July 23 – July 25, 2015 Bucknell University Lewisburg, PA

14 th MERCURY Conference on Computational Chemistry

Table of Contents

Welcome Section 1 Conference Program Section 2 Speaker Abstracts Section 3 Poster Session Abstracts Section 4 List of Conference Participants Section 5 Note Paper Section 6

14 th MERCURY Conference on Computational Chemistry

WELCOME Welcome to the 14th MERCURY Computational Chemistry conference! We are delighted to see so many of you here again. We have assembled an outstanding list of speakers, and about 50 students will present posters of their research. Undergraduates will have opportunities to learn about the breadth of research being done in the field, and students and faculty will be able to meet and discuss their work with other computational chemists.

This conference has been organized by the MERCURY (Molecular Education and Research Consortium in Undergraduate Computational chemistRY) group, an organization of investigators at predominately undergraduate institutions, funded by a National Science Foundation NSF-MRI grant. NSF, Bucknell University and Aspen Systems provided support for this conference.

After the conference, we will ask for your feedback via a web survey to help us improve the conference. We hope you will take the time to respond to the survey.

George Shields, Bucknell University Maria Gomez, Mt. Holyoke College Carol Parish, University of Richmond Marc Zimmer, Connecticut College

Maria Nagan, Adelphi University Tricia Shepherd, St. Edward’s University Kelling Donald, University of Richmond Eric Patterson, SUNY-Stony Brook Becky Eggimann, Wheaton College Mauricio Cafiero, Rhodes College Cliff Padgett, Armstrong Atlantic State U. Adam Van Wynsberghe , Hamilton College

Kelly Anderson, Roanoke College Sudeep Bhattacharyay, U. Wisconsin-Eau Clare Daqing Gao, Central State University Aimée Tomlinson, University of North Georgia Jim Phillips, U. Wisconsin- Eau Clare

CONFERENCE PROGRAM Section 2

14 th MERCURY Conference on Computational Chemistry

CONFERENCE PROGRAM

Thursday, July 23 2:00PM-5:00PM Check-in, registration

Location: McDonnell Hall Lobby

6:00-8:00PM Dinner (Picnic)

Location: Behind McDonnell Hall (2nd floor of McDonnell Hall in case of rain)

8:00PM-12:00AM Gathering at Uptown

Location: Uptown (on campus, across from McDonnell Hall)

Friday, July 24 7:00-8:30AM Breakfast

Location: Bostwick Cafeteria (Elaine Langone Center) 8:30-8:40AM Opening remarks

Location: Gallery Theatre (R301 - Elaine Langone Center)

Session I 8:40-9:40AM Keynote Speaker: Greg Tschumper University of Mississippi Title: Predictive Getting Down to the Fundamentals of Hydrogen Bonding with Convergent Quantum Chemistry

Location: Gallery Theatre (R301 - Elaine Langone Center)

9:40-10:40AM Keynote Speaker: Daniel Lambrecht University of Pittsburgh Title: Efficient Electronic Structure Approaches for Organic & Nano Materials

Location: Gallery Theatre (R301 - Elaine Langone Center)

10:40-11:00AM Break

Location: Arches Lounge (R304 - Elaine Langone Center)

Session II 11:00-12:00PM Keynote Speaker: Niri Govind Pacific Northwest National Lab (PNNL) Title: Valence and Core-Level Excited State Studies of Molecular Complexes and Materials

Location: Gallery Theatre (R301 - Elaine Langone Center)

12:00-1:00PM Lunch Location: Bostwick Cafeteria (Elaine Langone Center) (Serving line closes at 1:00PM)

1:00-1:15PM Conference group photo

Location: Steps of the Rooke Biology/Chemistry building

Session III 1:30-2:50PM Poster session oral presentations

Location: Gallery Theatre (R301 - Elaine Langone Center)

2:50-4:10PM Poster session I (A-K)

Location: Terrace Room (R276 Elaine Langone Center)

4:10-5:30PM Poster session II (L-Z)

Location: Terrace Room (R276 Elaine Langone Center)

5:30-7:00PM Dinner

Location: Bostwick Cafeteria (Elaine Langone Center) (Serving line closes at 6:30PM)

8:00PM-12:00AM Gathering at Uptown

Location: Uptown (on campus, across from McDonnell Hall)

Saturday, July 25 Session IV

7:00-9:00AM Breakfast

Location: Bostwick Cafeteria (Elaine Langone Center)

9:00-10:00AM Keynote Speaker: Qiaolin Deng, Merck Pharmaceuticals Title: Molecular Modeling & Rational Drug Design

Location: Gallery Theatre (R301 - Elaine Langone Center)

10:00-11:00AM Keynote Speaker: Joan-Emma Shea, University of California, Santa Barbara Title: Effect of surfaces in modulating protein folding and aggregation mechanisms

Location: Gallery Theatre (R301 - Elaine Langone Center)

11:00-11:20AM Break

Location: Arches Lounge (R304 - Elaine Langone Center)

Session V 11:20-12:20PM Keynote Speaker: Elena Jakubikova, North Carolina State University Title: Toward Computational Design of Iron-Based Chromophores for Solar Energy Conversion

Location: Gallery Theatre (R301 - Elaine Langone Center)

12:20-1:00PM Lunch

Location: Bostwick Cafeteria (Elaine Langone Center) (Serving line closes at 1:00PM)

Session VI 1:30-2:30PM Conference checkout (Please return keys)

Location: McDonnell Hall lobby

SPEAKER ABSTRACTS Section 3

14 th MERCURY Conference on Computational Chemistry

Getting Down to the Fundamentals of Hydrogen Bonding with Convergent Quantum Chemistry

Gregory S. Tschumper Department of Chemistry and Biochemistry, University of Mississippi

Non-covalent interactions, such as hydrogen bonding and London dispersion forces, play vital roles in a wide range of chemical, physical and biological phenomena. Examples include, but are certainly not limited to, molecular recognition, crystal engineering and self-assembly pro- cesses as well as the structure and function of biological macromolecules. Reliable ab initio electronic structure calculations have provided much insight into the strength, anisotropy and nature of these relatively weak interactions [1]. The first part of this talk will provide a very basic introduction to fundamental principles and concepts associated with quantum mechan- ical (QM) electronic structure computations. It will focus on computaional strategies that systematically converge toward exact numercial solutions of the electronic Schr¨odingerequa- tion via methodical application of correlated wave function methods and Gaussian atomic orbital basis sets. This “crash course” in convergent computational quantum chemistry will use simple H2O clusters to provide concrete examples, and it will also set the stage for the remainder of the lecture. As time permits, some recent applications of these strategies to hydrogen bonding and other non-covalent interactions will be highlighted [2,3].

References

[1] G.S. Tschumper in Reviews in Computational Chemistry; K.B. Lipkowitz and T.R. Cun- dari, Eds; Wiley: Hoboken, 26, 39–90 (2009). http://dx.doi.org/10.1002/9780470399545.ch2

[2] J.C. Howard, J.L. Gray, A.J. Hardwick, L.T. Nguyen and G.S. Tschumper, J. Chem. Theory Comput., 10, 5426–5435 (2014). http://dx.doi.org/10.1021/ct500860v (open access via ACS AuthorChoice)

[3] J.C. Howard and G.S. Tschumper, J. Chem. Theory Comput., 11, 2126–2136 (2015). http://dx.doi.org/10.1021/acs.jctc.5b00225 (open access via ACS AuthorChoice)

1 Efficient Electronic Structure Approaches for Organic & Nano Materials

Keith A. Werling and Daniel S. Lambrecht University of Pittsburgh, Department of Chemistry, 219 Parkman Ave, Pittsburgh PA, 15206

We present a hierarchy of hybrid approaches for embedded many-body expansions combined with local approximations to enable expedited first principles calculations of organic materials properties. We demonstrate calculations of deformation in response to external electric and mechanical stimuli within this framework, as is required to assess applications as responsive materials (e.g. piezoelectric sensors, shape-shifting electromechanical materials). Treating these crystalline or semi-crystalline molecular materials under the influence of external perturbations requires a balanced description of subtle intermolecular forces. We show that the presented approaches provide CCSD(T)-quality results while appealing with reduced scaling with system size as low as O(N) as well as embarrassing parallelism. Different variants for the many-body embedding, ranging from point charges to quantum mechanical embedding, are analyzed with respect to the quality of the results and computational efficiency. We then present applications of our approaches to finding improved hydrogen-bonded organic piezoelectric materials [1-2]. We conclude with an outlook on excited states of metallic nanoparticles [3] and present our development of fast approaches for excited state calculations.

[1] K. A. Werling, G. R. Hutchison, and D. S. Lambrecht, “Piezoelectric Effects of Applied Electric Fields on Hydrogen-Bond Interactions: First-Principles Electronic Structure Investigation of Weak Electrostatic Interactions”, J. Phys. Chem. Lett. 4, 1365-1370 (2013).

[2] K. A. Werling, M. Griffin, G. R. Hutchison, and D. S. Lambrecht, "Piezoelectric Hydrogen Bonding: Computational Screening for a Design Rationale", J. Phys. Chem. A 118, 7404-7410 (2014).

[3] M. J. Hartmann, H. Häkkinen, J. E. Millstone, and D. S. Lambrecht, "Impacts of copper position on the electronic structure of [Au25-xCux(SH)18]- nanoclusters", J. Phys. Chem. C 119, 8290-8298 (2015).

2 Valence and Core-Level Excited State Studies of Molecular Complexes and Materials

Niri Govind, Pacific Northwest National Lab (PNNL), Environmental Molecular Sciences Laboratory, Richland, WA 99354

In this seminar recent and ongoing developments geared towards studying excited states with linear-response (LR) and real-time (RT) time-dependent density functional theory (TDDFT) will be presented. Over the last decade or so LR-TDDFT has emerged as the method of choice for the calculation of valence excited states of large molecular complexes and materials. However, the method is limited to the linear-response regime. In recent years, RT-TDDFT has begun to emerge as a very promising and computationally efficient approach to model ultrafast electron dynamics and properties that go beyond linear-response perturbations like strong laser pulses. In addition to being able to provide detailed time resolved information, RT-TDDFT can also be used to calculate the optical response of systems with high densities of states like transition metal oxides and large chromophores. I will cover our recent developments in this area and provide an outlook. Finally, recent and ongoing developments and applications in the calculation of pre and near-edge X-ray spectra of molecular complexes and materials will also be presented.

