Chemical Physics and Physical Chemistry

Research in the Subotnik group focuses on the intersection of static quantum chemistry methods (especially for excited states) with nonadiabatic dynamics methods (specifically surface hopping). The focus is quantifying electron transfer, energy transfer, and electronic relaxation. Applications are to almost all photo-induced processes!

Computationally designed protein complex containing a nonbiological cofactor, designed and studied in collaboration with the DeGrado and Therien groups in the Department of Chemistry. On the left is the computationally designed protein scaffold (magenta) and two abiotic porphyrin cofactors (yellow). On the right is a model of the computationally designed sequence and structure.

Our research interests involve theoretical chemistry, particularly as it applies to biopolymers, macromolecules, condensed phases, and disordered systems. We are developing computational methods for understanding and designing molecular sytems having many physical and chemical degrees of freedom. Molecular simulation techniques are used both to study molecular systems in detail and to test and illustrate our theories. 

A current thrust of the group involves developing computational tools for understanding the properties of protein sequences consistent with a chosen three-dimensional structure. The group works closely with experimental groups at Penn and at other universities; some group members are involved in joint theoretical/experimental projects. Recent projects involve the design of soluble and membrane bound proteins, discerning the origins of conservation in naturally occurring proteins, biomolecular simulation, and the design of nonbiological folding molecules.

My research group creates and uses new theoretical and computational approaches to study complex systems in materials science, condensed-matter physics, and physical chemistry.

We look for new phenomena that occur when different components are brought together. For example, we examine molecules adsorbing on metal surfaces, in order to understand the effect of surface composition and structure on preferred adsorption sites, dissociation pathways, and vibrational dynamics. We also study how the compositions of oxide solid solutions lead to Angstrom-scale chemical structure, nanometer scale structural disorder, and long-range ferroelectric and piezoelectric properties. These studies find real-world applications in catalysis, corrosion, SONAR, fuel cells and other important technologies. Whenever possible, we model systems analytically, in order to extract general principles and simple pictures from complex systems. We recently derived general expressions for the vibrational lifetimes of molecules on surfaces, revealing the dependence of lifetime on molecular coverage and arrangement. Our recent exploration of quantum stress fields has helped to link chemical and mechanical effects in materials.

We are constantly developing methods for computing new properties, and for making quantum-mechanical calculations more accurate and more efficient. We tailor computational algorithms to maximize performance on modern computing platforms such as Beowulf clusters. Wherever possible, we also model systems analytically, in order to extract general principles and simple pictures from complex systems. This combination of theoretical and computational tools enables us to identify new phenomena in complex systems, like multicenter bonds between methyl radicals and the rhodium surface. ( See figure below )

Converting the 5d wavefunction of gold to a smoother pseudowavefunction results in a dramatic reduction in the required basis set size for converged calculations.

In my research group, we use a combination of analytical theory and numerical simulation to study problems in soft matter physics ranging from jamming in glassforming liquids, foams and granular materials, to biophysical self-assembly in actin structures and other systems. The idea of jamming is that slow relaxations in many different systems, ranging from glassforming liquids to foams and granular materials, can be viewed in a common framework. For example, one can define jamming to occur when a system develops a yield stress or extremely long stress relaxation time in a disordered state. According to this definition, many systems jam. Colloidal suspensions of particles are fluid but jam when the pressure or density is raised. Foams and emulsions (concentrated suspensions of deformable bubbles or droplets) flow when a large shear stress is applied, but jam when the shear stress is lowered below the yield stress. Even molecular liquids jam as temperature is lowered or density is increased this is the glass transition. We have been testing the speculation that jamming has a common origin in these different systems, independent of the control parameter varied. On the biophysical side, our research has been motivated recently by the phenomenon of cell crawling. When a cell crawls, its cytoskeleton--a network of filaments (primarily composed of the protein actin) that maintains the mechanical rigidity of the cell and gives the cell its shape--must reorganize in structure. This reorganization is accomplished via polymerization, depolymerization and branching of actin filaments, as well as by crosslinking the filaments together with "linker" proteins. The morphology of the resulting structure is of special interest because it determines the mechanical properties of the network. We are developing dynamical descriptions that capture morphology. In addition, we are exploring models for how actin polymerization gives rise to force generation at the leading edge.

