2020 Graduate Student Symposium

Help us judge student presentations!

Please fill out this form to help us score the presenters in Presentation Session I/II.

Session 1

Sage Scheiwiller: Investigation of morphology and conformation of polythiophene/polystyrene blends with neutron and x-ray scattering

Prabhleen Kaur: Development of non-fouling and lubricated surfaces for orthopedicimplants using highly-reactive haloester surface initiators for ARGET-ATRP

Brittany Bishop: Minimizing Interfacial Defects in InP SphericalQuantum Wells via Strain-Engineering

Nida Janulaitis: Computational study of the kinetics of the Menshutkin reaction using explicit and implicit solvent representations

Kacper Lachowski: Towards Mechanistic Understanding of Peptide Mediated Plasmonic Nanoparticle Synthesis

David Juergens: A data-driven algorithm for de novo protein sequence design

Mike McKenna: Establishing multiple particle tracking as a viable tool for characterizing changes in the brain extracellular environment throughout development and in response to injury

Zang Le: Detecting Solid Oxide Fuel Cell Stack Failure from Impedance Spectroscopy Using Machine Learning

Session 2:

Evan Komp: Machine Learning Quantum Reaction Rate Constants

Andrea Joseph: Nanotechnology for therapeutic applications in the developing brain

Jaime Rodriguez: High-Throughput and Data-Driven Strategies for the Design of Deep Eutectic Solvent

Jason Cain: Agent-based models explore multi-scale dynamics of hypoxia-induced tumor angiogenesis

Mary O’Kelly Boit & Jenny Bennett: Protein-Based Coiled-Coil XTEN Hydrogels as a Cell Carrier for Injectable Therapies

Jeremy Filteau: Leveraging High Throughput Organotypic Brain Slice Culturing to Time Lapse Image Cellular Interactions

Hawley Helmbrecht: Comprehensive Brain Cell Morphology Pipeline for Feature Quantification in Fluorescent Images

Coleman Martin: A COVID-19 Nucleic Acid Amplification Self-Test for use in the Home

Dr. Samantha Johnson

Dr. Samantha Johnson is a Computational Scientist at Pacific Northwest National Laboratory. She studies molecular electrocatalysts for chemical energy storage and harvesting using theoretical and computational methods. In particular, she is interested in the role of a catalyst’s surroundings in its performance and behavior. Dr. Johnson has a B.S. in Chemical Engineering from University of Colorado, Boulder. She received her Ph.D in materials science from California Institute of Technology in 2017 and was a postdoctoral researcher in the Center for Molecular Electrocatalysis at Pacific Northwest National Laboratory. She was a National Science Foundation Graduate Fellow and a Resnick Fellow during her graduate studies. She has received several awards for her research, including a Clean Energy Education and Empowerment (C3E) Poster Award and Best Poster at the Northwest Theoretical Chemistry Conference and a Department of Energy Team Science Award. 

Dr. Matthew Wagner

Originally from St. Louis, Missouri, Matt Wagner traveled the country for his education before settling in Cincinnati, Ohio, as an employee of The Procter & Gamble Company. Matt earned his BE in Chemical and Biomedical Engineering at Vanderbilt University (1998) and PhD in Chemical Engineering at the University of Washington (2002) under Professor David G. Castner. Following a postdoctoral fellowship at the National Institute of Standards and Technology, Matt joined The Procter & Gamble Company in 2004. While at P&G, Matt has held various positions across R&D in analytical chemistry, new technology and product development, and fragrance development, in addition to earning a MBA at Xavier University (2010). In his various roles, Matt looks to leverage his technical mastery, consumer insights, and business acumen to create delightful household and personal care products that improve the daily lives of people around the world.

Dr. Dan Widmaier

Dan Widmaier is the co-founder and CEO of Bolt Threads. Dan earned his PhD in Chemistry and Chemical Biology from UC San Francisco, where his graduate research involved designing genetic circuits to control microbial organelles. In 2009, he and his two co-founders founded Bolt Threads, which is creating the next generation of materials using biomimicry and biotechnology. Harnessing his experience in both science, and business development, Dan has grown Bolt Threads from an incubator start-up to a biomaterials platform company with 100+ employees. He has led Bolt Threads through multiple fundraising rounds, created lasting partnerships with brands like Stella McCartney and Patagonia, and launched an increasing number of commercially available products using Bolt Threads materials.

Help us judge these presentations!

Please fill out this form to help us score the presenters!

Investigation of morphology and conformation of polythiophene/polystyrene blends with neutron and x-ray scattering

Sage Scheiwiller

Blends of conductive conjugated polymers and strong flexible commodity polymers help make many technologies possible, yet there are pieces missing from a holistic understanding of the effect of morphology and the effect on polymers performance. In order to probe microstructures between two different polymers, contrast variation small-angle neutron scattering is in partnership with x-ray scattering and electronic analysis methods. 

