REU Site Summer Research Program Chemical Engineering

Program Overview

This NSF-funded project establishes a Research Experience for Undergraduates (REU) in Interdisciplinary, Multiscale Materials Modeling (I3M) at the University of Louisville (UofL) specifically focusing on Computational Chemistry and Molecular Dynamics methods and their application to materials and biological and biomimetic systems. This site will host 9 undergraduate students (Fellows) each summer from 2021 to 2023 at the University of Louisville located in Louisville. KY. It is not expected that REU fellows would have this background prior to their summer internship. Fundamentals of computational chemistry and molecular simulations coupled with hands-on computer laboratory training will be provided at the start of the REU program. Additional training will be provided by faculty mentors and participating graduate students as pertains to the specific REU project. Eighteen faculty members from the SPEED School of Engineering (Chemical Engineering, Mechanical Engineering, Bioengineering), Chemistry, Physics, and Medicine have provided 15 projects from which the REU Fellows can select. Participating Faculty Mentors are identified in the section on REU Faculty Mentors and Departmental Affiliations.  All 15 projects are listed in the following section on Available REU Projects. REU Fellows will have the opportunity to become proficient in a particular method associated with their selected project.

Multiscale Materials Modeling

Figure 1. Computational methods and their location on the multi-scale map

The research focus of our REU Site is Multiscale Materials Modeling. The 15 projects utilize different computational approaches to address focused projects offered by experienced faculty members representing multiple disciplines. In addition to working with project faculty, REU Fellows will interact with graduate students pursuing PhD degrees. The REU Fellows will be introduced to a variety of different computational methods such as Computational Chemistry (quantum mechanics), Molecular Dynamics, and Finite Elements that span the range of size and time domains as illustrated in Figure 1 and discussed briefly in the following sections. Individual REU Projects that utilize one or more specific computational approach are identified in each of the following sections.

Figure 2. H-bonding between a drug and polymer nanoparticle

Computational Chemistry

uses different approaches to obtain an approximate solution to the classical SchrÖdinger wave equation that can provide valuable information about individual molecules or small molecular clusters (small scale). Computational Chemistry methods include atomistic approaches such as first principles or ab initio methods, density function theory (DFT), and semiempirical approaches that require experimental parameterization but have the advantage of being computationally quicker and can be used to characterize larger molecules, such as proteins; however, results may be less accurate than achieved using ab initio methods. Molecular properties that can be obtained using Computational Chemistry include molecular geometry, conformation energies, band gaps, UV and IR spectra, dipole and quadrupole moments, and charge distribution. Programs such as Gaussian provide a graphical user interface (GUI) that facilitates computations. Gaussian and its graphical user interface (GUI), GaussView, will be discussed with hands-on training during the first week of the REU program. Projects: 1, 2, 6, 11, 15

 

Figure 3. MD simulation (side & top views) of a gramicidin channel (purple) in a lipid membrane

Molecular Dynamics Simulation

(MD) lies at the next level of scale. MD simulations enables study of larger molecular structures whose conformational and diffusional properties cover a longer size and time range as was illustrated in Figure 1. In contrast to Computational Chemistry methods, Molecular Dynamics and related Monte Carlo methods can model a broad range of molecular size to enable molecular level characterization of polymers, zeolites, DNA, and proteins. Figure 3 shown a MD simulation of a gramicidin channel (colored purple) sitting in a lipid bilayer membrane (colored green). The solid spheres show water molecules passing through the membrane channel.

Projects: 1, 2, 3, 5

Finite Element

Methods lie at the far range of size and time domains as was illustrated in Figure 1.  These methods provide a bridge between Molecular Dynamics and experiment. They use computational methods distinct from Computational Chemistry and Molecular Dynamics. Figure 4 illustrates how medical imaging of arteries can be used to develop system models to study blood flow using Computational Fluid Dynamics.

Project: 12

Figure 4. Steps in the development of a model of an artery in which blood flow can be investigated using Computational Fluid Dynamics

• Coarse-Grained Simulation of Self-Assembly of Biomimetic Block Copolymers In Different and Mixed Solvents

  Faculty Mentors: Joel Fried and Vance Jaeger, Chemical Engineering

  Student Outcomes: The REU Fellow will gain experience in the use of Coarse-Grained (CG) simulation methods to study polymeric systems with long time and length scale that requires using information obtained from recent atomistic simulations in our laboratory to develop CG models.

  Vision: It is expected that preliminary work during the summer by the REU Fellow in collaboration with PhD students in the Fried and Jaeger groups will provide a beginning in the parameterization of a CG model for biomimetic block copolymers that can be later extended to investigate how membrane proteins incorporate and function in these synthetic membrane bilayers.