Criegee intermediates: Research in the Lester laboratory is currently focused on the photo-induced chemistry of Criegee intermediates.  Alkene ozonolysis is a primary oxidation pathway for alkenes emitted into the troposphere and an important mechanism for generation of atmospheric OH radicals, particularly in low light conditions, urban environments, and heavily forested areas.  Alkene ozonolysis proceeds through Criegee intermediates, R1R2COO, which eluded detection until very recently.  In the laboratory, the simplest Criegee intermediate, CH2OO, and methyl-substituted Criegee intermediates, CH3CHOO and (CH3)2COO, have now been generated by an alternative synthetic route and detected by VUV photoionization.  This laboratory has further shown that UV excitation of the Criegee intermediates on a strong π*←π transition induces photochemistry, which involves multiple coupled excited state potentials and yields both O3P and O1D products.  This group has also demonstrated that IR excitation of methyl-substituted Criegee intermediates in the CH stretch overtone region initiates unimolecular decay.  The latter enables direct examination of the hydrogen transfer reaction leading to OH products, which is a key non-photolytic source of OH radicals in the atmosphere.

Hydrogen trioxide radical: This laboratory obtained the first infrared spectrum of the hydrogen trioxide (HOOO) radical, an intermediate invoked in the H + O3 and O + HO2 atmospheric reactions as well as the efficient vibrational relaxation of OH radicals by O2. There had been much debate in the literature as to whether HOOO is stable or metastable with respect to the OH + O2 limit, as well as the relative stability of the cis and trans conformers. We have characterized the geometric structure, vibrational frequencies, and stability of the cis and trans conformers of HOOO and its deuterated analog. In particular, by measuring the OH product state distribution following IR excitation of HOOO, we have directly determined the stability of trans-HOOO and shown that is much greater than prior estimates. As a result, HOOO may act as temporary sink for OH radicals and be present in measurable concentrations in the Earth's atmosphere. The experimental stability indicates that 25% of the OH radicals in the vicinity of the tropopause may be bound to O2, rather than free OH radicals. Studies of combination bands in the fundamental OH stretch region reveal nearly all other vibrational modes of trans- and cis-HOOO.  We have subsequently derived a torsional potential from our spectroscopic data to obtain the relative stability of the cis and trans conformers and isomerization barrier, which are critical for atmospheric modeling of HOOO. 

IR action spectrum of cis- and trans-HOOO in the OH overtone region (left), and fraction of atmospheric OH predicted to exist as HOOO (right).

Dynamical signatures of quenching: Collisional quenching of electronically excited OH A2Σ+ radicals has been extensively investigated because of its impact on OH concentration measurements in atmospheric and combustion environments. Yet little is known about the outcome of these events, except that they facilitate the efficient removal of OH population from the excited A2Σ+ electronic state by introducing nonradiative decay pathways. The quenching of OH A2Σ+ by H2 and D2 has emerged as a benchmark system for studying the nonadiabatic processes that lead to quenching. Theoretical calculations indicate that a conical intersection funnels population from the excited to ground electronic surfaces. Our studies examined the Doppler profiles for the H/D-atom products of reactive quenching, which show that most of the excess energy results in vibrational excitation of ‘hot’ water products. Our work also focused on characterizing the nonreactive quenching process, where OH X2Π products are generated with a remarkably high degree of rotational excitation and lambda-doublet specificity. The OH quantum state distribution directly reflects the anisotropy and A′ symmetry of the conical intersection region. We also demonstrated for H2 and D2 collision partners that reaction accounts for nearly 90% of the quenched products.  These distinctive dynamical signatures of passage through a conical intersection region have sparked intense theoretical interest in this system.

