Theory of controlled molecules

We develop theoretical methods and computational tools for investigations of the nuclear and electronic dynamics of polyatomic molecules that are based on the software packages RICHMOL, TROVE, PECD, and DYSON. Our focus is on accurate methods that are quantitatively predictive but at the same time applicable to complex molecular systems encountered in real-world situations. Our current primary research interests are:

  • variational simulations of nuclear motion dynamics of weakly-bound molecular complexes with an emphasis on the dissociation
  • simulations of photoelectron imaging techniques, such as photoelectron circular dichroism and laser induced electron diffraction.

We are also interested in high-resolution spectroscopy of molecules, especially the study of forbidden transitions and symmetry-breaking phenomena like ortho-para nuclear spin interactions and molecular chirality.

Machine learning

Machine learning has made a significant inroad into the natural sciences in recent years, providing efficient computational technology of highly expressive parametric functions. In two DASHH-funded projects, we are developing deep neural network algorithms for calculating highly excited and pre-dissociation vibrational states of molecules, as well as long-range excursion dynamics of photoelectrons for modelling of imaging experiments.

We also apply machine learning approaches for modelling of molecular potential energy surfaces. Calculating potentials for weakly-bound complexes is especially difficult and computationally expensive due to the loosely bound nature of the intermolecular interactions, which results in a complicated shape of the potentials with local minima and saddle points. In our recently published article, "Active learning of potential-energy surfaces of weakly-bound complexes with regression-tree ensembles," we proposed and investigated an active learning algorithm that allows for the construction of molecular PESs with reduced computational costs without sacrificing accuracy.

Laser control of molecules

Laser-controlled rotational-vibrational molecular dynamics is a subject of active research in physics and chemistry. In particular, the control of molecular spatial alignment and orientation is highly leveraged in many ultrafast imaging experiments and stereochemistry studies to reduce the blurring of observables and increase the experimental resolution. Laser-field control of chiral molecules is of particular interest because of the challenges of detecting the enantiomeric excess and handiness in chiral mixtures at ultrafast timescales as well as chiral purification and discrimination.
Through comprehensive theoretical predictions, we provide essential analysis for laser alignment experiments of complex molecules and develop new methodologies for realistic laser experiments to detect chirality, spatially separate enantiomers, and induce chirality through spontaneous symmetry breaking.

Precision spectroscopy

Recent advances in high-intensity tunable laser sources and improvements of detection sensitivity enable laboratory detection of ultra weak transitions in molecules, such as due to molecular quadrupole moment or nuclear spin flips. Using variational methods TROVE and RICHMOL we making predictions such transitions in small molecules with unprecedented accuracy. As an example, our calculated linelist of quadrupole transitions in enabled first experimental detection of such transiitons in gas phase water at room temperature. Our predicted quadrupole transitions in CO2 were found in atmosphere of Mars and detected in the laboratory.