  Overview: It has been shown that membrane proteins such as gramicidin and alamethicin can assemble and function within membranes formed from triblock copolymers of poly(2-methyl-2-oxazoline) and polydimethylsiloxane. These biomimetic polymer membranes mimic lipid membranes and self-assemble to incorporate many membrane proteins that form channels supported by the biomimetic membrane fabric. Our experimental studies show that the characteristics of the solvent (e.g., boiling point, chemical structure, hydrophilicity) can affect membrane structure during the self-assembly process. There have been no reported simulations to our knowledge that can provide insight into the self-assembly process and the subsequent functioning of synthetic membrane-protein structures. The Fried group has recently concluded an atomistic study using the COMPASS force field that Fried and then PhD student Pengyu Ren helped to parameterize. This atomistic study will be submitted for publication shortly and will serve as a basis to design an effective CG model (using Martini-like or Shinoda-like force fields) for these triblock copolymers to initially study the self-assembly process of the biomimetic triblock copolymers in different solvent environments. Later, this study will be expanded to investigate how membrane proteins incorporate during an assembly process.

• Molecular Modeling of Delivery Vehicle Interactions with Viral Particles

  Faculty Mentors: Jill Steinbach-Rankins (BE) and Joel Fried (ChE)

  Student Outcomes: Students will learn underlying concepts surrounding drug delivery through the use of mathematical modeling and will apply these considerations to optimize delivery vehicles design parameters. Overall, students will gain experiences in mathematical modeling at the molecular level, computer programing, drug delivery vehicle design and synthesis, in vitro imaging, and virus infectivity assays.

  Vision: The success of this project will have far-reaching ramifications in obtaining a better understanding of the factors that most impact drug delivery vehicle design and their therapeutic success. In addition, this program will foster interactive collaborations between ChE and BE faculty, graduate students, and REU Fellows to bring together computational and experimental expertise.

  Overview: While a variety of antiretroviral drugs demonstrate clinical effectiveness for HIV prevention, daily adherence regimens, toxicity with long-term use, and increased risk of viral resistance remain major concerns. New biological active agents in combination with polymeric delivery vehicles highlight the preventative role that alternative physicochemical methods of protection may provide. In particular, delivery vehicles themselves can be utilized or adapted to serve as a physical modality to protect against virus infection; however, an important challenge to better understand these issues is the molecular-scale interactions between biological agents, their delivery vehicles, and the viral particles. The project goal for this REU student will be to contribute to a computational modeling framework using Coarse-Grained (CG) simulation approaches that enables systematic evaluation of the interactions between proteins (biological active agents) and delivery vehicles (e.g., polymeric nanoparticles, to streamline the development of delivery vehicles that provide physicochemical protection against viral infection. The REU student will apply computational modeling to predict the interactions between viral glycoproteins and fiber or protein interfaces and will identify the extent of protein-virus or fiber-virus interactions that enable physicochemical protection. To validate this modeling, the student will experimentally determine the effects of delivery vehicles on virus inhibition via in vitro imaging, transport, and efficacy assays. This experience will provide the REU student with computational modeling skills in parallel with chemistry, polymer, biochemistry, virology, and biological assay skills.

• Extracting Multibody Potentials to Describe Colloidal Self-Assembly

  Faculty Mentors: Vance Jaeger (ChE), Jerry Willing (ChE), and Stuart Williams (ME)

  Student Outcomes: Student researchers will gain experience in Python programming as applied to materials science, data analysis, Monte Carlo optimization, and molecular dynamics

  Vision: The success of this project will provide a better description of colloidal assembly, accelerating the discovery of functional materials with tunable properties.

  Overview: An accurate description of inter-particle potentials is essential for predicting the dynamics of colloidal particles and their self-assembly into functional structures. Typically, when conducting a particle dynamics simulation, particle interactions are described with a pairwise summation of two-body interactions. However, three-body and higher order interactions play a significant role in the preferred stable structures that colloidal crystals form. Yet, due to limited computer power and difficulties with empirical tabulation of multibody potentials, these higher-order terms have largely been ignored. Meanwhile, a confluence of recent advances now allows for the accurate, empirical measurement of multibody potentials and the inclusion of higher-order terms in computer simulations. This project focuses on the development of techniques for collecting and analyzing large sets of microscopic videos of the assembly of colloidal crystals, with close collaboration between the ChE and ME department faculty. From experimental videos, two- and three-body interactions will be calculated using two levels of theory. First, within dilute colloidal suspensions, the dynamics of small interacting clusters of particles will be used to derive interaction potentials via an inversion of the numerical integration chain typically used in Brownian dynamics simulations. Previous work has analyzed systems away from equilibrium, but it is hypothesized that with enough data, the same method can be used in near-equilibrium experiments. Second, an inverse Monte Carlo approach will be used in crowded colloidal environments. Pair and triplet correlation functions provide structural information about experimental systems which can be converted to forces.