Development of non-fouling and lubricated surfaces for orthopedicimplants using highly-reactive haloester surface initiators for ARGET-ATRP

Prabhleen Kaur

Orthopedic implants are usually made using various industrially available materials including ceramics, metals, and plastics. These materials possess the desired mechanical strength but lack biocompatibility which can be enhanced through surface modification techniques like ARGET ATRP (Activator Regenerated by Electron Transfer Atom Transfer Radical Polymerization). Coating the substrates with non-fouling materials including Zwitterionic polymers such as poly-sulfobetaine methacrylate (pSBMA) can create ultra-low fouling and highly lubricated surfaces, ideal for various orthopedic implants. This work focuses on using plasma deposited haloester monomers like methyl 3-bromopropionate (M3BP) and methyl 2-bromopropionate (M2BP) as initiators for ARGET ATRP of SBMA. This is a solvent-free method of immobilizing the initiator on the surface, which makes it ideal for materials like polyurethane which are not compatible with the organic solvents used for immobilization of commonly used ATRP initiators like bromoisobutyryl bromide. This technique can be used for substrates of any geometry and surface chemistry, creating surface coatings that are resistant to delamination while providing tunable surface density.  Initiator density of 30% was obtained using this method resulting in thick pSBMA coatings. The coated surfaces showed an 87% reduction in the protein adsorption in comparison to the pristine surfaces. The average value of friction coefficient for a dry pSBMA sample was measured to be 190. For a sample soaked in DI water for 24 h, the average value was measured to be 0.99. The two orders of magnitude reduction for the wet samples is suggestive of the formation of hydration sheath layers by pSBMA brushes. 

Minimizing Interfacial Defects in InP SphericalQuantum Wells via Strain-Engineering

Brittany Bishop

Strain-engineering techniques are employed in colloidal nanomaterials to control the number of defects and mitigate strain at the interfaces of core/shell and multi-shelled nanostructures. Spherical quantum wells (SQWs) are a type of multi-shelled nanostructure that utilizes strain-engineering by sandwiching a thin layer of a quantum emitter between a core and outer shell of a larger-band-gap material. If the quantum well layer is very thin and the outer shell is thick enough, the compressive and tensile forces will compensate one other and create a plane of neutral strain at the interface, effectively minimizing interfacial defects and maximizing quantum efficiencies. Here, the growth of InP SQWs via SILAR is presented. The impacts of precursor reactivity, core and outer shell material, and interfacial surface chemistry on nanocrystal morphology and interfacial defects were studied.

Computational study of the kinetics of the Menshutkin reaction using explicit and implicit solvent representations

Nida Janulaitis

Many of the solvents currently used in the chemical industry are hazardous to human health and detrimental to the environment.¹ Replacing toxic solvents frequently used to carry out organic chemistry reactions with less toxic “green” alternatives would have major implications for the sustainability of chemical production. In order to evaluate the role of solvent models on kinetics, we have used string search and eigenvector following with density functional theory (DFT) to find minimum energy pathways (MEPs) for the ground-state, well-studied Menshutkin reaction. We investigated the reaction in both the gas phase and the liquid phase, with both an entirely implicit and a mixed implicit-explicit description of the aqueous solvent. The MEPs, combined with classical transition state theory, were used to calculate the reaction rate constants. Comparison between our results and previous work validates our methods and provides insight regarding the construction of an initial data set for guiding the prediction of reaction rate constants in solvents.

Towards Mechanistic Understanding of Peptide Mediated Plasmonic Nanoparticle Synthesis

Kacper Lachowski

Plasmonic nanoparticles continue to be explored for applications ranging from surface enhanced raman spectroscopy and photothermal therapies, to photocatalysis and OLED devices. Control over morphology, composition, and arrangement of nanoparticles is key to enabling these applications, but there is still much to be learned about the mechanisms governing synthetic approaches. Biomimetic syntheses have shortcomings in mechanistic details relative to the potential degree of control over structure they afford, as well as the associated practical advantages such as mild reaction conditions. Therefore, we focused on screening the structures formed in the presence of three gold binding peptide sequences with and without a myristoyl tail by using UV-VIS, SAXS, and TEM. By screening a large parameter space of precursor, peptide, and reducing agent, we capture the complexity of the reaction space, and draw out details that foreshadow future experiments of peptide mediated nanoparticle synthesis.