We gratefully acknowledge financial support from the National Science Foundation under Grant No. NSF CHE-1362835 and the Chemical Sciences, Geosciences and Biosciences Division, Office of Basic Energy Sciences, Office of Science, U.S. Department of Energy under Grant No. DE-FG02-87ER13792.

The focus of our research is to study how proteins fold from random or quasi-random coils to their biologically functional conformations. We are particularly interested in the kinetic aspects of the folding mechanisms. Novel laser spectroscopic methods are being used and developed to study the early folding events and folding intermediates.

Fast events in protein folding

Understanding how folding proceeds at early time is apparently essential to the elucidation of the entire folding mechanism. To access and characterize the early folding events requires a fast initiation method and a probe that has structural specificity. Our general approach is to use novel laser-induced temperature-jump and fast-mixing techniques to initiate refolding/unfolding on nanosecond or microsecond timescales, and use time-resolved infrared and fluorescence spectroscopies to probe the subsequent folding dynamics and structural ordering along the folding/unfolding pathways. This approach provides not only fast time resolution, but also the necessary structural sensitivity, since both infrared and fluorescence are well-established conformation probes. Recent works involve the study of the helix-coil transition, helix-helix interaction, and ß-sheet formation. 

Single-molecule study of protein conformation dynamics

Recently, a new view of the kinetics of protein folding has emerged based on the new conceptual framework of statistical mechanical models, replacing the pathway concept with the broader notion of rugged energy landscapes. The heterogeneity in folding kinetics therefore can be realized as a result of the motions of an ensemble of protein conformations on the rugged energy hypersurface that is biased towards the native state, analogous to parallel diffusion-like processes. Studying folding dynamics statistically using single-molecule techniques will provide unique information regarding a protein's folding energy landscape, which may not be obtained by conventional ensemble studies since the conventional measurements of molecular dynamics in the condensed phase represent only averages over large numbers of molecules and events. Currently, confocal fluorescence spectroscopy and microscopy are being used to study protein spontaneous fluctuation and folding dynamics at single-molecule level. 

From a knowledge of the interactions among molecules, it is possible in principle to predict the structure and the thermodynamic properties of materials as well as the dynamics of molecular processes. The overall objective of our research program is twofold: to evaluate the potential energies of intermolecular interactions for various systems as accurately as possible and to study by means of statistical mechanics the influence of these potentials of intermolecular force on the structure and properties of macroscopic systems.

Our group is interested to study the effect of nano-confinement on structure, dynamics and other properties of materials. Materials behave differently on surfaces, interfaces or small length scales compared to their bulk properties.  Understanding such differences are crucial in many technological applications where materials are constrained in nanometer size spaces, such as organic electronics, polymer applications and drug delivery. One can take advantage of such difference to produce novel materials, such asexceptionally stable glasses or harvest light for various applications. In biological systems, most of the dynamics happens in nanometer size proximity of surfaces and interfaces, and understanding the properties in confinement is a key in predicting function. We focus our efforts on understanding the origins of such modified properties on a fundamental level as well as possible application of such phenomena in producing novel materials or experimental tools.

Enhanced Mobility at the Surface of Polymeric and Organic Glasses: 

We study the properties of glasses at the air/glass interface. Our studies show that below the glass transition temperature, where the bulk of the material is in an out of equilibrium state, the interfacial dynamics are many orders of magnitude faster that the bulk dynamics. As a result a layer close to the interface maintains equilibrium properties. We study the dynamics of this layer, its thickness, and its effect on the properties of the underlying glass. The interfacial layer can strongly modify properties of amorphous materials in nanometer length scales. They also allow one to produce near-equilibrium structures at temperatures well below bulk glass transition temperature, through physical vapor deposition.

Exceptionally Stable Glasses: 

The enhanced mobility of the interfacial layer allows us to produce near-equilibrium glasses at temperatures well below the bulk glass transition temperature, Tg, by means of physical vapor deposition (PVD). Exceptionally stable glasses are formed when the substrate temperature during PVD is maintained just below the glass transition temperature. We study the morphology and the kinetics of PVD films during formation and their relationship to the final properties of the stable glass. These studies provide information on mechanisms of rapid aging below Tg and stable glass formation. We also investigate exceptional material properties of these glasses and the role of the chemical structure in these properties such as the optical birefringence and electronic properties.

