Temperature-programmed desorption curves for ammonia (NH3-TPD) were predicted successfully using grand canonical ensemble Monte Carlo simulation methods and force field parameters derived from quantum mechanical ab initio data. This approach provides a means to study the relationship between the structure and acidity of zeolites at the molecular level. Analysis of the predicted NH3-TPD curves reveals that zeolite pore size is a critical factor influencing the curve shape. The so-called weak and strong acids of zeolites, which give rise to two peaks in the TPD curve, are roughly related to the interactions among ammonia molecules in the pores and to the interactions between ammonia molecules and the zeolite pore walls, respectively. However, the NH3-TPD curves may still show a double peak feature without any strong acid centers if the pore is such a size that additional strain is put on the NH3 hydrogen-bond network.

Calcium and magnesium ions play important roles in many physicochemical processes. To facilitate the investigation of phenomena related to these ions that occur over large length and time scales, a coarse-grained force field (CGFF) is developed for MgCl2 and CaCl2 aqueous solutions. The ions are modeled by CG beads with characteristics of hydration shells. To accurately describe the nonideal behavior of the solutions, osmotic coefficients in a wide range of concentrations were used as guidance for parametrization. The osmotic coefficients were obtained from the chemical potential increments of water calculated using the Bennett acceptance ratio (BAR) method. The result CGFF accurately reproduces experimental osmotic coefficients, densities, surface tensions, and cation–anion separations of calcium chloride and magnesium chloride solutions at molalities up to 3.0 mol/kg. As a preliminary application, the force field is applied to simulate aggregations of sodium dodecyl sulfate (SDS) in calcium chloride solution, and the simulation results are consistent with experimental observations.

The free energy based Lennard-Jones 12-6 (FE-12-6) coarse-grained (CG) force field developed for alkanes1 has been extended to model small molecules of light hydrocarbons (methane, ethane, propane, butane, and isobutane), nitrogen, oxygen, and carbon dioxide. The adjustable parameters of the FE-12-6 potential are determined by fitting against experimental vapor–liquid equilibrium (VLE) curves and heat of vaporization (HOV) data for pure substance liquids. Simulations using the optimized FE-12-6 parameters correctly reproduced experimental measures of the VLE, HOV, density, vapor pressure, compressibility, critical point, and surface tension for pure substances over a wide range of thermodynamic states. The force field parameters optimized for pure substances were tested on methane/butane, nitrogen/decane, and carbon dioxide/decane binary mixtures to predict their vapor–liquid equilibrium phase diagrams. It is found that for nonpolar molecules represented by different sized beads, a common scaling factor (0.08) that reduces the strength of the interaction potential between unlike beads, generated using Lorentz–Berthelot (LB) combination rules, is required to predict vapor–liquid phase equilibria accurately.

An automatic high-throughput computing (HTC) procedure is implemented for calculating vapor-liquid equilibrium (VLE) curves of binary systems using molecular dynamics simulation and coarse-grained force field. The HTC procedure builds simulation models, carries out the simulations, and validates the simulations automatically. The procedure is demonstrated by calculating the VLE curves for 16 binary mixtures of carbon dioxide and n-alkanes on 856 state points in the temperature range of 277.2–420.0 K and pressure range of 0.3–25.0 MPa. The averaged uncertainties in predictions are 0.119 MPa for pressure, 0.007 and 0.029 for liquid and vapor CO2 molar fractions, and 0.005 and 0.010 for liquid and vapor densities respectively. The Validation against experiment data on 10 binary systems and 316 state points shows that the predictions are accurate with average deviation of about 5% for CO2 mole fraction and 3.5% for saturated density at pressure range to P < 0.9Pc. This automatic procedure can only be used for the prediction of the vapor-liquid equilibrium. The computed data can be obtained from http://sun.sjtu.edu.cn/msd.Download high-res image (147KB)Download full-size image