• Assessing Natural Orbital Active Space Decomposition Strategies for Non-Adiabatic Processes in Composition Intermediates

  Faculty Mentors: Lee Thompson and Jinjun Liu, Chemistry

  Student Outcomes: Students will gain experience in applying computational and molecular spectroscopic methodologies to the study of non-adiabatic chemical process as well as an understanding of how computations and experiments are used in concert to elucidate the behavior of chemical systems. Additional skills include programming, quantum chemical method development, molecular simulation and spectral interpretation.

  Vision: The proposed work will elucidate vibronic interactions that govern combustion intermediate reactivity and assist in the development of cleaner combustion technologies. Additionally, completion of the proposed project will provide a basis for determining the utility of the natural orbital active space decomposition (NO-ASD) strategy as a reduced cost approach to studying nonadiabatic processes.

  Overview: Nonadiabatic processes occur when reaction pathways cross through regions of electronic near-degeneracy. Modelling of nonadiabatic processes is essential for understanding photochemistry, collisions of electronically excited species and electron transfer processes. Simulation requires computation of multiple electronic states that are treated equally as well as approaches to account for open-shell systems and strong-correlation. However, current computational methodologies that have the desired features are computationally expensive and difficult to apply to systems of chemical interest. The group of Prof. Thompson has developed a NO-ASD strategy that utilizes the diabatic properties of symmetry-broken mean-field wavefunction descriptions of excited states as an approach to reduce computational scaling. The REU student will investigate the performance of NO-ASD applied to a number of free radical species that play important roles in combustion and atmospheric chemistry. Results from simulations will be used to interpret results from the experimental high-resolution laser spectroscopy laboratory of Prof. Liu to unravel the intricate connection between electronic and nuclear motion in chemical reaction intermediates. The REU student will investigate the spin-vibronic structure of NO₂ at the conical intersection connecting its  and  electronic states and the “quantum chaos” due to the interplay between the spin-orbit and Renner-Teller interactions. Finally, the spin-vibronic structure of the  ground electronic state of the methoxy radical (CH₃O) and its asymmetrically deuterated isotopologues (CH₂DO and CHD₂O) will be investigated aiming at understanding its unimolecular dissociation reaction CH₃OàCH₂=O+H or CH₂+OH or HCO+H₂. In each case, quantum chemical calculations will be carried out to simulate and fit experimentally obtained high-resolution laser spectra, and quantitative understanding of vibronic interactions in the target systems will be achieved.

• Quadruplex DNA and Gene Promoters as Transcription Control Elements

  Faculty Mentor: John Trent, Medicine

  Student Outcomes: Students will gain experience in empirical simulations applied to proteins and nucleic acids and with combining biophysically derived parameters in simulations for understanding structures that are not amenable for traditional structural biology.

  Vision: The success of this project will provide a better understanding of the basic structures of gene promoters as potential transcription regulation by quadruplex DNA.

  Overview: Receptor-based drug discovery via computational virtual screening relies on knowledge of the structure of the target protein or nucleic acid. In many cases, traditional structural biology approaches cannot determine structures of interest. Therefore, a combination of molecular-dynamics simulations and biophysical approaches, such as analytical ultracentrifugation, can provide experimental and calculatable hydrodynamics parameters that can rule in or out structural models as candidates. The REU student will have the opportunity to work directly with Prof. Trent and a graduate student to use computational methods, such as explicitly solvated atomistic molecular-dynamics simulations combined with detailed hydrodynamics calculations, and biophysical techniques, including circular dichroism, differential scanning fluorimetry, and analytical ultracentrifugation, to investigate the structure of quadruplex DNA, as transcription regulation elements, in gene promotors. They will subsequently use several computational virtual screening methods, docking and scoring, to evaluate pre-filtered drug-like libraries for potential binders, coupled with free energy calculations as a final filter. These small molecule binding predictions will be experimentally verified as tools to further examine quadruplex DNA as potential transcription control elements.