3 Molecular Modeling & Rational Drug Design

Qiaolin Deng

Structural Chemistry, Merck Research Laboratories, 126 E. Lincoln Avenue, Rahway, NJ 07065, United States

In the talk, I will introduce basic ideas on drug discovery, molecular modeling, and rational drug design. I will use an example to show how we applied molecular modeling methods/tools to help decrease hERG liability in adenosine A2A receptor antagonist design (BMCL, 2015, 25, 2958-2962).

4

Effect of surfaces in modulating protein folding and aggregation mechanisms

Joan-Emma Shea Department of Chemistry and Biochemistry University of California, Santa Barbara, CA

Protein-surface interactions are ubiquitous in the crowded cytosol, where proteins encounter a variety of surfaces, ranging from membranes surfaces, to the surfaces presented by chaperone molecules. Protein-surface interactions are also at the heart of a number of emerging technologies, including protein micro-arrays, biosensors and biomaterials. The effect of surfaces on protein structure and stability can vary substantially depending on the chemical composition of the surface. In this talk, I will present coarse-grained as well as detailed atomistic simulations of the folding of small proteins in the presence of surfaces of relevance to biology and biotechnology. Examples will range from the underwater adhesion of marine-mussels to globular protein adsorption on membrane-mimics.

5 Toward Computational Design of Iron-Based Chromophores for Solar Energy Conversion

Elena Jakubikova

Department of Chemistry, North Carolina State University, Raleigh, NC 27695

Photoactive transition metal complexes anchored to semiconductor surfaces play an important role as chromophores in artificial systems for solar energy conversion, such as dye-sensitized solar cells (DSSCs). Fe(II)-polypyridines share many properties with Ru(II)-polypyridines, which have been successfully used as photosensitizers in DSSCs. Visible light excitation in both types of compounds results in the population of photoactive metal-to-ligand charge transfer states (MLCT). The main obstacle to the utilization of Fe(II)-based compounds as photosensitizers is the short lifetime of the initially populated MLCT states due to their de- activation by ultrafast intersystem crossing events into photo-inactive metal centered (MC) ligand-field states. We employ density functional theory and quantum dynamics simulations to investigate how various modifications to the polypyridine ligands as well as semiconductor anchor groups influence the relative energies of the MC and MLCT states and relative rates of the intersystem crossing and interfacial electron transfer events. The results obtained lead to a better understanding of structure-property relationships in these complexes and have implications for development of photosensitizers based on first-row transition metals.

6 POSTER SESSION ABSTRACTS Section 4

Students with last names beginning with

A – K à Poster Session I L – Z à Poster Session II

Students are required to stand by their poster during their session and are encouraged to visit others’ posters or remain by their posters in the other session.

Posters can be set up at the Terrace Room (R276 Elaine Langone Center) anytime starting 6PM on Thursday, July 23rd.

14 th MERCURY Conference on Computational Chemistry

Sulfate Aerosol Formation Rates in the Presence of Different Bases: Ammonia, Methylamine and Mixed Ammonia­Methylamine Nana Appiah­Padi *, Berhane Temelso, George C. Shields *Lewisburg Area High School, Lewisburg PA 17837 Department of Chemistry, College of Arts & Sciences, Bucknell University, Lewisburg PA 17837

Aerosol particles in the atmosphere reflect back sunlight and regulate the lifecycle of clouds. Our work focused on ternary sulfate aerosols containing methylamine (MA, CH 3 NH 2 ), ammonia (A, NH3 ), the combination of the two (MA­A) and sulfuric acid dimers (S 2 ) and compared their effectiveness in new particle formation in the atmosphere. We chose MA2 , A 2 , and MA­A because recent experiments have shown that the presence of heterogeneous bases

(MA­A) enhances aerosol formation significantly more than homogeneous bases (A2 and MA2 ). This work attempts to explain this phenomena using computational tools. To sample the possible configurations, we used genetic algorithm (GA) with the PM7 and SCC­DFTB semiempirical methods. The unique low energy structures from the final pool of 1500 structures for PM7 and DFTB were minimized using the PW91/ 6­31G* density functional method. These structures were finally run through PW91, M06­2X and wB97X­D methods with 6­311++G** basis sets. Although experimental measurements showed that there was a greater new particle formation rate for the MA­A system, followed by MA 2 and A2 , the computed binding energy of the clusters follows a different order: MA 2 > MA­A > A2 . The reasons for these differences between computation and experiment will be explored.

Figure 1: Aerosol formation rates are higher for ternary systems with two different bases than those with same bases.

1 Monte Carlo simulations of the adsorption from solution of alcohol/alkane mixtures near an explicit platinum surface

Jacob Barfield and Dr. Kelly Anderson

Department of Chemistry

Roanoke College, Salem, Virginia

The goal of this project is to study the way that the molecules in an alcohol/alkane mixture interact near an explicit platinum surface. Understanding adsorption is important in many different fields ranging from rheology to catalysts to lubrication. Starting with octanol/alkane mixtures that have been studied before, both experimentally and using molecular simulations with an implicit surface, we hope to validate the parameters that we are using in the simulations. We will examine the way that the size of the molecule affects the ability for that molecule to be adsorbed to the surface, and we will pay special attention to the role that hydrogen bonding from the alcohols plays in the adsorption of molecules to the surface. We plan to explore how the ratio of alcohol size to alkane size impacts the ability for that molecule to be adsorbed to the surface.

2 Accelerated Piezoelectric Evaluation (APE) Miranda Boca, Keith A. Werling, end Daniel S. Lambrecht University of Pittsburgh, Department of Chemistry, 219 Parkman Ave, Pittsburgh, PA 15260

Piezoelectric materials produce an electric charge when mechanically deformed and can be deformed by an electric charge. This dual property can be used to harvest power from movement and provide a shock sensor; uses that range from recharging pacemakers with a heartbeat to firing an airbag when a collision occurs. Normally, piezoelectric materials have been inorganic, but this project looks to investigate organic monomers as they have a different potential set of uses and can more easily be varied. Calculating the coefficient was tedious and time consuming; automating that process opens up the ability to find trends and monomers that give a high piezoelectric response. I created the accelerated piezoelectric evaluator (APE) which is an automated script that uses the coordinates from a dimer system to return a piezoelectric coefficient. This process takes a fraction of the time it takes to do the calculations by hand and is repeatable. APE finds two monomers from the set of coordinates and then distances them from one another along a hydrogen bond between them. It then uses the Q- Chem program package to calculate potential energy curves. From this information, the piezoelectric coefficient along the hydrogen bond axis is calculated via the second derivative and dipole moment. Preliminary results are consistent with previously calculated values and initial tests with a pool of candidates revealed some monomers that have high piezoelectric coefficients. APE can be interfaced with other approaches such as genetic algorithms for new material discovery.

3 Hydration of Sulfuric Acid­Methylamine Clusters in the Troposphere Bobby Cao, Berhane Temelso, George C. Shields ​ Department of Chemistry, College of Arts & Sciences, Bucknell University, Lewisburg, PA 17837

Sulfate aerosols in the atmosphere have a cooling effects on the earth’s climate. Sulfuric acid (H2SO4) is the main driver of these aerosols and the presence of bases like ammonia and ​ ​ ​ ​ methylamine (CH3NH2) stabilize molecular clusters and enhance aerosol formation through a ​ ​ ​ ​ process called ternary nucleation. We are studying the formation of particles that include sulfuric acid dimer (H2SO4)2, methylamine and up to three water (H2O) molecules. Thousands of ​ ​ ​ ​ ​ ​ ​ ​ clusters of the form (H2SO4)2(CH3NH2)(H2O)n, n=0­3, were initially generated by using genetic ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ algorithm (GA) search on semi­empirical (PM7 and SCC­DFTB) potential energy surfaces. Then the most stable clusters were studied more rigorously at the PW91/6­311++G** level of theory.

Comparing the results from the current system [(H2SO4)2(CH3NH2)(H2O)n, n=0­3] with ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ that of a binary sulfuric acid­water system [(H2SO4)2(H2O)n n = 0­3] and a ternary sulfuric acid­ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ methylamine­water system [(H2SO4)(CH3NH2)(H2O)n n = 0­3], we concluded that the addition of ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ methylamine stabilizes the sulfuric acids substantially, but it does not affect the subsequent hydration thermodynamics significantly. The addition of methylamine to a two sulfuric acid system stabilizes the cluster by 20 kcal/mol compared to a system without methylamine. Also, acid dissociation of one or both sulfuric acids is accelerated in the presence of methylamine compared to binary (H2SO4­H2O) systems. ​ ​ ​ ​ ​ ​

Figure 1: The role of bases like methylamine in the early stages of atmospheric aerosol ​ formation Figure adapted from Curtius, J. EPJ Web of Conferences 2009, 1, 199­209. ​ ​ ​ ​ ​

4 Measuring the Change in Free Energy of Mutated Glucose/Galactose Binding Protein Using Molecular Dynamics Simulations

Ross R. Cirincione, Christopher L. McCoy, Kathleen J. Jedreski, and Amil A. Anderson

Department of Chemistry | Wittenberg University | Springfield, Ohio

GGBP is found in the bacteria Escherichia coli where it is in the periplasmic space and is used for chemotaxis. As the binding of glucose to this protein can be directly related the amount of glucose in an environment, the use of GGBP as a biosensor in humans has been proposed. In this research, molecular dynamics were used to determine the binding affinity of glucose to mutants of the glucose galactose binding protein. With the binding of the sugar to the protein, the two protein domains close around the sugar in a hinge-like manner, leading to the closed conformation of the protein. GROMACS software was used to simulate interactions using the umbrella sampling method, where the glucose molecule was pulled from the binding pocket of GGBP. Values for the potential of mean force were calculated using weighted histogram analysis method, and differences from the resulting plot were used to determine the change in the binding free energy, and then the difference of binding affinity of the sugar to GGBP and its mutant. A thermodynamic cycle (Figure 1) is being investigated to contribute to a better profile of the sugar-protein system. A lowered binding affinity is favored as human physiological levels of glucose would saturate the wild-type GGBP. Lower binding affinity can be achieved through mutagenesis of residues located near the ligand binding pocket of GGBP; in this case, Alanine 213 was mutated to Arginine 213 (A213R).

Wild-type A213R ΔG binding ΔG binding

Figure 1. Thermodynamic cycle of GGBP. The change in free energy for binding is the difference in energy between any two corners of the same side in the diagram (ΔG).

