By applying hierarchal formalism to multi-electric states, we can investigate laser excitation and photoisomerization process described by multi-electric state numerically rigorously. As an example, we computed Wigner distributions for excited, ground, and coherent states. We then investigated excited state dynamics involving transitions among these states by analyzing electronic spectra with various values of the heat bath parameters. Our results provide predictions for spectroscopic measurements of photoexcitation and photoisomerization dynamics.

Y. Tanimura, Real-Time and Imaginary-Time Quantum Hierarchal Fokker-Planck Equations, J. Chem. Phys. 142, 144110 (2015).

T. Ikeda and Y. Tanimura, Probing photoisomerization processes by means of multi-dimensional electric spectroscopy: The Multi-state quantum hierarchal Fokker-Planck Equation approach, J. Chem. Phys. 146, 014102 (2017)

In this talk, I will give an overview of the theory of mixed quantum-classical nuclear-electronic motion both in solution and at metal surfaces, focusing on both stochastic and mean-field descriptions. This theory gives a clear and intuitive picture of electronic relaxation in a condensed environment, e.g. how photo-excitation can lead to macroscopic charge transfer and how the second law of thermodynamics (i.e. entropy growth) emerges from microscopically reversible dynamics. I will attempt to give an introductory but also deep account of this rich field of study. Towards the end, I will highlight open questions, especially at metal surfaces, where propagating ab initio calculations remains very daunting.

Recent advances in computational materials science have made ab initio/quantum simulations, at the density functional level, increasingly feasible from a computational standpoint. Unfortunately, the accessible length and time scales are still limited, of the order of a few nanometers in length and tens of picoseconds in time, necessitating the development of empirical force-fields that can capture the structure and dynamics at significantly longer scales. The focus of my research is the development of these types of classical models or force-fields, both atomistic and coarse-grained, that are computationally efficient yet capture the essential physics of the problem. Furthermore, my group also works on developing reactive force-fields that allow for changes in bond topology. We use a bottom-up approach to develop these models based on reference data from higher level simulations. The first part of my talk will outline the development of these types of models. The second part of my talk will focus on some interesting applications that my group is working on. There are three main thrusts: solvation and aggregation of biomimetic polypeptoids, gas-liquid nucleation in the presence of acidic defects (relevant to atmospheric chemistry), and finally the study of novel electrolytes for next generation rechargeable battery technologies like sodium-air batteries.

How proteins fold into native structures in physiological conditions is one of the most fundamental questions in molecular biology and biophysics. Until now, a huge number of theoretical, computational, and experimental studies have been carried out to answer this question. In the lecture, I first introduce standard computational approaches, such as coarse-grained simulations based on the Go-model and enhanced conformational sampling algorithms like replica-exchange molecular dynamics (REMD) methods. Then, I explain our machine learning approach connecting time-series data of single-molecule experiments with molecular dynamics simulations. This method is free from the force-field bias and can provide the conformational ensembles of proteins that match with the single-molecule experimental data. I also compare the usage of single-molecule time-series data to the ensemble-average data in the machine learning approach.

Density-functional theory and its descendants are the workhorse methods of computational materials science, despite a number of now well-known shortcomings. In a different direction, I will provide a broad overview of systematically-improvable wavefunction-based methods for condensed-phase electronic structure calculations. I will review the challenges of electronic structure in bulk materials, touching on aspects of many-body theory, diagrammatic techniques, and open-source software. With motivating examples drawn from recent work on low-dimensional semiconductors, I will describe our ongoing research agenda aimed at using deterministic methods (especially coupled-cluster theory) and stochastic methods (quantum Monte Carlo) for the accurate description of condensed-phase electronic structure, excited-state properties, and spectroscopic observables.

A general goal in our quest to control matter and energy is the design of strategies to control electronic properties and electron dynamics using coherent laser sources. In addition to its interest at a fundamental level, lasers permit manipulation on an ultrafast timescale opening the way to control the ability of matter to chemically react, conduct charge, absorb light, or other properties, in a femto to attosecond timescale.

In this talk, I will summarize our efforts to understand electronic decoherence processes in molecules that are deleterious to interference-based scenarios for the laser control. In addition, I will discuss how, through Stark effects, non-resonant light of intermediate intensity (non-perturbative but non-ionizing) can be used to generate “laser-dressed” molecules and materials with non-equilibrium properties that can be very different from those observed by matter near thermodynamics equilibrium. In particular, I will discuss how Stark effects can be employed to turn transparent nanomaterials into broadband absorbers, and to generate currents in nanoscale junctions.

This lecture will start with a review of the statistical thermodynamic basis for the conventional treatment of chemical equilibrium, and a discussion of how this treatment breaks down for small systems. This fundamental derivation inspired a “brute-force” strategy based on explicit enumerations of integer partitions, to allow ensemble-averaged equilibrium concentrations to be related to cluster free energies even for small systems. A more elegant and computationally cheaper approach called “PEACH” (Partition-Enabled Analysis of Cluster Histograms) formulated in collaboration with members of the number theory group at Emory, will then be presented. Evidence from coarse-grained simulations of an anionic surfactant will be presented to validate the PEACH approach. A case where the method worked suspiciously well (nanodroplet formation from methyl t-butyl ether) will then be discussed, along with how it led us to a simple strategy to reduce complications from non-ideal effects in aggregation equilibria. Finally, PEACH analysis of equilibria involving NaCl clustering in three solvent environments will be used to illustrate qualitatively different pathways towards formation of an ordered precipitate from supersaturated solution.

I will review recent progress in reduced-scaling electronic structure methodology. Many-body electronic structure methods, like coupled-cluster in chemistry and Green's function methods in solid state physics, are capable of significantly higher accuracy than the density functional theory, but their application is hindered by their significantly higher computational cost. Recent rapid advances in reduced-scaling formulation of the coupled-cluster and other many-body theories, in combination with numerical regularization of Coulomb Hamiltonian singularities, reduces or eliminates the cost gap between the many-body methodology and DFT. I will review the key ideas (most of them quite old) behind the modern reduced-scaling electronic structure methods that allow to expose and utilize the sparsity of the electronic wave function, and discuss some of our recent work in this area.

In this seminar I’ll provide an overview for the status and recent developments of the Den- sity Functional Tight Binding (DFTB) model, especially those motivated by biophysical and solid/liquid interface applications. Following an introduction to DFTB, I’ll discuss the treatment of non-covalent interactions with DFTB3, a third-order variant of the DFTB model. In particular, I’ll present improvement of electronic polarization in DFTB3 and benchmark of the model using both gas phase and condensed phase examples. Next, I’ll present progress and challenges for the development of DFTB3 for (transition) metal ions, which play important roles in many metal- loenzymes. I’ll discuss several applications to illustrate the value and remaining issues of DFTB based simulations. If time permits, I’ll also comment on the use of DFTB3(/MM) as the low-level method to drive sampling in multi-level QM/MM free energy calculations.

Intermolecular interactions are ubiquitous in chemistry, biochemistry, and materials. Theoretical approaches to the study of intermolecular interactions that are based on accurate fragmentation methods will be discussed, followed by consideration of interesting applications, including ion solvation, excited state hydrogen transfer and ionic liquids.