.; Knudsen, K.D.; Janek, M.; Plivelic, T.S.; da Silva, G.J. Intercalation and retention of carbon dioxide in a smectite clay promoted by interlayer cations. Scientific reports 2015, 5, 8775. 21 ACS Paragon Plus Environment


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

Page 22 of 36

(25) Lee, M. S.; McGrail, B. P.; Glezakou, V. A. Microstructural response of variably hydrated Ca-rich montmorillonite to supercritical CO2 . Environ. Sci. Technol. 2014, 48 (15), 8612-8619. (26) Myshakin, E. M.; Saidi, W. A.; Romanov, V. N.; Cygan, R. T.; Jordan, K. D. Molecular dynamics simulations of carbon dioxide intercalation in hydrated Na-montmorillonite. J. Phys. Chem. C 2013, 117 (21), 11028-11039. (27) Botan, A.; Rotenberg, B.; Marry, V.; Turq, P.; Noetinger, B. Carbon dioxide in montmorillonite clay hydrates: thermodynamics, structure, and transport from molecular simulation. J. Phys. Chem. C 2010, 114 (35), 14962-14969. (28) Jin, Z.; Firoozabadi, A. Effect of water on methane and carbon dioxide sorption in clay minerals by Monte Carlo simulations. Fluid Phase Equilib. 2014, 382, 10-20. (29) Makaremi, M., Jordan, K.D., Guthrie, G.D. and Myshakin, E.M., 2015. Multiphase Monte Carlo and molecular dynamics simulations of water and CO2 intercalation in montmorillonite and beidellite. J. Phys. Chem. C 2015, 119 (27), 15112-15124. (30) Sena, M.M.; Morrow, C.P.; Kirkpatrick, R.J.; Krishnan, M. Supercritical carbon dioxide at smectite mineral-water interfaces:

molecular dynamics and

adaptive biasing force investigation of CO2 /H2 O mixtures nanoconfined in Namontmorillonite. Chem. Mater. 27, 20, 6946-6959. (31) Rao, Q.; Leng, Y. Molecular understanding of CO2 and H2 O in montmorillonite clay interlayer under CO2 geological sequestration conditions. J. Phys. Chem. C 2016, 120, 2642-2654. (32) Rao, Q.; Leng, Y. Effect of layer charge on CO2 and H2 O intercalations in swelling clays. Langmuir 2016, 32, 11366-11374.


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


(33) Yang, W.; Zaoui, A. Capture and sequestration of CO2 in the interlayer space of hydrated calcium montmorillonite clay under various geological burial depth. Physica A 2016, 449, 416-425. (34) Kumar, N.A.; Seidel, C. Polyelectrolyte brushes with added salt. Macromolecules 2005, 38 (22), 9341-9350. (35) Nair, A.K.N.; Uyaver, S.; Sun, S. Conformational transitions of a weak polyampholyte. J. Chem. Phys. 2014, 141 (13), 134905. (36) Kadoura, A.; Nair, A. K. N.; Sun, S. Adsorption of carbon dioxide, methane and their mixture by montmorillonite in the presence of water. Microporous Mesoporous Mater. 2016, 225, 331-341. (37) Kadoura, A.; Nair, A. K. N.; Sun, S. Molecular dynamics simulations of carbon dioxide, methane, and their mixture in montmorillonite clay hydrates J. Phys. Chem. C 2016, 120, 1251712529. (38) Middleton, R.S.; Keating, G.N.; Stauffer, P.H.; Jordan, A.B.; Viswanathan, H.S.; Kang, Q.J.; Carey, J.W.; Mulkey, M.L.; Sullivan, E.J.; Chu, S.P.; Esposito, R.; Meckel, T. A. The cross-scale science of CO2 capture and storage: from pore scale to regional scale. Energy Environ. Sci. 2012, 5, 7328-7345. (39) Rotenberg, B.; Marry, V.; Salanne, M.; Jardat, M.; Turq, P. Multiscale modelling of transport in clays from the molecular to the sample scale. C. R. Geosci. 2014, 346, 298-306. (40) Lowenstein, W. The distribution of aluminum in the tetrahedra of silicates and aluminates. Am. Miner. 1954, 39, 92-96. (41) Frenkel, D.; Smit, B. Understanding Molecular Simulation: From Algorithms to Applications; Academic Press: London, 2002.