5 Synthesis of 6-Ethenyldopamine for Analysis in SULT1A3

C. Skyler Cochrane, Jennifer C. Rote, Gabrielle E. Bailey, Diana J. Bigler, Larryn W. Peterson, Mauricio Cafiero

Department of Chemistry, Rhodes College, Memphis, TN 38112

Within the human body, there is an important class of enzymes known as sulfotransferases (SULTs). SULTs are responsible for the metabolic regulation of many endogenous compounds and xenobiotic substances through catalyzing the transfer of a sulfate - group (-SO3 ) from 3´-phosphoadenosine-5´-phosphosulfate (PAPS) to various substrate molecules. There are two main classes of SULTs: cytosolic SULTs and membrane-associated SULTs. Of the 13 known human cytosolic SULTs, SULT1A3 is of particular interest due to its affinity to sulfate catecholamines, like the neurotransmitter dopamine. Even though SULT1A3 shares over 90% sequence similarity with other SULTs, like SULT1A1, both select very different substrates. The goal of this study is to understand the molecular basis for SULT1A3’s substrate selectivity. Computational results show that 6-ethenyldopamine has highly interesting binding characteristics and has therefore become the target of this research. The synthesis of 6- ethenyldopamine will be discussed.

6

Investigation of Sialic Acid Association Kinetics to H274Y Neuraminidase Using Molecular Dynamic Simulations

David F. Dacres1, Jordan F. Graziadei1, Patrick F. Marris1, and Adam W. Van Wynsberghe1

1Department of Chemistry, Hamilton College, Clinton, NY 13323

Influenza is an extremely common and contagious respiratory illness that has the potential for epidemic and pandemic outbreaks. The major drug target for most influenza antivirals is neuraminidase, a homotetrameric viral capsid surface protein. The role of neuraminidase is to recognize and cleave terminal sialic acid moieties from cell surface receptors, allowing for nascent virus release. The purpose of our research is to understand how a mutant neuraminidase,

H274Y, displays resistance to oseltamivir, currently one of the most-prescribed antivirals, but still binds to its natural substrate sialic acid even though the two molecules are very similar. To better understand H274Y ligand binding we investigated the pathways of sialic acid-protein complexation using multi-scale simulation methods. First, Brownian Dynamics (BD) was used to collect clusters representing the ligands’ diffusional approach to neuraminidase, which allowed for the use of Molecular Dynamics (MD) to get detailed frame-by-frame trajectories as sialic acid approached the active and secondary binding sites. Data obtained from the MD simulations were analyzed using Molecular Mechanics/Generalized Born Surface Area (MM/GBSA) free energy calculations to highlight the favorable pathways and binding sites of the ligand.

Figure 1. H274Y neuraminidase monomer7 and sialic acid DFT design of inhibitors of the LPXC enzyme

Carolyn Dishuck, Kayla Wilson, Allison J.L. Dewar1, Larryn Peterson1, Mauricio Cafiero1 1Rhodes College, Department of Chemistry, 2000 N. Parkway

In recent years bacterial infections have become more resistant to treatments, posing a challenge for both researchers and health professionals. It has become imperative that novel, effective therapies against these resistant bacterial infections be discovered. Gram-negative bacteria present an additional challenge due to the presence of a selectively permeable outer membrane. Among the components of the outer membrane is Lipid A, which is responsible for the growth and pathogenicity of Gram-negative bacteria. The enzyme LpxC is responsible for catalyzing the first committed step in the biosynthetic pathway of Lipid A. The inhibition of LpxC would therefore, prevent the production of Lipid A, and hence result in a corrupted outer membrane. Starting from a LpxC crystal structure with a natural substrate bound in the active site, we have docked several novel ligands in the active site. The structure for these ligand-protein complexes were optimized using m06l and the 6-31G basis set (and lanl2dz for zinc) both in vacuo and in solution phase. Interaction energies for the ligand and protein complex were calculated using m06l and mp2 with the 6-311+G* basis set (and lanl2dz for zinc). Initial suitability studies were done to confirm that our model chemistry described the zinc binding in the protein appropriately. In addition, the synthesis of components of the proposed ligands is underway.

Figure One: Molecule SA-001 docked in the active site.

8 Confined diffusion of monovalent electrolytes Alfredo Dominguez and Tricia D. Shepherd Dept. of Chemistry, St. Edward's University, Austin TX

The behavior of monovalent electrolytes in confined aqueous environments is of significant interest in order to advance the knowledge of biological ion channels and material applications involving nanomembranes. In this work, molecular dynamics simulations were employed to investigate the diffusion patterns of monovalent electrolytes through a cylindrical carbon nanotube. Using LAMMPS, MD simulations were performed to a system composed of an atomistic carbon nanopore with coarse-grained water and ions solvents. It was determined that narrow cylindrical pathways act as ion filters dictating the diffusion of specific electrolytes through porous carbon membranes based on the radii size of the nanopore. Based on the results, it is suggested that the ion selectivity of carbon nanotubes can be accomplished by varying the radii size in order to enhance the “hydrophobic gating” effect that can arise in continuous water-filled pores.

9 Water transport through carbon nanotubes with defects Eduardo Dominguez and Tricia D. Shepherd Dept. of Chemistry, St. Edward's University, Austin TX

Understanding the transport of water through hydrophobic channels is vital to the understanding of proteins in biological systems and systems involving carbon nanotubes. By using a carbon nanotube, we can observe similar properties within more complex systems like pore structure within cells. Although, these systems have been studied using carbon nanotubes both through molecular dynamics and experimentally, there are few studies done on water transport through carbon nanotubes with defects. It is important to see how defects can affect the diffusion of water through these hydrophobic channels since they are prevalent in synthesized carbon nanotubes and can occur in biological systems. Throughout this investigation, we utilized LAMMPS in order to perform molecular dynamic simulations with an atomistic carbon nanotube and coarse-grained water model (mW) in order to observe the diffusion of water through carbon nanotubes with defects. Preliminary results indicate that the diffusion of water through hydrophobic carbon nanotubes decreases as the number of defects increases. These observations are depicted by the diffusion coefficients tabulated based on the mean squared displacement of the water atoms in the nanotube.

10

Structural Effects of Tertiary Structure on the Catalytic Dimer in Form 1B ribulose-1, 5-bisphosphate carboxylase/reductase Anna Donovan, Yuewei Wen, and Nicholas Boekelheide Department of Chemistry, Colby College, Waterville ME

Ribulose-1, 5-bisphosphate carboxylase/reductase (RuBisCO) is the primary catalyst of biological carbon fixation. Although the active site is structurally similar across autotrophs, the tertiary structure of the enzyme varies from a simple dimer of catalytic large subunits (below, right) to a hexadecamer that includes four catalytic dimers and eight structural small subunits (below, left). Catalytic properties including the catalytic rate, substrate specificity, and isotopic fractionation are influenced by the tertiary structure suggesting a conformational effect on the active site1. Here we use long-time molecular dynamics simulations to characterize the effect of the tertiary structure on the catalytic dimer in five hexadecameric RuBisCOs and derive a biasing potential that mimics the interaction between the full enzyme and the catalytic dimer. These results highlight similarities and differences in structural flexibility between RuBisCOs and inform a simulation strategy for catalytic dimers in the absence of the small subunits.

Loss of structural support from the small subunit causes an increase in average distance between α- carbons in the catalytic dimer (right) compared to the catalytic dimer in the full hexadecamer (left). This is demonstrated by the probability distribution function of distances for the two simulations (orange and blue, respectively). To compensate for the average potential differences between the dimer system, V2(x),

and the hexadecamer system, V16(x), a biasing potential, Vbias(r(x)), is applied.

1. Karkehabadi, S., Peddi, S.R., Anwaruzzaman, M., Taylor, T.C., Cederlund, A., Genkov, T., Andersson I., and Spreitzer R.J. Biochemistry. 44 (2005): 9851-61.

11 The Influence of a Grignard Reagent on Acyl Group Addition to Bipyrrole Fiona Evans, Jason M. Keith Department of Chemistry, Colgate University, Hamilton, NY

Porphyrins are cyclic molecules that contain four pyrrole groups, with various substituents and connections, and can serve as important metal ion transporters in nature. In an organic synthesis to produce a porphyrin called N-confused corrole, an acetyl chloride was added in excess to bipyrrole in the presence of the Grignard reagent ethylmagnesium bromide. Because of an understanding of conjugation, experimenters expected a particular diacyl product (Figure 1), but did not observe the formation of that product. DFT energy calculations of all possible isomers of products and reactants were performed using a 6-311+G basis set and a B3LYP functional. These calculations revealed that the symmetric diacyl product should have the lowest energy (Figure 2) and the Grignard reagent could provide an energy lowering effect if it is bound the nitrogen atoms of the bipyrrole system. Transition state calculations using DFT have also been performed on the lowest energy isomers of the system, which have indicated the mechanism of the reaction. During the addition of the acetyl chloride, the chlorine atom gets pushed away from the acetyl carbon as it approaches the biphenyl system and the chloride ion is lost to form a tetrahedral intermediate. Next, the chloride ion removes the hydrogen atom from the tetrahedral carbon, leading to somewhat planar product. Energy calculations for the transition states, using DFT, are still being performed to determine if kinetic or thermodynamic control is the determining factor in the formation of the diacyl product. Additionally, NMR calculations using both DFT and HF methods are being performed to determine if experimental NMR spectra are being interpreted correctly. !

!

! Figure!1.!Expected!diacyl!product.! ! Figure!2.!Diacyl!product!with!the!lowest!calculated!energy.!

!

12 Characterization of Arginine-Rich Binding in Protein-RNA Interactions Zachary J. Fallon, Megan T. Hoffman, Wesley T. Hodges, Shane M. Bernard, and Maria C. Nagan* Chemistry Department, Adelphi University, Garden City, New York Chemistry Department, Truman State University, Kirksville, Missouri

The Human T-Cell Leukemia Virus Type-1 (HTLV-1) is a complex retrovirus and causative agent of adult T-cell leukemia and tropic spastic paraparesis. The Rex protein in HTLV-1 is essential for the reproduction of new viral particles, acting post-transcriptionally in the nucleocytoplasmic transport of the unspliced and partially spliced viral mRNAs, allowing for the formation of new virions. The Rex protein functions by binding to a region on the viral RNA called the Rex Response Element (RxRE), and belongs to a family of proteins that identify their RNA targets utilizing arginine-rich motifs (ARMs). A previous study characterized the NMR coordinates of a truncated 15-mer Rex peptide containing the ARM bound to an RNA aptamer (shown in Figure 1), and identified three arginine residues that, when mutated to lysine, resulted in a thirty-fold loss in binding affinity between the Rex peptide and RNA aptamer (ARG5, -7, and -13). In this study, two 100ns MD simulations were ran for each mutant peptide complex, as well as two 100ns simulations of a triple mutant complex—giving two 100ns simulations for each of the four different systems. The mutated Rex peptides exhibit displacements from the RNA aptamer due to loss of key water molecules at sites of high water density, presumably due to the lacking of the arginine “fork”. The change in binding affinity due to the mutations is characterized by hydrogen bonding analysis, water density and lifetime analyses, and by monitoring displacement of the mutated peptide along the RNA aptamer. Simulations were run using modifications of the Cornell et al. force field and TIP4P-Ew water; analyses and visualization were conducted using AmberTools and VMD.

