Quantum Dynamics and Control of Surface Ultra-fast Processes

金属表面的光化学：为了了解表面上的光化学反应，应该明确电子激发机制，众所周知，基质介导的(间接)激发发生在大多数金属表面的反应中。利用非平衡态格林函数理论，我们发展了一种新的理论方法来计算底物激发引发的反应几率对光子能量的依赖关系。我们进行了密度泛函理论计算，得到了二聚体的吸附结构，并根据我们的NEGF理论计算了光解吸几率。我们发现，当光活性物种是NO二聚体时，合适的共振轨道只有一个，而且这种单一的共振足以重现实验数据。界面电子量子输运的NEGF方法：为了模拟电子通过分子结的输运，我们提出了一种有效的方法，使用从头算自洽NEGF和密度泛函理论相结合。由于它的灵活性，我们采用了一个接近扩展分子方法的模型，但通过与格林函数相同的从头计算水平的系统程序，改进了与分子-表面耦合和长程势有关的问题。由此产生的算法包括三个主要步骤：(I)嵌入势的构造；(Ii)格林函数在分子轨道基中的微扰展开；以及(Iii)通过将分子轨道空间分为不活跃空间、活跃空间和虚拟空间来截断分子轨道空间。该算法适用于大分子体系。

Photochemistry on Metal Surfaces: To understand photochemical reactions on surfaces, electronic excitation mechanisms should be specified, and it is well known that the substrate-mediated (indirect) excitation occurs in the majority of reactions on metal surfaces. We have developed a new theoretical method to calculate photon energy dependence of reaction probability triggered by substrate-mediated excitation using Nonequilibrium Green's function (NEGF) theory. We have performed DFT calculations to obtain the dimer adsorption structure and calculated the probability of photodesorption based on our NEGF theory. We have found that while the photoactive species is the NO dimer, a suitable resonant orbital is only one, and this single resonance is enough to reproduce the experimental data.A NEGF approach to interfacial electronic quantum transport: To model electron transport through a molecular junction, we propose an efficient method using an ab initio self-consistent NEGF combined with density functional theory. We have adopted a model close to the extended molecule approach, due to its flexibility, but have improved on the problems relating to molecule-surface couplings and the long-range potential via a systematic procedure for the same ab initio level as that of the Green's function. The resulting algorithm involves three main steps: (i) construction of the embedding potential; (ii) perturbation expansion of the Green's function in the molecular orbital basis; and (iii) truncation of the molecular orbital space by separating it into inactive, active, and virtual spaces. The algorithm is suitable for application to large molecular systems.