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

(42) Cygan, R. T.; Liang, J. J.; Kalinichev, A. G. Molecular models of hydroxide, oxyhydroxide, and clay phases and the development of a general force field. J. Phys. Chem. B 2004, 108 (4), 1255-1266. (43) Cygan, R. T.; Romanov, V. N.; Myshakin, E. M. Molecular simulation of carbon dioxide capture by montmorillonite using an accurate and flexible force field. J. Phys. Chem. C 2012, 116 (24), 13079-13091. (44) Berendsen, H. J. C.; Grigera, J. R.; Straatsma, T. P. The missing term in effective pair potentials. J. Phys. Chem. 1987, 91, 6269-6271 (45) 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 (31), 12021-12024. (46) Koneshan, S.; Rasaiah, J. C.; LyndenBell, R. M.; Lee, S. H. Solvent structure, dynamics, and ion mobility in aqueous solutions at 25 ◦ C J. Phys. Chem. B 1998, 102 (21), 4193-4204. (47) ˚ Aqvist, J. Ion-water interaction potentials derived from free energy perturbation simulations. J. Phys. Chem. 1990, 94 (21), 8021-8024. (48) Errington, J. R.; Panagiotopoulos, A. Z. A fixed point charge model for water optimized to the vapor-liquid coexistence properties. J. Phys. Chem. B 1998, 102 (38), 7470-7475. (49) Plimpton, S. Fast parallel algorithms for short-range molecular dynamics. J. Comput. Phys. 1995, 117, 1-19. (50) Ferrage, E.; Lanson, B.; Michot, L. J.; Robert, J. L. Hydration properties and interlayer organization of water and ions in synthetic Na-smectite with tetrahedral layer charge Part 1. Results from X-ray diffraction profile modeling. J. Phys. Chem. C 2010, 114, 4515-4526.


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


(51) Dazas, B.; Lanson, B.; Delville, A.; Robert, J.L.; Komarneni, S.; Michot, L.J.; Ferrage, E. Influence of tetrahedral layer charge on the organization of interlayer water and ions in synthetic Na-saturated smectites. J. Phys. Chem. C 2015, 119 (8), 4158-4172. (52) Young, D. A.; Smith, D. E. Simulations of clay mineral swelling and hydration: Dependence upon interlayer ion size and charge. J. Phys. Chem. B 2000, 104, 9163-9170. (53) Myshakin, E. M.; Makaremi, M.; Romanov, V. N.; Jordan, K. D.; Guthrie, G. D. Molecular dynamics simulations of turbostratic dry and hydrated montmorillonite with intercalated carbon dioxide. J. Phys. Chem. A 2014, 118 (35), 7454-7468. (54) Saharay, M.; Balasubramanian, S. Evolution of intermolecular structure and dynamics in supercritical carbon dioxide with pressure: An ab initio molecular dynamics study J. Phys. Chem. B 2007, 111 (2), 387392. (55) Ishii, R.; Okazaki, S.; Okada, I.; Furusaka, M.; Watanabe, N.; Misawa, M.; Fukunaga, T. Density dependence of structure of supercritical carbon dioxide along an isotherm J. Chem. Phys. 1996, 105 (16), 7011-7021. (56) Glezakou, V. A.; Rousseau, R.; Dang, L. X.; McGrail, B. P. Structure, dynamics and vibrational spectrum of supercritical CO2 /H2 O mixtures from ab initio molecular dynamics as a function of water cluster formation. Phys. Chem. Chem. Phys. 2010, 12 (31), 8759-8771. (57) Criscenti, L. J.; Cygan, R. T. Molecular Simulations of Carbon Dioxide and Water: Cation Solvation. Environ. Sci. Technol. 2013, 47, 8794. (58) Versteeg, G. F.; Van Swaalj, W. Solubility and Diffusivity of Acid Gases (CO2 , N2 O) in Aqueous Alkanolamine Solutions. J. Chem. Eng. Data 1988, 33, 29-34. (59) Holz, M.; Heil, S. R.; Sacco, A. Temperature-dependent self-diffusion coefficients