13 Rotation about Aromatic Amide Bonds II: A Computational Project

Evan Grassi, Donald D Clarke, Department of Chemistry, Fordham University & James B Foresman, Department of Physical Sciences, York College of Pennsylvania

Previously we showed that the solvent induced change in the proton NMR spectrum of 2-nitroacetanilide could not be explained by a change in H bonding in CDCl3 vs DMSO. Rather a change in the angle of the acetamido group to the plane of the benzene ring explains this. Calculations using Gaussian can trace the change of 1H chemical shifts of the ring protons as the angle of the acetamido group to the plane of the benzene ring is varied. These suggest that in CDCl3 the energy minimum for this rotation is about 50o while in DMSO it is close to 90o. In N- methyl-2-nitroacetanilide, where H bonding is removed there also is a 13 change in spectrum in going from CDCl3 to DMSO. The C spectrum shows this more clearly. Here rotation is along the amide bond rather the C1C2NH axis. Since amides have partial double bond character, separate signals are observed for cis and trans conformers. Calculations using Gaussian confirm that the trans [E] conformer is the major one in both solvents. The ratio of these conformers goes from approx 3:1 in CDCl3 to 3:2 in DMSO. This corresponds to a difference in ∆G between conformers of about 2.2 KJ/mole in CDCl3 and 1KJ/mole in DMSO.

14 Design and synthesis of novel inhibitors for the catechol-O-methyltransferase enzyme

A. Katherine Hatstat, Mallory Morris, Larryn W. Peterson, and Mauricio Cafiero Rhodes College, Department of Chemistry, 2000 N. Parkway, Memphis, TN 38112

L-DOPA is commonly used as a xenobiotic for patients with conditions such as Parkinson’s disease. L-DOPA is transformed into dopamine by DOPA-decarboxylase. Dopamine derived from L-DOPA is deactivated via metabolism by the COMT enzyme. The targeted inhibition of the COMT enzyme prolongs the effectiveness of L-DOPA, resulting in a net increase in pharmacological efficiency. By selectively designing an inhibitor for the catechol-O- methyltransferase enzyme, the effectiveness of the L-DOPA can be extended by regulating the metabolism of dopamine derived from L-DOPA. The effectiveness of these dopaminergic derivatives has been measured via in silico models in which the strength of interaction between each substrate and the enzymatic active site was analyzed. A crystal-structure of the COMT enzyme active site, docked with a known COMT inhibitor, BIA 8-176, was isolated from the Protein Data Bank (PDB ID: 2CL5). Novel dopaminergic derivatives were optimized using M062X/6-31G in vacuum and in implicit solvation with rigid amino acid side-chains. Interaction energies between the ligands and the protein were calculated using M06L and MP2 with the 6- 311+G* basis set. Interesting differences were noted between in vaccuo and solvated calculations. For example, ligands with a nitrile substituent were favored over other substituent variations in vacuum, but this preference was not retained when the same ligands were optimized with implicit solvation. Preliminary results on synthesis of nitrile substituent derivate will be presented as well.

Figure 1 Dopamine docked in the active site of the COMT enzyme

15 DFT Benchmarking of Eumelanin-Inspired Small Molecules

Jessica Holmes, David Wheeler, and Dr. Aimée Tomlinson

University of North Georgia, Dahlonega, GA 30597

The high conjugation within the natural pigment molecule, eumelanin, is what gives it the ability to absorb UV radiation. Thus, molecules that apply eumelanin as a core component allow for the preparation of an optimal semiconductor. In order to determine the optimal functional and basis- set pairings that can be used to later calculate HOMO, LUMO, and optical band gap information about these molecules, a benchmarking process took place by combining a variety of different GGA and GH functionals and basis-sets. The results of this study are presented here.

16 Teaching with Technology: A More Effective Method of Learning Organic Spectroscopy Dan Hopkins, Liam Feehery, Joshua Schrier Department of Chemistry, Haverford College, Haverford, PA 19041

Hundreds of thousands of undergraduate chemistry students learn to interpret infrared spectra during their collegiate study of organic chemistry. Traditionally, students learn to identify functional groups in spectra primarily through unsupervised practice exercises. A significant number of students struggle with this task and must seek supervised practice with faculty and teaching assistants so that they may receive immediate feedback and useful direction toward appropriate supplementary information.

We propose to apply learning psychology strategies – commonly used by popular websites like Memrise and DuoLingo for the purpose of foreign language acquisition – to accelerate and improve learning of spectral analysis without the presence of faculty members or TAs. These strategies include: (i) Testing Effect: frequent low-stakes testing with immediate feedback and direction to appropriate resources. (ii) Spacing Effect: avoid short-term memory saturation through spaced practice intervals – robustly enforced by the website. (iii) Scheduling: minimize the time needed for a student to reach proficiency by testing weaker material with more relative frequency.

With these strategies in mind, we’ve developed a website that aims to accomplish this task by using: (i) HTML5 + JavaScript – no plugins necessary! (ii) Open source spectral data from ChemSpider (iii) Parse database of molecules and spectral data indexed by SMILES strings

Our poster will go into detail in regards to the database schema, technical specifications, and show screenshots of the website in action. Additionally, Dan Hopkins will run a live demo of the most current build of the website. We plan to test the use of this website during Haverford College’s introductory chemistry courses in the fall of 2015, and are interested in finding other Colleges and Universities that are willing to help us test this product.

17 Condensed-Phase Effects on the Structural Properties of Nitrile—SiF4 Complexes: A Low-Temperature IR and Computational Study

N. J. Hora and J. A. Phillips (mentor)  Department of Chemistry, UWEC

Our research involves the identification of molecular complexes that change structure when the chemical environment is altered, e.g., gas-phase to solid-state. This project stems from a previous study on CH3CN–SiF4, in which condensed-phase structural changes were predicted but not observed [1]. In the present case, we are dealing with four specific complexes: CH3CH2CN–SiF4, C6H5–SiF4, (CH3)3CCN–SiF4 and pyridine-SiF4. We expect the larger carbon groups to enhance the bonding interaction and lead to more significant structural change in the condensed phase. Using M06, MP2, ωB97XD, and the 6-311G+(2df,2pd) basis set, we computed equilibrium geometries, binding energies, frequencies, N-Si potential curves, in the gas phase and in bulk dielectric media (PCM/M06/6-311G+(2df,2pd)). In our search for the most stable structure we considered both axial and equatorial coordination (two conformers for each). Structures and binding energies were very similar to that of CH3CN–SiF4. Nitrile complexes were found to be weak (ΔE~4-5 kcal/mol) with long N-Si bonds (~2.8 Å) and the pyridine complex was stronger (ΔE= -12.3 kcal/mol) with a shorter N-Si bond (2.1 Å). Gas phase potentials show a long bond length in the gas phase and a very slight energy rise towards shorter donor-acceptor distances. For C6H5–SiF4, the shape of the PE curve changes dramatically with increasing dielectric constant, and the minimum energy point shifts from 2.8 Å at ε=2 to 2.0 Å at ε=10. Future work will involve Infrared Spectroscopy of nitrile/SiF4 thin-films, in an attempt to observe condensed-phase effects.

#$ ,-./#)%$ %$ ,-./*)%$ ,-./"%$ !#$ 012$-3124$ M06/6-311+(2df,2pd) !($ H H F F C C !'$ H C C C N Si F C C 2.878 Å !&$ H H !"#$"#%&'#()%*&+,-./012/3& F 106.1° ΔE (M06)=-4.8 kcal/mol !"%$ ΔE (MP2)=-4.5 kcal/mol !"#$ ")*$ #$ #)*$ +$ +)*$ ($ 4"56&7"89.#-(&+:3&

Figure 1. Equilibrium geometry of C6H5–SiF4 (left) via M06, with binding energies for M06 and MP2. Also shown are potential energy curves for the gas phase as well as in bulk dielectric media using M06 (right).

1. Helminiak, H. M.; Knauf, R. R.; Danforth, S. J.; Phillips, J. A. J. Phys. Chem. A 2014, 118 (24), 4266–4277.

18 Molecular Dynamics Studies of Retinoid Ligands in the Fatty Acid Binding Protein FABP5 as a Potential Cancer Therapy

Nathanael Hunter, Chrystal D. Bruce John Carroll University Department of Chemistry, John Carroll University, University Heights, OH 44118

Treatment of cancer cells with retinoic acid, a fatty acid, can either lead to proliferation or inhibition of cancer cells depending on the relative ratio of two proteins in the cancer cell: FABP5 and CRABP-II. Inhibiting the action of FABP5, or fatty-acid binding protein by binding with retinoic acid may lead to the observed reduction in cell proliferation. The goal of this project was to examine the binding of retinoic acid and its isomer, 9-cis-retinoic acid in the binding pocket of FABP5 in a molecular dynamics simulation. Results of the molecular dynamics production run, performed using AMBER12 software, as well as free energy approximations, calculated using molecular mechanics/generalized born surface area and normal mode harmonic entropy approximations, will be presented. These will be compared to preliminary docking results found using AutoDock Vina docking software.

19 Measuring the Free Energy Change of the Glucose/Galactose Binding Protein Utilizing Molecular Dynamic Simulations

Kathleen J. Jedreski, Christopher L. McCoy, Ross R. Cirincione, and Amil. G. Anderson

Department of Chemistry | Wittenberg University | Springfield OH

Glucose/Galactose Binding Protein (GGBP) is an α/β protein found in bacteria that mediates chemotaxis and is involved in the active transport of glucose and galactose. The binding of the sugar to the protein occurs through hydrogen bonding with the OH groups on these sugars. The formation of these bonds is important in determining the binding energy of the sugar to the protein. Utilizing Molecular Dynamic simulations, a thermodynamic cycle is being performed to determine the difference in binding affinity for the wild type and a mutant form of GGBP (Ala213 to Arg213) of glucose. The cycle is broken down into four steps and four ∆G calculations (illustrated below). The focus of this research is to test the ∆G of mutation between the wild type and mutated form (∆G2 on illustration). Gromacs-4.6.5 is being used to run the Molecular Dynamic simulations and a double annihilation. “Dummy” atoms were added in the wild GGBP which were then replaced by the mutated residue topology (Arg213). The change in free energy will be calculated using the Weighted Histogram Analysis Method (WHAM) to extract the Potential of Mean Force (PMF) and thus to calculate the ∆GMutation (∆G2). The ∆∆Gbinding can be calculated by ∆G1-∆G3 which is equal to ∆G2-∆G4, collecting three of the four ∆G’s we can calculate the missing ∆G and ∆∆G, since ∆G1+∆G4 - ∆G3-∆G2 = 0 (Fig.1).

