Marcus G. Martin

Intermolecular potentials for branched alkanes and the vapour-liquid phase equilibria of n-heptane, 2-methylhexane, and 3-ethylpentane

Configurational bias Monte Carlo calculations in the Gibbs ensemble have been used to perform direct simulations of the vapour-liquid phase equilibria of three heptane isomers: n-heptane, 2-methylhexane, and 3-ethylpentane. The simulations were carried out using isotropic united-atom representations of the Lennard-Jones type for the CH3, CH2 and CH groups. The aim of these calculations is to extend our force field, previously derived for linear alkanes, to branched alkanes by fitting new interaction parameters for methyl and ethyl branches.

Effect of pressure, membrane thickness, and placement of control volumes on the flux of methane through thin silicalite membranes: A dual control volume grand canonical molecular dynamics study

The flux of methane through the straight channels of thin silicalite membranes is studied via dual control volume grand canonical molecular dynamics. The adsorption layers on the surfaces of the thin membranes are found to provide a significant resistance to the flux of methane. This strong surface effect for thin membranes requires that the control volumes (where insertions and deletions are performed) must be placed far enough away from the membrane surface that they do not overlap with the surface adsorption layer. The permeance (flux/pressure drop) of methane through the surface layer is shown to be insensitive to both the average pressure and the pressure drop. In contrast, the permeance through the interior of the membrane increases with decreasing average pressure. These results are explained using a model which treats the transport through the surface barrier as driven by the pressure gradient and transport through the zeolite as driven by the chemical potential gradient. A new force field named D...

Using arbitrary trial distributions to improve intramolecular sampling in configurational-bias Monte Carlo

A new formulation of configurational-bias Monte Carlo that uses arbitrary distributions to generate trial bond lengths, angles and dihedrals is described and shown to provide similar acceptance rates with substantially less computational effort. Several different trial distributions are studied and a linear combination of the ideal distribution plus Gaussian distributions automatically fit to the energetic and ideal terms is found to give the best results. The use of these arbitrary trial distributions enables a new formulation of coupled–decoupled configurational bias Monte Carlo that has significantly higher acceptance rates for cyclic molecules. The chemical potential measured via a modified Widom insertion is found to be ill-defined in the case of a molecule that has flexible bond lengths due to the unbounded probability distribution that describes the distance between any two atoms. We propose a simple standard state that allows the computation of consistent chemical potentials for molecules with flexible bonds. We show that the chemical potential via Widom insertion is not computed properly for molecules with Coulombic interactions when the number of trials for any of the nonbonded selection steps is greater than one. Finally, we demonstrate the power of the new algorithms by sampling the side-chain conformations of a polypeptide.

A novel Monte Carlo algorithm for polarizable force fields: Application to a fluctuating charge model for water

In this Monte Carlo algorithm for polarizable force fields, the fluctuating charges are treated as special degrees of freedom subject to a secondary low-temperature thermostat in close analogy to the extended Lagrangian formalism commonly used in molecular dynamics simulations of such systems. The algorithm is applied to Berne’s SPC-FQ (simple point charge–fluctuating charge) model for water. The robustness of the algorithm with respect to the temperature of the secondary thermostat and to the fraction of fluctuating-charge moves is investigated. With the new algorithm, the cost of Monte Carlo simulations using fluctuating-charge force fields increases by less than an order of magnitude compared to simulations using the parent fixed-charge force fields.

Theory of persistent, p-type, metallic conduction in c-GeTe

It has been known for over twenty years that rhombohedral c-germanium telluride is predicted to be a narrow gap semiconductor. However, it always displays p-type metallic conduction. This behaviour is also observed in other chalcogenide materials, including Ge2Sb2Te5, commonly used for optically and electrically switched, non-volatile memory, and so is of great interest. We present a theoretical study of the electronic structure of the perfect crystal and of the formation energies of germanium/tellurium vacancy and antisite defects in rhombohedral germanium telluride. We find that germanium vacancies are by far the most readily formed defect, independent of Fermi level and of growth ambient. Moreover, we predict that the perfect crystal is thermodynamically unstable. Thus, the predicted large equilibrium densities of the germanium vacancy of ~5 × 1019 cm−3 results in a partially filled valence band and in the observed p-type conductivity.