A data-driven algorithm for de novo protein sequence design

David Juergens

Machine learning has driven major scientific advances in many fields through the ability to accurately and efficiently model complex phenomena. These algorithms continually prove to be great supplements or replacements to traditional approaches in molecular modeling and engineering. Here I will discuss my efforts to build one such algorithm for the de novo design of amino acids onto protein backbones of unknown or non-optimal sequences. The algorithm comes in two parts: (1) A deep 3D CNN classifier that predicts the probability distribution of amino acids that can be placed at a position in a protein, conditioned on the local atomic environment of the position. (2) A stochastic sampling and mutation loop to iteratively optimize a sequence onto a backbone from any initial state. The algorithm shows logical chemical intuition, and significant native sequence recovery when designing sequences from scratch.

Establishing multiple particle tracking as a viable tool for characterizing changes in the brain extracellular environment throughout development and in response to injury

Mike McKenna

The brain extracellular matrix (ECM) plays critical roles throughout neural development and in response to injury. Unfortunately, the majority of work aimed at characterizing brain ECM relies on static immunofluorescent imaging, which fails to provide information on local physical properties that alter neuronal function. We present the methodology we have established to overcome this limitation: multiple particle tracking (MPT). By tracking the diffusion of individual nanoparticles in the brain extracellular space, we can extract relevant physical and mechanical properties such as ECM mesh size and local viscosity and unveil how specifically brain ECM changes in response to injury.

Detecting Solid Oxide Fuel Cell Stack Failure from Impedance Spectroscopy Using Machine Learning

Zang Le

Solid oxide fuel cells (SOFC) offer great potential for efficient use of renewably-generated storage fuels. However, commercialization of SOFCs is currently limited by a lack of long-term durability. This work explores the potential of using electrochemical impedance spectroscopy (EIS) as in-situ diagnostic of SOFC stack and system health based on machine learning. A multi-physics model of SOFC, operating with humidified hydrogen, is used to simulate impedance data at various nominal operating conditions. The model is also used to simulate impedance responses under different critical failure modes, such as uneven fuel distribution that leads to high local fuel utilization, electrode delamination or deactivation. Based on impedance simulations, classification models are synthesized to recognize patterns and distinguish different degradation mechanisms. Based on these models, we believe that impedance measurements can provide better indications of SOFCs system health compared to polarization curves.

Machine Learning Quantum Reaction Rate Constants

Evan Komp

The ab initio calculation of exact quantum reaction rate constants comes at a high cost due to the required dynamics of reactants on multidimensional potential energy surfaces. In turn, this impedes the rapid design of the kinetics for large sets of coupled reactions. In an effort to overcome this hurdle, a deep neural network (DNN) was trained to predict the logarithm of quantum reaction rate constants multiplied by their reactant partition function – rate products. The training dataset was generated in-house and contains ~1.5 million quantum reaction rate constants for single, double, symmetric and asymmetric one-dimensional potentials computed over a broad range of reactant masses and temperatures. The DNN was able to predict the logarithm of the rate product with a relative error of 1.1%. Further, when comparing the difference between the DNN prediction and classical transition state theory at temperatures below 300K a relative percent error of 31% was found with respect to the exact difference. Systems beyond the test set were also studied, these included the H + H₂ reaction, the diffusion of hydrogen on Ni(100), the Menshutkin reaction of pyridine with CH₃Br in the gas phase, the reaction of formalcyanohydrin with HS^–  in water and the F + HCl reaction. For these reactions, the DNN predictions were accurate at high temperatures and in good agreement with the exact rates at lower temperatures. This work shows that one can take advantage of a DNN to gain insight on reactivity in the quantum regime.

Nanotechnology for therapeutic applications in the developing brain

Andrea Joseph

Drug delivery to the brain remains a major challenge in medicine despite decades of research and large financial investments. One promising solution is nanotechnology, which can overcome the brain’s uniquely restrictive biological barriers. In this poster, we describe the design and evaluation of nanoparticles for the treatment of developmental brain injury.

High-Throughput and Data-Driven Strategies for the Design of Deep Eutectic Solvent