The COMPASS II force field has been developed by extending the coverage of the COMPASS force field (J Phys Chem B 102(38):7338–7364, 1998) to polymer and drug-like molecules found in popular databases. Using a fragmentation method to systematically construct small molecules that exhibit key functional groups found in these databases, parameters applicable to database compounds were efficiently obtained. Based on the same parameterization paradigm as used in the development of the COMPASS force field, new parameters were derived by a combination of fits to quantum mechanical data for valence parameters and experimental liquid and crystal data for nonbond parameters. To preserve the quality of the original COMPASS parameters, a quality assurance suite was used and updated to ensure that additional atom-types and parameters do not interfere with the existing ones. Validation against molecular properties, liquid and crystal densities, and enthalpies, demonstrates that the quality of COMPASS is preserved and the same quality of prediction is achieved for the additional coverage.

The mechanisms of aluminophosphate oligomerization were investigated using density functional theory with the SMD solvation model. Two aluminum species, Al(OH)4– and Al(H2O)63+, and four phosphorus species, H3PO4, H2PO4–, HPO42–, and PO43–, were considered as the monomers for polycondensation reactions. It was found that the most favorable pathway to dimerization was a Lewis acid–base reaction: the aprotic oxygen of phosphoric acid (P═O) performs a nucleophilic attack on the central aluminum atom of Al(OH)4–. Using this mechanism as a pattern, plausible dimerization mechanisms were investigated by varying the proticity and hydration of the phosphorus and aluminum monomers, respectively. The relative reaction rates of each mechanism were estimated under different pH conditions. The chain growth of aluminophosphates to trimers, tetramers, and pentamers and the cyclization of a linear tetramer were also investigated. For oligomerization reactions beyond dimer formation, it is found that cluster growth favors the addition of the phosphoric monomers rather than aluminum monomers.

Whether or not a coarse grained force field (CGFF) can be made to be transferrable is an important question to be addressed. By comparing potential energy with potential of mean force (PMF) of a molecular dimer, we proposed to use a free energy function (FE-12-6) with the parameters in entropic and energetic terms explicitly to represent the nonbond interactions in CGFF. Although the FE-12-6 function cannot accurately describe the PMF curves, a cancelation of short radii and strong repulsion makes the function a good approximation. For nonpolar molecules represented by linear alkanes, FE-12-6 is demonstrated to be highly effective in representing the nonbond interactions in CGFF. The force field parameters are well transferrable among different alkane molecules, in different thermodynamic states and for predicting various thermodynamic properties including heats of vaporization, vapor–liquid-equilibrium coexistence curves, surface tensions, and liquid densities.

Accelerated molecular dynamics (aMD) is a popular method for biomolecular simulations. The estimator to recover the true free energy profile, termed “reweighting”, constitutes a critical part of aMD. Recently the second order cumulant expansion has been recommended as an improved reweighting method. Here we examine the validity of this reweighting method and provide an insight into the need for selecting appropriate boost potentials.

Molecular simulation is a promising tool for the study of zeolite formation. However, sufficient sampling remains a grand challenge for the practical use of molecular simulation for this purpose. Here, we investigate the initial stage of zeolite synthesis under realistic conditions by using the replica-exchange method and the ReaxFF reactive force field. After a total simulation time of 480 ns, both energetic and structural properties approach convergence. Analyses of data collected at 600 K show that the inorganic structure directing agent NaOH promotes the aggregation of silicate, the formation of branched Si atoms and the formation of 5-membered rings. With the trajectories collected simultaneously at different temperatures, the effect of temperature is discussed.

The surface-bulk partition of nonionic surfactants was predicted by calculating chemical potential differences using two simulation boxes representing the surface and the bulk phases separately. A published coarse grained force field was modified and validated for this application. Thermodynamic integration (TI) was applied to compute excess chemical potentials. The high concentration surface was stabilized by applying an external harmonic potential, and the bulk was treated as ideal solution, which was confirmed by simulation using a lattice model at conditions near the critical micelle concentration (∼10–5 mol/L). Based on the calculated chemical potential differences with precision of ca. 1 kJ/mol, the equilibria of surface-bulk concentration was predicted well in comparison with the experimental data.

