Wassja A. Kopp

Automated Discovery of Reaction Pathways, Rate Constants, and Transition States Using Reactive Molecular Dynamics Simulations

We provide a methodology for deducing quantitative reaction models from reactive molecular dynamics simulations by identifying, quantifying, and evaluating elementary reactions of classical trajectories. Simulations of the inception stage of methane oxidation are used to demonstrate our methodology. The agreement of pathways and rates with available literature data reveals the potential of reactive molecular dynamics studies for developing quantitative reaction models.

Hydrogen Abstraction from n-Butyl Formate by H• and HO2•

The combustion chemistry of esters has been elucidated in the past through the study of smaller formates and acetates. Hydrogen abstraction from the fuel as an initiation step is mostly modeled based on estimations for similar abstractions from nonoxygenated hydrocarbons. This study reports computed ab initio rates for abstractions by H˙ and HO₂˙ radicals from the recently proposed biofuel candidate n-butyl formate. The energies are evaluated with a double hybrid density functional that performs especially well for barrier heights (B2KPLYP/aug-cc-pvtz). Hindered rotation of HO₂˙ with respect to n-butyl formate is treated using accurate eigenvalue summation and shows large influence on the rates. Transition states at the γ and δ positions are still influenced by the formate group. The abstraction from the γ carbon by HO₂˙ is slowest, although proceeding over the lowest barriers, due to the important influence of transition state entropies. A comparison with smaller esters and n-butanol shows that estimated rates deviate within 1 order of magnitude from the ab initio computations for similar groups in n-butyl formate.

Efficient yet accurate approximations for ab initio calculations of alcohol cluster thermochemistry.

We have calculated the binding enthalpies and entropies of gas phase alcohol clusters from ethanol to 1-decanol. In addition to the monomers, we have investigated dimers, tetramers, and pentamers. Geometries have been obtained at the B3LYP/TZVP level and single point energy calculations have been performed with the Resolution of the Identity-MP2 (RIMP2) method and basis set limit extrapolation using aug-cc-pVTZ and aug-cc-pVQZ basis sets. Thermochemistry is calculated with decoupled hindered rotor treatment for large amplitude motions. The results show three points: First, it is more accurate to transfer the rigid-rotor harmonic oscillator entropies from propanol to longer alcohols than to compute them with an ultra-fine grid and tight geometry convergence criteria. Second, the computational effort can be reduced considerably by using dimerization energies of longer alcohols at density functional theory (B3LYP) level plus a RIMP2 correction obtained from 1-propanol. This approximation yields results almost with the same accuracy as RIMP2 - both methods differ for 1-decanol only 0.4 kJ/mol. Third, the entropy of dimerization including the hindered rotation contribution is converged at 1-propanol with respect to chain length. This allows for a transfer of hindered rotation contributions from smaller alcohols to longer ones which reduces the required computational and man power considerably.

The furan microsolvation blind challenge for quantum chemical methods: First steps

Herein we present the results of a blind challenge to quantum chemical methods in the calculation of dimerization preferences in the low temperature gas phase. The target of study was the first step of the microsolvation of furan, 2-methylfuran and 2,5-dimethylfuran with methanol. The dimers were investigated through IR spectroscopy of a supersonic jet expansion. From the measured bands, it was possible to identify a persistent hydrogen bonding OH-O motif in the predominant species. From the presence of another band, which can be attributed to an OH-π interaction, we were able to assert that the energy gap between the two types of dimers should be less than or close to 1 kJ/mol across the series. These values served as a first evaluation ruler for the 12 entries featured in the challenge. A tentative stricter evaluation of the challenge results is also carried out, combining theoretical and experimental results in order to define a smaller error bar. The process was carried out in a double-blind fashion, with both theory and experimental groups unaware of the results on the other side, with the exception of the 2,5-dimethylfuran system which was featured in an earlier publication.

Automated Chemical Kinetic Modeling via Hybrid Reactive Molecular Dynamics and Quantum Chemistry Simulations

An automated scheme for obtaining chemical kinetic models from scratch using reactive molecular dynamics and quantum chemistry simulations is presented. This methodology combines the phase space sampling of reactive molecular dynamics with the thermochemistry and kinetics prediction capabilities of quantum mechanics. This scheme provides the NASA polynomial and modified Arrhenius equation parameters for all species and reactions that are observed during the simulation and supplies them in the ChemKin format. The ab initio level of theory for predictions is easily exchangeable, and the presently used G3MP2 level of theory is found to reliably reproduce hydrogen and methane oxidation thermochemistry and kinetics data. Chemical kinetic models obtained with this approach are ready to use for, e.g., ignition delay time simulations, as shown for hydrogen combustion. The presented extension of the ChemTraYzer approach can be used as a basis for methodological advancement of chemical kinetic modeling schemes and as a black-box approach to generate chemical kinetic models.

Assessing Statistical Uncertainties of Rare Events in Reactive Molecular Dynamics Simulations

Reactive molecular dynamics (MD) simulations are a versatile tool which allow for studying reaction pathways and rates simultaneously. However, most reactions will be observed only a few times in such a simulation due to computational limitations or slow kinetics, and it is unclear how this will influence the obtained rate constants. Therefore, we propose a method based on the Poisson distribution to assess the statistical uncertainty of reaction rate constants obtained from reactive MD simulations.

General formulation of rovibrational kinetic energy operators and matrix elements in internal bond-angle coordinates using factorized Jacobians

We show how inverse metric tensors and rovibrational kinetic energy operators in terms of internal bond-angle coordinates can be obtained analytically following a factorization of the Jacobian worked out by Frederick and Woywod. The structure of these Jacobians is exploited in two ways: On one hand, the elements of the metric tensor as well as its determinant all have the form ∑rmsin(αn)cos(βo). This form can be preserved by working with the adjugate metric tensor that can be obtained without divisions. On the other hand, the adjugate can be obtained with less effort by exploiting the lower triangular structure of the Jacobians. Together with a suitable choice of the wavefunction, we avoid singularities and show how to obtain analytical expressions for the rovibrational kinetic energy matrix elements.