• Multilayered Nanoparticle-Fiber Composites for the Prevention of Bacterial Infections

  Faculty Mentors: Jill Steinbach-Rankins and Hermann Frieboes, Bioengineering

  Student Outcomes: Students will learn to fabricate probiotic-containing 3D bioprinted and electrospun fiber delivery vehicles. They will learn how to characterize these platforms for morphology, loading, and release, and apply these vehicles to biological assays that relate to the in vitro efficacy against bacteria-vaginosis infection.

  Vision: This project will explore how the features of different bioprinted or fiber platforms impact their ability to provide therapeutic effect to bacterial vaginosis infection.

  Overview:  Bacterial vaginosis is a highly recurrent imbalance of the vaginal microbiota that affects 1/3 of U.S. women, and is associated with numerous adverse health outcomes; however, antibiotic-sparing treatments such as the delivery of probiotic bacteria often require daily administration for effectiveness. We will optimize the design of a novel delivery platform composed of either 3D-bioprinted scaffolds or electrospun fibers that will deliver lactobacilli, then test their efficacy in in vitro models of BV. This study will provide a foundation for improved therapeutic outcomes, while providing new insight into the effects of probiotic-containing devices on BV disease markers. The project goal for this REU student will be to contribute to the development of bioprinted or electrospun fiber (EF) delivery platforms that can prolong the release of probiotics to prevent and treat infection in bacterial co-culture experiments. The student will help develop a computational modeling framework that systematically evaluates the interactions between probiotics incorporated within delivery vehicles and the host tissue, to enable the streamlined development of sustained-delivery platforms. This experience will provide the REU student with mathematical and computational modeling skills in parallel with polymeric delivery platform, microbiology, and biological assay skills.

• Modeling of Cancer Growth and Treatment Combining Engineering and Biology Principles

  Faculty Mentor: Hermann Frieboes, Bioengineering

  Student Outcomes: Students will learn how principles from engineering and the physical sciences can be applied to study disease progression

  Vision: The success of this project provides for a better understanding of cancer growth and treatment.

  Overview: This project pursues an improved understanding of disease progression and response to treatment by applying principles from engineering and the physical sciences. The work is focused on the development and integration of mathematical modeling, computational simulation, and experimental biology techniques to study cancer. This effort is part of the burgeoning field of “Physical Oncology,” in which cancer is studied not only from a biological standpoint but also as a physical system using mathematics and physics. This interdisciplinary study of cancer requires that experimental and clinical data drive the computational and modeling work. The aim is to predict tumor behavior from molecular- and cellular- scaled events, with the ultimate goal to help guide the treatment of individual patients. This research intersects the fields of cancer biology, scientific computing, data visualization, mathematical biology, and physical oncology. Activities related to this project may include prediction of tumor progression as a function of molecular- and cell-scale events, characterization of treatment response by integration of modeling and experimentation advancement of mathematical modeling and computational techniques to characterize tumor growth, and modeling and simulation of cancer nanotherapy and immunotherapy.

• Modeling of Drug Transport and Efficacy in Living Tissues

  Faculty Mentors: Hermann Frieboes and Jill Steinbach-Rankins (Bioengineering)

  Student Outcomes: Students will gain experience into drug delivery principles to living tissue.  Students will also be exposed to modeling and simulation techniques applied in the field of biomedical engineering

  Vision: The success of this project provides a better understanding of how drug delivery platform parameters affect transport and pharmacodynamics, in diseases including viral infections and cancer.

  Overview:  Identifying key design parameters that can prolong drug delivery and maximize effectiveness against a diversity of diseases is difficult to achieve, as complex drug-drug, drug-vehicle, and drug-tissue interactions often preclude systematic evaluation via iterative empirical processes. Mathematical modeling enables the testing and optimization of theoretical drug delivery systems, complementary to experimental in vitro and in vivo validation, which allows for increased precision and greater understanding of kinetic and chemical principles governing drug pharmacokinetics. This project explores how changing delivery vehicle design parameters can modulate the release of drugs into living tissue. The modeling includes analysis of parameters specific to particular drugs and drug delivery vehicle formulations. Students will contribute to the proposed project by simulating the transport and pharmacodynamics of drugs targeting various diseases, including viral infections and cancerous tumors, and thus help to rationally design delivery platforms for more effective prophylactic and treatment approaches.

• Enhancing Electrocatalyst Activity and Acid Stability for the Oxygen Evolution Reaction with Controlled Oxygen Vacancies

  Faculty Mentors: Badri Narayanan (ME) and Joshua Spurgeon (ChE)

  Student Outcomes: Students will learn methods for catalyst synthesis, electrochemical characterization techniques, and materials characterization tools such as electron microscopy and X-ray photoelectron spectroscopy. They will also learn modern scientific programming, and computational materials modeling techniques and gain expertise in relating computational predictions with experimental observations.

