Accurate Atomistic Modeling of Solutions

点击翻译按钮获取中文摘要

Peter Pulay of the University of Arkansas is supported by an award from the Chemical Theory, Models and Computational Methods program to develop improved electronic structure methods for simulating reactions in solutions, with possible extension to a protein environment. Accurate simulation of these reactions requires the evaluation of thermodynamic averages, increasing the computational burden by many orders of magnitude. Pulay's recent General Polarizability Quantum Mechanics / Molecular Mechanics (GP-QM/MM) methodology has already been implemented in a Metropolis Monte Carlo code and shown to speed up conventional QM/MM calculations by four orders of magnitude with barely any sacrifice in accuracy. Pulay and his coworkers will develop analytical gradients for the new method, allowing the determination of reaction paths in solutions. They will also implement improved solvent models, to match the accuracy of the description of the reactive system. They will extend the current Monte Carlo simulations to the alternative Molecular Dynamics method which can supply information about the time scale of the changes. The new techniques will be applied to the prediction of pKa values, equilibria between zwitterionic and neutral forms of amino acids and small peptides, conformational preferences of peptides, and reactions known to be strongly affected by polar solvents. They will also investigate the factors determining whether anions preferentially adsorb on the surface in water microdroplets.During the last three decades, both computer power and theoretical methods advanced tremendously, finally allowing accurate routine calculation of the electronic structures and properties of medium-sized chemical systems in the gas phase. However, most chemistry (and virtually all biochemistry) takes place in solution or in a protein environment. Polar solvents like water have a profound effect on chemical reactivity, and require very large numbers of simulation runs to obtain accurate averages. As a result, significant approximations are frequently made in the methods, although this reduces their accuracy. Instead, the methods under development in the Pulary group aim to correct this situation and have been shown to reduce the computational cost by 4 orders of magnitude; larger savings are expected for bigger systems. This will allow the use of accurate theoretical methods for solutions. Pulay has contributed a number of algorithmic advances in the field of electronic structure theory, beginning with the analytical gradient technique which is the cornerstone of most current quantum chemical applications. Other widely used contributions are molecular geometry optimization, acceleration of the basic method to calculate wavefunctions (SCF), the calculation of NMR chemical shifts, and Local Electron Correlation. Pulay and his coworkers also work on utilizing computer simulation of molecules to improve undergraduate and graduate chemistry education, using sophisticated computer graphics to convey the information.