Influence of structural heterogeneity of nanoporous sorbent walls on hydrogen storage

Heterogeneity is an ubiquitous aspect of adsorption, often modifying substantially the observed behaviour of the adsorbate-adsorbent system. In this paper, the influence of heterogeneity is explicitly analyzed for the case of the adsorption of molecular hydrogen onto nanoporous carbon. Grand Canonical Monte Carlo simulations were used to study the mechanism of adsorption in the models of the adsorbate that include both energetic and structural modifications of graphene-based slit pores. In particular, a partial substitution of carbons by boron modifies both the symmetry of the energy landscape and the strength of hydrogen physisorption; which results in considerable increases of the amount of adsorbed gas without major modification of the mechanism of adsorption. Additional heterogeneity arises from structural modifications of the adsorbent by neutron irradiation of boron-doped samples, where the boron fission products generate additional surface area for adsorption. Simulations of adsorption in such pores show that hydrogen uptake is strongly dependent on the chemical nature of the modified pore walls.     

Quantum states of rigid linear rotors confined in a slit pore: quantum sieving of hydrogen isotopes and extreme one dimensional confinement

The quantum states of molecular hydrogen isotopes confined in carbon slit pores of varying width are calculated by direct diagonalization of the single particle Hamiltonian, evaluated with a model potential. The results are used to calculate the zero-pressure isotopic selectivity that can be expected from adsorption on these nanostructures, and the contribution from the rotational degrees of freedom is analysed in detail. It is shown that the role of the rotational degrees of freedom is far from negligible, resulting in values of the isotopic separation 8 to 50 times higher than those calculated using spherically symmetric models for the hydrogen molecule. It is also shown that if the pores are narrow enough, the molecular axis in the ground state is found preferentially aligned along the slit planes, and the conditions when this happens will be analysed in detail. This behaviour, termed “extreme one dimensional confinement”, is analogous to a similar behaviour observed in carbon nanotubes.     

Phase behavior and fluid-solid surface tension of argon in slit pores and carbon nanotubes

Phase behaviors of argon in several types of cylindrical and slit pores are examined by grand-canonical Monte Carlo simulations. Condensation processes in single- and multi-walled carbon nanotubes along with those in hard-wall tubes are compared. Effects of the pore size on pressure-tensor components, the fluid-wall surface tension, and the adsorption are also compared for the different fluid-pore interactions. The chemical potential at which the fluid begins to condense in the single-walled nanotube is greater than that in the multi-walled nanotube by an amount nearly equal to the difference in the potential-well depth of the fluid-pore interaction, and the adsorption isotherms overlap each other almost completely for narrow pores and partially for wider pores. Similar analyses are performed for slit pores of two different hydrocarbon models.     

Hydrogen adsorption in slit-shaped pores of carbon adsorbents

The hydrogen adsorption in slit-shaped pores of carbon adsorbents is investigated using the density functional theory. Hydrogen adsorption in the gap between two monocarbon (graphene) walls is calculated for 20.33, 77, and 200 K. At T = 20.38 and 77 K, our data on the hydrogen storage capacities in slit pores are in good agreement with the results of A.A. Fomkin and V.A. Sinitsyn [23], obtained using Dubinin’s volume pore filling theory and the standard data on benzene adsorption. Under supercritical conditions, the adsorption capacity in the modeling adsorbents is underestimated by half.     

Characterization of PSD of activated carbons by using slit and triangular pore geometries

