Abstracts from the College of Engineering, Physics, Computer Science, and Mathematics
Undergraduate Summer Research Symposium August 8, 2007

Ordered alphabetically by student's last name

Bah
 Cherian Gangloff Knieriem Myrick Paul Schmiedel Stirparo Van Wie
Bosco
 Fleagle Gao Lort Pagels
 Petroff Selekman Sweriduk Williams
Boyle Fisher Hoffmann Miller Parkhurst Puzio Stern Ulissi Woods


Noise Challenges for Achieving Pico-Tesla Sensitivity in Tunneling Magnetoresistive Sensors
Mohamed Bah and Edmund Nowak
Department of Physics and Astronomy

Magnetic tunnel junctions (MTJs) are thin film solid structures with magneto-electronic properties that are favorable for advancing low-power, very high resolution magnetic field sensors.  Room temperature tunneling magnetoresistance (TMR) of nearly 400% is one of the main characteristics that make MTJs capable of achieving Pico-Tesla resolution.  To optimize signal-to-noise ratio, high response must be coupled with low intrinsic sensor noise.  We report low frequency noise measurements in MTJ devices configured in a Wheatstone bridge where two legs of the bridge are shielded from small magnetic fields.  The resulting bridge imbalance voltage is linear for fields ranging from -4 to +4 G.  Mechanisms that contribute to intrinsic sensor noise include: thermal or Johnson noise, shot noise due to the discreteness of the electron charge, thermal magnetic noise, and resistance fluctuations.  The latter typically exhibit a 1/f power spectrum, where f is frequency, but occasionally discrete switching between resistance states (telegraph noise) is also observed.  Measurements as a function of field reveal that the 1/f noise has both electronic and magnetic origins that will limit the ultimate resolution of sensors below 1 kHz.  The electronic component does not depend on field and is associated with charge traps in the tunnel barrier.  The magnetic component can be large and sometimes is found to scale with the slope of the resistive response.  The origin of this noise is discussed.  A noise equation is presented that takes into account the various noise sources and can be used to examine the tradeoffs in designing sensors to achieve Pico-Tesla sensitivity.

Characterization of Supported PtNi and PtCo Bimetallic Catalysts
using FT-IR Gas-Phase Reactions Studies and CO Chemisorption
Jeff Bosco and Jingguang Chen
Center for Catalytic Science and Technology (CCST), Department of Chemical Engineering

Through extensive DFT modeling and single-crystal surface studies under ultra-high vacuum conditions, the Chen research group has demonstrated that platinum-nickel (PtNi) and platinum-cobalt (PtCo) bimetallic surfaces show enhanced activity over single metal surfaces (Pt, Ni, Co) towards the hydrogenation of both carbon-carbon and carbon-oxygen double bonds.  The purpose of our current research is to employ Fourier-transform Infrared (FT-IR) spectroscopy as a method of characterizing PtNi and PtCo bimetallic nanoparticles supported on γ-alumina.  Methods of characterization include catalytic gas-phase reaction studies of the hydrogenation of cyclic-alkenes, aromatics, and simple un-saturated aldehydes as well as surface CO chemisorption studies.  Areas of catalyst characterization include the verification of the presents of truly bimetallic particles, overall hydrogenation activity evaluated by kinetic rate parameter fitting, and the effect of impregnation sequencing during bimetallic catalyst preparation. Funding provided by the Department of Energy.


Creating Patterned Surfaces Using Particle Spreading at Air/Water and Oil/Water Interfaces
Michael Boyle and Eric Furst
Department of Chemical Engineering

Repulsive forces of colloidal particles at the 2D interface allows for them to self assemble into patterns along the boundary between two distinct phases.  These patterns are basically a hexagonal close packed system of the colloidal particles along the oil-water interface.  These patterned arrays can be imaged using scanning electron microscopy or simply a camera attached to an upright microscope.  Once the pattern of the colloidal particles is generated, the subphase, which is water, has to be gelled to fix the colloidal particles in place.  Next, with the subphase gelled, an epoxy resin is substituted in for the oil or air, and the surface of the epoxy resin has the colloidal particle pattern when it is extracted.  These colloidal particles can then be heated and subject to a magnetic field.  This magnetic field will orient the colloidal particles so that each of them has a magnetic dipole moment.  Once the particles are magnetized, they can be used in microfluidic channels to separate other magnetic particles in a fluid flow.