We develop here the methodology for dramatically accelerating the ReaxFF reactive force field based reactive molecular dynamics (RMD) simulations through use of the bond boost concept (BB), which we validate here for describing hydrogen combustion. The bond order, undercoordination, and overcoordination concepts of ReaxFF ensure that the BB correctly adapts to the instantaneous configurations in the reactive system to automatically identify the reactions appropriate to receive the bond boost. We refer to this as adaptive Accelerated ReaxFF Reactive Dynamics or aARRDyn. To validate the aARRDyn methodology, we determined the detailed sequence of reactions for hydrogen combustion with and without the BB. We validate that the kinetics and reaction mechanisms (that is the detailed sequences of reactive intermediates and their subsequent transformation to others) for H2 oxidation obtained from aARRDyn agrees well with the brute force reactive molecular dynamics (BF-RMD) at 2498 K. Using aARRDyn, we then extend our simulations to the whole range of combustion temperatures from ignition (798 K) to flame temperature (2998K), and demonstrate that, over this full temperature range, the reaction rates predicted by aARRDyn agree well with the BF-RMD values, extrapolated to lower temperatures. For the aARRDyn simulation at 798 K we find that the time period for half the H2 to form H2O product is ∼538 s, whereas the computational cost was just 1289 ps, a speed increase of ∼0.42 trillion (1012) over BF-RMD. In carrying out these RMD simulations we found that the ReaxFF-COH2008 version of the ReaxFF force field was not accurate for such intermediates as H3O. Consequently we reoptimized the fit to a quantum mechanics (QM) level, leading to the ReaxFF–OH2014 force field that was used in the simulations.

Single-crystalline mesostructured zeolite nanosheets (SCZN) are synthesized by using designed surfactants with an aromatic group and only single quaternary ammonium head. Both the number of benzene rings and the length of the carbon chain play important roles in the ordered self-assembly of these alternating MFI (mordenite framework inverted) nanosheets. The surfactants, only with two benzene rings and a carbon chain larger than 4, lead to the formation of SCZN because of their highly ordered orientation through π–π stacking; the interlamellar spacing of SCZN could be controlled in the range of 1.7–2.1 nm through variation of the carbon chain length from 6 to 10. A combination of X-ray diffraction patterns and electron microscopy provides visible evidence for the mesostructural transformation from two amorphous aluminosilicate layers to one MFI sheet. The highly ordered orientation of the aromatic groups through π–π stacking geometrically matches the MFI framework to form the crystallographically correlated mesostructure. The low binding energies for the self-assembly of this synthesis system, obtained by molecular mechanics calculations, provide theoretical evidence of the feasibility of our strategy. In addition, this strategy is successfully verified using bolaform cationic surfactants, which also result in the crystallographically ordered MFI nanosheets owning to the similar π–π stacking interactions.

Using force field parameters developed and validated for zeolitic imidazolate frameworks (ZIFs) and carbon dioxide (CO2) independently from adsorption data, we predicted CO2/ZIFs adsorption isotherms using the Grand Canonic Monte Carlo (GCMC) method. The results are in sharp contradiction: the calculated adsorption data agree well with the experimental data for SOD-type ZIF-8, but are more than 100% higher than the experimental data for GME-type ZIFs. Using non-equilibrium molecular dynamics simulations and potential of mean force (PMF) calculations, we reveal that the discrepancies are due to the kinetic blockage which is significant for ZIF-68 but negligible for ZIF-8. This study demonstrates that a force field developed independently from the adsorption data can be used to predict the adsorptions accurately; and the kinetic factor must be considered if the bottlenecks exist in the adsorption paths due to geometric and energetic features of adsorbate and adsorbent. It could be very misleading if the force field parameters are adjusted by fitting the GCMC simulation data to experimental data without considering the kinetic factors.