A mixed geometry model for activated carbons, representing the porous space as a collection of an undetermined proportion of slit and triangular pores, is developed, evaluated theoretically and applied to the characterization of a controlled series of samples of activated carbon obtained from the same precursor material. A method is proposed for the determination of the Pore Size Distribution (PSD) for such a mixed geometry model, leading to the unique determination of the proportion of pores of the two geometries fitting adsorption data. By using the Grand Canonical Monte Carlo (GCMC) simulation method in the continuum space, families of N2 adsorption isotherms are generated both for slit and triangular geometry corresponding to different pore sizes. The problem of the uniqueness in the determination of the PSD by fitting an adsorption isotherm using the mixed geometry model is then discussed and the effects of the addition of triangular pores on the PSD are analyzed by performing a test where the adsorption isotherm corresponding to the known PSD is generated and used as the “experimental” isotherm. It is found that a pure slit geometry model would widen the PSD and shift it to smaller sizes, whereas a pure triangular geometry model would produce the opposite effect. The slit and triangular geometry families of isotherms are finally used to the fit experimental N₂ adsorption data corresponding to a family of activated carbons obtained from coconut shells through a one-step chemical activation process with phosphoric acid in air, allowing for the determination of the micropore volume, the proportion of slit and triangular pores and the PSD corresponding to the mixed geometry. The same experimental data were fit using both the conventional slit pore model and the mixed geometry model. From the analysis of the effect of different preparation procedures on the resulting PSDs, it is concluded that the proposed mixed geometry model may probably better capture the morphology and energetics of activated carbons prepared by chemical activation under mild temperatures.     

Adsorption of water—methanol mixtures in carbon and aluminosilicate pores: a molecular simulation study

A theoretical study is reported of the adsorption behaviour of water—methanol mixtures in slit carbon and in uncharged alumino-silicate micropores. The adsorption isotherms are obtained for a pore of width of 2 nm and at a temperature of 298 K from grand canonical ensemble Monte Carlo simulations. The results show that the graphite and uncharged silicate surfaces are covered by a dense layer of flatly adsorbed water and methanol molecules having weaker hydrogen bonding. In the interior of the pore, the fluid exhibits bulk-like behaviour with a stronger hydrogen bonded structure. Solvation forces are also calculated as a function of pore size. The positive values found for the solvation force for all pore sizes reflect the hydrophobic interactions of the mixture with the carbon and uncharged alumino-silicate walls.     

Molecular simulation study of triangle-well fluids confined in slit pores

Grand canonical Monte Carlo simulations are used to study the behaviour of triangle-well (TW) fluids with variable well widths confined inside slit pores. The effect of individual factors influencing the properties of confined fluids such as fluid–fluid interactions, pore size and pore wall–fluid interactions are obtained using simulations as it is difficult to experimentally determine the same. An interesting observation of this study is that inside the narrow pore of slit height h* = 5 at the high-pressure condition of P* = 0.8, for the TW fluid with long-range attraction or for the fluid at a low temperature for even a short-range attraction, the density profiles show layering such that there is a sticking tendency of the particles at centre, while there is a depletion of particles near the wall (as the layers at the centre have higher density peak heights than near the walls).     

The adsorption of water in finite carbon pores

Grand canonical Monte Carlo simulations were applied to the adsorption of SPCE model water in finite graphitic pores with different configurations of carbonyl functional groups on only one surface and several pore sizes. It was found that almost all finite pores studied exhibit capillary condensation behaviour preceded by adsorption around the functional groups. Desorption showed the reverse transitions from a filled to a near empty pore resulting in a clear hysteresis loop in all pores except for some of the configurations of the 1.0?nm pore. Carbonyl configurations had a strong effect on the filling pressure of all pores except, in some cases, in 1.0?nm pores. A decrease in carbonyl neighbour density would result in a higher filling pressure. The emptying pressure was negligibly affected by the configuration of functional groups. Both the filling and emptying pressures increased with increasing pore size but the effect on the emptying pressure was much less. At pressures lower than the pore filling pressure, the adsorption of water was shown to have an extremely strong dependence on the neighbour density with adsorption changing from Type IV to Type III to linear as the neighbour density decreased. The isosteric heat was also calculated for these configurations to reveal its strong dependence on the neighbour density. These results were compared with literature experimental results for water and carbon black and found to qualitatively agree.     

