Quantum chemical investigations of surface chemistry with a hierarchy of cluster models

Krishnan Raghavachari得到了理论和计算化学计划的支持，该计划通过开发和应用一套分层理论模型来开发研究表面化学的理论方法。分层概念涉及在小型系统上使用高度精确的方法，以验证用于研究复杂系统的更接近和计算成本效益更高的方法。在PI研究小组先前工作的基础上，Raghavachari及其同事正在三个领域开发更复杂和更广泛适用的方法：(1)适用于3D和4D过渡金属元素和第四排主族元素的精确热化学模型。这些方法是基于成功的Gauss-n(Gn)模型，但将包括许多改进，以克服已知的缺陷；(2)二价伪原子参数，以使得能够准确和高效地计算四面体键合系统的(100)表面的化学；(3)新的电子嵌入方法，用于精确处理基于流行的ONIOM框架的QM/QM多层计算。这些进展被应用于半导体和金属氧化物表面化学的一系列应用，包括(1)Si(100)表面的可控功能化，(2)化合物半导体表面的合金生长，(3)生物分子在二氧化硅表面的反应，以及(4)钛氧化物和钒氧化物表面的催化机理。科学界可在该计划中开发的计算机代码作为高斯软件包的一部分获得，也可以在PI网站上的一套免费代码中获得。

Krishnan Raghavachari is supported by an award from the Theoretical and Computational Chemistry program to develop theoretical methods for studying surface chemistry by developing and applying a set of hierarchical theoretical models. The hierarchical concept involves the use of highly accurate methodology on small systems to validate more approximate and computationally cost effective methodologies for investigating complex systems. Building on prior work in the PI's research group, Raghavachari and coworkers are developing more sophisticated and broadly applicable methods in three areas: (1) accurate thermochemical models applicable for 3d and 4d transition metal elements and fourth-row main group elements. These methods are based on the successful Gaussian-n (Gn) models, but will include many improvements to overcome known deficiencies; (2) divalent pseudoatom parameters to enable accurate and efficient calculations for investigating chemistry on (100) surfaces of tetrahedrally bonded systems; (3) new electronic embedding methods for an accurate treatment of QM/QM multilayer calculations based on the popular ONIOM framework. These developments are applied to a range of applications on semiconductor and metal-oxide surface chemistry, including (1) controlled functionalization of Si(100) surfaces, (2) alloy growth of compound semiconductor surfaces, (3) reactions of biomolecules on silica surfaces, and (4) catalytic mechanisms of titanium oxide and vanadium oxide surfaces. The computer codes developed in this program are available to the scientific community as part of the Gaussian software package as well as in a set of freely available codes on the PI's website.