Two posts in one day!?!
Jonathan’s paper has also been published in JCP (directly after Naresh’s!). In this paper, we derive the full multicomponent MP3 and MP4 equations for the first time. Multicomponent MP4 performs surprising well considering the well known failures of multicomponent HF. This paper is also the first time any connected triple-excitation contribution has been included in a many-body multicomponent method and these triple excitations are shown to increase the accuracy of proton affinities significantly.
Kurt did most of the derivations of this paper, while Jonathan did the herculean job of coding it all up. The reason I (Kurt) mention this is that I can still remember being a sophomore undergraduate just starting research and being handed a copy of Szabo and Ostlund and thinking the Hugenholtz and Goldstone diagrams in it and on the cover were the COOLEST thing ever (not that I understood them until a few years ago). While I may no longer think they are the “coolest thing ever,” I still am pretty happy in a silly way that this is the first paper I have been a part of that has made use of them.
Link is here: https://doi.org/10.1063/5.0071423
It has been a while since we have had a post on the group website. Part of the reason for this is that Kurt has been teaching General Chemistry for the first time this semester (his review of the experience: lots of work, but fun otherwise).
But while he was doing that, the rest of the group continued working hard. Today, Naresh’s first paper in the group was published in JCP! In this paper, Naresh takes our previous implementation of multicomponent HBCI and (1) extends it to excited states and (2) implements the perturbative energy correction. Both of these are important extensions for a fully featured multicomponent HBCI method and appear to perform well.
Kurt also added some results to the study about how to best benchmark multicomponent methods for excited states using FGH. He has had these ideas since at least his first semester at Mizzou. Basically, in order to get appropriate FGH reference results you need to restrain the electronic basis to not follow the quantized particle when performing FGH.
Link to paper: https://doi.org/10.1063/5.0076006
The group has seen some turnover in the last month.
First, Jonathan has taken Horace Greeley’s advice and headed west to Stanford for graduate school. We wish him the best and are excited for his future.
We also have two new members of the group!
The first is Naresh Alaal, who is a postdoctoral researcher coming to us from Kaust in Saudi Arabia. Naresh was supposed to start in April of 2020, but then the pandemic happened, so we were unable to get him a visa. We are thrilled that he is finally able to start working with us.
The second is Gabbie Tucker, who is a Stevens Scholar for the summer and is currently an undergraduate at the College of the Ozarks. Already in her short time here, she has shown a lot of potential as a future theoretical chemist and we are enjoying seeing her grow as a researcher.
Abu’s paper has been published in Molecular Physics! In this study, we take the formulation of vibrational heat-bath configuration interaction by the Berkelbach group and implement an alternative formulation by explicitly calculating and storing all Hamiltonian matrix elements. Like a lot of quantum chemistry methods, which one of the formulations is best seems to depend on the system and the level of accuracy required. Our technique appears to work better for ethylene oxide, while the previous formulation does better for naphthalene. Link is here.
Jonathan’s paper has been published in JCTC. In this paper, we implement a new multicomponent CASSCF method using the multicomponent heat-bath CI algorithm as the CI solver. We previously used this CI solver in our multicomponent CISDTQ study. This allows us to use extremely large active spaces relative to previous multicomponent CASSCF studies. It turns out dynamical correlation is important for obtaining accurate protonic densities (which really isn’t that great of a surprise), as we find that very large active spaces are required for even qualitatively accurate protonic densities. We hope to have plenty more to come on this and similar topics in the future. Link is here: https://pubs.acs.org/doi/full/10.1021/acs.jctc.0c01191