Soluble Aggregates in Nonnative Aggregation of aCgn and a-La
Deepthi Cherian, William Weiss, Christopher Roberts
Department of Chemical Engineering

Heat induced aggregation of two proteins, α-chymotrypsinogen (aCgn) and α-lactalbumin (a-LA), was conducted to in an attempt to create soluble aggregates using variations in pH and salt content, for subsequent biophysical characterization to determine their structure and morphology. Resulting precipitates were also treated with acid in attempt to recover soluble aggregates. The concentration of these aggregates were determined by High Performance Liquid Chromatography. While aCgn has been previously shown to create soluble aggregates when heated at a pH of 3.5, soluble aggregates were also recovered by the addition of acid to the protein after precipitation.  Similarly, soluble aggregates were present in the re-acidified solutions of a-LA. However, the concentration of these aggregates with respect to monomer is presently too small to allow for further characterization under different assays. Supported by the HHMI Undergraduate Science Education Program.


Predicting the Lifetime Stability of Photovoltaic Modules Using Accelerated Environmental Testing
Grant L. Fisher and Steven S. Hegedus
Department of Mechanical Engineering, Department of Physics and Astronomy 

Solar energy through photovoltaics is the only renewable energy source that could feasibly supply the world’s electrical needs. Currently, photovoltaic energy represents less than 0.1% of the global power usage because of its high cost relative to oil, gas, and coal generated electricity.   Recent developments have made it possible to mass produce solar cells using cheaper materials, but in order for these cells to be economically viable, their efficiencies must not sufficiently degrade over a period of decades.  Some of these new technologies have had no field experience. The only realistic way of evaluating a product’s performance over such a long period of time is to conduct accelerated environmental tests (AETs) over a much shorter time interval.  These tests must be carefully designed because the complexity of the photovoltaic systems allows for many possible failure mechanisms. The photovoltaic device and the encapsulant that protects it can fail very differently depending on temperature, humidity, and electrical bias.  Another problem is accurately scaling the failure data from the AETs to predict lifetime stability.  We will be performing AETs on emerging solar cell technologies to determine their stability over long time periods. Funding for this project is provided by the Science and Engineering Scholars program.

Aggregation of a Model β-helical Protein in Solutions of Varying pH and Ionic Strength
Carly Fleagle, Michelle Spatara, and Anne Skaja Robinson
Department of Chemical Engineering

Protein aggregation occurs between monomeric proteins that often consist largely of non-native secondary structures that associate to form oligomers.  In the biopharmaceutical industry, finding conditions that minimize aggregation during production and storage is critical.  Protein aggregation has also been implicated in over 20 human diseases including Huntington’s, Alzheimer’s, and Parkinson’s diseases. In this work, a truncation of the P22 tailspike protein comprising the β-helix domain is being used as a model system to study aggregation.  This truncation, termed bhx, was expressed recombinantly in E.coli and purified using a series of chromatography steps.  Purified protein was then exposed to elevated temperature (40°C) to induce aggregation under different solution conditions.  The rate of monomer loss of the bhx in various pH conditions and salt concentrations was analyzed using native PAGE.  We found that changes in ionic strength had no effect on the rate of monomer loss. In contrast, altering the pH of the buffer drastically affected the aggregation propensity of bhx.  Funding provided by HHMI (CF) and Merck (MS)


Design of a Wearable Upper Extremity Orthotic Exoskeleton
John J. Gangloff Jr.1, Venky Dubey2, Sunil K. Agrawal1
1Department of Mechanical Engineering, University of Delaware, Newark, DE
2The School of Design, Engineering and Computing, Bournemouth University, UK