∆G2 Wild-type GGBP Arg213 GGBP w/bound w/bound Glucose Glucose (closed) (Closed)

∆G1 ∆G3

∆G Wild-type 4 Arg213 GGBP (open) GGBP (open) + + Free Glucose Free Glucose

Figure 1. Thermodynamic Cycle of the Wild-Type GGBP/Mutated GGBP binding with glucose with the red arrows indicating the ∆G being measured in this research.

20 The structure of ancestral form IB RuBisCO Shangcheng Jiang1, Victor Hanson-Smith2,Nicholas Boekelheide1

1Colby College, Department of Chemistry 2University of California San Francisco Microbiology and Immunology

Ribulose-1,5-bisphosphate carboxylase/oxygenase (RuBisCO) catalyzes CO2 fixation during photosynthesis. The enzyme’s ancestral history is complicated by evolutionary bottlenecks due to changes in global atmospheric oxygen concentration and due to the oxic or anoxic conditions to which a specific lineage has adapted. In order to better understand the biochemical role of ancestral RuBisCOs, we have developed a homology-based algorithm to reconstruct the 3D-structure RuBisCOs inferred from a reconstructed phylogenetic tree of the RuBisCO protein family. Here we present the 3D-structure of the ancestral form IB RuBisCO representing the common ancestor linking cyanobacteria and plants.

S. elongatus ...NSYFAFIAYPLDLFEEGSV... Th. elongatus ...NQFIAYIAYPLDLFEEGSV... Tobacco ...DQYIAYVAYPLDLFEEGSV... Algae ...NQYIAYVAYPIDLFEEGSV... Spinach ...NQYICYVAYPLDLFEEGSV...

Predicted ancestor ...NSYFAFIAYPLDLFEEGSV...

Top left: Catalytic dimer of large subunits from cyanobacteria S. elongatus. The blue, grey and red subunit is colored by similarity to four other form IB RuBisCOs. Bottom: A multiple sequence alignment colored by similarity. The bottom line is the predicted sequence of the ancestor to the cyanobacteria/algae/plant lineage. Top right: The predicted structure of the ancestral sequence following our protocol. The residues of the displayed sequence are shown as ball-and-stick models.

21 Explaining the Increased Sulfate Aerosol Formation Rates in Mixed Ammonia­Dimethylamine Systems

Grace Kim, Berhane Temelso, George C. Shields Department of Chemistry, College of Arts & Sciences, Bucknell University, Lewisburg PA 17837

Composed of gas phase molecules found in the atmosphere, aerosols form through a nucleation process before serving as cloud condensation nuclei (CCNs) that seed clouds. Aerosols have been found to have a net cooling effect on the global climate, but the process of their formation is still not well­understood.

Acid­base reactions between sulfuric acid [H 2 SO4 , S] and ammonia [NH3 , A] promote aerosol formation and the effect is even greater for amines. Recent studies have shown that reactions between acids and stronger bases, such as amines, may further enhance aerosol formation by preventing evaporation. Furthermore, the presence of ammonia and amines increases aerosol formation rates significantly more than ammonia or an amine alone. For this study, dimethylamine [(CH3 ) 2 NH, DMA] was chosen due to its stronger basicity and larger size.

The possible configurations of (H 2 SO4)2 ­((CH 3 )2 NH)­(NH 3 ) and (H2 SO4)2 ­((CH 3 ) 2 NH) 2 were sampled using a genetic algorithm applied on semi­empirical (PM7, SCC­DFTB) potential energy surfaces. The low energy structures were subject to quantum mechanical calculations such as PW91, M06­2X, and wB97X­D with 6­311++G** basis set. From our data, we have come to the conclusion that binding energies strongly correlate with the basicity of the base:

S2 ­DMA 2 > S2 ­DMA­A > S2 ­A2 . These conclusions disagree with experimental nucleation rates which increase in the order S2 ­DMA­A > S2 ­DMA­DMA > S 2 ­A2 . The reason for these discrepancies are explored.

Figure 1: How does the presence of two different bases enhance aerosol formation rates compared to single base systems?

22 Calculating of UV-Vis Absorbance/Fluorescence Spectra of Indole and Tryptophan

Elijah B. Kofke, Kristine M. Vorwerk, William W. Kennerly Department of Chemistry Skidmore College Saratoga Springs, NY

We are characterizing the UV-Vis absorption and fluorescence spectra of gaseous indole and tryptophan, using quantum computational modelling methods available in Gaussian 09. We have settled on using time- dependent density functional theory (TD-DFT), after initially attempting to use complete active space (CAS) calculations. Using DFT at the B3LYP/aug-cc-pVTZ level of theory, we calculated optimized ground-state geometries for both indole and tryptophan. We then used these geometries to calculate vertical excitation data using a variety of basis sets and functionals. The CAM-B3LYP and ωB97XD functionals showed a high degree of consistency, across a variety of augmented and polarized basis sets (both double and triple split valence). In both indole and tryptophan, the S1 and S2 states are nearly degenerate. Gaussian has struggled to distinguish their minimum energy geometries, converging the second excited state geometry to that of the first. Our predicted vertical absorption energies at the ωB97XD/6-31+G(d) level of theory are 5.08 eV (S1) and 5.13 eV (S2) for indole, and 4.98 eV (S1) and 5.08 eV (S2) for tryptophan. Our predicted vertical fluorescence energy (from S1) for indole and tryptophan is 4.50 eV and 4.30 eV, respectively. We were partially successful characterizing indole using smaller basis sets in CAS calculations, but were unable to generate results for larger basis sets or for tryptophan. With successful characterization, we hope to move on to investigating tryptophan in solvated environments, and for protein-embedded tryptophan moieties.

23

A Computational Study of Friedel-Crafts Intermediates: RX’–MX3 J.R. Lanska, J.A. Phillips (Mentor) Department of Chemistry, University of Wisconsin- Eau Claire

Friedel-Crafts processes are an important class of carbon-carbon bond forming reactions and among the most common involve an alkyl halide and a Lewis acid catalyst. We are interested in the intermediates of these processes, which are alkyl halide acid-base complexes: RX’–MX3. Using a variety of computational methods (ω- B97X-D/aug-cc-pVTZ preferred) we obtained equilibrium structures and binding energies for RX’–MX3 compounds where R = CH3; M = B, Al, or Ga; and X = F or Cl. We considered four distinct conformations for each complex, with the conformation shown below consistently being the most stable. Binding energies from aluminum compounds were the most stable, followed by gallium and finally boron, which were generally weak. Using an NPA charge analysis, we assessed the charge on the R group as well as the extent of charge transfer from the R-X’ group to the MX3 group. In addition, we mapped the M-X’ potential curves for several representative complexes in this family, including CH3Cl-AlCl3, CH3F-BF3, CH3Cl-AlF3, CH3F-AlCl3 via ω-B97X-D/aug-cc-pVTZ. Future work will assess the effect of dielectric media on the potential curves. 10# B3LYP# CH3F-BF3: 8# ΔE = -4.4 kcal/mol B3PW91# mp2/aug-cc-pVTZ 6#

4# MP2# 2.362 Å 108.2° 2# 1.402Å X3LYP# 0# M06# !2# 91.0° CCSD# !4#

Binding#Energy#(kcal/mol)# !6# wb97xd# !8# 1# 1.5# 2# 2.5# 3# 3.5# 4# Bond#Distance#(Å)#

Figure 1: On left is equilibrium structure for complex CH3F-BF3 in the most stable conformation. On right is a curve of bond potential vs. distance for the complex using nine different methods to determine most accurate method as compared to CCSD.

24 Examining the Binding Pathways of Oseltamivir to H274Y Neuraminidase via Molecular Dynamics Simulations and MM/GBSA Analysis

Erin M. Lewis, Richard W. Wenner, Patrick F. Marris,

and Adam W. Van Wynsberghe

Department of Chemistry, Hamilton College, Clinton, NY 13323

Neuraminidase (NA) is essential to the proliferation of the influenza virus. The homotetrameric enzyme cleaves terminal sialic acid moieties from host cell surface receptors enabling nascent viral release. Small molecules that act as NA inhibitors, including the widely prescribed oseltamivir, effectively prevent proliferation of the wild type influenza virus. However, a strain of mutated influenza (H274Y) has displayed resistance to oseltamivir. In order to fully understand the molecular basis of this resistance, we analyzed the complete binding trajectories of oseltamivir and H274Y NA. Pathways were sampled using a multi-scale methodology utilizing

Brownian Dynamics (BD) for the diffusional approach, and Molecular Dynamics (MD) to examine the close-range interactions. Molecular mechanics/generalized Born surface area

(MM/GBSA) free energy calculations were employed to analyze the favored binding pathways.

Figure 1. Neuraminidase monomer and oseltamivir ligand.

25 Investigating binding pathways to neuraminidase using MM/GBSA free energy analysis

1 1 2 1 Patrick F. Marris , Leah M. Krause , Jesper Sørensen , Adam W. Van Wynsberghe

1Department of Chemistry, Hamilton College, Clinton, NY 13323 2Department of Chemistry and Biochemistry, University of California-San Diego, La Jolla, CA 92093

Influenza neuraminidase is a homotetrameric viral enzyme that is integral to the influenza virus’ replication cycle. A successful strategy for combatting the influenza virus has been to inhibit the catalytic activity of this enzyme using such drugs as peramivir, zanamivir (Releza®), and oseltamivir (Tamiflu®). These inhibitors compete with terminal cell-surface receptor sialic acid moieties for access to neuraminidase’s binding site. To better understand how these ligands interact with neuraminidase, we employed Molecular Mechanics/Generalized Born Surface Area (MM/GBSA) calculations in order to elucidate specific binding pathways and areas of binding favorability for both the wild-type neuraminidase and sialic acid complex and the wild-type neuraminidase and oseltamivir complex. The MM/GBSA calculations analyzed frames from 194 five-nanosecond close-range molecular dynamics (MD) simulations for the oseltamivir system and 261 five-nanosecond close-range MD simulations for the sialic acid system. After visual comparison of the MM/GBSA free energy maps from both complexes, alanine mutation scanning and residue decomposition calculations were performed in order to better understand specific receptor-ligand interactions.

