Assigning infrared spectra, particularly high-frequency stretching vibrations, often requires the incorporation of higher-order anharmonic effects. The most pervasive of these effects are Fermi resonance effects in both CH and OH stretch spectra. For large systems such as molecular clusters, the brute force assignment of experimental spectra by computationally iterating over all possible candidate structures is prohibitive.<br/> <br/>In this work, we describe our efforts to automate spectroscopic assignments (which we call “inverse spectroscopy”) through a data-driven approach. We will demonstrate methods that automatically assign anharmonic spectra at harmonic cost. This approach employs a Bayesian-optimization-derived algorithm as its driver and does not require a large volume of initial training data. Instead, only a flexible, a priori form of the clusters’ vibrational Hamiltonian is required. We will demonstrate the performance of our approach on model problems, higher-level calculations of benchmark systems, and real experimental spectra found in the literature. We then show our efforts to model multi-component spectra in an automated fashion and provide uncertainty quantification for assignments.

Inverse infrared spectroscopy via a data-driven approach

Fall 2022 abstracts, posters, and presentations will launch in mid-September.

ACS Fall 2022

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