Computationally Intensive Research Project
Accurate ab initio Determinations of Thermochemistry of Novel Molecular Radicals, Cations and Neutrals of Relevance to Combustion, Environmental, and Atmospheric Chemistry
MSC News
Additional Information
Cheuk-Yiu Ng,1 Kai-Chung Lau,2 See-Wing Chiu3
1University of California, Davis, 2University of Chicago, 3National Center for Supercomputing Applications
FY07 Allocation - 80,000
Abstract
This proposal is to carry out high-level ab initio electronic structure calculations to study the chemistry of novel molecular systems such as radicals, cations and neutrals, which are relevant to the mission of the Department of Energy and are of interest to the nation. We plan to use the CCSD(T)/CBS method to predict ionization energies (IEs) and heats of formation (HOF) for C4 - C7 hydrocarbon radicals. The CCSD(T)/CBS method involves the approximation to the complete basis set (CBS) limit at the coupled cluster with single and double excitations plus perturbative triple excitations [CCSD(T)] level together with other high-level correction terms such as the anharmonic zero-point vibration energy, the core-valence electronic correction due to core electrons, the scalar-relativistic corrections, the diagonal Born-Oppenheimer correction and the higher order correction beyond the CCSD(T) wavefunction. Our previous study shows that the CCSD(T)/CBS method are capable to predict to the adiabatic IEs for the C1 - C3 hydrocarbon radicals to an accuracy of +/-10 meV. This benchmarking would allow us to explore the scaling effect of the CCSD(T)/CBS method with respect to the molecular size, and potentially discover the limitation and shortcoming of the theory. The results would provide a new set of thermochemical data and allow a better understanding of photoionization energetics and dynamics, especially for molecules with poor Franck-Condon factor near the ionization threshold.
We propose to use the density functional theory to obtain the bonding and structural properties of organometallic compounds such as metal carbonyls, metal carbides/dicarbides, metal methylidynes and carbenes. We predict the IE and HOF values for transition metal-ligated species and their cations using the CCSD(T)/CBS method. Currently, much structural and energetic data for these species and their cations are lacking. This part is expected to provide insight into the metal-carbon bonding, which plays an important role in physical and biochemical catalytic reactions.
We plan to carry out multi-reference based theoretical prediction on the structures and energetics for carbon clusters Cn/Cn+ (n=2-10). The carbon clusters have been known to be important species in hydrocarbon flames. Theoretical calculations on Cn and Cn+ species have been challenging due to their dense electronic states and the existence of isomeric structures. The part of proposed work is to provide important energetics and dynamics information to the formation of soot during hydrocarbon and fuel combustion.
This last part of the proposal is to examine (i) the performance of several CBS extrapolation schemes including the 3-point exponential form and 2-point power function and (ii) the performance of different theoretical approaches for predicting anharmonic zero-point vibrational energy contribution.
All the computational results are benchmarked and fully validated with the state-of-the-art experimental works carried out in the laboratory of the team leader. The experimental investigation features single-photon vacuum ultraviolet (VUV) as well as two-color infrared-VUV laser photoionization and photoelectron techniques for the spectroscopic studies of transition metal complexes, carbon clusters and hydrocarbon radicals. The resolutions achieved for VUV-photoionization and VUV-photoelectron measurements are 0.12 cm-1 and 1 cm-1, respectively. Previous studies indicate that spectroscopic and energetic data for polyatomic species can be determined to within 0.2 kJ mol-1. Such unrivaled precisions from experiments provide an ideal benchmarking tool to any high-level ab initio calculations. A variety of quantum mechanical models (such as perturbation theory, density functional theory, coupled cluster, and multi-reference based configuration interaction methods) together with correlation-consistent basis sets will be employed to provide reliable computed results and the best possible comparisons with the experimental measurements. This proposal work is primarily based on and evolved from the findings we obtained in the pilot project (EMSL9591) from 2004 to 2006.