Vapor–liquid phase equilibria of triacontane isomers: Deviations from the principle of corresponding states

Abstract Coupled–decoupled configurational–bias Monte Carlo (CD-CBMC) simulations in the Gibbs ensemble were carried out to determine the vapor–liquid coexistence curves for n-triacontane and 2,6,10,15,19,23-hexamethyltetracosane (squalane). The transferable potentials for phase equilibria-united atom (TraPPE-UA) force field was used for these simulations. The simulated systems consisted of 200 molecules and the production periods extended to 100,000 Monte Carlo cycles, a system size about twice as large and a simulation length about one order of magnitude longer than used in previous simulations. The simulation results are in satisfactory agreement with the available experimental data. Examination of the coexistence curves in reduced units for the two triacontane isomers and for n-octane and 2,5-dimethylhexane shows that both molecular weight and branching can lead to deviations from the principle of corresponding states. Analysis of the molecular structures in the vapor and liquid phases points to a partial collapse (self-solvation) of the triacontane isomers as the likely origin of the deviations from the principle of corresponding states.

Molecular studies of the structural properties of hydrogen gas in bulk water

We report on our studies of the structural properties of a hydrogen molecule dissolved in liquid water. The radial distribution function, coordination number and coordination number distribution are calculated using different representations of the interatomic forces within molecular dynamics (MD), Monte Carlo (MC) and ab initio molecular dynamics (AIMD) simulation frameworks. Although structural details differ in the radial distribution functions generated from the different force fields, all approaches agree that the average and most probable number of water molecules occupying the inner hydration sphere around hydrogen is 16. Furthermore, all results exclude the possibility of clathrate-like organization of water molecules around the hydrophobic molecular hydrogen solute.

Simulating Retention in Gas–Liquid Chromatography: Benzene, Toluene, and Xylene Solutes

Accurate predictions of retention times, retention indices, and partition constants are a long sought-after goal for theoretical studies in chromatography. Although advances in computational chemistry have improved our understanding of molecular interactions, little attention has been focused on chromatography, let alone calculations of retention properties. Configurational-bias Monte Carlo simulations in the isobaric–isothermal Gibbs ensemble were used to investigate the partitioning of benzene, toluene, and the three xylene isomers between a squalane liquid phase and a helium vapor phase. The united-atom representation of the TraPPE (transferable potentials for phase equilibria) force field was used for all solutes and squalane. The Gibbs free energies of transfer and Kovats retention indices of the solutes were calculated directly from the partition constants (which were averaged over several independent simulations). While the calculated Kovats indices of benzene and toluene at T=403 K are significantly higher than their experimental counterparts, much better agreement is found for the xylene isomers at T=365 K.

Final report :LDRD project 84269 supramolecular structures of peptide-wrapped carbon nanotubes.

Carbon nanotubes (CNT) are unique nanoscale building blocks for a variety of materials and applications, from nanocomposites, sensors and molecular electronics to drug and vaccine delivery. An important step towards realizing these applications is the ability to controllably self-assemble the nanotubes into larger structures. Recently, amphiphilic peptide helices have been shown to bind to carbon nanotubes and thus solubilize them in water. Furthermore, the peptides then facilitate the assembly of the peptide-wrapped nanotubes into supramolecular, well-aligned fibers. We investigate the role that molecular modeling can play in elucidating the interactions between the peptides and the carbon nanotubes in aqueous solution. Using ab initio methods, we have studied the interactions between water and CNTs. Classical simulations can be used on larger length scales. However, it is difficult to sample in atomistic detail large biomolecules such as the amphiphilic peptide of interest here. Thus, we have explored both new sampling methods using configurational-bias Monte Carlo simulations, and also coarse-grained models for peptides described in the literature. An improved capability to model these inorganichiopolymer interfaces could be used to generate improved understanding of peptide-nanotube self-assembly, eventually leading to the engineering of new peptides for specific self-assembly goals.

High performance computing for the application of molecular theories to biological systems

In this paper we summarize our work to bring modern numerical methods to bear in computations for density functional theories (DFTs) of inhomogeneous fluid systems. We present the general mathematical structure of the problem, and briefly discuss different strategies for solving the problems. Finally, we present a few recent results from calculations on complex peptide assemblies in lipid bilayers to demonstrate the application of the methods in one complex 3-dimensional system. We find that while solver strategies developed and optimized for partial differential equations (PDEs) can be applied to these systems of equations, they do not provide optimal solutions with respect to speed, memory use, or parallel partitioning.