The surface–bulk partition of surfactants is a piece of important information for understanding the performance of surfactants. In this work, we predicted surface–bulk partitions for three ionic surfactants (sodium hexyl sulfate, sodium nonyl sulfate, sodium dodecyl sulfate) using molecular dynamics simulations and a coarse-grained force field. The simulations were performed with the bilayer model that represents thin films. By setting initial configurations using chemical potentials calculated for infinitely dilute solutions, we obtained a sufficient number of independent samples for each simulation which ran at least 2 μs. Predicted surface concentrations and surface tensions are in good agreement with the experimental data. The surface adsorption isotherms, surface structures, and critical micelle concentrations (CMCs) derived from the simulated data are in reasonable agreement with the experimental data and previously calculated data. For surfactants with a short alkyl chain (SHS), stable oligomers are formed before micelles in the bulk.

Molecular dynamics simulations and thermodynamic integration method are used to calculate free energy differences when replacing Na+ with NH4+ in hydrated faujasite (Y-type) zeolite and in an aqueous solution of these cations. Force field parameters are optimized by quantum mechanics density functional theory calculations and then validated by calculating cation distributions in dehydrated zeolite and differences in hydration free energies of the cations. Using force field and Monte Carlo simulations, we predict cation distributions at different sites of hydrated zeolite with different mole fractions of NH4+. The free energy changes upon replacement of Na+ with NH4+ at different sites are calculated based on the predicted distributions. The statistically weighted average of the free energy differences of Na+/NH4+ exchange in zeolite is compared with that in an aqueous solution of the cations. The equilibrium isotherm is predicted and found to agree well with experimental data.

Highly stable salt functional groups consisting of lithium cation and aromatic anions (CnHnN5−n−Li) are studied for hydrogen storage using ab initio calculations, force field development, and grand canonical Monte Carlo simulations. Second-order Møller–Plesset perturbation theory with the resolution of identity approximation calculations are calibrated at the CCSD(T)/complete basis set (CBS) level of theory. The calibrations on different types of binding sites are different, but can be used to correct the van der Waals interactions systematically. The anion and salt functional groups provide multiple binding sites. With increased number of nitrogen atoms in the aromatic anion, the number of binding sites increases but the average binding energy decreases. Among the functional groups considered, CHN4-Li exhibits the largest number of binding sites (14) and a weak average binding energy of 5.7 kJ mol–1 with CCSD(T)/CBS correction. The calculated adsorption isotherms demonstrate that the introduction of the functional group significantly enhances hydrogen uptake despite relatively weak average binding energy. Therefore, it is concluded that searching for functional groups with the larger number of binding sites is another key factor for enhancing the hydrogen storage capacity, given that other conditions such as free volume and surface area are fixed.

The stability of Newton black films (NBFs) under lateral mechanical stretch was investigated by nonequilibrium molecular dynamics (NEMD) simulations using force field parameters validated by accurate prediction of surface tensions. The applied strains accelerated film ruptures, enabling efficient measurements of the critical thicknesses of the films. Two representative surfactants, sodium dodecyl sulfate (SDS) for ionic surfactant and pentaethylene glycol monododecyl ether (C12EO5H) for nonionic surfactant, were investigated and compared. The predicted critical thickness of C12EO5H-coated film is smaller than that of the SDS-coated film, which is consistent with previously reported experimental observations. Our simulation results show that while the two surfactant-coated films exhibit similar dynamic properties attributed to the Marangoni–Gibbs effect, their surface structural characteristics are quite different. Consequently the two films demonstrate distinct rupture mechanisms in which rupture starts at uncovered water domains in the SDS-coated film, but at lateral surfactant/water interfaces in the C12EO5H-coated film. Our findings provide new insights into the stabilization mechanisms of NBFs and will facilitate the design and development of new films with improved properties.

