arXiv subjects

Junping Shao

Publications and source records attributed to Junping Shao.

4 recordsLinked to original sources

arXiv2018

Pomeranchuk Instability of Composite Fermi Liquids

Nematicity in quantum Hall systems has been experimentally well established at excited Landau levels. The mechanism of the symmetry breaking, however, is still unknown. Pomeranchuk instability of Fermi liquid parameter $F_{\ell} \le -1$ in the angular momentum $\ell=2$ channel has been argued to be the relevant mechanism, yet there are no definitive theoretical proofs. Here we calculate, using the variational Monte Carlo technique, Fermi liquid parameters $F_\ell$ of the composite fermion Fermi liquid with a finite layer width. We consider $F_{\ell}$ in different Landau levels $n=0,1,2$ as a function of layer width parameter $η$. We find that unlike the lowest Landau level, which shows no sign of Pomeranchuk instability, higher Landau levels show nematic instability below critical values of $η$. Furthermore, the critical value $η_c$ is higher for the $n=2$ Landau level, which is consistent with observation of nematic order in ambient conditions only in the $n=2$ Landau levels. The picture emerging from our work is that approaching the true 2D limit brings half-filled higher Landau-level systems to the brink of nematic Pomeranchuk instability.

cond-mat.str-el↗

arXiv2017

Stratified construction of neural network based interatomic models for multicomponent materials

Recent application of neural networks (NNs) to modeling interatomic interactions has shown the learning machines' encouragingly accurate performance for select elemental and multicomponent systems. In this study, we explore the possibility of building a library of NN-based models by introducing a hierarchical NN training. In such a stratified procedure NNs for multicomponent systems are obtained by sequential training from the bottom up: first unaries, then binaries, and so on. Advantages of constructing NN sets with shared parameters include acceleration of the training process and intact description of the constituent systems. We use an automated generation of diverse structure sets for NN training on density functional theory-level reference energies. In the test case of Cu, Pd, Ag, Cu-Pd, Cu-Ag, Pd-Ag, and Cu-Pd-Ag systems, NNs trained in the traditional and stratified fashions are found to have essentially identical accuracy for defect energies, phonon dispersions, formation energies, etc. The models' robustness is further illustrated via unconstrained evolutionary structure searches in which the NN is used for the local optimization of crystal unit cells.

cond-mat.mtrl-sci↗

arXiv2016

BaSn$_2$: A new, wide-gap, strong topological insulator

BaSn$_2$ has been shown to form as layers of buckled stanene intercalated by barium ions~\cite{Kim_2008}. However, despite an apparently straightforward synthesis and significant interest in stanene as a topological material, BaSn$_2$ has been left largely unexplored, and has only recently been recognized as a potential topological insulator. Belonging to neither the lead nor bismuth chalcogenide families, it would represent a unique manifestation of the topological insulating phase. Here we present a detailed investigation of BaSn$_2$, using both {\it ab initio} and experimental methods. First-principles calculations demonstrate that this overlooked material is a indeed strong topological insulator with a bulk band gap of 360meV, among the largest observed for topological insulators. We characterize the surface state dependence on termination chemistry, providing guidance for experimental efforts to measure and manipulate its topological properties. Additionally, through {\it ab initio} modeling and synthesis experiments we explore the stability and accessibility of this phase, revealing a complicated phase diagram that indicates a challenging path to obtaining single crystals.

cond-mat.mtrl-sci↗

arXiv2014

Entanglement entropy of the $ν=1/2$ composite fermion non-Fermi liquid state

The so-called ``non-Fermi liquid'' behavior is very common in strongly correlated systems. However, its operational definition in terms of ``what it is not'' is a major obstacle against theoretical understanding of this fascinating correlated state. Recently there has been much interest in entanglement entropy as a theoretical tool to study non-Fermi liquids. So far explicit calculations have been limited to models without direct experimental realizations. Here we focus on a two dimensional electron fluid under magnetic field and filling fraction $ν=1/2$, which is believed to be a non-Fermi liquid state. Using the composite fermion (CF) wave-function which captures the $ν=1/2$ state very accurately, we compute the second Rényi entropy using variational Monte-Carlo technique and an efficient parallel algorithm. We find the entanglement entropy scales as $L\log L$ with the length of the boundary $L$ as it does for free fermions, albeit with a pre-factor twice that of the free fermion. We contrast the results against theoretical conjectures and discuss the implications of the results.

cond-mat.str-el↗