Research projects

I offer research projects in computational chemistry and scientific computing.

Example projects and presentations


During the Covid crisis Jeremy David (center) continued to work on two projects.

  • Characterization of temporary Be32- and Mg32- cluster cluster dianions. Some electronic states of these short-lived clusters are σ or π aromatic, but aromaticity isn't associated with increased stability or lifetimes. The results were published in J. Phys. Chem..

  • Comparison of the three workhorse L2 methods for characterizing temporary anion states: Complex absorbing potentials, Howard-Taylor stabilization, and regularized analytic continuation. Each method has in turn variants, but the computational protocol is deceptively method/variant independent. This is a long-term project, which is available as a github repository, and the current state has been accepted for publication in Eur. Phys. J. D.

These two projects were presented at three on-line meetings. At the 2021 Louisiana Academy of Sciences meeting, Jeremy gave a talk, and at the 2021 Spring national ACS and the 2021 POSMOL meetings Jeremy presented posters.


Jeremy Davis continued a project started by Megan Davis (no relation) and presented it at the 2019 Southwest Theoretical Chemistry Conference (SWTCC) held in Norman, OK.

Electron binding energies of dipole-bound states can, of course, be simply predicted by doing high quality ab initio calculations, however, we would like to be able to understand and predict trends within and maybe even beyond classes of related molecules. In this project, we try to identify descriptors of the neutral molecule to do so. Apart from the dipole and the polarizability, the so-called excluded volume turns out to be essential.


Joshua Melugin, Prakash Hamal, Megan Davis, and Januka Khanal investigated the intermolecular interaction between the Lewis acids B(OH)3, BH3, and BF3 with the Lewis base NH3.

This project is a collaboration with our organic chemist Dr. Dolliver, who uses a Boron-based catalyst in her hydroboration reactions. In the crystalline catalyst the boron atom is bound to three O and one N atoms, but in solution one of these bonds has to open up, so that the B atom can play the role of an active site. We studied the small systems first, because the most popular density functionals are unsuitable for this type of Lewis acid-base bond, and established an efficient and reliable computational protocol to study the catalyst. Joshua had the honor to present these findings at the University of Louisiana System Day, which was held at the State Capitol in Baton Rouge (May 2015).


Nicole Blondeau and Robin Joshi studied extrapolation methods as a means of predicting the energies of temporary anions.

Extrapolating bound energies is attractive because it is much cheaper and easier than true continuum methods, yet extrapolations as such are always problematic, and, in particular, extrapolating through several avoided crossings with discretized continuum remains a challenge. Nicole presented her findings at the 2013 Spring meeting of the ACS.


Katelyn Dreux investigated the fate of Lithium's 2s electron when the Lithium atom is solvated by one to four ammonia molecules.

Katelyn showed that the electron occupies a Rydberg-like orbital, that is, a 2s-like orbital, of the positive Li(NH3)4 core. In contrast to previous analyses, Katelyn established that the excess electron density near the N atoms is small and that the small amount close to the N atoms is independent almost independent of computational method employed. The work was presented the 20th Conference on Current Trends in Computational Chemistry in Jackson (October 2011) and published in the Journal of Chemical Physics .


Bijay Bhattarai investigated the binding of and "excess" electron to sodium chloride clusters (NaCl)n with n = 1, 2, 3, and 4.

Bijay found that the smallest salt cube possible, (NaCl)4, can bind an electron. However, this electron is not bound by electrostatics, but by electron-correlation effects. In other words, the (NaCl)4- cube represents a true correlation-bound state that is predicted to be unstable by Hartree-Fock method, Moeller-Plesset perturbation theory, and even standard coupled-cluster methods. However, binding is predicted by direct methods like EA-ADC and EA-EOM-CCSD. Bijay's findings suggest a new interpretation of observed ion counts in mass spectrometric experiments, and his results together with more calculations done by collaborators at the University of Heidelberg have been published published in the Journal of Chemical Physics.

[Item] Becky Weber investigated empirical methods for the study of metastable intermediates in electron-induced reactions, and presented her findings at the 18th Conference on Current Trends in Computational Chemistry in Jackson (October 2009). Her results were published in the Journal of Physical Chemistry.

Max McCray (at the computer) studied doubly and triply-charged negative ions. These ions are unknown in solids or solutions, but have been detected using mass spectrometry.

From the mass spectra obtained at the University of Arizona we knew that the dianion C5O22- existed, but its structure and electron binding energy were unknown. Max identified several likely candidates for the observed ions, and computed their relative energies as well as their electron binding energies. His work has been published in the International Journal of Mass Spectrometry.


Becky Weber investigated the mechanism of an azide synthesis studied by our organic chemist Dr. Dolliver.

Becky found the structures and energies of the reactant, two possible products, various intermediates, and two transition states connecting the intermediates. Her work established a mechanism for these reactions, and enabled us to interpret the experimental yields obtained in different solvents. She presented her results at the National ACS meeting in New Orleans (April 2008), and she was a coauthor on a paper published with Dr. Dolliver's group in the Journal of Physical Organic Chemistry.