Greater Boston Area Theoretical Chemistry Lecture Series

2015-2016 Speaker Schedule

On systems with and without excess energy in environment: ICD and other interatomic mechanisms

10/14/15 4:00pm

MIT Building 4, Room 163

Lorenz Cederbaum

Universitat Heidelberg




Lorenz Cederbaum

How does a microscopic system like an atom or a small molecule get rid of the excess electronic energy it has acquired, for instance, by absorbing a photon? If this microscopic system is isolated, the issue has been much investigated and the answer to this question is more or less well known. But what happens if our system has neighbors as is usually the case in nature or in the laboratory? In a human society, if our stress is large, we would like to pass it over to our neighbors. Indeed, this is in brief what happens also to the sufficiently excited microscopic system. A new mechanism of energy transfer has been theoretically predicted and verified in several exciting experiments. This mechanism seems to prevail “everywhere” from the extreme quantum system of the He dimer to water and even to quantum dots. The transfer is ultrafast and typically dominates other relaxation pathways.
Can there be interatomic/intermolecular processes in environment when the system itself (again, an atom or small molecule) does not possess excess energy? The answer to this intriguing question is yes. The possible processes are introduced and discussed. Examples and arguments are presented which make clear that the processes in question play a substantial role in nature and laboratory. Work on the interatomic processes discussed can be found in the Bibliography.

Statistical Physics of Adaptation

10/21/15 4:00pm

MIT Building 4, Room 163

Jeremy England

MIT




Jeremy England

Many-body systems that are driven far from thermal equilibrium can exhibit a seemingly endless range of different "self-organization" phenomena, whether during long periods of transient relaxation over a hierarchy of timesacles, or in an ergodic steady-state. Indeed, the range of possible behaviors is so diverse that it includes (but is not limited to) everything that living things do! In the face of such phenomenological diversity, it is difficult to articulate any thermodynamic commonality that might be analogous to the tendency to minimize free energy observed in equilibrated systems. Here, we try to exploit recent fundamental progress in our understanding of far-from-equilibrium dynamics to develop predictive thermodynamic principles for a general class of driven self-organized systems. We find there is a language in which Darwinian selection in biological systems may be thought of as a special case of a more general physical tendency for "dissipative adaptation" that arises from the correlation between irreversible changes in shape and the absorption of external work. We close by exploring this hypothesis in different simulation frameworks.

Theoretical spectroscopy using density functional theory: new ideas for long-standing problems

11/17/15 4:00pm

MIT Building 4, Room 163

Leeor Kronik

Weizmann Institute of Science




Leeor Kronik

Accurate prediction of the electronic structure and optical properties is essential for rational design of materials for novel (opto)electronic applications. Quantities of interest include, e.g., the bandgap, band dispersion and band width, optical absorption, exciton binding energies, and more. Preferably, we would like to predict such quantities using density functional theory (DFT), because the relative computational simplicity afforded by DFT allows us to attack realistic problems. Unfortunately, despite many other successes, DFT has traditionally struggled with prediction of the above quantities. Specifically, research has been fraught with very difficult questions as to the extent to which spectroscopic conclusions can be drawn from DFT even in principle, followed by serious concerns as to the reliability of typical DFT approximations in practice. In this lecture, I will start with a tutorial overview on DFT. I will then focus on new formal and practical approaches which offer fresh answers to the above long-standing questions. In particular, I will show that DFT can, in many cases, mimic successfully the quasi-particle picture of many-body theory, allowing for quantitative calculations of both single- and two-particle excitations. I will show how this is achieved for finite systems, present initial generalizations to solids, and discuss limitations and remaining challenges.