The adsorption and decomposition of ammonia–borane (AB) on the surface of silicon carbide nanotubes (SiCNTs) was investigated using density functional theory. Five adsorption types and four reaction channels were identified. The most favorable reaction channel that generates a H2 molecule is slightly endothermic; furthermore, the energy barrier for the decomposition of the AB molecule is only 9.6 kcal mol−1. The side reactions that generate NH3 or BH3 are highly endothermic; therefore, the generation of side products can be depressed by decreasing the temperature. However, desorption of hydrogen atoms from the surface appears to be a more difficult step. The energy-barrier height for generation of a H2 molecule and its subsequent desorption from the surface is approximately 34.2 kcal mol−1. The migration of hydrogen atoms on the surface of SiCNTs involves lower energy than the desorption process, indicating that the desorption of H2 molecules from the surface may be more complicated.

The adsorption of ethanol vapor on a mica surface at 298 K and different relative humidities (RHs) are studied using grand canonical Monte Carlo and molecular dynamics simulations. The simulations show that the adsorbed ethanol molecules form a monolayer on the mica surface, sharply contrasting the behavior of water, which forms multiple adsorption layers on the mica surface. Simulations of an ethanol and water mixture reveal that the adsorbed molecules are segregated into a water-rich domain near the mica surface and an ethanol-rich domain on top of the water-rich domain. The water-rich domain exhibits multilayers unless the RH is extremely low (<1%), whereas the ethanol-rich domain exhibits a monolayer. These findings are supported by calculations of the isosteric heats of adsorption and analyses of configurations, concentrations, and diffusivities of molecules in different layers.

Thermal conductivities of graphene-like silicon and carbon hybrid nanostructures with silicon atom percentages varying from 0 % (graphene) to 100 % (silicene) are investigated using the reserve non-equilibrium molecular dynamic (RNEMD) method and Tersoff bond order potentials. The thermal conductivity of graphene is dramatically reduced with increasing silicon concentration, and the reduction appears to be related more to the topological structures formed than the amount of doped silicon atoms present. The reduction is collectively contributed to by reduced phonon group velocities (v), phonon free paths (l∞), and the specific heat capacity (c) of the material. For systems with high symmetry, thermal conductivity is mainly influenced by v and c. For systems with low symmetry, thermal conductivity is dominated by l∞; such materials are also more direction-dependent on thermal flux than highly symmetric materials.

We report the kinetic analysis and mechanism for the initial steps of pyrolysis and combustion of a new fuel material, 1,6-dicyclopropane-2,4-hexyne, that has enormous heats of pyrolysis and combustion, making it a potential high-energy fuel or fuel additive. These studies employ the ReaxFF force field for reactive dynamics (RD) simulations of both pyrolysis and combustion processes for both unimolecular and multimolecular systems. We find that both pyrolysis and combustion initiate from unimolecular reactions, with entropy-driven reactions being most important in both processes. Pyrolysis initiates with extrusion of an ethylene molecule from the fuel molecule and is followed quickly by isomerization of the fuel molecule, which induces additional radicals that accelerate the pyrolysis process. In the combustion process, we find three distinct mechanisms for the O2 attack on the fuel molecule: (1) attack on the cyclopropane, ring expanding to form the cyclic peroxide which then decomposes; (2) attack onto the central single bond of the diyne which then fissions to form two C5H5O radicals; (3) attack on the alkyne-cyclopropane moiety to form a seven-membered ring peroxide which then decomposes. Each of these unimolecular combustion processes releases energy that induces additional radicals to accelerate the combustion process. Here oxygen has major effects both as the radical acceptor and as the radical producer. We extract both the effective activation energy and the effective pre-exponential factor by kinetic analysis of pyrolysis and combustion from these ReaxFF simulations. The low value of the derived effective activation energy (26.18 kcal/mol for pyrolysis and 16.40 kcal/mol for combustion) reveals the high activity of this fuel molecule.