Novel Emergent Optical Properties in Disordered Nanoparticle Clusters: 

Using simple synthetic routs we can produce dielectric core-gold shell nanoparticles decorated with randomly packed nanoparticles of various shapes and sizes. Spiky nanoparticles are a good example of such nanoparticles. Broadband and tunable structure of spiky gold nanoshells makes them ideal for various applications such as enhanced Raman scattering, temperature and index sensing and sensors for biological and light harvesting applications. Exceptional properties, such as higher order quadrupoloar scattering and magnetic dipole plasmons in these nanoparticles are due to inherent disorder in their structure and random packing arrangements. We explore optical properties of these nanoshells, using various theoretical and experimental tools. We also develop new techniques that allow us to study properties of meta materials formed from these types of particles.

Surface Mediated Self-assembly of Amyloid Aggregrates: 

Surface self-assembly provides an alternative pathway for amyloid aggregation that is not available in bulk solutions. We us high-resolution atomic force microscopy and other imaging techniques to study the adhesion and diffusion of peptides on various surfaces and their role in facilitating amyloid fibril formation through self-assembly routs. We also use our exceptional capabilities in high-resolution imaging to study the conformation of amyloids formed under various conditions in aqueous conditions.

Our lab is developing chemical and biophysical tools to study and manipulate complex biological systems. Projects span many areas of synthetic organic, inorganic, and biophysical chemistry; molecular, cell, and developmental biology; and bioengineering. We are particularly interested in developing new technologies for biomolecular imaging and the fabrication of functional bio-nanomaterials.

Hyperpolarized 129Xe Biosensors for Early Cancer Detection

Molecular imaging technologies hold great promise for early cancer diagnosis and intervention. Our goal is to develop new reagents that extend the capabilities of magnetic resonance imaging (MRI) for monitoring multiple cancer markers simultaneously in vivo. 129Xe has found increasing use for biological imaging applications, due to its biological compatibility (xenon is an anesthetic at high concentrations), hyperpolarizability (this enhances signals 1,000-fold), and high affinity for organic cages such as cryptophanes. The chemical shift of 129Xe varies by a remarkable 200 ppm, depending on its molecular environment: Thus, a 129Xe atom encapsulated inside a cryptophane is a sensitive reporter of perturbations outside the cage. Based on this principle, our lab is generating new biosensors that will identify biomarkers associated with cancers of the breast, lungs, brain, and pancreas. The long-range goal of this project is to use MRI to detect aberrant proteins that cause cancer in humans, years before the formation of a tumor.

Ferritin Templates for Nanoparticle Synthesis and Assembly

The goal of this project is to use ferritin proteins as templates for synthesizing and assembling inorganic nanoparticles with nanometer precision. Ferritins contain 24 four-helix bundle subunits that self-assemble to create a large central cavity. We have made water-stable, 10-12-nm gold and silver nanoparticles inside ferritin (gray sphere). Particles are fully characterized using facilities at the UPenn Laboratory for Research on the Structure of Matter (LRSM). We are functionalizing the surface of these ferritin-metal nanoparticles for sensing and nano/biomaterials applications. We are also performing computational protein design, in collaboration with the Saven lab, to mutate residues inside the ferritin cavity to enhance their metal-binding properties. Methods for organizing ferritin metal nanoparticles in 2- and 3-dimensions are being developed, in order to build very small conducting circuits. 

Laser-Activated Chemical Biology: Controlling Genes with Light

The goal of this project is to develop methods for turning genes "on" and "off" with light inside neurons and developing zebrafish embryos with high spatial and temporal control. As a first step, we have developed methods for incorporating a photoactive blocking group in the middle of a DNA or RNA oligonucleotide. In one application, we modulated primer extension by DNA polymerase (KF) using UV light. Photoactivation was monitored using a fluorescent reporter. We are now developing methods to control protein translation by the ribosome using similarly caged RNA. Blocking groups mask the messenger RNA start codon, and are designed to prevent translation until photocleavage. We will control complex gradients of proteins involved in cell signaling during zebrafish development and wound healing, using a state-of-the-art UV confocal microscope in the lab.

