Posted on by

Splitting Molecules in the (Quantum?) Computer

Following up our previous work that uses an idea from coupled cluster but in DFT calculations, we reported further application of this concept to compute the binding energies of diatomic molecules and to illustrate the elimination of the well-known problem of fractional charges. This work is the result of productive efforts by Greta Jacobson (2021 REU student at the time, now at U Washington Chemistry), Dr. Marmolejo-Tejada, and myself. DFT simulations struggle sometimes to predict properties of systems where one or more bonds are “stretched” significantly in such a way that fragments with unpaired electrons emerge. Our work shows a practical way to avoid these problematic charges. In this paper we also discuss procedures that save computational resources for future applications. Big thanks to the #LatinXChem 2021 organizers and ChemPhysChem for their invitation to contribute to this special issue in Latin American Physical Chemistry. Link to publication.

Posted on by

Propagating Arbitrary Quantum States Within Coupled-Cluster Theory

Second response couple cluster theory and standard unitary propagations for a model system yield strongly similar results for the prediction of observable evolution (such as a dipole moment), suggesting the possibility of a new way of propagating the state of quantum systems. Our work also discusses the computation of wave function amplitudes from different theory angles.

This paper published in Physical Review A shows our theory based on coupled cluster to propagate in principle arbitrary superpositions of quantum states, while maintaining the desired property of size-extensivity. It, size-extensivity, is a fundamental property that physical “large” quantum systems have, and that is difficult to satisfy theoretically speaking. The idea is based on the concept of “second response”: If a system that is initially a linear combination of ground and excited state is propagated, where the “weight” of the excited state is small, by varying such weight one can infer how the excited state would evolve alone (free of the ground state). This idea is somewhat trivial, but when applied to the time-dependent coupled cluster wave function, we obtain a set of equations that describe how the excited state, or a combination of quantum states, change as a function of time. Link

Posted on by

Raman Signals and Electronic Features of 2D Heterostructures

In this paper published last year at JPCC, we teamed up with Pierce Fix, Nicholas Borys (MSU Physics), and Patrick Kung (U. Alabama), to show that theory and experiment work together to quantify and interpret spatial non-resonant Raman spectra of 2D MoxW1-xS2. We track the evolution of the dominant modes as a function of composition in the heterostructure, and this includes the characterization of S-vacancy states. This allowed us to reverse-engineer the interface and observe that the averaged variation of composition in terms of position, indicating a smooth transition between the non-hybridized regions. Link

Posted on by

Size-Consistency and the Exponential Ansatz

We recently showed that the exponential Ansatz applied to DFT calculations eliminates fractional charges in spin-unrestricted calculations. It also offers ways to study, within single-particle formalisms, states where multiple electrons are excited (not just a single electron as in standard linear response methods). Looking forward to a more extended implementation of the method. This work is also compatible with the addition of two-body interactions. More info (Journal of Physical Chemistry A link): bit.ly/3oGgYq1

Posted on by

Germanium Fluorides

Our recent work on Germanium Fluoride nanocages was accepted by the Journal of the American Chemical Society. In this work, we studied with DFT a series of high-symmetry molecules with Formula GenFn, where n is between six and sixty. The molecules we studied show optical transparency or semi-transparency, electron delocalization, LUMO energies below -4 eV, and gaps in the range 1.6-3.2 eV. Although the formation energies are high, the fullerene-like systems have isomers with lower ground-state energies, where some Fluorine atoms are located inside the cages. So the fullerene below could undergo a transformation with F atoms preferring to be inside; but the barrier for such a transition should be very high. Link https://pubs.acs.org/doi/10.1021/jacs.8b11259

Posted on by

Releasing space in a MacBook Pro

Many of the applications I use require the gigantic package Xcode, which is often updated. An issue with these updates is that they always re-install a set of platforms. These are used I think to simulate several devices. Unfortunately they take up a lot of space. The platforms are located in /Applications/Xcode.app/Contents/Developer/Platforms. `du -hd 1` shows
1.4G ./AppleTVOS.platform
2.2G ./iPhoneOS.platform
1.1G ./WatchOS.platform
 46M ./MacOSX.platform
5.9M ./WatchSimulator.platform
4.5M ./AppleTVSimulator.platform
6.3M ./iPhoneSimulator.platform
4.7G
I don’t use any of these except MacOSX (I tried to delete it the other day and learned it’s actually required by Xcode). So erasing them (except again MacOSX.platform) saves about 4.7GBs in space, much needed in a laptop with only 128GBs in SSD 🙁