A classical force field capable for accurately predicting surface tensions, surface concentration, and other interfacial properties is reported for sodium dodecyl sulfate (SDS). This force field is proposed by combining parameters from well established force fields for components of the air/SDS/water interface and optimized by adjusting the van der Waals diameter of sulfate oxygen to fit experimental data of surface tension. The force field parameters are transferable; as good agreement with experimental data is achieved from independent calculations on activity derivatives of aqueous solution of sodium methyl sulfate using the Kirkwood−Buff theory. The adjusted parameter effectively modulates the electrostatic interactions of solvated ions in solutions. This modification has a strong impact to surface tension and location and mobility of sodium cations but minimal impact to properties such as density profiles of bulk phase and conformations and orientations of surfactant chain, for which consistent results compared with previous studies are obtained.

Two-layer ONIOM calculations were carried out to study initial reactions of catalytic cracking of 1-butene to produce propene and ethene on HZSM-5 and HFAU zeolites. Direct cracking and dimerization cracking mechanisms were evaluated. The calculated data indicate that the dimerization cracking is more favorable than the direct cracking based on both kinetic and thermodynamic considerations. HZSM-5 is catalytically more effective than HFAU for the cracking reactions. ONIOM energy analysis shows that the effectiveness of zeolite is due to long-range van der Waals interaction energies that stabilize all reaction intermediates, and short-range interaction between the zeolite and the reacting species that reduce the activation energies. For the dimerization process, stepwise and concerted mechanisms are similar in energy changes. Generally speaking, dimerizations appear to be exothermic reactions with modest activation energies. The activation energies for isomerization, β-scission, and deprotonation are lower for larger molecular fragments than for smaller ones.

We propose to incorporate a lithium tetrazolide group into porous materials for enhancing hydrogen storage capacity. The lithium tetrazolide group is much more stable and polarized than the models made by doping aromatic groups with lithium atoms. More importantly, each of the lithium tetrazolide provides 14 binding sites for hydrogen molecules with modest interaction energies. The advantage of multiple binding sites with modest binding energies is partially demonstrated by constructing a new porous aromatics framework (PAF-4) with the lithium tetrazolide moieties and predicting its hydrogen uptake using first-principles GCMC simulations. The predicted hydrogen uptake reaches 4.9 wt % at 233 K and 10 MPa, which exceeds the 2010 DOE target of 4.5 wt %.Keywords (keywords): adsorption; hydrogen; isosteric heats; lithium tetrazolide; porous organic framework;

The octanol–water partition coefficients (Ko/w) and infinite-dilution activity coefficients (γ∞) of 1-ethylpropylamine and 3-methyl-1-pentanol were predicted based on free energy calculations. The multi-configuration thermodynamic integration (TI) method was applied with molecular dynamics simulations to calculate free energy changes. With optimized simulation options and force field parameters the calculated free energy changes were highly precise (uncertainties <0.52 kJ/mol) and accurate (deviations <2.7 kJ/mol from the experimental data). Although this accuracy level is comparable with the best reported in the literature, it was still too low to yield accurate predictions of Ko/w and γ∞ coefficients. Nevertheless, the predictions were close to those obtained using general QSPR methods.

An ab initio force field that describes interactions between hydrogen molecules and IRMOF materials is proposed. The force field parameters were derived by fitting to ab initio data that includes higher-order electron correction and extended basis-set effects and validated by calculating adsorption isotherms and isosteric heats of adsorption of H2 in IRMOF-1 using GCMC simulations performed at 77 and 298 K, in a broad range of pressure from 0.0 to 8.0 MPa. Excellent agreements with experimental data were obtained. The force field was then applied to predict hydrogen-storage capacities for eight additional IRMOF materials. It was identified that the void fraction of volume (VFV) has a strong impact on the adsorption capacity, and its impacts on gravimetric and volumetric adsorption uptakes exhibit opposite trends. An overall optimal VFV is ca. 87% for IRMOFs at 77 K and 8.0 MPa.

Density functional theory was employed to study the hydrothermal stability of P-modified ZSM-5 zeolites using cluster models. The calculations of hydrolysis energies indicated that the introduction of phosphorus increases the hydrothermal stability of ZSM-5 zeolites. The initial paths of dealumination were studied with explicit water molecules. It was found that the framework Al-O coordination bond can be replaced by coodination bonds between water molecules and the aluminium. One to three water molecules can form coordination bonds with framework Al and release energies. The P-modification restrain the dealumination. The calculated 27Al NMR chemical shifts for the obtained structures are consistent with the experimental measurements.

