Developing the MCTDH Quantum Dynamics Code: Accurate Direct Dynamics of Non-Adiabatic Phenomena

Understanding the energy flow in molecules after absorbing a photon of light is important to understand their light-activated properties. These are important in a range of technologies, such as harvesting solar energy and data storage, as well as understanding photo-stability and deriving new molecules via photochemical pathways. This energy flow is complex, as a number of different pathways (channels) are available. Of particular importance are what are termed non-adiabatic pathways that allow a molecule to undergo a change of chemical character (electronic state) effectively instantaneously and so dominate the molecular evolution. Unfortunately it is impossible to say a priori which pathways are most important for a particular molecular, and computer simulations have a large role to play in answering this question by modelling the potential energy surfaces and molecular dynamics after excitation.

Computer codes for simulating excited-state dynamics are becoming more powerful and useful. However, there is still no code available that can capture all of the quantum mechanical effects required for the often large molecules of interest. The MCTDH package is one of the few codes that is able to treat these problems accurately in a general, user-friendly way. A development part of the code implements a novel method, the DD-vMCG algorithm. This promises to be able to accurately treat the molecules of interest using what is termed direct dynamics, which couples the dynamical simulation of the changes in molecular structure to a separate program that calculates the potential energy surfaces only when required, thus saving much of the work. Initial studies support this.

This DD-vMCG code has shown its potential in a number of studies. It now needs to be developed so that it is efficient, easy to use, and can be developed in a sustainable manner. The sustainable development of scientific software is crucial for it to survive and adapt as science progresses. The planned work will ensure that this method will be able to fulfill its potential after many years of development and be used by the scientific community to help understand and engineer how molecules behave and can ultimately be controlled in the presence of light.

The short-term impact of the work in the academic field should be substantial. The new code will give a tool that can provide accurate calculations of interest to many, and also a calibration of the methods used by others. This will help put the growing field of non-adiabatic direct dynamics on a stable foundation. This impact will be ensured by dissemination of the code and ideas through literature, conferences and training. The fact that over 200 researchers already use the MCTDH package for other types of calculations means that the dissemination of the new code will happen quickly. We will run a training workshop towards the end of the project.

It is of course difficult to predict the longer term impact of new code. The long term aim of the project is that quantum dynamics simulations will be as easy to do as quantum chemistry calculations. These already have made a huge impact on the way chemists think and go about research due to the availability of programs such as Gaussian and Molpro that allow accurate calculations on the energetics of molecules that can support observations and guide experiments. The addition of being able to simulate the dynamical behaviour of fundamental chemical reactivity should add another dimension to the role computers play in science.