An upper extremity orthotic exoskeleton has been designed in order to aid in the rehabilitation of stroke patients’ arm movements. The exoskeleton will assist patients in moving their gleno-humeral and elbow joints in an adaptive manner to relearn independent usage. The exoskeleton is designed to be lightweight and wearable, while providing an improved rehabilitation experience. In order to establish the underlying concepts, a plastic human skeleton model has been used to test the exoskeleton before clinical trials can be performed on patients. Research in outfitting the skeleton with freely movable gleno-humeral and elbow joints is in progress so to reflect the actual conditions of patients in reference to realistic joint movements. The skeleton is outfitted with digital encoders and a force-torque sensor to provide necessary experimental data, so that range of motion is achievable and external forces to the patients are minimized. In order to further validate the design idea and control options, a three-dimensional CAD model of the skeleton with the exoskeleton has been created. The underlying mathematics for the kinematics and dynamics of the exoskeleton system is being developed for optimized design performance. The CAD model, in conjunction with the experimental exoskeleton outfitted with sensors, is used to optimize the exoskeleton performance. Work in building and testing the exoskeleton with the skeleton model is in progress with clinical trials to follow.


Ammonia Decomposition Catalysis for Hydrogen Generation for Fuel Cell Technology
Belinda Gao, Elizabeth M. D’Addio, and Jochen A. Lauterbach
Department of Chemical Engineering

The decomposition of ammonia to produce hydrogen for fuel cell technology, while a beneficial alternative to other current methods of hydrogen production, must be sufficiently catalyzed without much sacrifice in conversion or reaction conditions. Aided by high-throughput experimentation, the Lauterbach group is now working on developing and characterizing catalysts that effectively convert ammonia to hydrogen at lower temperatures. The catalytic activity depends on many factors, such as support materials, precursors, dispersion, and promoters; this particular study has focused on some of the properties of supports and precursors. Past experiments have determined Ru to have the highest catalytic activity, but K-promoted catalysts have shown significantly more activity than their single metal counterparts; experimentation was done to determine if this could be attributed to the leeching of leftover chlorine (a known catalytic inhibitor). In addition to RuCl3, two other ruthenium precursors were examined (Ru(C5H7O2)3 and KRuO4). For supports, carbon nanotubes (CNTs) have shown promising results as support materials from other researchers due to their high surface area and good electron conductivity; since conducting supports had not yet been tested, we were interested in comparing a CNT-supported catalyst with ones previously studied. Used as is, however, the use of CNTs resulted in poor conversion in comparison to the support usually used (γ-Al2O3). Examination from the SEM showed degradation of the catalyst after reaction. Future work will be done to explore pretreatment options for the nanotubes to determine if the CNTs can be thermally stabilized and the conversions improved. Funded by the University of Delaware’s Undergraduate Research Program and the Department of Energy.


Mathematical Models for the Simulation of Gene Expression
Kyle Hoffmann, Tobin Driscoll2, Prasad Dhurjati1
1Department of Chemical Engineering2, Department of Mathematical Sciences

The ability to simulate networks of genes will allow researchers to better study existing networks and to create more efficient experiments. The existing C based simulation ExPatGen allows a user to create a gene network via an on-line interface and to view the results as they would appear in a microarray.  This program was repaired and placed on-line.  Its code was then translated into the Matlab programming language.  An interface was created for inputting gene networks and for displaying the data either as it would appear in a series of microarrays, or the mRNA concentration as a function of time.  This program was then used to simulate several simple gene networks.  Currently, there is a substantial set of microarray data for various organisms under different contexts.  Methodologies for interpretation of this information can lead to new diagnostic techniques and also aid in discovery of new drugs.  Some microarray data from simple simulated gene networks were interpreted to address the inverse problem of going from data to gene networks.  This provided insight into the enormous complexity of the problem.  This work was funded by HHMI. 


