Pablo G. Debenedetti
Dean for Research
Class of 1950 Professor in Engineering and Applied Science
Professor of Chemical and Biological Engineering
The Phase Behavior of Supercooled Water: A Computational Perspective
Ever since the pioneering work of Speedy and Angell (J. Chem. Phys., 65, 851, 1976), there has been continued interest in the properties of deeply supercooled water. The possibility that this substance can have a second critical point has inspired a large body of experimental, computational and theoretical work (e.g., Mishima and Stanley, Nature, 392, 164, 1998; Debenedetti, J. Phys.: Condens. Matter., 15, R1669, 2003). Computational investigations have been instrumental in defining the boundaries of contemporary knowledge on this subject (e.g., Poole, Sciortino, Essmann and Stanley, Nature, 360, 324, 1992; Limmer and Chandler, J.Chem.Phys., 135, 134503, 2011; Liu, Palmer, Panagiotopoulos and Debenedetti, J. Chem. Phys., 137, 214505, 2012). In this talk I will illustrate how advanced free energy techniques provide definitive evidence of the existence of a second critical point in an atomistic model of water.
Swanlund Professor of Chemistry
University of Illinois at Urbana-Champaign
Hydrogen Tunneling, Electrostatics, and Conformational Motions in Enzyme Catalysis
The roles of hydrogen tunneling, electrostatics, and conformational motions in enzyme catalysis will be discussed. We have developed hybrid quantum/classical molecular dynamics methods that include the quantum mechanical effects of the active electrons and transferring proton(s), as well as the motions of the entire solvated enzyme. These methods have been used to study the proton and hydride transfer reactions catalyzed by the enzymes dihydrofolate reductase (DHFR) and ketosteroid isomerase (KSI). The free energy profiles are generated along a collective reaction coordinate, and the changes in hydrogen bonding and electrostatic interactions are analyzed along the entire reaction pathway. An analysis of the simulations resulted in the identification and characterization of a network of coupled motions that extends throughout the enzyme and represents equilibrium conformational changes that facilitate the chemical reaction. Mutations distal to the active site are shown to significantly impact the catalytic rate constant by altering the conformational sampling of the entire enzyme, thereby changing the probability of sampling configurations conducive to the catalyzed reaction. We have also developed quantum mechanical/molecular mechanical methodology to calculate the vibrational Stark shifts of thiocyanate probes incorporated into the active site of an enzyme. This methodology is shown to reproduce the experimentally measured vibrational shifts upon binding of an intermediate analog to KSI for two different nitrile probe positions. Analysis of the simulations provides atomistic insight into the roles that key residues play in determining the electrostatic environment and hydrogen-bonding interactions experienced by the nitrile probe. Currently, this approach is being used to study the vibrational shifts of nitrile probes for intermediates along the reaction pathway for DHFR to elucidate the conformational changes occurring during catalysis.
Professor of Biology
Estimating Recent Evolutionary History by Integrating over Genealogy Space
When multiple genes are genomes are sequenced from one or more closely related populations or species, the pattern of variation holds information about the recent evolutionary history. To estimate that history one approach is to join a mutation model and a demographic coalescent model, both of which are functions of an unknown branching genealogy. Then by integrating over genealogies, the demographic history can be inferred from the pattern of variation. Integration approaches, and the families of models and histories that can be addressed this way, will be discussed.
Senior Scientific Computing Expert
International Centre for Theoretical Physics
Observations on Optimizing Scientific Computing Applications
Computational tools are common in many areas of research and researchers are working on increasingly complex problems. High-performance computing resources are in high demand and many researchers worry about parallelizing and optimizing their applications.
However, making an application for a given problem run faster may not always be straightforward or as complicated as expected. Also, optimization and parallelization tightly dependent on each other. This talk will discuss two example cases with two very different approaches: one with no change to the original application(s) at all, and a second where only a complete rewrite from scratch resulted in the desired improvement.
Assistant Professor of Biology
Genomics and the Origin of Species
How populations diverge and form separate species still remains one of life's greatest mysteries. Fortunately, new genomic data and tools are beginning to provide us with a better understanding of how species have multiplied to generate the extraordinary diversity of life that currently exists. My lab uses functional and evolutionary genomic approaches to examine genetic systems across a range of species--from insects to primates--so that we can identify the very evolutionary forces important in speciation. We provide evidence across multiple taxa that, through sexual selection, an interplay between drift and selection on reproductive systems plays a significant role in early species divergence.
