For the screening novel ionic liquids (ILs), it’s important to predict the crystal structure of solid-state ILs so that the melting-point of the ILs can be calculated using already developed methods. We introduce a method combining genetic-algorithm-based structure evolution and forcefield-base molecular dynamics to predict the most energetically stable structure of an IL. Our example showed the structure evolution can be successfully, and the time cost of this method is quite short compared with DFT approach. We believe that this is a feasible method to predict the crystal structure of novel ILs which can facilitate the prediction of their melting-points.
Continue reading[PHYS-20]Elucidating the co-existence of classical and non-classical nucleation pathways in heterogeneous ice nucleation
We employ Markov States models (MSMs) and transition path theory to elucidate the ensemble kinetic pathways of heterogeneous ice nucleation (HIN) simulated by Molecular dynamics simulations. Interestingly, our MSM reveals the coexistence of classical and non-classical nucleation pathways with comparable fluxes at T=230K. Specifically, the classical HIN pathway shown by purple arrows, follows a typical one-step ice nucleation process, corresponding to a direct transition from liquid to hexagonal ice (green spheres). While the non-classical nucleation shown in red and orange arrows proceeds via two-step pathways with square ice (violet spheres) formed as an intermediate state.
Continue reading[PHYS-19]Singlet fission modulated by electron-phonon interactions in crystalline pentacene
[PHYS-18]Can we predict molecular properties with limited experimental data?
[PHYS-17]Development and implementation of OSV-MP2 theory in many-body expansion for structural and dynamical properties of large molecules
OSV approximation has been proved to bring significant improvement in scaling performance and memory usage for post Hartree-Fock methods. To bring multiprocessing to the OSV-MP2 computations, we developed a many body partitioning scheme which can divide a molecule system into one-body clusters having one occupied orbital and corresponding OSVs. Due to the small size of one-body clusters, two- and three body correlations are considered with a proper screening process. The current poster demonstrates the algorithm and implementation of MBE(3)-OSV-MP2. and scaling evaluations are carried out with geometry optimization and molecular dynamics simulation.
Continue reading[PHYS-16]Modelling the diffusion and degradation of superoxide in the electrolyte of aprotic Na-O2 batteries
Previously we have developed a model to study the degradation of NaO2(sol.) in aprotic Na-O2 batteries. However, we did not consider the spatial distribution of NaO2. Here, a new model is developed that involves the diffusion of NaO2(sol.) along axial direction, which is closer to actual experimental condition. This work helps understand the behavior of NaO2(sol.) and provides guidance for designing better Na-O2 batteries.
Continue reading[PHYS-15]High rate capability of BiOCl electrode for aqueous batteries
BiOCl on carbon cloth was synthesized by solution combustion method by impregnating the precursor solution into carbon cloth. Different oxygen acceptors are responsible for BiOCl reduction in neutral and acidic media as shown in equations (1) and (2), while the kinetics of (2) was proved to be much better. Here we propose that presence of hydrated Al3+ increases the availability of proton favorable for BiOCl reduction as well as the overall redox reaction.
Continue reading[PHYS-14]Visible-light-driven alloy nanomotor by composition programming
In this study, an inorganic core-shell nanomotor with high-quality epitaxial junction is demonstrated with continuous tunability in the visible spectrum due to composition variation. The spectrum response of the nanomotor migration matched with single nanowire solar cell experiment and optical measurement data. Also, the speed of the nanomotor migration also showed correlation with the spectrum.
Continue reading[PHYS-13]Developing downfolded ab-initio Hamiltonian method for molecular excited states
We present a formulation and algorithm for targeting an individual excited state in which a downfolded Hamiltonian is diagonalized in a designated embedded subspace of much reduced dimension built from Feshbach projection. A relatively smaller active space will be pre-calculated for excited state tracing in a multi-reference manner.
Continue reading