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

of water and six selected molecular liquids for calibration in accurate 1 H NMR PFG measurements. Phys. Chem. Chem. Phys. 2000, 2, 47404742. (60) Harned, H. S.; Owen, B. B. The Physical Chemistry of Electrolytic Solutions; Reinhold Publishing Corporation: New York, 1950. (61) Bastug, T.; Kuyucak, S. Temperature dependence of the transport coefficients of ions from molecular dynamics simulations. Chem. Phys. Lett. 2005, 408, 84-88.


Page 26 of 36

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


Table 1: Approximate relative humidity (RH) ranges and the corresponding hydration states/basal d−spacings observed experimentally and employed in our GCMC simulations for clay minerals exposed to wet scCO2 at T = 323.15 K and a bulk pressure of 90 bar.

Clay type

RH (%)

Na-montmorillonite

Ca-montmorillonite

Mg-montmorillonite

hydration state/basal d−spacing Experiment 10,13

Simulation (˚ A)

0 – 20

0W

10

20 – 60

1W

12

60 – 100

2W

15

0 – 20

1W

12

20 – 100

2W

15

0 – 20

1W

12

20 – 100

2W

15



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

Page 28 of 36

Table 2: Coordination numbers of water oxygen, clay oxygen, CO2 oxygen, and CO2 carbon around CO2 carbon for clay minerals exposed to wet scCO2 at T = 323.15 K and a bulk pressure of 90 bar.

Clay type

RH (%)

Coordination number Cc –Ow

Cc –Os

Cc –Oc

Cc –Cc

30 (1W)

4.02

13.36

3.69

1.81

60 (1W)

4.07

13.03

–

–

80 (2W)

9.10

5.17

–

–

5 (1W)

3.57

13.03

5.01

2.46

60 (2W)

7.33

4.72

–

–

80 (2W)

9.03

5.45

–

–

5 (1W)

3.41

12.45

5.13

2.52

Mg-montmorillonite 60 (2W)

7.72

5.21

–

–

80 (2W)

9.55

5.23

–

–

Na-montmorillonite

Ca-montmorillonite


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


Table 3: Coordination numbers of water oxygen, clay oxygen, and CO2 oxygen around corresponding cations for clay minerals exposed to wet scCO2 at T = 323.15 K and a bulk pressure of 90 bar.

Clay type

RH (%)

Coordination number Ion–Ow

Ion–Os

Ion–Oc

30 (1W)

3.53

1.56

0.09

60 (1W)

3.64

1.46

0.06

80 (2W)

5.03

0.62

0.04

5 (1W)

4.70

0.64

0.00

60 (2W)

7.77

0.06

0.00

80 (2W)

7.51

0.42

0.00

5 (1W)

4.54

1.43

0.00

Mg-montmorillonite 60 (2W)

5.99

0.01

0.00

80 (2W)

6.00

0.00

0.00

Na-montmorillonite

Ca-montmorillonite



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: Equilibrium snapshot of Mg-montmorillonite in contact with wet scCO2 (RH of 60%) at T = 323.15 K and a bulk pressure of 90 bar. The basal d-spacing is 15 ˚ A. Color code: O, red; H, white; Si, yellow; Al, light blue; Mg, light green; C, black.


Page 30 of 36

12

(a)

8

4

0 0

20

40

60

80

100

Relative humidity (%) CO2 uptake per O20 (OH)4

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


Water uptake per O20 (OH)4

Page 31 of 36

1

(b)

0.8 0.6 0.4 0.2 0 0

20

40

60

80

100

Relative humidity (%) Figure 2: Adsorbed amounts of interlayer (a) H2 O and (b) CO2 as a function of RH observed experimentally 10,13,31 (open symbols) and calculated from GCMC simulations (corresponding solid symbols) for Na- (circles), Ca- (squares), and Mg-montmorillonite (diamonds) each exposed to wet scCO2 at T = 323.15 K and a bulk pressure of 90 bar. If not shown, simulation error bars are smaller than the size of the symbols.