Further observations on the polarized absorption spectrum of benzene crystals at 2000 Å

The structure of a hard sphere fluid confined by model slit and cylindrical pores is investigated. Results from grand canonical Monte Carlo (GCMC) simulations and from the hypernetted chain/mean spherical approximation (HNC/MSA) equation are reported. GCMC results are compared with those from the HNC/MSA equation, and agreement is good. The effect of confinement on liquids at the same chemical potentials is that the absorption of the hard sphere fluid into the pores decreases with increasing confinement, i.e., when going from planar to cylindrical geometry or by narrowing the pores. The adsorption on the pore walls has, in general, the opposite behaviour. For high bulk concentrations and certain values of cylindrical pore diameter the concentration profile is higher at the centre of the pore than next to the pore wall. A very strong, but continuous, transition occurs in the concentration profile, as a function of the cylinder's diameter. These results could be of some interest in catalysis studies.     

The effects of curvature and surface heterogeneity on the adsorption of water in finite length carbon nanopores: a computer simulation study

The effects of pore curvature and surface heterogeneity on the adsorption of water on a graphitic surface at 298 K were investigated using a Grand Canonical Monte Carlo (GCMC) simulation. Slit and cylindrical pores are used to study the curvature effects. To investigate the surface heterogeneity the functional group and the structural defect on the surface were specifically considered. The hydroxyl group (OH) is used as a model for the functional group and the water potential model proposed by Müller et al. is used to calculate the water interaction. For the homogeneous cylinder, the pore filling occurs at a pressure lower than the saturation pressure of the water model, while it is greater in the case of homogeneous slit pore. The size of hysteresis loop is more sensitive to the length of cylinder than that of the slit, and it increases with decreasing pore length. The isotherms of water in cylindrical pores are found to depend on the position and the concentration of the functional group. The pore filling pressure is lower with an increased number and/or with the position of the functional group. The structural defect shows significant effects on the adsorption isotherm in shifting to a lower pore filling pressure when it is located at a position away from the pore entrance. The adsorption of water on the heterogeneous surface was studied and it was found that the simulated isotherm can describe the behaviour of water on Graphitized Thermal Carbon Black (GTCB) satisfactorily. The water cluster grows mostly along the surface for the case of finite extent surface, while for the slit the pore grows in all directions but the preference is a direction perpendicular to the pore wall. Reasons for the direction of growth will be discussed.     

Simulation of hydrogen adsorption in carbon nanotubes

Computer simulations are reported of hydrogen adsorption in multi-walled carbon nanotubes (MWNTs) and single-walled carbon nanotubes (SWNTs). The gas-solid interaction was modelled both as pure dispersion forces and also with a hypothetical model for chemisorption introduced in a previous paper (CRACKNELL, R., F., 2001, Phys. Chem. chem. Phys., 3, 2091). A two-centre model for hydrogen was employed and the grand canonical Monte Carlo methodology was used throughout. Uptake of hydrogen in the internal space of a carbon nanotube is predicted to be lower than in the optimal graphitic nanofibre with slitlike pores (provided the gas-solid potential is consistent). Part of the difference arises from the assumption of pore surface area used in converting the raw simulation data to gravimetric adsorption; however, the majority of the differences can be attributed to the curvature of the pore. This reduces the uptake of hydrogen (on a gravimetric basis) in spite of deepening the potential minimum inside the pore associated with dispersion forces. It is concluded that for the uptake of hydrogen in SWNTs of 5–10% reported by Heben (DILLON, A. C., JONES, K. M., BEKKEDAHL, T. A., KIANG, C. H., BETHUNE, D. S., AND HEBEN, M. J., 1997, Nature, 386, 377), gas-solid forces other than dispersion forces are required and most of the adsorption must occur in the interstices between SWNTs.     

氢及其同位素宽区状态方程的核量子效应

应用大规模第一性原理分子动力学数值模拟方法,系统研究了氢及其同位素在极端条件下的分子解离规律及状态方程,并给出了参数化的解函数拟合公式。结合分子、原子流体量子振动的一阶修正,揭示了氢及其同位素的核量子效应,解析获得了氢、氘分子在温稠密区域分子解离规律的差别。由第一性原理状态方程给出了雨贡纽曲线,与气炮、化学炸药、磁驱动、高能激光等实验数据相符合,并详细讨论了由核量子效应导致的氢、氘雨贡纽曲线的同位素效应。     

Applicability of molecular dynamics method to the pressure-driven gas flow in finite length nano-scale slit pores