Our research program falls into two areas, namely nano-scale materials science and fundamental biophysics (or biophysical chemistry). The materials science effort is directed toward the development of novel electro-optical devices, both single particle and 2-D to 3-D ensemble based, utilizing the unique microscopic properties of designed cofactor-artificial peptide complexes. The cofactors are based primarily on extended conjugated chromophores designed to exhibit light-induced electric charge transfer over large nano-scale distances and possess minimal HOMO-LUMO bandgaps. The highly stable, artificial α-helical peptides are based on n-helix bundle structural motifs designed to vectorially incorporate the cofactor within the core of the bundle and to order the assembly of the peptide-cofactor complexes at the liquid-gas, solid-liquid or solid-gas interface. Ensembles of these complexes have potential for both photovoltaic devices applications relevant to solar energy conversion and non-linear optical device applications relevant to broad-band communications. The structures and properties of both the isolated complexes and the ensembles thereof are determined by cutting-edge techniques, including molecular dynamics simulation, synchrotron x-ray and cold neutron scattering, and polarized CW and transient spectroscopies. The biophysics effort is directed toward understanding the mechanism of volatile general anesthetic action on membrane ion channels. To better access the physical chemistry of the anesthetic-protein interaction, we are utilizing artificial membrane ion channels based on an amphiphilic 4-helix bundle motif, the hydrophilic domain designed to possess the anesthetic binding cavity and the hydrophobic domain designed as a membrane-spanning cation channel. The same techniques mentioned above are utilized to probe the nature of the anesthetic-peptide interaction and its effect on the conformation of the ion channel domain. Future work will also be directed toward understanding the mechanism of electro-mechanical coupling in the substantially more complex, natural voltage-gated ion channels under the control of the applied transmembrane electrochemical potential. These studies will employ cutting-edge time-resolved synchrotron x-ray and cold neutron scattering techniques coupled with molecular dynamics simulation, and will lead to the investigation of the effects of anesthetic binding on this mechanism. All of the work mentioned above involves both extensive collaborations with other faculty in the Department, at Penn, and elsewhere, as well as experimental work at the National Laboratories on a regular basis.

Figure Legend #1: An instantaneous configuration from a molecular dynamics simulation of the structure of an extended conjugated chromophore (a butadiyne-bridged Zn-porphyrin dimer: red/yellow) incorporated into the core of the hydrophilic domain of an amphiphilic 4-helix bundle peptide (green ribbon representation) and vectorially-oriented at the water-octane (gray-pink) interface. The time-averaged structure of the chromophore-peptide complex within a monolayer ensemble at the interface has been determined experimentally. The designed coiled-coil structure of the 4-helix bundle induces a twist in the chromophore that is key to optimizing its non-linear optical polarizability.

Figure Legend #2: Instantaneous configurations from molecular dynamics simulations of the structures of a computationally designed, model anesthetic-binding membrane ion channel vectorially-oriented at the water-octane interface (water/octane not shown; see Figure 1). The hydrophilic domain (blue ribbon representation) of the amphiphilic 4-helix bundle peptide contains the anesthetic-binding cavity with the volatile anesthetic halothane (CPK representation) in the cavity shown on the right-side, with the ion channel hydrophobic domain (red ribbon representation). The helices are relatively straight and un-coiled in the side-on view (upper), as more readily seen in the end-on view shown below. Removal of the anesthetic from the cavity is seen to induce a coiled-coil structure extending from the cavity into the ion channel hydrophobic domain, shown in the side-on and end-on views on the left side. Importantly, this anesthetic-dependent conformational change depends upon the amino acid composition of the cavity. Experimental verification of these predictions from the computational design and molecular dynamics simulations are underway.