Figure 1. H274Y neuraminidase monomer with MM/GBSA map of oseltamivir binding favorability (red points indicate relatively favorable free energies while blue points indicate relatively unfavorable free energies)

26 The Effect of Point Group Symmetry on QSPR of Boiling Point

Drew Marshburn NCSSM Online North Carolina School of Science and Mathematics Durham, North Carolina

This study was conducted to determine the effect of point group symmetry on the QSPR of the boiling points of haloalkanes. Once the point group symmetries of 90 haloalkanes were identified, all the molecules were imported as CML files into Bioclipse- R, an open source QSPR/QSAR software. Bioclipse-R was then used to generate values for 44 descriptors; as many had multiple outputs, this resulted in over 170 descriptor values for each molecule. All of this information was imported into Mathematica for statistical analysis. Linear regression analysis was used to compare the descriptors to the boiling points of the haloalkanes and the descriptors with the five highest R-squared values were reported and compared for all the molecules together, the C1 point group, the CS point group, and the C3v point group. The polarizability of a given haloalkane was the most correlated property to the boiling point regardless of point group; however, the ranking of the descriptors which in some way reported the polarizabilities of the molecules did change between the point groups. Thus, point group symmetry does have a slight effect on the QSPR of boiling point for haloalkanes. Predictive QSPR models that incorporate differences between point groups for molecules should be more accurate than those that do not, but the increase in accuracy is likely marginal at best.

27 Measuring the Free Energy Change of the Glucose/Galactose Binding Protein using Molecular Dynamics and Umbrella Sampling with Limited Position Restraints

Christopher L. McCoy, Kathleen J. Jedreski, Ross R. Cirincione, and Amil G. Anderson

Department of Chemistry | Wittenberg University | Springfield, OH

The Glucose/Galactose Binding Protein (GGBP) plays a significant role in bacterial chemotaxis towards glucose and galactose. There are two distinct conformations for GGBP; the open-state and the closed-state when either glucose or galactose is bound between the two domains. GGBP has recently been evaluated as a possible biosensor to non-invasively measure glucose levels in diabetes mellitus patients. The binding free energy of GGBP and glucose can be calculated computationally with umbrella sampling. This research attempts to calculate the ΔGbinding of GGBP and glucose using the umbrella sampling pull method with a single positional restraint on GGBP. Previous research has performed the umbrella sampling pull method for GGBP and glucose using a more involved six degrees of freedom restraint on the protein/glucose system during simulations. This research will contribute to a larger project in creating a thermodynamic cycle for the mutation of residue Ala-213 to Arg in GGBP to measure ΔΔGbinding. These simulations also help us determine if a two-step process (Closed GGBP with bound glucose to open GGBP with bound glucose to open GGBP with free glucose) is required to calculate the

ΔGbinding.

Wild-type GGBP Arg-213 GGBP (Closed) (Closed) With bound Glucose With bound Glucose

wt A213R ΔGBinding ΔGBinding

Wild-type GGBP Arg-213 GGBP (Open) (Open) + free Glucose + free Glucose

28 Comparing Single and Multiple Base Systems in Secondary Aerosol Formation Fern Morrison , Berhane Temelso, George C. Shields Department of Chemistry, College of Arts & Sciences, Bucknell University, Lewisburg PA 17837

Atmospheric aerosols have a great impact on our global climate. They cool the planet by directly reflecting light away from the earth, and indirectly by acting as cloud condensation nuclei that brighten and extend the lifespan of clouds. Despite their importance in the earth’s radiation balance, the lack of understanding about their formation has led to uncertainties about their effect on the global climate. Previous experimental works have shown that sulfate systems with ammonia and an amine tend to have higher formation rates than those with ammonia or an amine alone. This study attempts to explain that behavior by exploring the clustering of sulfuric acid [H2 (SO)4 , (S)], ammonia [NH 3 , (A)], and tr imethylamine [(CH3 ) 3 N , (TMA)]. Initial configurational sampling was performed using genetic algorithm (GA) interfaced to semi­empirical methods (PM7 and SCC­DFTB) to find a large number of low energy configurations. These structures were then subject to quantum mechanical calculations using PW91, M06­2X and wB97X­D functionals with large basis sets. The thermodynamics of formation for the most stable structures was then reviewed to determine why multiple base systems have higher rates of aerosol formation than single base ones. The combined data showed that the binding energy of the cluster is related to the pK b value of the base; the more basic the amine group the stronger the binding. This trend was consistent for dimethylamine (DMA) and methylamine (MA) groups as well. Further research is being conducted to assess the evaporation rate of sulfuric acid from single and multiple base systems.

Figure 1: Why is the formation rate of aerosols higher for multiple base systems?

29 DFT analysis of water clusters, dopaminergic derivatives, and their desolvation energies Mallory Morris, Katie Hatstat, Larryn Peterson, Mauricio Cafiero Rhodes College Chemistry Department

The catechol-O-methyltransferase enzyme is responsible for the metabolism of the neurotransmitter dopamine, a catecholamine involved in the degenerative disorder known as Parkinson’s Disease. One treatment for Parkinson’s disease is L-DOPA therapy, where this dopamine precursor is transformed into dopamine by DOPA decarboxylase. The dopamine derived from L-DOPA is degraded by COMT; therefore, inhibiting COMT would be ideal to prolong the effectiveness of L-DOPA and to increase pharmacological efficiency by preventing the premature metabolism of the medication. Computational models of dopaminergic analogs were used to examine the substrates’ binding in the enzymatic active site. The binding of a ligand to an enzyme not only involves the interaction between the ligand and the enzyme but also the energy lost or gained by desolvation of the ligand. Desolvation of dopaminergic derivatives was examined using a series of hydration shells that increase in size. The desolvation energies were calculated using M062X with the aug-cc-pvdz, cc-pvdz, and cc-pvtz basis sets. Ligands with the carboxylic acid and nitro substituents exhibited the least favorable energies, whereas the nitrile substituents exhibited the most favorable desolvation energies in each of the explicit water models. The implicit Polarizable Continuum Model was also used together with explicit solvation to calculate desolvation energies of dopaminergic ligands. The use of implicit and explicit models was compared. This information will be combined with prior research done on ligand/enzyme interaction in order to get a more comprehensive understanding of ligand binding in this system.

Benzonitrile

Acetonitrile

Figure 1. Dopaminergic derivatives, Benzonitrile and Acetonitrile, each surrounded by a shell of hydration containing thirteen water molecules explicitly placed around the ligands.

30 The Free Energy Effects of Harmonic Restraints on a Protein-ligand Complex Natalie Nguyen ’17 and Amil Anderson Wittenberg University, Springfield, OH The use of umbrella sampling in molecular dynamics simulations to measure free energy changes is a widely known method in computational chemistry. When applying a harmonic potential to a ligand, in order to induce a pulling simulation, we are additionally applying a restraint to the system. Restraints can restrict the ligand from its full entropic potential and ultimately contribute to the calculated free energy of the system. This free energy penalty may or may not have a significant effect on the overall results obtained from umbrella sampling. The goal of this investigation is to quantify the difference in free energy when such a harmonic force is applied to and removed from a ligand in a protein-ligand complex. Our method involves using thermodynamic integration implemented with Gromacs (Version 4.6.5) to measure free energy values associated with the addition and removal of the harmonic restraining potential. This will be applied to a glucose-bound Glucose-Galactose Binding Protein (GGBP) system as part of a larger project that involves characterizing the binding site of the protein. The results of this study have the potential to improve previous and future results obtained through umbrella sampling in order to mirror results obtained experimentally. In addition, these results can be compared to the analytically calculated effect of imposing restraints as dictated by the Virtual Bond Algorithm, which allows the ligand even lesser degrees of freedom.

The figure above is a visual representation of the imposition of the harmonic restraint potential with respect to lambda states. Not all intermediate states are being represented for simplification.

31 Model development of enzymatic biofuel cells Kim Pham and Tricia D. Shepherd Dept. of Chemistry, St. Edward's University, Austin TX

Biofuel cells are batteries that employ biocatalysts. Enzymes are excellent fuel cell electron shuttles due to their high substrate specificity, being very selective in terms of what is oxidized or reduced. The ability of enzymes to utilize biologically-derived fuels, such as sugars, ascorbate, dopamine, and alcohols makes them viable as electric power sources in implantable devices in living organisms. Recently, scientists have been seeking to modify enzymes by adding graphitic attachments or various linker groups to increase their efficiency and longevity. Using molecular dynamic simulations, one can predict how an enzyme would interact with graphitic attachments in a system involving the transport of ions, gas and water - specifically biofuel cells. This project seeks to construct an atomistic model of the blue multi-copper oxidase laccase, run molecular dynamics simulations in water to calculate bond and angle parameters and the energy involved. Data from these simulations will be used to develop a coarse-grained model to study the transport of oxygen, ions and water at the interface between a fuel cell membrane, a graphitic surface, and the electrocatalytic enzyme laccase.

32 Computational Screening for High Performance Aqueous Redox Flow Battery Materials Richard Phillips, Joshua Schrier Department of Chemistry, Haverford College 370 Lancaster Ave, Haverford PA

Redox flow batteries using water-soluble quinone derivatives have significant potential as a means of environmentally clean and cost-effective energy storage. However, to maximize energy density, there must be a larger range of half-cell potentials and improvements must be made to solubility, to increase power density, and cost before redox flow batteries become a viable means of energy storage. Reduction potentials for affordable and synthetically feasible thiophenoquinone derivative frameworks were calculated with model chemistry B3LYP/6-311+G(d,p) for thermochemical calculations (using the SMD solvation model). Our previous work shows that this estimates the redox potential to within +/- 0.03 V. Reduction potentials for the same molecules were calculated with an attached quaternary ammonium tail. Calculations assert that the addition of the quaternary ammonium tail greatly reduces the ΔG of solvation (predicted difference of ~240 kJ/mol for each species). Among the molecules tested, a notable minimum voltage was naphtho(2,3-b)thiophene-4,9-dione with an attached quaternary ammonium tail (NAP) with a predicted reduction potential of 0.15 V.

Calculations were made for the redox potentials for NAP with 9 different functional groups (‐NCH3CH3, ‐NHCH3, ‐NH2, ‐CCH, ‐CCCH3, ‐SH, ‐OH, ‐OCH3, ‐SCH3) each tested in two different positions. From this data, eight molecules are predicted to have lower voltages than NAP, including a molecule with a calculated reduction potential of -0.02 V.