  Overview: Efficient water oxidation holds the key to the advancement of a number of renewable energy technologies, including solar fuels and metal–air batteries. Until now, only a few noble metal oxide catalysts are known to be highly active and corrosion-resistant in acidic media where efficiency is maximized. This project will advance the scalability of electrolytic H₂ by developing high activity, low-platinum-group-metal (PGM) oxygen evolution reaction (OER) catalysts for acid-based devices. Catalysts will be synthesized with an innovative non-equilibrium plasma oxidation process that produces phase homogeneous multi-metallic oxide nanomaterials of controllable composition. Our recent work shows that compositions of such multi-metallic oxides can be tuned to achieve favorable oxygen evolution activity; however, the underlying physical factors that control OER activity in such materials remain unclear. In particular, the role played by concentration, location, and distribution of oxygen vacancies, as well as presence of metastable mixed metal oxide phases on OER activity is still not well understood. Here we will integrate electrochemical experiments with density functional theory (DFT) calculations and molecular dynamics (MD) simulations to address these outstanding questions. Such knowledge, alongside precise experimental control over incorporation of Ovs in the bulk and at the surface of plasma-synthesized metastable phases of mixed metal oxides will enable the production of acid-stable, low-PGM catalysts with highly competitive activity. For the REU project, we will focus on W_1-xIr_xO_3-δ catalysts. The bulk OV concentration will be controlled using dilute H₂ in the plasma during growth; while surface OV concentration will be tuned via variable post-fabrication exposure to a reducing agent. Electrochemical activity will be characterized and correlated to the catalyst oxygen vacancies. The electrochemical data will be compared to computational modeling in which density functional theory (DFT)/molecular dynamics (MD) calculations will be performed for Ovs at all sites near the Ir metal at the surface and in the bulk to determine the energetics of the intermediate reaction steps of the electrochemical oxygen evolution.

• Receptor-Based Bayesian Structure–Activity Relationship Models ror Non-Mutagenic Carcinogens and Anticarcinogens

  Faculty Mentor: Al Cunningham, Medicine

  Student Outcomes: Student researchers will gain experience in computational toxicology and chemistry, data analysis, model development and validation, and Bayesian.

  Vison: The successful development of this project will enable us to explore the next dimension of structure–activity relationship modeling with the inclusion of biological descriptors and chemical-biomolecular interactions along with traditional chemical-defined model descriptors.

  Overview:  A fundamental problem exists for carcinogen determination and anticancer drug development. Few targets, aside from DNA and a small number of receptors, are known that can (1) be used to assess/predict the carcinogenic/anticarcinogenic potential of chemicals and (2) serve as targets for assessing chemical safety or for the discovery of cancer chemopreventive and therapeutic agents. To address this problem, we are developing a Bayesian structure-activity relationship (SAR) modeling algorithm that uses the significant power of virtual ligand-receptor interactions to model the carcinogenic and anticarcinogenic activity of chemicals. Preliminary studies indicate receptor-based models for non-mutagenic carcinogenesis and anticarcinogenesis have a good degree of predictiveness based on cross-validations. We propose to develop a Bayesian modeling scheme to these data in order to help identify the most important receptors that may be involved in the carcinogenic/anticarcinogenic activity of chemicals. The rationale for this proposal is: There is a paucity of efficient and systematic methods to (1) predict the carcinogenic potential of non-genotoxic chemicals, and (2) predict/analyze the anticarcinogenic effects of chemicals (noting that anticarcinogenesis is an important and overlooked component of the National Toxicology Program’s two-year rodent cancer bioassay). Our hypothesis is that non-genotoxic carcinogens and anticarcinogens interact with molecular targets aside from DNA and these small molecules modulate tumorigenesis through those targets. That is to say, more nuanced events aside from DNA interactions may be considered for cancer/anticancer mechanism exploration and drug discovery (e.g., pathway analysis and receptor-mediated drug development). The project goal for the REU student will be to contribute to the development of Bayesian receptor-based SAR models for carcinogens/anticarcinogens. The student will help in the process to combine rationally the vast amount of structural and pathology data for non-genotoxic carcinogens and anticarcinogens with a large library of molecular targets to ultimately model agents with targets that perturb tumorigenesis. These models/data can be may lead to a better predictive system for carcinogens/anticarcinogens, facilitate the development of safer consumer products and industrial chemicals, the identification of natural products as chemopreventive agents, starting points for anticancer drug discovery/development, and the repurposing of existing drugs for cancer treatment.