Plasmonic Nanostructures: Artificial Molecules

12/02/15 4:00pm

MIT Building 4, Room 163

Peter Nordlander

Rice University




Peter Nordlander

The recent observation that metallic nanoparticles possess plasmon resonances that depend sensitively on the shape of the nanostructure has led us to a fundamentally new understanding of the plasmon resonances supported by metals of various geometries. This picture-“plasmon hybridization”, reveals that the collective electronic resonances in metallic nanostructures are mesoscopic analogs of the wave functions of simple atoms and molecules, interacting in a manner that is analogous to hybridization in molecular orbital theory. The new theoretical insight gained through this approach provides an important conceptual foundation for the development of new plasmonic structures that can serve as chemical and biosensors, substrates for surface enhanced spectroscopies, and subwavelength plasmonic waveguides and active optical devices. In the first part of the lecture I will introduce the field of plasmonics and the theoretical techniques that we use to model plasmonic phenomena. In the second part of the talk, I will review several hot applications including aluminum, graphene, molecular plasmonics, quantum plasmonics, plasmonic Fano resonances, and plasmon induced hot carrier generation for photodetection and photocatalysis.

Targeting Excited States with Quantum Monte Carlo

02/10/16 4:00pm

MIT Building 4, Room 163

Eric Neuscamman

University of California Berkeley




Eric Neuscamman

Essentially all of our current methods for modeling electronically excited states are constrained either by their algorithmic cost or by their connection to the ground state. I will discuss two developments in Monte Carlo methods relevant to this situation. The first is the extension of linear response equation of motion theory to variational Monte Carlo, which can be seen as fitting neatly into the current methodological paradigm alongside other linear response theories and which shares most of their advantages and weaknesses. The second development, the unlocking of a new excited state variational principle via Monte Carlo integration, is not so easily categorized and appears to present an altogether new avenue in excited state modeling in which the entire variational freedom of an ansatz is molded around an individual excited state. Results for both methods will be presented, many of which compare favorably against both equation of motion coupled cluster and multireference benchmark data. Finally, potential future uses and improvements for these new methods will be discussed.

Thermodynamics and Kinetics of Deeply Supercooled Water: a Computational Perspective

02/24/16 4:00pm

MIT Building 4, Room 163

Pablo Debenedetti

Princeton University




Pablo Debenedetti

Water, like any other liquid, can be cooled below the equilibrium freezing temperature and still remain in the liquid state: it is then said to be supercooled. Large quantities of supercooled water exist in clouds and play an important role in ice formation, latent heat release, and in the atmosphere’s overall radiative balance. The physical properties of supercooled water have been a source of continued interest since the early ‘70s, when sharp increases in compressibility and heat capacity upon cooling were first reported. One intriguing hypothesis that has been formulated to explain this behavior is the existence of a metastable phase transition between two different liquids at deeply supercooled conditions. The preponderance of experimental evidence is consistent with this hypothesis, although no definitive proof exists to date. State-of-the-art free energy techniques provide clear evidence of a metastable transition between two distinct liquid phases in a molecular model of water. The fact that a phase transition is metastable implies that the possibility of observing it, whether in the computer or in experiments, depends on system size and on the duration of the observation. Understanding the manner in which force field details influence the existence and observability of liquid-liquid transitions is currently a subject of intense study. Although freezing is a ubiquitous phenomenon, large gaps in understanding persist regarding the detailed microscopic mechanism and the rate of ice formation at atmospherically-relevant conditions. Using state-of-the-art computational methods designed to probe rare events, we are able to study the early stages of ice nucleation at deeply supercooled conditions. We observe a competition between cubic and hexagonal ice polymorphs. Transition states are rich in the kinetically-favored cubic ice, rather than in the thermodynamically stable hexagonal ice. These examples illustrate the power of modern computational techniques rooted in statistical mechanics, as well as the considerable challenges that still lie ahead in the quest for accurate and predictive depictions of complex phenomena.

Path Integral Methods for Simulating Nonadiabatic Charge and Energy Transfer Dynamics

03/09/16 4:00pm

MIT Building 4, Room 163

Nandini Ananth

Cornell University




Nandini Ananth

Simulating inherently quantum mechanical charge and energy transfer processes in the condensed phase remains an outstanding challenge to theory. The high computational costs associated with exact quantum dynamic simulations of large systems necessitate the development of approximate methods that (i) exhibit low computational cost and near-linear scaling with system dimensionality; (ii) capture important quantum effects including tunneling, zero-point energy, and coherence effects; and (iii) employ a consistent framework to describe transitions between electronic states and the dynamics of nuclei to accurately capture nonadiabatic processes. In this talk, we will discuss several well-established, approximate methods derived from the real- time and imaginary-time path integral formulations of quantum mechanics. We will explore their regimes of applicability based on the criteria outlined above. We then introduce a novel method developed by our group for the direct dynamic simulation of photo-induced chemical reactions in the condensed phase, Mapping-Variable Ring Polymer Molecular Dynamics (MV-RPMD). Finally, we will numerically demonstrate the accuracy and applicability of this method in simulations of photo-induced excited electronic state dynamics.