Ronald M. Levy
Board of Governors Professor of Chemistry, Rutgers University
Laura H. Carnell Professor of Biophysics and Computational Biology, Temple University (January 1, 2014)
Exploring Landscapes and Timescales for Protein-ligand Binding, Functional Motions, and Fitness
Advances in computational biophysics depend critically on the development of accurate effective potentials and powerful methods to sample the complex energy landscapes that proteins must traverse in order to function. In my talk I will describe work in my lab over the last several years concerning the construction of all-atom effective potentials and multi-scale methods for simulating protein-ligand binding , their folding and functional motions [2,3,4], and protein fitness .
 Gallicchio, E., and R.M. Levy. Advances in all atom sampling methods for modeling protein-ligand binding affinities. Current Opinion in Structural Biology, 21 , 161-166 (2011) doi.10.1016/j.sbi.2011.01.010
 Zheng, W., E. Gallicchio, N. Deng, M. Andrec, and R.M. Levy. Kinetic Network Study of the Diversity and Temperature Dependence of TRP-Cage Folding Pathways: Combining Transition Path Theory with Stochastic Simulations. J. Phys. Chem. B, 115, 1512-1523 (2011) doi.10.1021/jp1089596
 Deng, N. ,W. Dai, and R. M. Levy. How Kinetics Within the Unfolded State Affects Protein Folding.. J. Phys. Chem. B, in press (2013)
 Deng, N., W. Zheng, E. Gallicchio, and R.M. Levy. Insights into the Dynamics of HIV-1 Protease: a Kinetic Network Model Constructed from Atomistic Simulations, J. Am. Chem. Soc., 133, 9387-9394 (2011) doi.10.1021/ja2008032
 Haq, O, A. Morozov, and R.M. Levy. Correlated Electrostatic Mutations Provide a Reservoir of Stability in HIV Protease, PLoS Computational Biology, 8, 1-10 (2012)
Associate Professor of Physics
Single-Spin Asymmetries and the 3-D Structure of the Nucleon
Many high-energy scattering processes of microscopic particles can be very well described in the framework of Quantum Chromodynamics (QCD), the generally accepted quantum field theory of the strong interaction. However, the (large) single-spin asymmetries observed in reactions like proton-proton scattering or electron-proton scattering are much harder to understand in QCD. In this talk some recent developments in this field are presented. In particular, it is pointed out how single-spin asymmetries can provide information about the structure of the proton and the neutron in three dimensions.
Professor of Computer and Information Sciences
Laura H. Carnell Professor of Data Analytics
Predictive Modeling of Patient State and Therapy Optimization
Uncontrolled inflammation accompanied by an infection that results in septic shock is the most common cause of death in intensive care units and the 10th leading cause of death overall. In principle, spectacular mortality rate reduction can be achieved by early diagnosis and accurate prediction of response to therapy. This is a very difficult objective due to the fast progression and complex multi-stage nature of acute inflammation. Our ongoing DARPA DLT project is addressing this challenge by development and validation of effective predictive modeling technology for analysis of temporal dependencies in high dimensional multi-source sepsis related data. This lecture will provide an overview of the results of our project, which show potentials for significant mortality reduction in severe sepsis patients.
John P. Perdew
Professor of Physics
Laura H. Carnell Professor of Physics and Chemistry
Density Functionals that Recognize Covalent, Metallic, and Weak Bonds
Much of computational science in physics, chemistry, and biology requires finding the ground-state energy and electron density of a many-electron system such as an atom, molecule, liquid, or solid. The orbital-based density functional theory of Kohn and Sham, which combines computational efficiency with useful accuracy, is the most widely-used approach to this problem. It can also be the basis for the construction of effective potentials for molecular dynamics. Until recently, it has not been possible to find an approximate density functional with satisfactory simultaneous performance for covalent, metallic, and weak (van der Waals) bonds. The key to the solution lies in identifying the right dimensionless ingredient combining the electron density, its gradient, and the orbital kinetic energy density .
 J. Sun, B. Xiao, Y. Fang, R. Haunschild, P. Hao, A. Ruzsinszky, G.I. Csonka, G.E. Scuseria, and J.P. Perdew, arXiv:1303.5688.