• Assessing Electronically Excited States of Vitamin B12 Derivatives via Time-Dependent Density Functional Theory

  Faculty Mentor: Pawel Kozlowski, Chemistry

  Student Outcomes: Students will gain experience in applying state-of-the-art computational techniques to study electronically-excited states of complex bioinorganic molecules.

  Overview: The biologically active forms of vitamin B₁₂, methylcobalamin (MeCbl) and adenosylcobalamin (AdoCbl), are important cofactors in a variety of enzymatic processes. Typical structure of cobalamins (Cbls) consists of a corrin ring with a central cobalt (Co) atom which is equatorially connected with the four nitrogen atoms of the corrin macrocycle. The metal center is axially ligated to the intramolecular 5,6-dimethylbenzimidazole base (DBI) at the lower axial position (α face) and the upper axial position can be ligated with various ligands such as adenosyl (Ado), methyl (Me), or aqua (H₂O). In addition to their roles as cofactors, Cbls have unique photolytic properties based on the light-sensitivity of the organometallic Co-C bond. Their photolysis is mediated by low-lying excited states, where photodissociation of the Co-C bond leads to formation of singlet-born alkyl/cob(II)alamin radical pairs (RPs). Also, the geometric and electronic characteristics of Cbls are different based on whether the DBI base is bound as the lower axial ligand (base-on) or replaced by water (base-off) in strongly acidic conditions. The principal objective of this research project is to elucidate photolytic properties of cobalamins via analysis of electronically excited states. To accomplish this, we will employ methods of computational chemistry that include Density Functional Theory (DFT), dispersion-corrected DFT (DFT-D), time-dependent DFT (TD-DFT) and the corresponding combined molecular mechanics (MM) approaches, i.e. QM(DFT)/MM, or QM(TD-DFT)/MM, respectively. Potential energy surfaces (PESs) associated with low-lying excited states, constructed as functions of both axial bond lengths, provide the most reliable tool for understanding the photodissociation mechanisms. Based on such constructed PESs, energy pathways connecting metal-to-ligand charge transfer (MLCT) and ligand field (LF) electronic states, can be associated with the light induced photo-homolysis of the Co-C bond that is observed experimentally. Likewise, the crossing of the S₁/S₀ surfaces can be used to describe internal conversion (IC) to the ground state. Particular emphasis will be placed on the differences observed in the photodissociation mechanisms and the photolytic properties of the base-on versus base-off forms.

• Non-Invasive Assessment of Coronary Artery Disease Using Patient Imaging and Computational Fluid Dynamics

  Faculty Mentor: Eric Benson, Chemical Engineering

  Student Outcomes: Student researchers will gain experience in applied computational fluid dynamics as applied to blood flow modeling.

  Vision: The success of this project provides a better understanding of the relationship between blood flow phenomena/physiology and severity of blocked arteries.

  Overview: One million invasive coronary angiography (ICA) procedures are performed every year in patients who present with chest pain or are known to have stable coronary artery disease (CAD). The goal of the procedure is to determine if there is any significant blockage (stenosis) that limits blood flow to the heart muscle in the coronary arteries. The cardiologist performing the procedure in the catheterization lab determines the significance of the stenosis by one of two methods: either by visually estimating the degree of stenosis (‘eyeballing’ the stenosis), which is the routine practice and is done in the majority of patients, or by invasively measuring fractional flow reserve (i-FFR), which is the Gold Standard test that has been demonstrated to both improve patient outcomes and diminish the cost of healthcare. However, i-FFR is only performed in 10-20% of patients because it is invasive, expensive, and time-consuming, and requires more radiation and contrast exposure. In this project, we combine computational fluid dynamic (CFD) modeling with patient-specific imaging of the arteries to visualize blood flow in and around stenotic regions. Patient renderings will be provided by Cardiologists in the University of Louisville Department of Medicine, providing interdisciplinary collaboration. Students in the REU program will utilize CFD to characterize the altered blood flow dynamics that is indicative of physiologically significant stenosis. By characterizing in terms of fluid dynamics, transport, and mixing concepts, students will apply fundamental chemical engineering concepts to develop quantifiable metrics instrumental in cutting edge non-invasive diagnostic medicine. Retrospective validation studies will be conducted to compare developed metrics with data from patients with known coronary arterial disease who have already undergone coronary angiography and i-FFR measurement.