Understanding evolution on multiple scales: from protein physics to population genetics

03/23/16 4:00pm

MIT Building 4, Room 163

Eugene Shakhnovich

Harvard University




Eugene Shakhnovich

Biological phenomena unfold in a broad range of scales ranging from molecules to cells to populations and ecosystems. Variation of molecular properties of biomolecules profoundly impact the ability of cells to survive and propagate (fitness). Finally, the fate of a mutation is decided by Darwinian selection on the level of the population, where three outcomes are possible: fixation in the population, elimination by purifying selection or separation in the population in a subdominant clone (polymorphism). In this lecture I will outline my lab’s and others efforts in an emerging new field which merges molecular mechanism with evolution. I will review basic concepts and models that contributed to our multiscale understanding of physical-chemical basis of biological phenomena. First, I will discuss physical chemistry of protein folding. I will present the fundamental heteropolymer model of protein folding and briefly outline the statistical mechanical analysis, which uncovered the energy gap criterion - the necessary and sufficient conditions for a heteropolymer sequence to encode a foldable protein. I will highlight the analogy and fundamental differences between heteropolymer and spin glass models. I will also discuss how understanding of basic principles of protein folding helps in our efforts to design new proteins and decipher the ‘’messages’’ hidden in multiple sequence alignment. I will then discuss the analogy between sequence selection for energy gaps and statistical mechanics of a class of generalized spin models. The statistical mechanical view of sequence selection enjoyed renaissance with the development of statistical methods to derive structural information about proteins from the analysis of variation in multiple sequence alignment. Finally I will discuss the relation between selection for foldable sequences and thermodynamic and kinetic mechanisms of protein folding such as first-order-like cooperativity. Next, I will present recent efforts at modeling evolutionary dynamics that merges molecular mechanisms with population genetics. Traditional population genetics models are agnostic to the physical-chemical nature of mutational effects. Rather they operate with an a’priori assumed distributions of fitness effects (DFE) of mutations from which evolutionary dynamics are derived. Alternatively some population genetics models aim to derive DFE from evolutionary observations. In departure with this tradition the novel multiscale models integrate the molecular effects of mutations on physical properties of proteins into physically intuitive yet detailed genotype-phenotype relationship (GPR) assumptions. I will present a range of models from simple analytical diffusion-based model on biophysical fitness landscapes to more sophisticated computational models of populations of model cells where genetic changes are mapped into molecular effects using biophysical modeling of proteins and ensuing fitness changes determine the fate of mutations in realistic population dynamics. Examples of insights derived from biophysics-based multiscale models include the scale-free character of Protein Universe, the fundamental limit on mutation rates in living organisms, physics of thermal adaptation, co-evolution of protein interactions and abundances in cytoplasm and related results, some of which I will present and discuss. Finally I will briefly present “bottom-up: experimental efforts based on genome editing approaches to test basic assumptions of multiscale biophysics-based models of evolution.

Canceled

03/30/16 4:00pm

Ivet Bahar

University of Pittsburgh




Ivet Bahar

Temporary yet fatal: metastable electronic states as a gateway for electron-attachment induced chemistry.