CO2 density (˚ A−3 )

0.04

(a)

RH = 30 % (1W) RH = 60 % (1W) RH = 80 % (2W)

0.03 0.02 0.01 0 0

2

4

6

8

A) z (˚


0.04

(b)

RH = 5 % (1W) RH = 60 % (2W) RH = 80 % (2W)

0.03 0.02 0.01 0 0

2

4

6

8

A) z (˚ 0.04


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

Page 32 of 36

(c)

RH = 5 % (1W) RH = 60 % (2W) RH = 80 % (2W)

0.03 0.02 0.01 0 0

2

4

6

8

A) z (˚ Figure 3: Equilibrium distributions of CO2 molecules in the interlayers of (a) Na-, (b) Ca-, and (c) Mg-montmorillonite each exposed to wet scCO2 at T = 323.15 K and a bulk pressure of 90 bar. The origin corresponds to the clay surface oxygen. 32 ACS Paragon Plus Environment

Page 33 of 36

2

g (r)

(a)

RH = 30 % (1W) RH = 60 % (1W) RH = 80 % (2W)

1

0 0

2

4

6

8

10

A) r (˚ 2

g (r)

(b)

RH = 5 % (1W) RH = 60 % (2W) RH = 80 % (2W)

1

0 0

2

4

6

8

10

A) r (˚ 2 (c)

g (r)

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


RH = 5 % (1W) RH = 60 % (2W) RH = 80 % (2W)

1

0 0

2

4

6

8

10

A) r (˚ Figure 4: RDFs of CO2 carbon−clay oxygen for (a) Na-, (b) Ca-, and (c) Mgmontmorillonite each exposed to wet scCO2 at T = 323.15 K and a bulk pressure of 90 bar. 33 ACS Paragon Plus Environment


3 (a)

RH = 30 % (1W) RH = 60 % (1W) RH = 80 % (2W)

g (r)

2

1

0 0

2

4

6

8

10

A) r (˚ 3 (b)

RH = 5 % (1W) RH = 60 % (2W) RH = 80 % (2W)

g (r)

2

1

0 0

2

4

6

8

10

A) r (˚ 3 (c)

RH = 5 % (1W) RH = 60 % (2W) RH = 80 % (2W)

2

g (r)

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

Page 34 of 36

1

0 0

2

4

6

8

10

A) r (˚ Figure 5: RDFs of ion−CO2 oxygen for (a) Na-, (b) Ca-, and (c) Mg-montmorillonite each exposed to wet scCO2 at T = 323.15 K and a bulk pressure of 90 bar.


Page 35 of 36

10 0 10 -1

Dxy /Dbulk

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


10 -2 10 -3 10 -4 0

20

40

60

80

100

Relative humidity (%) Figure 6: Normalized diffusion coefficients of ions (black-filled symbols), H2 O (open symbols), and CO2 (gray-filled symbols) as a function of RH for Na- (circles), Ca(squares), and Mg-montmorillonite (diamonds) each exposed to wet scCO2 at T = 323.15 K and a bulk pressure of 90 bar.



CO2 uptake per O20 (OH)4

Clays exposed to wet scCO2 at T = 323.15 K & P = 90 bar H2 O uptake per O20 (OH)4

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

Page 36 of 36

12

8

4

0 0

20

40

60

80

100

1 0.8 0.6 0.4 0.2 0 0

Na-montmorillonite

20

40

60

80

100

Relative humidity (%)

Relative humidity (%)

Ca-montmorillonite

Mg-montmorillonite

GCMC

GCMC

GCMC

Expt.

Expt.

Expt.

TOC Graphic


Molecular Simulation Study of Montmorillonite in Contact with Variably

Recommend Documents