This study investigates the applicability of the molecular dynamics (MD) method to the pressure-driven gas flow in finite length nano-scale slit pores. The reflecting particle membrane is introduced to induce a pressure difference between the inlet and outlet. The flow properties are compared with those of the Burnett equations. The inlet and outlet pressures, as well as the mass flow rate in these two simulations are maintained the same by adjusting the tangential momentum accommodation coefficient in the Burnett simulation, which is found to be between 0.4 and 0.5. Qualitative and quantitative agreements are observed between the MD and Burnett simulation results in the bulk of the pore for both streamwise distributions and cross-section profiles. The MD simulation shows an advantage in the near-wall region, in which the wall force field dominates flow behaviour. This study indicates that MD simulation can be used to describe the pressure-driven gas flow characteristics in finite length nano-scale slit pores.     

Molecular dynamics simulations of surface-controlled phase separation of binary fluid mixtures confined in slit pores - Online only

When binary mixtures are confined into nanoscopic slit pores, an intricate interplay between surface enrichment (wetting) of one component and lateral phase separation occurs. After a brief review of the static equilibrium phase diagram of such systems, a discussion of the kinetics of phase separation is given. Considering quenches from an initially homogeneous distribution of the two species in the slit, it is shown by molecular dynamics simulation that typically in the initial stages a stratified structure develops, with enrichment layers of the preferred component at the walls of the slit pore. Then this laterally homogeneous structure breaks up into domains, which coarsen with time according to a power law with a 2/3 exponent. This growth law must be attributed to a hydrodynamic mechanism, since corresponding simulations of a diffusive Ginzburg-Landau model yield an exponent of 1/3 only. The relation to spinodal decomposition in d=2 space dimensions is briefly discussed.     

Adsorption isotherm of a Lennard-Jones nitrogen in a carbon slitlike pore

A density functional perturbation approximation based both on second-order perturbation theory and on the pore average density has been proposed to study the adsorption hysteresis of nitrogen in a carbon slit pore. The main advantage of the present approximation is that it is computationally much simpler than the original density functional approximation based on the second-order perturbation theory of liquids, and can be applied to several model fluids confined in a strong external field in order to study their structural and thermodynamic properties. The calculated adsorption hysteresis for the confined Lennard-Jones nitrogen is in very good agreement with computer simulation, even if its accuracy slightly deteriorates for the desorption branch. The calculated equilibrium particle density distributions also compare well with computer simulations, and are better than those of a density functional theory based on the so-called mean-field approximation.     

Quantum effect induced reverse kinetic molecular sieving in microporous materials

We report kinetic molecular sieving of hydrogen and deuterium in zeolite rho at low temperatures, using atomistic molecular dynamics simulations incorporating quantum effects via the Feynman-Hibbs approach. We find that diffusivities of confined molecules decrease when quantum effects are considered, in contrast with bulk fluids which show an increase. Indeed, at low temperatures, a reverse kinetic sieving effect is demonstrated in which the heavier isotope, deuterium, diffuses faster than hydrogen. At 65 K, the flux selectivity is as high as 46, indicating a good potential for isotope separation.     

Effect of confinement in nano-porous materials on the solubility of a supercritical gas

By combining Gibbs Ensemble Monte Carlo simulations and density functional theory, we investigate the influence of confinement in a slit-shaped carbon pore on the solubility of a supercritical solute gas in a liquid solvent. In the cases studied here, competing adsorption of the solvent and solute determines whether the solubility is enhanced or suppressed for larger pores. We find that the solubility in the confined system is strongly dependent on pore width, and that molecular packing effects are important for small pore widths. In addition, the solubility decreases on increase in the temperature, as for the bulk mixture, but the rate of decrease is greater in the pore due to a decrease in the partial molar enthalpy of the solute in the pore; this effect becomes greater as pore width is decreased. The solubility is increased on increasing the bulk pressure of the gas in equilibrium with the pore, and obeys Henry's law at lower pressures. However, the Henry constant differs significantly from that for the bulk mixture, and the range of pressure over which Henry's law applies is reduced relative to that for the bulk mixture. The latter observation indicates that solute–solute interactions become more important in the pore than for the bulk at a given bulk pressure. Finally, we note that different authors use different definitions of the solubility in pores, leading to some confusion over the reported phenomenon of ‘oversolubility’. We recommend that solubility be defined as the overall mole fraction of solute in the pores, since it takes into account the increase in density of the solvent in the pores, and avoids ambiguity in the definition of the pore volume.     