• Characterizing The Interactions of Peptides and RNA with Silica Nanoparticle for Biopreservation

  Faculty Mentors: Vance Jaeger and Gautum Gupta, Chemical Engineering

  Student Outcomes: Students researchers will gain experience in non-equilibrium molecular dynamics methods and become familiar with concepts of adsorption and self-assembly

  Vision: The success of this project will lead to a better understanding of the interactions of bio-macromolecules with inorganic surfaces with the goal of encapsulation of proteins and RNA.

  Overview: Biomineralization is the process by which biomacromolecules template the growth of inorganic structures. The Stöber process is a method to synthesize silica nanoparticles from small precursor molecules. Recent work in the lab of Dr. Gupta has indicated that a modified Stöber-like process can be enhanced through the use of microwaves to produce biomineralized silica nanoparticles that protect the biomacromolecule from digestion by enzymes (patent in process). This technology is of interest for the long-term storage and recovery of biological samples. To gain molecular level insight into the thermodynamics of protein and RNA-silica interactions a set of non-equilibrium molecular dynamics simulations will be conducted. Alchemical mutation of protein or RNA sidechains will be performed in the GROMACS molecular dynamics package with the assistance of the pmx python package. Fast-growth thermodynamic integration will be analyzed using the Crooks Gaussian Intersection method and the Bennett Acceptance Ratio. Students will calculate differences in the free energy of amino acid and nucleotide binding to theoretical silica surfaces at various pHs. The CHARMM force field will be used with the compatible Interface Force Field (IFF) that has been specifically designed to reproduce interfacial properties of inorganic materials such as silica. Students and mentors will develop and test hypotheses about the mechanisms by which biomineralized silica nanoparticles protect biological samples.

• Modeling of Tandem Semiconductor Single-Particle Solar Water-Splitting Performance

  Faculty Mentors: Josh Spurgeon and Eric Berson, Chemical Engineering

  Student Outcomes: Students will learn the criteria for efficient solar photoelectrochemical water splitting and principles that describe the energy-conversion behavior. They will develop skills with finite element method simulation software, and build a predictive model to identify optimal performance parameters.

  Vision: This project will fit within the context of PI Spurgeon’s NSF CAREER award and contribute insight to understanding the mesoscale phenomena of product generation and dissolved gas concentration profiles in a solar-driven semiconductor particle slurry for low-cost hydrogen generation.

  Overview: Hydrogen generation via solar water-splitting is an ideal pathway to convert and store the energy of intermittent sunlight into a reliable, energy-dense, portable form. A semiconductor particle slurry reactor holds promise as one of the most cost-effective approaches to achieve this, by obviating the need for many expensive components of photovoltaics and electrolyzers. The tandem single-particle design in this work allows for maximum efficiency, but the co-evolution of H₂ and O₂ leads to concerns regarding safety and back reactions. Thus, slurry reactor conditions are desired which permit high solar-to-hydrogen efficiency while maintaining the H₂/O₂ concentrations either above or below the flammability limits. To connect observable slurry metrics (e.g., H₂/O₂ production rate) to fundamental particle behavior as a function of the device parameters, we will perform mesoscale simulations of the particle photogeneration, reaction kinetics, and product transport. PI Spurgeon has previously been involved with multiphysics modeling for solar fuels, including PEC system modeling as a function of electrode geometry, catalyst properties (exchange current density, Tafel slope, etc.), and material and solution conductivity. Current distribution and ohmic and activation overpotentials can be modeled, along with the back-reaction rates and H₂ and O₂ steady-state concentration profiles. COMSOL Multiphysics will be the modeling software, which uses a mesh approach with the finite element method to iterate interconnected systems of differential equations. Such PEC models have been described which solve for charge transport and conservation in the liquid electrolyte and solid semiconductor. The kinetics of the electrochemical reaction are modeled with Butler-Volmer expressions, species transport by an advection/diffusion equation, charged species transport by the Nernst-Plank equation, and the tandem semiconductor performance by a detailed balance limit and fitted diode equation. The PEC model will simulate the two-dimensional profiles for H₂ and O₂ concentration, back reaction rates, and ohmic potential drop around a single particle as a function of device geometry (e.g., wire height, catalyst surface area, etc.) and photocurrent density. The boundary conditions around this single-particle unit cell will be varied to model the effects of packing density in a slurry.

• Exploring Two-Dimensional Layered Materials using Computer Simulations

  Faculty Mentors: Chakram Jayanthi and Ming Yu, Physics

  Student Outcomes: Students participating in this project will gain experience in first-principles and semi-empirical simulations as applied to two-dimensional (2D) layered materials with van der Waals (vdW) interaction between layers. They will also gain experience with programming skills, computer algorithms, and materials simulation techniques.