Associate Professor of Physics
Medium Response Dynamics in Femtosecond Laser Filamentation: Modeling and Simulation
Filamentation of intense femtosecond laser pulses in atmospheric-pressure gases is accompanied by considerable shortening of the pulse, up to a few optical cycles with the coherent bandwidth ranging from terahertz to UV, and this is primarily responsible for the current intense research interest. Filaments result from dynamic balance of self-focusing via Kerr lensing and defocusing due to ionization-generated free-electron gas. We discuss modifications of these basic effects on microscopic level as the filamenting laser pulse shortens and increases in intensity. Essentially new dynamical features arise in the nonlinear optical response of the medium, which may call for re-evaluation of the currently accepted picture of filament formation. Further, we consider nonlinear optical manifestations of the excited medium in the filament wake channels, as it evolves toward equilibrium on subnanosecond timescale and can be accessed in pump-probe experiments.
Peter J. Rossky
Marvin K. Collie-Welch Regents Chair in Chemistry
Professor of Chemical Engineering
University of Texas at Austin
Exciton Migration and Dissociation in Conjugated Molecular Materials
In order to develop a working chemical intuition about electronically active organic materials, and particularly with the goal of developing design principles for organic photovoltaic materials, it is imperative to understand the relationship between molecular-level structure and the electronic excited state phenomena of exciton migration and charge separation dynamics both within conjugated polymers and at organic donor/acceptor interfaces. In this presentation, recent progress in simulating these processes, using a mixed quantum/classical molecular dynamics approach that employs an all-atom description of the intermolecular interactions coupled with semi-empirical electronic structure will be described. The results of exploring several systems at ambient temperature will be discussed, including phenylene-vinylene and thiophene oligomers, as well as cyanine and fullerene components. The roles of molecular structural fluctuations and intermolecular electronic couplings, as well as the roles of donor and acceptor excited state energy alignments will be discussed in the context of exciton transport and dissociation. The development of a molecular-level interpretation of experimental ultrafast time-resolved spectroscopic probes of these processes in a layered phtalocyanine-C60 system will illustrate the mechanistic insight accessible via such simulations.
The results reported here are based on work supported as part of the Energy Frontier Research Center "Understanding Charge Separation and Transfer at Interfaces in Energy Materials" (EFRC:CST), funded by the US Department of Energy, Office of Science, Office of Basic Energy Sciences under Award Number DE-SC0001091.
Assistant Professor of Physics
Van Der Waals Interactions in Materials Science
The weak van der Waals (vdW) interaction is a many-body effect arising from instantaneous charge fluctuations. Even if this effect is much weaker than a normal chemical bond, it plays an important role in materials science and biochemistry. High level wavefunction methods account for vdW interactions at a significant computational cost. The computationally tractable density functional theory however misses the long-range part of the vdW interactions. Different approaches to compute the vdW energy include non-local density functionals, pairwise approximations or the non-local random-phase approximation [1,2,3]. In this talk I will present improvements and limitations of these methods as exemplified by molecules, metallic systems and nanostructures [1,3,4].
 M. Dion, M, H. Rydberg, E. Schroeder, D.C. Langreth, B.I. Lundqvist, Phys. Rev. Lett., 92, 246401 (2004).
 J. Tao, J.P. Perdew, and A. Ruzsinszky, Proc. Nat. Acad. Sci. (USA) 109, 18 (2012).
 A. Ruzsinszky, J.P. Perdew, J. Tao, G.I. Csonka, and J.M. Pitarke, Phys. Rev. Lett. 109, 233203 (2012).
 J.F. Dobson, A. White, and A. Rubio, Phys. Rev. Lett. 96, 073201 (2006).
Associate Professor of Chemistry
Molecular Metaprogramming: Developing the Software and Hardware for Organic Nanotechnology
My group has developed a radical new approach to the construction of shape-programmable, nanometer scale organic structures that display functional groups in three-dimensional constellations that mimic active sites in proteins. These molecules are being designed accelerate chemical reactions (to create molecules that make other molecules) and to bind proteins (as new nanoscale therapeutics). The "hardware" consists of stereochemically pure, cyclic bis-amino acid building blocks that we couple through pairs of amide bonds to create ladder molecules with well defined shapes. We are developing the chemistry to connect multiple ladder molecules together into networks that resemble enzyme active sites to accelerate reactions and bind proteins. This chemical "hardware" is complex and while it is much more amenable to design than proteins - choosing the right building blocks, the right functional groups and the order in which to assemble them to solve a specific chemical problem is very hard. Hence, we are developing software to assist in this problem. This consists of about 0.5 million lines of C++ and a built in compiler to allow a chemist to specify complex molecular design problems and have the software search through chemical space to identify the best macromolecules to solve the problem. I will describe catalysts that we have made and the software that we are using to design them.