Jaime Rodriguez

Within the framework of green chemistry, Deep Eutectic Solvents (DES) have been identified as promising candidates for use in many applications, including battery electrolytes. DES are characterized by two or three materials that associate with each other through hydrogen bond interactions, resulting in a eutectic mixture whose freezing point is below that of the individual materials. This design space is overwhelmingly large and poses a challenge for screening a vast and diverse set of materials. Here we present a strategic approach consisting of high throughput experimentation (HTE) coupled with data science driven analysis to identify exceptional DES candidates based on key physiochemical and electrochemical properties. Much of our HTE adopts methods that are already used frequently in the biotech and pharmaceutical industries, most notably performing parallel syntheses and analyses in 96-well-plate formats. First,a basis set of promising DES candidates was constructed from a large database of materials that have been screened according to certain engineering metrics.  The identified candidates are then synthesized using an open-sourced automated liquid handling robot. DES melting points are then determined by monitoring the melting process with an infrared camera and identifying the temperature at which the thermal conductivity of the samples changes abruptly. The solubility of battery redox-species is determined via UV-VIS well-spectrophotometers. Finally, the electrochemical stability window and cycling properties of DES electrolytes are measured in high-throughput by using screen-printed electrodes on 96-well plates adapted for use with a standard potentiostat. The ability to rapidly and efficiently collect data also creates a need for the development and use of automated processes for data analysis, which have been developed in an open-sourced format by our group. This approach to HTE also allows for the incorporation of data science techniques, such as feature extraction and machine learning, that further aid in probing a design space that is ultimately too large for experimental methods alone.

Agent-based models explore multi-scale dynamics of hypoxia-induced tumor angiogenesis

Jason Cain

Hypoxia, a hallmark of most solid tumors, results from heightened nutrient competition and contributes to drug resistance, blood vessel formation, and metastasis during tumor development. For this reason, hypoxia is instrumental in the design and execution of therapeutic interventions targeting the tumor microenvironment. Unfortunately, the tumor microenvironment is difficult, if not impossible, to observe and experimentally modulate across time and space. Agent-based models are an intuitive modeling framework that can inform the emergent dynamics of this niche. I am developing comprehensive agent-based models to characterize key features of the hypoxic tumor microenvironment and identify clinically tractable targets through inclusion and analysis of in silico hypoxia-induced angiogenesis.

Protein-Based Coiled-Coil XTEN Hydrogels as a Cell Carrier for Injectable Therapies

Mary O’Kelly Boit & Jenny Bennett:

While the global burden of heart disease and the heart’s inability to regenerate is well-known, it has been recently found that stem-cell-derived-cardiomyocytes, or the beating cells of the heart, significantly improve cardiac function better than any other therapy developed to date. Although promising, only 10% of injected cardiomyocytes persist long-term. Thus, this research will elucidate the potential for an injectable protein-based hydrogel system, programmable at the amino acid level, to improve cell survival and retention of current stem cell therapies for the heart. The choice of using a protein-based injectable biomaterial, specifically our coiled-coil XTEN platform, was chosen due to its simple design and flexible nature that can be delivered in a minimally invasive method (through a catheter or needle). Coiled-coil XTEN is a protein composed of a flexible unstructured linker (XTEN) flanked by coil domains (based on rat cartilage oligomeric matrix protein). Its material properties can be optimized by altering the length of XTEN resulting in varied degradation and mutating single amino acids in the coil domains to change self-healing behavior. 

Comprehensive Brain Cell Morphology Pipeline for Feature Quantification in Fluorescent Images

Hawley Helmbrecht

Brain cell morphology is altered by the state of the brain microenvironment in response to many factors including disease, age, and region. We developed a data science pipeline for comprehensive cell shape quantification and analysis. The pipeline has been applied to quantify cell feature changes during oxygen-glucose deprivation, and it is currently being expanded to quantify features of specific brain cell types.

Leveraging High Throughput Organotypic Brain Slice Culturing to Time Lapse Image Cellular Interactions

Jeremy Filteau

Organotypic brain slices are an ex vivo platform that retain 3D architecture and native cell populations. Current organotypic brain cell culture technique is limited in terms of throughput, causing studies with high animal numbers to be difficult to process. We have created a fully customizable, 3D-printed device that facilitates the culture of 9 brain slices at a time with live imaging compatibility. This will allow the imaging of cellular interactions, which are paramount to understanding diseases of the central nervous system, over the course of multiple days with high throughput and experimental efficiency.

A COVID-19 Nucleic Acid Amplification Self-Test for use in the Home

Coleman Martin

In the absence of a vaccine, rapid home based testing may be a valuable tool key to reopening society and reduce impacts of the SARS-COV2 (COVID-19) pandemic. The current gold standard for diagnostic testing, reverse transcription, quantitative Polymerase Chain Reaction (rt-qPCR), is highly specific and sensitive; however, it requires specialized equipment and personnel resulting in delays from collection, diagnosis, and reporting.  Point-of-care nucleic acid tests have received emergency use authorization; however, these platforms still require significant investment in instrumentation and not simple enough to be used in the home.  We are developing a COvid19 nucleic acid Amplification Self-Test (COAST) that is simple enough to be operated at home by untrained users and is disposable. COAST uses Recombinase Polymerase Amplification isothermal amplification that can detect low copy number sensitivity results in 30 minutes.