Institute for Physical Chemistry, Christian-Albrechts-University, Olshausenstr. 40, 24098 Kiel, Germany.
J Chem Phys. 2013 Dec 14;139(22):224303. doi: 10.1063/1.4837237.
We have fitted reactive force fields of the ReaxFF type to the ground and first excited electronic states of azobenzene, using global parameter optimization by genetic algorithms. Upon coupling with a simple energy-gap transition probability model, this setup allows for completely force-field-based simulations of photochemical cis→trans- and trans→cis-isomerizations of azobenzene, with qualitatively acceptable quantum yields. This paves the way towards large-scale dynamics simulations of molecular machines, including bond breaking and formation (via the reactive force field) as well as photochemical engines (presented in this work).
我们使用遗传算法进行全局参数优化，为偶氮苯的基态和第一激发电子态拟合了 ReaxFF 型反应力场。通过与简单的能隙跃迁概率模型耦合，该设置允许使用完全基于力场的方法模拟偶氮苯的光致顺式-反式和反式-顺式异构化，其量子产率具有定性的可接受性。这为包括键断裂和形成（通过反应力场）以及光化学发动机（在本工作中提出）在内的分子机器的大规模动力学模拟铺平了道路。