George C. Schatz
Morrison Professor of Chemistry
Professor of Chemical and Biological Engineering
Coupling Self-Assembly to Plasmons
Self-assembly of amphiphilic molecules provides a well-known way to make nanoscale (and larger) supramolecular structures including micelles, ribbons, sheets and aggregates. Recently there has been growing interest in the coupling of this self-assembly chemistry with silver and gold nanoparticles, leading to a new generation of materials of interest for optical devices and biodetection. This talk describes the self-assembly modeling and plasmonic properties of two classes of these materials: DNA-linked nanoparticle superlattices and peptide amphiphile ribbon nanoparticle helices. The presentation will present a novel coarse-graining strategy for describing the assembly of DNA-linked superlattices, and models of the chiral and metamaterials optical properties of the resulting structures.
Gustavo E. Scuseria
Robert A. Welch Professor of Chemistry
Professor of Physics and Astronomy
Symmetry Breaking and Restoration
Symmetries play a crucial role in electronic structure theory. I will discuss our recent developments regarding the self-consistent variation-after-projection optimization of symmetryprojected wave functions [1,2] for number, spin S2 and Sz, complex conjugation, point group, and lattice translation. The resulting method yields a comprehensive black-box treatment of static correlation with mean-field computational cost. The ensuing wave function is of high quality multireference character competitive with CASSCF. The method can be applied to excited states and spectral functions  and has been extended to non-orthogonal multireferences . Applications to both molecules and lattice systems will be presented. The curse of the thermodynamic limit and the quest for a low-cost treatment of residual correlations will also be addressed.
 Projected quasiparticle theory for molecular electronic structure, G. E. Scuseria, C. A. Jimenez-Hoyos, T. M. Henderson, J. K. Ellis, and K. Samanta, J. Chem. Phys. 135, 124108 (2011).
 Projected Hartree-Fock theory, C. A. Jimenez-Hoyos, T. M. Henderson, and G. E. Scuseria, J. Chem. Phys. 136, 164109 (2012).
 Symmetry-projected variational approach for ground and excited states of the two-dimensional Hubbard model, R. Rodriguez-Guzman, K. W. Schmid, C. A. Jimenez-Hoyos, and G. E. Scuseria, Phys. Rev. B 85, 245130 (2012).
 Multi-reference symmetry-projected variational approaches for ground and excited states of the one-dimensional Hubbard model, R. R. Rodriguez-Guzman, C. A. Jimenez-Hoyos, R. Schutski, and G. E. Scuseria, Phys. Rev. B 87, 235129 (2013).
Assistant Professor of Mathematics
Efficient Computational Methods for Interface Tracking, Radiative Transfer, and Atherosclerotic Plaque Growth
This talk highlights three problems that are governed by highly active research efforts with regards to the best methodologies for their computational solution.
In multiphase fluid computations, the tracking of the interface between fluid phases with high accuracy, while allowing for topology changes (pinch-off events), is a fundamental mathematical challenge. We demonstrate how the gradient-augmented level set method can achieve this goal.
Radiative transfer problems, which are governed by a high-dimensional phase space, are frequently computed by Monte-Carlo codes that run on large supercomputers. However, for the purpose of optimization and control (such as for radiation dose optimization in cancer-therapy), MC methods are not an appropriate framework. We outline how deterministic moment methods, computed with high order discontinuous Galerkin methods, can provide an attractive alternative that is intrinsically compatible with optimization and control.
In the study of atherosclerosis (a major cause of heart attacks), the growth process of a plaque on an arterial wall is far from well understood, largely because of the significant separation of time scales between the plaque growth rate (years) and the heart rate. We present an ongoing research effort on how to bridge the gap between these two time scales to enable a computational resolution of the evolution of the disease.
