The exponential scaling nature of quantum mechanics is very clearly seen from the strong computational constraints when nuclear quantum effects are to be determined in conjunction with electronic structure. Here we present two methods: (a) The first approach allows us to compress the potential energy surface data through estimates to the local Nyquist frequency of the potential which then leads to an adaptive tessellation scheme of multi-dimensional surfaces. (b) We also discuss methods to compress nuclear wavefunction information that two different types of tensor-network schemes. The first topic is closely related to ideas in information theory. The second topic allows us to tune the entanglement and in turn compress the amount of information in quantum nuclear wavefunctions.<br/><br/>The methods are demonstrated through the study of hydrogen transfer reactions of prevalence (i) in specific atmospheric systems, and (ii) in biological enzymes. In both cases reaction coordinate dependent quantum nuclear effects are efficiently captured and accurately describe the hydrogen transfer (tunneling) process.

Shannon entropy and tensor networks: Large scale data compression methods for quantum nuclear dynamics simulations

256th National Meeting (2018)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.