Molecular Simulation of H<Subscript>2</Subscript>O,CO<Subscript>2</Subscript>, and CH<Subscript>4</Subscript> Adsorption in Coal Micropores

Based on the chemical model of coal, slit micropores with different pore sizes are established and structures are optimized in the software of materials studio. As the temperature rises, absolute adsorption capacities of H₂O are slightly affected, while absolute adsorption capacities of CO₂ and CH₄ gradually decrease. As the fugacity rises, excess adsorption curves of CO₂ experience increase-decrease-gentle three stages, while the curves of CH₄ gradually decrease. With the increase of pore size, adsorption capacities of H₂O increase, while adsorption capacities of CO₂ and CH₄ gradually decrease. H₂O firstly adsorbs on the oxygen-containing functional group, so the walls of pore are the preferential area for H₂O, while CO₂ and CH₄ choose to adsorb on–C–C–, therefore the walls are the primary area for CO₂ and CH₄. Strong potential in micropores and hydrogen bond among water molecules will promote the water adsorption, while the adsorptions of CO₂ and CH₄ are only induced by the Van der Waals interaction, but the difference between adsorption density and bulk density of CO₂ and CH₄ decides the change of excess adsorption capacity.     

Importance of pore length and geometry in the adsorption/desorption process: a molecular simulation study

Discrete potentials can describe properly the liquid vapour boundary that is necessary to model the adsorption of gas molecules in mesoporous systems with computer simulations. Although there are some works in this subject, the simulations are still highly time-consuming. Here we show that an efficient alternative is to use the three-dimensional Ising model, which allows one to model large systems, with geometries as complex as required that accurately represent the liquid vapour boundary. In particular, we report molecular simulations of cylindrical pores of two different geometries, using a discrete potential. The effect of the length of the pore in the hysteresis loop for a finite pore and infinite one is studied in detail. Lastly, we compare our predictions with experimental results and find excellent agreement between the area of the hysteresis loop predicted for the finite pore and that found in adsorption/desorption experiments.     

Density profile of water confined in cylindrical pores in MCM-41 silica

Recently, water absorbed in the porous silica material MCM-41-S15 has been used to demonstrate an apparent fragile to strong dynamical crossover on cooling below ～220 K, and also to claim that the density of confined water reaches a minimum at a temperature around 200 K. Both of these behaviours are purported to arise from the crossing of a Widom line above a conjectured liquid-liquid critical point in bulk water. Here it is shown that traditional estimates of the pore diameter in this porous silica material (of order 15 ?) are too small to allow the amount of water that is observed to be absorbed by these materials (around 0.5 g H(2)O/g substrate) to be absorbed only inside the pore. Either the additional water is absorbed on the surface of the silica particles and outside the pores, or else the pores are larger than the traditional estimates. In addition the low Q Bragg intensities from a sample of MCM-41-S15 porous silica under different dry and wet conditions and with different hydrogen isotopes are simulated using a simple model of the water and silica density profile across the pore. It is found the best agreement of these intensities with experimental data is shown by assuming the much larger pore diameter of 25 ? (radius 12.5 ?). Qualitative agreement is found between these simulated density profiles and those found in recent empirical potential structure refinement simulations of the same data, even though the latter data did not specifically include the Bragg peaks in the structure refinement. It is shown that the change in the (100) peak intensity on cooling from 300 to 210 K, which previously has been ascribed to a change in density of the confined water on cooling, can equally be ascribed to a change in density profile at constant average density. It is further pointed out that, independent of whether the pore diameter really is as large as 25 ? or whether a significant amount of water is absorbed outside the pore, the earlier reports of a dynamic crossover in supercooled confined water could in fact be a crystallization transition in the larger pore or surface water.