Frank C. Spano
Professor of Chemistry
Interplay Between Intrachain and Interchain Interactions in Semiconducting Polymers: Morphology Driven "Jekyll-Hyde" Behavior
Understanding how the photophysical properties of conjugated polymer films depend on film morphology is essential in designing and optimizing devices such as organic solar cells and light emitting diodes. It was previously shown that absorption and photoluminescence in poly(3-hexylthiophene) -or P3HT - spin-cast films could be understood by treating the polymer pi-stacks as weakly-coupled H-aggregates
, the latter being consistent with Kasha's designation for aggregates of "side-by-side" oriented chromophores. By contrast, the photophysical response of isolated chains of conjugated polymers, such as the red form of polydiacetylene, can be understood in terms of linear J-aggregates i.e. as a linear array of "head-to-tail" coupled chromophores (repeat units). In this talk we present the HJ-aggregate model, an extension of the Kasha model to include exciton motion along the polymer chain as well as across chains. The model shows how competition between interchain and intrachain electronic interactions impacts the photophysical response. Surprisingly, a given polymer can display both J-like and H-like photophysics i.e. "Jekyll-Hyde" behavior - depending on the morphology. Strong disorder and the associated small conjugation lengths weaken the intrachain interactions while strengthening the interchain interactions, thereby resulting in H-aggregate behavior. Conversely, intrachain interactions are stronger and interchain interactions are weaker in well-ordered long polymer segments, favoring J-aggregate behavior. The theory accounts for recent observations of J-aggregate-like photophysical behavior in P3HT nanofibers, in marked contrast to the H-aggregate behavior displayed by P3HT spin-cast films. The HJ-aggregate model is also successful in accounting for the photophysics of other conjugated polymer assemblies such as red-phase MEH-PPV aggregates.
 M. Kasha, Radiation Research 20 (1), 55 (1963).
 M. Schott, in Photophysics of molecular materials: from single molecules to single crystals, edited by G. Lanzani (Wiley-VCH, Weinheim, 2006), pp. 49. H. Yamagata and F. C. Spano, J. Chem. Phys. 135, 054906 (2011). H. Yamagata and F. C. Spano, J. Chem. Phys. 136 (18), 184901 (2012). E. T. Niles, J. D. Roehling, H. Yamagata, A. J. Wise, F. C. Spano, A. J. Moule, and J. K. Grey, J. Phys. Chem. Lett. 3 (2), 259 (2012). C. J. Collison, L. J. Rothberg, V. Treemaneekarn, and Y. Li, Macromolecules 34, 2346 (2001). A. Kohler, S. T. Hoffmann, and H. Bassler, J. Am. Chem. Soc. 134 (28), 11594 (2012).
John C. Tully
Sterling Professor of Chemistry, Physics and Applied Physics
Molecular Dynamics beyond Born-Oppenheimer
The adiabatic (Born-Oppenheimer) approximation, the cornerstone of chemical reaction rate theory, is inapplicable in cases involving electronic transitions, including non-radiative transitions, electron transfer, reactions of open shell species, energetic reactions, and chemistry at metal surfaces. Mixed quantum-classical dynamics (MQCD) has been a successful strategy for introducing the effects of electronic transitions into molecular dynamics simulations. A crucial concern in MQCD is feedback between the classical and quantum motions. The time-dependent motion of the classical nucleii induces quantum transitions among electronic states. Electronic transitions, in turn, alter the forces that govern the motion of the classical particles. Proper treatment of this "quantum backreaction" has been a subject of controversy for more than 40 years. Aspects of this controversy will be examined, including quantum decoherence, entanglement, detailed balance, tunneling, and conical intersections.
Assistant Professor of Chemistry
New Insights into Protein Folding and Design of Peptidomimetics
Proteins are nature's original "nanotechnology" - molecular machines built from chains of amino acids that have the amazing ability to self-assemble into unique three-dimensional structures to perform their functions. New computational advances such as GPU hardware and kinetic networks models of molecular dynamics, called Markov State Models (MSMs), have made it possible to simulate protein folding over longer timescales than ever before. This talk will describe how we are using these new tools to examine how sequence variations perturb folding dynamics, with the long-term goals of understanding the role of mutations in human disease, and designing therapeutic proteins and peptidomimetics from scratch.
Assistant Professor of Physics
Ab Initio Studies of Ionization Potentials of Hydrated Hydroxide and Hydronium
The ionization potential distributions of hydrated hydroxide and hydronium are computed with many-body approach for electron excitations with configurations generated by ab initio molecular dynamics. The experimental features are well reproduced and found to be closely related to the molecular excitations. In the stable configurations, the ionization potential is mainly perturbed by solvent water molecules within the rst solvation shell. On the other hand, electron excitation is delocalized on both proton receiving and donating complex during proton transfer, which shifts the excitation energies and broadens the spectra for both hydrated ions.