Development of a General Strategy of Optimal Control of Photochemical Reactions

随着飞秒化学中激光技术的发展和完善，光化学机制现在可以在单个分子振动的空间（+）和时间（fs）尺度上观察到。基于成形激光脉冲的最佳控制技术可以选择性地产生一种光产物，通常涉及在正常条件下永远不会产生的反应途径。然而，这样的实验并不容易实现。人们经常需要建造特殊的实验设备，因此可能需要数年时间来研究所需的反应。理论方法对于最优控制方法是必不可少的，因为必须通过实验探索的可能性是天文数字。在使用波包动力学的理论计算中，人们可以通过创建直接模拟实验的振动态的相干叠加来进行实验。如果知道势能面，就可以在反应路径的早期设计一个目标，知道它最终会导致所需的目标，从而将可能性的数量减少到可管理的水平。因此，人们可以在设计实验之前向实验者提出一个策略。该项目涉及方法开发，以提高我们的量子动力学方法的效率，以便可以研究现实的化学系统。这种理论发展的重点是方法，使基态和激发态的同时表示。我们的应用工作有两个方面：基准将运行的例子，其中最佳控制已经实现（花青染料）和新的系统，最佳控制似乎是可能的，但它可以实现的条件是未知的。

With the development and refinement of laser technology in femtochemistry, photochemical mechanisms are now observable on the spatial (+) and temporal (fs) scale of a single molecular vibration. Optimal control techniques based on shaped laser pulses can selectively generate one photoproduct, often involving reaction pathways that would never have been generated under normal conditions. However, such experiments are not easy to implement. One often needs to construct special experimental apparatus and thus it can take years to study a desired reaction. Theoretical methods are essential for optimal control methods because the number of possibilities that must be explored experimentally is astronomical. In a theoretical computation using wavepacket dynamics, one can experiment by creating coherent superpositions of vibrational states that directly mimic experiment. Given a knowledge of the potential energy surface, one can design a target early in the reaction path that one knows will lead ultimately to the desired objective and so reduce the number of possibilities to a manageable level. Thus one can suggest a strategy to the experimentalist before experiments are designed. This project involves method development to increase the efficiency of our quantum dynamics approach so that realistic chemical systems can be studied. This theoretical development is focussed on methods that enable the simultaneous representation of ground and excited states. Our application work has two aspects: benchmarks will be run against examples where optimal control has already been achieved (cyanine dyes) and on new systems where optimal control seems possible but where the conditions under which it can be achieved are unknown.