33 Coarse-grained simulations of confined water in nanotubes SeungYoun Shin and Tricia D. Shepherd Dept. of Chemistry, St. Edward's University, Austin TX

Water-filled nanotubes can serve as a model to better understand ion, water, or other solute diffusion through membranes in biological systems. We performed coarse-grained molecular dynamics simulations in order to investigate larger system sizes and longer simulation times possible for common atomistic models. In this work, we studied the impact of different interactions between the nanotube walls and coarse-grained water. By varying the hydrophilic attraction of the wall, we observed changes in flow rates of water through the nanotube. The results of this experiment help us understand how changing the nature of the interactions of water with the nanotube walls impacts the dynamics and the structure of confined water.

34 A Comparison of Nitrogen-donor-HCl Complexes C. Soares and J.A. Phillips (mentor) Department of Chemistry, UW-Eau Claire Eau Claire, WI 54702

We have been investigating and comparing properties of 1:1 H-bonded complexes between hydrogen chloride (HCl) and pyridine (C5H5N), acetonitrile (CH3CN), formaldimine (CH2NH), ammonia (NH3), methylamine (CH3NH2) and hydrogen cyanide (HCN). We are particularly interested in systems for which the H- bonding interaction is affected by inert, low-dielectric media (e.g. NH3-HCl), which can be assessed via matrix-IR experiments. The key to predict the condensed phase effects is to map the intermolecular potential along the N---H and N---Cl coordinates. At this point we are using M06, MP2, B3LYP, wB97XD methods with the aug-cc- pVTZ basis-set to compute potential curves and compare the features of them across this range of complexes, which span a range from strong H-bonds to partial H+ transfer systems. Recent results from charge analyses, which convey the extent of H+ transfer, as well as the effect of solvation energies on the N---Cl potential, may also be presented.

c c 1.336 Å Cl N H c 1.748 Å c c Pyridine-HCl M06/aug-cc-pVTZ ΔE = -8.7 kcal/mol

Figure 1: On the left is a potential curve for the N-H distance of the pyridine-HCl complex calculated using four different methods and on the right you see a equilibrium structure of the pyridine-HCl complex.

35 Computational Electrochemistry of Bio-Derived Quinone Derivatives for Organic Redox Flow Batteries

Julian Taylor, Joshua Schrier

Department of Chemistry, Haverford College

370 Lancaster Avenue Haverford, PA 19041

Large-scale, low-cost electrical energy storage is necessary for using intermittent renewable energy sources such as solar and wind. Redox flow batteries, particularly organic aqueous redox flow batteries, may be the solution to this problem. There are several factors to be accounted for when trying to decide which molecules to use for the redox flow batteries. The molecules need to have high solubility and extreme half-cell voltages. They also need to be synthesized in “green” and low cost ways. To overcome the problem of synthesizing molecules at low cost without too much environmental impact, we observed that Streptomyces coelicolor metabolites produce four quinone molecules (actinorhodin, gamma actinorhodin, DMAC, and aloe) in high yield. Using quantum chemistry calculations, we examined the capabilities of these molecules to serve as electroactive carriers in redox flow batteries. Several combinations of sulfonic acids and primary amines were added to these molecules to increase solubility. Of the 100 molecules tested with cheminformatics for high solubility, 74 molecules were determined to have a solubility of exceeding 1 M. Thermochemical calculations using the B3LYP/6- 311+G (d, p) with the SMD solvation model were used to predict the voltages. So far, the lowest predicted voltage is 0.28 V and the highest predicted voltage is 1.04 V. This combination of high solubility, wide voltage range, and bio-based production offers a new strategy for high-performance redox flow battery materials.

36 A Thermodynamic Study of 3’ Overhangs in the RNAi Mechanism Stephen Telehany PI: Dr. Maria Nagan

RNA interference (RNAi) is the mechanism of post-transcriptional gene repression conserved in eukaryotes. Because of the outstanding specificity of the RNAi mechanism and our ability to engineer double stranded RNA (dsRNA), RNAi can be used as a novel means of “turning off” malevolent genes. The objective of this study is to computationally characterize 10 sets of 3’ overhangs previously studied in a wet bench environment. 100 nanosecond molecular dynamics simulations of 10 sets of 3’ overhangs on a GGCC core were ran at 200 mM and 1 M NaCl ion concentrations. The force field utilized includes the ff99 force field, the Bsc0 and χOL3 backbone modifications, the OPC water model, the TIP4PEW ion model, additionally, backbone oxygens O3’, O5’, O1P and O2P utilized Case’s phosphate parameters for phosphate oxygens. Preliminary RMSD calculations indicate low deviations from both starting and average structures. RDF calculations of the 200 mM system reveal non-aggregating sodium and chloride ions, which was expected at the lower ion concentration. The future direction of this work includes RDF calculations of the 1 M systems, constructing base stacking plots to map out the interactions between the overhanging bases and the terminal base pair and RMSD’s for the full 100 ns simulations. The conclusions drawn from this work will give insight into the RNAi mechanism, specifically the way in which Dicer cleaves long dsRNA into the 21-25 base pair segments characteristic of the RNAi pathway.

Indicates a well equilibrated PDB image taken on VMD system (Black from Avg., Red from highlighting the overhanging bases. Starting).

37 The Effects of Solvent Choice on the Regioselectivity of Ketene Radical Polymerization

Lauren E. Tragesser, Benjamin J. Albrecht, and Daniel S. Lambrecht

University of Pittsburgh, Department of Chemistry, 219 Parkman Ave, Pittsburgh PA 15260

1 Ketenes (H2C=C=O) are versatile bifunctional molecules that have been recently proposed as monomers for synthesizing conjugated polymers via radical polymerization. The radical chain propagation can in principle alternate between all three centers, Cα, Cβ, and O, which may lead to very different polymer products. It is therefore important to understand the regioselectivity of ketene radical polymerization and to identify ways to control it. The purpose of this project is to determine the effect that solvents have on the regioselectivity of the radical system. This work studies the methyl silylacetate radical as a model of methyl trimethylsilyl-acetate, which has been shown to be promising in experiment. All data was obtained using the Q-Chem2 program package and the ωB97X- D density functional3 in combination with the 6-311++G(3df, 3pd) basis set. Solvent effects were modeled by polarizable continuum models, which are dependent upon the dielectric constant (ε) of a solvent. A range of 0 ≤ ε ≤ 80 was sampled to approximate apolar solvents such as oxirane through polar solvents such as water. Frequency calculations and CHELPG charge analysis were performed to determine the effect that solvent has on vibrational frequency, dipole moment, spin density and atomic partial charge. These analyses were chosen to obtain an answer to the question “where is the radical?”: CHELPG charges and dipole moment provide direct information about the charge distribution, and vibrational frequencies are sensitive to bond strengths which is affected by the location of the radical electron. Importantly, these vibrational frequencies can be easily compared to experiment. Finally, the spin density yields a more qualitative picture of where there is unpaired electron density, or where the radical resides. Our results reveal a common pattern: the largest changes occurred for ε=1-10, with very little change for ε=11-80. Frequency was noted to have shifted by about 3 cm-1 within this range, and the partial charge on oxygen decreases by just under -0.05e, while only 0.02e of the extra negative charge are accounted for by the Cβ the rest appears to be contributed by Cα. The magnitude of the dipole moment increases by about 0.5 Debye, which is quite significant. These findings indicate that choosing solvents in the ε=1-10 range should allow one to significantly alter the regioselectivity, and these solvents should be explored experimentally for further study of the radical chain polymerization.

References:

1Tidwell, T. T., Ketene Chemistry: The Second Golden Age. Acc. Chem. Res. 1990, 23, 273-279.

2Y. Shao et al., Mol. Phys. 113, 184-215 (2014) DOI: 10.1080/00268976.2014.952696

3Chai, Jeng-Da and Head-Gordon, Martin, Long-range corrected hybrid density functionals with damped atom-atom dispersion corrections, PCCP, 10, 6615-6620(2008) DOI: 10.1039/B810189B

38 Get the Most Accurate Geometry Optimization Christina Verdi North Carolina School of Science and Mathematics

This project focused on finding the best combination of basis sets, theories, and engines to run an optimization that is the most accurate with experimental data. The jobs for this experiment were run on the North Carolina High School Computational Chemistry Server and the Shodor Server. The engines used were Guassian and Mopac. The theories used were AM1, PM3 PBE, Hartree-Fock, and B3LYP. The basis sets used were minimal, regular, basic, and accurate. It was found that the different combinations between basis sets, theories, and engines have an influence on how accurate the molecule is when compared to the experimental data. Jobs run using the basic basis set were found to have to most accurate data. Hartree-Fock was the most accurate for lengths, but for angles PBE was the most accurate.

39 Investigations into the Mechanisms of Nucleophilic Substitution Reactions with M06-2X Phuc Vo, and Justin Houseknecht Department of Chemistry, Wittenberg University Springfield, OH This computational project investigates the nucleophilic acyl substitution (NAS) reactions between acetic anhydride and either m-methylphenolate or m-nitrophenolate. One of the main goals of this investigation is to computationally model the kinetics of these NAS reactions in both gas and rigid aqueous phase in order to determine whether they occur through step-wise or concerted processes. This is achieved mainly through studying transition state energies, reaction profiles, as well as changes in partial charge distribution as the reactions progress. The project is performed with Gaussian ’09 using the M06-2X density functional method and the 6-31+G** basis set, which, according to previous work on this topic, will most accurately describe these reactions. Results obtained from several conformers so far have revealed that the general reaction profiles vary significantly between conformers across both types of phenolates, and that both step-wise and concerted mechanisms have been observed within each phenolate group. However, a general trend is seen, and is expected, where the m-nitro reaction is concerted and the m-methyl reaction is step-wise. By comparing these results with those produced by SMD solution phase investigations as well as those using the MP2 method, we hope to shed more light on the mechanistic properties of these NAS reactions.

An m-nitrophenolate tetrahedral intermediate (left) and the appropriate reaction profile generated with Gaussian ‘09

40 What is the Most Stable Isomer of Ammonia – Methyl Trifluorogermane Complex (H3N–GeF3­CH3)? Ben Wahl, James A Phillips Department of Chemistry, University of Wisconsin-Eau Claire Eau Claire, Wisconsin 54702

We have undertaken a computational study (m06/6‐31+G(d)) of the stability of various isomers of H3N ‐ GeF3‐CH3, considering not only coordination site, axial or equatorial, but also the various conformations for each of four possible configurations (Figure 1). We will assess the stability by relative energies and the occurrence of imaginary frequencies. Another key issue is the Lewis acid strength of GeF3CH3 relative to GeF4. We expect the methyl group to reduce the acid strength, and we will compare to results for H3N–GeF3‐CH3 to those for H3N–GeF4. Recent results for analogous compounds and those from higher levels of theory will also be discussed.