  Vision: The success of this project provides a better understanding on the role played by vdW interactions in two-dimensional layered materials. Specifically, a successful outcome of this project will shed light on different ways of manipulating electronic behaviors of 2D materials that include effects of strain due to twisting, external strain, and different stacking orders that can lead to an understanding of materials design of layered heterojunctions.

  Overview: An accurate description of van der Waals (vdW) interactions, is essential for predicting the structure and properties of layered two-dimensional materials and many soft matter materials. A successful modeling of vdW interactions must correctly reproduce its influence at both medium and long-range distances in two-dimensional crystalline materials. In the density functional theory (DFT) based methods, the vdW interactions are addressed by adding atom-pairwise dispersion functional to the DFT total energy expression. However, the implementation of DFT-based methods to large complex nanostructures and biomolecules is restricted due to the computational memory and time required to study large systems, even with present day computer architectures. To overcome the size-limitations of modern density functional methods, we will adopt a semi-empirical Hamiltonian, referred as SCED-LCAO, which was developed by Condensed Matter Theory group at University of Louisville (Jayanthi and Yu). The SCED-LCAO method uses parameterized interaction Hamiltonian that goes beyond the traditional two-center tight-binding methods by including environment-dependent (ED) functions and allowing self-consistent (SC) charge calculations. The framework used for developing this Hamiltonian is the linear combination of atomic orbitals (LCAO). We will modify this SCED-LCAO Hamiltonian to include weak dispersive interaction correction for performing computational studies of layered materials. The projects that we will undertake in the NSF/REU proposal include: (i) bi-layers of graphene, (ii) 2D bi-layer systems such as silicene/silicene, germanene/germanene, (iii) bi-layers of hetero-compound systems such as SiC or GeC, and (iv) 2D heterojunctions of SiC/GeC with strong in-plane chemical bonding but weak inter-planar vDW interaction. These projects will shed light on how electronic behaviors of layered materials can be manipulated by strain, and how the competition between inter-layer coulomb interaction and vDW interactions manifest in layered materials.

• Molecular Simulation of m/p-Polybenzimidazole Membranes for Extended Electrochemical Devices

  Faculty Mentors: Joel Fried and Vance Jaeger, Chemical Engineering

  Student Outcomes: The REU Fellow will gain experience in the use of molecular dynamics simulations to study the properties of high-polymer content phosphoric acid-doped copolymers of polybenzimidazole (PBI) that have shown significant potential for extended lifetime electrochemical devices. Refs. Pingitore at al., ACS Appl. Energy Mater., 2019, 2,

  1720-1726; Li et al., Polym. Eng, Sci., 2013 (DOI 10.1002/pen.23295.)

   

  Vision: It is expected that simulation will provide significant information in areas such as polymer conformation, density, and hydrogen bonding between the copolymer and phosphoric acid. It is expected this research will leave to further improved membrane systems that can be used for large-scale hydrogen production. Results obtained during the summer REU coupled with preliminary simulations already completed in our facilities will be suitable for publication.

  Overview:

  The work of Benicewicz cited under Outcomes (ACS Appl. Energy Mater., 2019, 2, 1720) has shown that high-performance PBI membranes can be prepared by a new process using poly(phosphoric acid) (PPA). Preliminary studies have shown that such membranes can be used in high-temperature polymer electrolyte membrane fuel cells (HT-PEMFCs). These membranes exhibit lower compressive creep compliance, high proton conductivities, and long-term performance. One important potential application is the electrolyzer for producing hydrogen via a solar-driven hybrid-sulfur process (J. Appl. Electrochem, 2016, 46, 829).

• Eligibility Information

  To be eligible for the REU Site Summer Research Program for summer 2022, you must:

  • US. Citizen, U.S. national, or permanent resident of the United States
  • Majoring in in Engineering, Chemistry, Physics or a closely related field and has completed at least one year of study
  • Minimum GPA of 3.0

  Women, underrepresented minorities, and students from institutions with limited research opportunities are especially encouraged to apply!

• Stipend, Housing, and Travel

  REU participants will receive a stipend of $600 per week and will be provided with housing in campus dorms, costing approximately $1,800 over the 10-week session for a shared-bedroom dormitory. In addition, REU participants will receive a $10/day supplement for food. Airfare to and from Louisville and transportation to and from the airport is budgeted at $500 per student. A supplemental $90/week is available to cover the cost of structured outings, working lunches, and other related costs.