04/06/16 4:00pm

MIT Building 4, Room 163

Ksenia Bravaya

Boston University




Ksenia Bravaya

Electronic states metastable with respect to electron ejection are ubiquitous in highly energetic environment, chemical, and biological systems, and often lead to chemical destruction. Prediction of the energetics and lifetimes of the metastable states, resonances, is crucial for understanding the processes of electron capture and the resulting chemical conversion. In this talk I will give a general introduction to the phenomenon of metastable states and to the theoretical tools for treatment of resonances. Specifically, the focus will be on non-Hermitian quantum mechanics approaches for calculating energies and lifetimes of metastable electronic states from the first principles [1]. Non-Hermitian formalisms allow one to exploit quantum chemistry methods developed for conventional bound electronic states for treatment of resonances, which belong to continuous spectrum. I will also discuss the role of the metastable states in chemical and biological processes. In particular, I will present the result of our recent computational studies of electronic structure para-benzoquinone, prototypical biological electron acceptor, and highlight the role of resonances in electron capture by the molecule [2].

1. D. Zuev, T.-C. Jagau, K.B. Bravaya, E. Epifanovsky, Y. Shao, E. Sundstrom, M. Head-Gordon, and A.I. Krylov. Complex absorbing potentials within eom-cc family of methods: Theory, implementation, and benchmarks. J. Chem. Phys., 141:024102, 2014.

2. A.A. Kunitsa and K.B. Bravaya; First-principles calculations of the energy and width of the 2Au shape resonance in p-benzoquinone, a gateway state for electron transfer. J. Phys. Chem. Lett., 6:1053–1058, 2015.

Feel the interactions: Achieving chemical accuracy through many-body representations

04/13/16 4:00pm

MIT Building 4, Room 163

Francesco Paesani

University of California, San Diego




Francesco Paesani

Two of the most challenging problems at the intersection of electronic structure theory and computer simulations are the accurate representation of intermolecular interactions and the development of reduced-scaling algorithms applicable to large systems. To some extent, these two problems are antithetical, since the accurate calculation of intermolecular interactions requires correlated methods whose computational scaling with respect to system size precludes the application to large systems. In this talk, I will describe our many-body molecular dynamics (MB-MD) methodology that has been shown to enable computer simulations with chemical and spectroscopic accuracy from the gas to the condensed phase. MB-MD combines many-body (MB) potentials derived entirely from “first principles” with quantum molecular dynamics (MD) methods based on the path-integral formalism. Our MB potentials are built upon the many-body expansion of the interaction energy and contain explicit 2-body and 3-body terms represented by permutationally invariant polynomials obtained from the application of machine learning techniques to correlated electronic structure data. These terms smoothly transition into a sum of classical expressions for the electrostatic and dispersion energies that reproduce the correct asymptotic behavior. Similarly, many-body representations of the dipole moment and polarizability surfaces are derived. The combination of these many- body representations with quantum dynamics methods enables the simulation of (linear and nonlinear) infrared, Raman, and sum-frequency generation spectra, which can be directly compared with experiment without requiring any empirical frequency shift or ad hoc scaling of the spectral intensity. I will illustrate the accuracy and predictive power of our MB-MD methodology through the analysis of the properties of aqueous systems from the gas to the condensed phase, with a particular emphasis on nuclear quantum effects and vibrational spectroscopy.

Topological phase effects in molecular dynamics beyond the Born-Oppenheimer approximation

04/20/16 4:00pm

MIT Building 4, Room 163

Artur Izmaylov

University of Toronto




Artur Izmaylov

The Born-Oppenheimer approximation for molecules introduces natural separation between dynamics of electrons and nuclei. This separation allows one to consider nuclear chemical dynamics independently from that of the electronic subsystem. However, there are also a few complications associated with cases when several electronic potential energy surfaces become similar in energy or even cross. The latter case often presents itself in the form of a conical intersection. For example, in many photochemical, charge and electronic energy transfer processes nuclear molecular dynamics proceeds near conical intersections. In this talk, I will discuss how non-trivial topological phases associated with conical intersections of electronic surfaces affect chemical dynamics of molecular systems and what intuitive picture can be extracted from exact numerical simulations for simple models.