N

N Ge C Ge C

Ammonia and Methyl Axial Ammonia Equatorial

C Ge C Ge N

N

Ammonia and Methyl Equatorial Methyl Equatorial

Figure 1: Four possible structural isomers for H3N–GeF3CH3.

41 Geometric Influence on Metal-Ligand Covalency in Diphosphine Complexed of Pd and Ti Haochuan Wei Colgate University

Density functional theory(DFT) has been combined with X-ray absorption spectroscopy(XAS) to obtain K-edge behaviors of transition metal diphosphine compounds, and to demonstrate the connection between coordination geometry and metal-ligand bonding. The impact of coordination chemistry, diphosphine bite angle, and the phosphine trans influence on covalency were studied on NiCl2 and ​ ​ PdCl2 complexes that contain PPh3 or Ph2P(CH2)nPPh2. In this project, we use Gaussian DFT and ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ time-dependent DFT to study the K-edge features of TiCl4 and other PdCl2 species containing ​ ​ ​ ​ Ph2P(CH2)nPPh2 or C6H11P(CH2)nPC6H11. The geometric properties of these species are analyzed ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ ​ mathematically to suggest quantification of Ti- or Pd-phosphorous covalency in terms of the metal-P-C angles, instead of bite or twist angles. The Ti or Pd species were optimized by Gaussian DFT to first be compared with crystallography results. Optimized geometries and wave functions were entered into TDDFT to calculate K-edge X-ray absorption spectra. Predicted peaks and intensities were matched with data from X-ray absorption spectroscopy. Given the degree of agreement between XRD and DFT on geometry, and between XAS and DFT results on spectroscopy, we seek to extend our calculation in an attempt to establish more evidences on how the metal-ligand molecular orbitals can be quantitatively linked to metal-P-C angles.

42 Thermodynamic and Kinetic Interactions of Ligands in the SULT1A1 Active Site

Danielle Wilson, Amelie Weems, Larryn Peterson, Mauricio Cafiero

Rhodes College 2000 N. Parkway Memphis, TN 38112

We have studied the substrate selectivity of the sulfotransferase enzyme (SULT1A1) by identifying important protein-ligand interactions in the active-site through electronic structure calculations. The sulfotransferase enzymes (SULTs) catalyze the addition of a sulfate group to a variety of small molecules, including neurotransmitters and xenobiotics. This reaction can activate or deactivate bio-active molecules or change their pharmacokinetic behavior. A variety of ligands analogous to known substrates of the SULT were chosen for study. M062X/6-31G optimization of the ligands was used to find the structures of the ligand-protein complexes in three ways: assuming a static active-site, a static active site with implicit solvent, and a relaxed active site with implicit solvent. Interaction energies between the ligands and the amino-acids of the active-site were calculated using MP2 and M062X with 6-311+g*; these energies can be used to determine the thermodynamic stability of the ligand in the active site. The addition of the sulfuryl group to the ligand depends on deprotonation of a phenol group on the ligand. Thus, the activation energies for proton extraction from the ligand to the histidine residue were calculated.

Figure 1. Dopamine Docked in SULT1A1 enzyme.

43 Inhibiting Lipid A biosynthesis in Gram-negative bacteria through the design and synthesis of natural substrate analogues of LpxC. Kayla A. Wilson, Carolyn F. Dishuck, Gene G. Lamanilao, Sarah N. Malkowski, Larryn W. Peterson, Mauricio Cafiero Dept. of Chemistry, Rhodes College, Memphis TN, 38112

Bacterial infections, including those that lead to septicemia, the 10th leading cause of death in the United States, have become an increasingly serious problem. The emphasis of this study is the development of novel antibacterial compounds which combat Gram-negative bacterial infections via the inhibition of LpxC. LpxC, a zinc-dependent deacetylase, is involved in the biosynthesis of Lipid A, an important part of lipopolysaccharide, which makes up the outer cell membrane of Gram-negative bacteria. When LpxC is inhibited, the production of Lipid A is halted and the virulence of the bacteria is significantly affected. Using key information provided by the crystal structure of LpxC and the work done by our computational collaborators, this study focuses on the design and synthesis of molecules that mimic the enzyme’s natural substrate. The proposed molecules are composed of a nucleoside, a linker, and a zinc binding motif as shown in Figure 1. The synthesis of substrates with a uracil base and a ribosugar linked to a serine containing either a hydroxamic acid or carboxylic acid will be discussed.

Figure 1. General structure of proposed inhibitors

44 Nucleophilic Acyl Substitution of Acetic Anhydride with Substituted Phenolates: MP2

Harrison J. Wood & Justin B. Houseknecht*

Chemistry Department, Wittenberg University, Springfield, Ohio 45501

This project is a computational investigation of the nucleophilic acyl substitution, NAS, reaction mechanism using the MP2/6-31+G** level of theory. There is some disagreement as to whether these reactions occur in a concerted or a stepwise process. We are hoping to show that NAS reactions between weak nucleophiles and molecules that contain a good leaving group will occur in a concerted step, while those with a stronger nucleophile, will proceed through a two step process. The systems that we will be investigating are m-nitropheol (weak nucleophile) reacting with acetic anhydride and m- methylphenol (stronger nucleophile) reacting with acetic anhydride. We hope to show that the reaction involving m-nitrophenolate is concerted, while the reaction involving the m- methylphenol is not. Preliminary results show that as expected the m-methylphenolates react in a two-step mechanism as shown in the figure below.

X X O O + + OO –O O– O O

X O O – OO

45 The effect of yttrium concentration on the proton conduction pathways in barium zirconate

Wanshu Zhu, Gillian Kwan, Xintong Zuo, and Maria A Gomez Department of Chemistry, Mount Holyoke College

Doped perovskites show good promise as fuel cell proton conducting membranes. Proton conduction in 6.25 % yttrium doped barium zirconate is explored. The dopant increases octahedral distortions as it did in the 12.5% yttrium doped system. However, the distortion is barely perceptible in the region between the dopants. The proton binding site energy roughly increases with distance away from the dopant. Rotational and intra­octahedral transfer barriers have been found between proton binding sites using the Nudged Elastic Band (NEB) method. Inter octahedral transfer barriers are found for some sites but when the octahedral distortion is barely perceptible, the proton prefers to make two intra­octahedral transfers rather than a single inter­octahedral transfer. Graph theory and kinetic Monte Carlo are used to find multi­step proton conduction pathways in this lower concentration dopant system. The pathways are compared and contrasted with higher dopant concentration system from our earlier studies..

46 LIST of CONFERENCE PARTICIPANTS Section 5

Last Name First Name Affiliation E-mail Address

SPEAKERS Deng Qiaolin Merck Pharmaceuticals [email protected] Pacific Northwest National Govind Niri Lab (PNNL) [email protected] North Carolina State Jakubikova Elena University [email protected] Lambrecht Daniel University of Pittsburgh [email protected] University of California at Shea Joan-Emma Santa Barbara [email protected] Tschumper Greg University of Mississippi [email protected]

FACULTY Anderson Amil Wittenberg University [email protected] Anderson Kelly Roanoke College [email protected] Boekelheide Nicholas Colby College [email protected] Bruce Chrystal John Carroll University [email protected] Cafiero Mauricio Rhodes College [email protected] Clarke Donald Fordham University [email protected] Gomez Maria Mount Holyoke College [email protected] North Carolina School of Gotwals Robert Science and Mathematics [email protected] Houseknecht Justin Wittenberg University [email protected] Keith Jason Colgate University [email protected] Kennerly William Skidmore College [email protected] Nagan Maria Adelphi University [email protected] Peterson Larryn Rhodes College [email protected] Phillips Jim UW - Eau Claire [email protected] Schrier Joshua Haverford College [email protected] Shepherd Tricia St. Edward’s University [email protected] Shields George C. Bucknell University [email protected] Van Wynsberghe Adam Hamilton College [email protected] Varner Mychel Iona College [email protected] Zachary Karl Mary Baldwin College [email protected]

NON-FACULTY Temelso Berhane Bucknell University [email protected] Young Steve Hamilton College [email protected]

UNDERGRADUATES Appiah-Padi Nana Bucknell University [email protected] Barfield Jacob Roanoke College [email protected] Boca Miranda University of Pittsburgh [email protected] Cao Bobby Bucknell University [email protected] Cirincione Ross Wittenberg University [email protected] Cochrane Colleen Rhodes College [email protected] Dacres David Hamilton College [email protected] Dishuck Carolyn Rhodes College [email protected] Dominguez Eduardo St. Edwards University [email protected] Dominguez Alfredo St. Edwards University [email protected] Donovan Anna Colby College [email protected] Evans Fiona Colgate University [email protected] Fallon Zachary Adelphi University [email protected] Grassi Evan Fordham University [email protected] Hatstat Katie Rhodes College [email protected] Holmes Jessica University of North Georgia [email protected] Hopkins Dan Haverford College [email protected] University of Wisconsin-Eau Hora Nicholas Claire [email protected] Hunter Nathanael John Carroll University [email protected] Jedreski Kathleen Wittenberg University [email protected] Jiang Shangcheng Colby College [email protected] Kim Grace Bucknell University [email protected] Kofke Elijah Skidmore College [email protected] Lanska John UW-Eau Claire [email protected] Lewis Erin Hamilton College [email protected] Marris Patrick Hamilton College [email protected] North Carolina School of Marshburn Drew Science and Mathematics [email protected] McCoy Christopher Wittenberg University [email protected] Morris Mallory Rhodes College [email protected] Morrison Elizabeth Bucknell University [email protected] Nguyen Natalie Wittenberg University [email protected]

Pham Kim St. Edward's University [email protected]

Phillips Richard Haverford Colloege [email protected]

Ritter Abby Rhodes College [email protected]

Shin SeungYoun St. Edward's University [email protected] University of Wisconsin Eau- Soares Camilla Claire [email protected]

Taylor Julian Haverford College [email protected]

Telehany Stephen Adelphi University [email protected]

Tragesser Lauren University of Pittsburgh [email protected] North Carolina School of Verdi Christina Science and Mathematics [email protected]

Vo Phuc Wittenberg University [email protected]

Wahl Benjamin UW - Eau Claire [email protected]

Wei Haochuan Colgate University [email protected]

Wen Yuewei Colby College [email protected]

Wilson Danielle Rhodes College [email protected]

Wilson Kayla Rhodes College [email protected]

Wood Harrison Wittenberg University [email protected]

Zhu Wanshu Mount Holyoke College [email protected]

NOTE PAPER Section 6

14 th MERCURY Conference on Computational Chemistry