Computational prediction of adsorption of small molecules in porous materials has great impact on the basic and applied research in chemical engineering and material sciences. In this work, we report an approach based on grand canonical ensemble Monte Carlo (GCMC) simulations and ab initio force fields. We calculated the adsorption curves of ammonia in ZSM-5 zeolite and hydrogen in MOF-5 (a metal-organic-framework material). The predictions agree well with experimental data. Because the predictions are based on the first principle force fields, this approach can be used for the adsorption prediction of new molecules or materials without experimental data as guidance.

Isothermal–isobaric ensemble (NPT) molecular dynamics (MD) simulation was applied to calculate the heats of mixing for binary fluids based on an all-atom force field. The calculation protocol was tested on mixtures of 1-propanol/n-heptane, n-butylamine/n-heptane and n-butylamine/water at various temperatures and compositions. Based on a simple error analysis and comparisons of calculated and experimental data, we propose that a necessary but not sufficient condition in the predictions of the heats of mixing is to have a force field highly accurate in evaluating energetic properties for pure substances. More work is required in order to find out the sufficient conditions.

Ab initio methods were used to investigate the α-fluorination of simple phosphonic acid molecules. We found that fluorine atoms are not strong hydrogen bond acceptors, and the substitution marginally enhances the hydrogen bonds between the phosphonic group (–PO(OH)2) and its surrounding polar groups. However, profound effects are observed in the stability of ionic species due to the substitution of fluorine atom(s). The proton transfer from phosphonic acids to proton-accepting groups such as water or ammonia molecules is energetically plausible in water solution to generate anions, and the α-fluorination significantly stabilizes the anions. This result explains the established correlation between bioactivities and the pKa values and sheds light on the significant enhancement in the bioactivity of phosphatase inhibitors due to their fluoro-substitution.

Automatic parameterization procedures, which are important components in our latest approach of making transferable and extensible general force fields, were demonstrated by deriving parameters from ab initio and empirical data, and subsequently predicting densities for nine molecular fluids. Excellent agreements with available experimental data were obtained for most of the systems studied. The automated procedure for deriving van der Waals parameters, constructed by combining the molecular dynamics simulation and statistical mechanical perturbation theory, is described.

The periodic perturbation method is applied to predict shear viscosities of liquid 1,4-butanediol, 1,3-butanediol, 1,2-butanediol, 2-methyl-1,3-propanediol and 1,2,4-butanetriol under two thermodynamical conditions: T = 373 K, P = 0.1 MPa and T = 373 K, P = 250 MPa. A linear dependence of the calculated shear viscosities with respect to the applied perturbation forces is identified. Based on this finding, extrapolation of the calculated viscosities to zero perturbation force is applied to estimate the shear viscosities for the “undisturbed” fluids. The uncertainties of the estimates are calculated using the block average method. The predicted values compare favorably with the experimental data. Although the force field is optimized using equilibrium properties (liquid densities and vaporization enthalpies), calculated results demonstrate that the force field can be used to predict the kinetic properties accurately. The force field parameters are well transferable among different state points. However, transferring parameters among different molecules should be executed with caution.

Molecular simulation is a promising tool for the study of zeolite formation. However, sufficient sampling remains a grand challenge for the practical use of molecular simulation for this purpose. Here, we investigate the initial stage of zeolite synthesis under realistic conditions by using the replica-exchange method and the ReaxFF reactive force field. After a total simulation time of 480 ns, both energetic and structural properties approach convergence. Analyses of data collected at 600 K show that the inorganic structure directing agent NaOH promotes the aggregation of silicate, the formation of branched Si atoms and the formation of 5-membered rings. With the trajectories collected simultaneously at different temperatures, the effect of temperature is discussed.