04/27/16 4:00pm

MIT Building 4, Room 163

Joel Eaves

University of Colorado at Boulder




Joel Eaves

Porous two dimensional crystals offer many promises for applications in water desalination, but for computer simulation to play a predictive role in this area, one needs to have reliable methods for simulating an atomistic system in hydrodynamic currents.In this talk I will describe our methodology for simulating nonequilibrium steady states in atomistic simulations and show how statistical mechanical models give microscopic insights into the interactions that control aqueous flows through atomically thin and narrow channels. I will devote the first part of the talk to developing methods, models, concepts and interpretative tools that include Gaussian constraint dynamics and thermostat development in molecular simulation, Markov Models, and continuous time random walks. In the second part of the talk, I will discuss water passage through porous two-dimensional materials in detail, with particular emphasis on the hydrophobic effect. In this class of materials, hydrophobicity plays a static and dynamical role that can both help and hinder water transport.

Accelerating Materials Discovery with Data-Driven Atomistic Computational Tools

05/04/16 4:00pm

MIT Building 4, Room 163

Chris Wolverton

Northwestern University




Chris Wolverton

Many of the key technological problems associated with alternative energies (e.g., thermoelectrics, advanced batteries, hydrogen storage, etc.) may be traced back to the lack of suitable materials. Both the materials discovery and materials development processes may be greatly aided by the use of computational methods, particular those atomistic methods based on density functional theory (DFT). Here, we present an overview of our recent work utilizing high-throughput computation and data mining approaches to accelerate materials discovery, specifically highlighting three new approaches: (i) We describe our high-throughput DFT database, the Open Quantum Materials Database (OQMD)1, which contains over 280,000 DFT calculations and is freely available for public use at http://oqmd.org. (ii) We show how computational crystal structure solution may be addressed via a new hybrid approach, the First-Principles Assisted Structure Solution (FPASS) approach2, which combines experimental diffraction data, symmetry information, and first-principles-based evolutionary algorithmic optimization to automatically solve crystal structures. (iii) We also describe a newly-developed machine learning approach3,4 to construct a materials screening model based on an extensive set of thousands of DFT calculations. The resulting model, which has “learned” rules of chemistry from these many examples, can predict the stability of arbitrary compositions without requiring any a priori knowledge of crystal structure, at about six orders of magnitude lower computational expense than the original QM tools. We use this model to scan—in a matter of minutes—roughly 1.6 million candidate compositions for novel ternary compounds (AxByCz), and predict roughly 4,500 new stable materials.

1) J. E. Saal, S. Kirklin, M. Aykol, B. Meredig, and C. Wolverton "Materials Design and Discovery with High-Throughput Density Functional Theory: The Open Quantum Mechanical Database (OQMD)", JOM 65, 1501 (2013).

2) B. Meredig and C. Wolverton, "A Hybrid Computational-Experimental Approach for Crystal Structure Solution" Nature Materials 12, 123 (2013).

3) B. Meredig and C. Wolverton, “Dissolving the Periodic Table in Cubic Zirconia: Data Mining to Discover Chemical Trends” Chem. Mater. 26, 1985 (2014).

4) B. Meredig, A. Agrawal, S. Kirklin, J. E. Saal, J. W. Doak, A. Thompson, K. Zhang, A. Choudhary, and C. Wolverton, "Combinatorial screening for new materials in unconstrained composition space with machine learning", Phys. Rev. B 89, 094104 (2014).

Theory and Simulation of Functional Dynamics of Biomolecules

05/11/16 4:00pm

MIT Building 4, Room 163

Gerhard Stock

University of Freiburg




Gerhard Stock

Driven by recent experimental and computational progress, structural dynamics has been recognized as an essential player for the functionality of proteins. The importance of functional dynamics, e.g., in folding, aggregation, binding and signaling, extends the classic structure-function paradigm of biochemistry to include the 'dynamical personality' of proteins. As an overview, I start with a brief introduction to classical all-atoms molecular dynamics simulations, with an emphasis on the dynamical description of biomolecules. In particular, I introduce dimensionality reduction methods to construct the system's free energy landscape, that characterizes its metastable conformational states as well as the barriers connecting these states. The second part is concerned with two topics. First I consider hierarchical dynamics of biomolecular processes which accounts for the nonlinear coupling of fast (picosecond) fluctuations to slow (microseconds) structural rearrangement. Then, I present recent extensive molecular dynamics simulations of the propagation of conformational change in intramolecular signaling in a photoswitchable PDZ domain.

Past Schedules