The Structure of Social Networks and Its Role in Modeling the Spread of Infectious Disease
Donald Knieriem and Richard Braun
Department of Mathematics

The modeling of the spread of infectious disease in humans relies both on the properties of the disease in question and the interpersonal contact within the host population. This study focuses on the importance of the structure of the social network in which a disease spreads. Older models assume that the population is homogenous, and thus results depend only on the relative virulence of the disease. A novel model is used, which incorporates a social network. My results show that the behavior of diseases is as much dependent on the social structure of the population. Various social network types were used in simulating infectious disease outbreaks, and results confirm that social networks with different properties  exhibit different behavior for the same disease. Because the research is based on a new model for the spread of disease, the properties of the model itself were fully explored as well. Supported by a Howard Hughes Medical Institute’s Undergraduate Science Education Award.

 
Use of Resonant Systems to Achieve Hovering Flight in Flapping Wing Micro Aerial Vehicles
Richie Lort,  Sunil Agrawal, Zaeem Khan
Department of Mechanical Engineering

The goal of producing a flapping wing MAV presents many challenges.  In order to achieve sufficient lift to produce hovering flight, it is necessary to find a way to transfer power very efficiently from the motor to the wings.  This allows a very small lightweight motor and gear box to be utilized, which is essential if you wish to create a prototype light enough to achieve flight.  In order to do this we have developed the idea of combining a four bar mechanism with a spring system in order to maximize the power transferred to the wings.  The four bar’s rocker link is connected to the base of the wing via a passive spring connection.  By matching the frequency of the flapping to the natural resonant frequency of the spring, it is possible to greatly increase the amplitude of the flapping motion.  The lift generated is increased proportional to the flapping amplitude.  In order to accomplish this we must find a way to create an extremely light weight, low friction system capable of withstanding the dynamic forces generated by high frequency flapping.   Thanks to the National Science Foundation for sponsoring this project.


The Effects of Surfactant Mixtures and Phase Behavior on Membrane Protein Stabilization
Dan Miller, Kelley Kearns, Abraham M. Lenhoff, and Eric W. Kaler
Chemical Engineering Department

<>Membrane proteins represent a pivotal area in the field of pharmaceutical research, as it is estimated that over half of all pharmaceuticals target membrane proteins. Despite the importance of membrane proteins, they account for fewer than one percent of the known structures in the Protein Data Bank. Membrane proteins are relatively hydrophobic, and are not water soluble. This characteristic adds a new layer of complexity to protein studies, because surfactants are needed to extract the proteins from their native membrane, and provide the means for protein solubilization by the formation of a protein detergent complex (PDC).  There are several important considerations to take into account when surfactants are employed, especially that many surfactants have very interesting thermodynamic phase behavior, including the possibility of a liquid-liquid phase separation. Therefore, choosing an appropriate surfactant that preserves the protein in its stable native form is an essential step in any membrane protein study. Mixtures of pairs of surfactants in different concentrations were studied using a wide range of techniques, with the main goal of discovering the factors that contribute to optimum membrane stability of the membrane protein diacylglycerol kinase (DGK).  Different mixtures of n-decyl-β-D-maltopyranoside (DM) and n-decyl-β-D-glucopyranoside (DG) were used to determine the effects of phase behavior on stability, using an activity assay.  Circular dichroism (CD) was employed in combination with an activity assay to correlate changes in secondary structure with a loss of activity in mixtures of DM and n-hexyl--D-maltopyranoside (HM). Isothermal titration calorimetry (ITC) was used to measure the critical micelle concentration (CMC) of the HM-DM mixtures to assure that micelles were present to solubilize the protein. Surprisingly, complex phase behavior had little effect on membrane protein stability, and secondary structure was highly maintained in inactivated protein samples. However, there was considerable activity loss in HM-DM mixtures with high concentrations of HM. This research was funded by HHMI.

Investigating the Utilization of Matrix Metalloproteinase-1 (MMP-1) as a Tool
for Responsive Nanoparticle Surface Modification During DNA Delivery
Stephanie L. Myrick1, Peter G. Millili2, Millicent O. Sullivan2
Department of Chemistry and Biochemistry1, Department of Chemical Engineering2,
 
The design of efficient DNA delivery systems is currently a popular area of research due to its potential for therapeutic benefit.  We are interested in formulating non-viral DNA delivery vehicles containing multiple functional layers that can be sequentially cleaved off during the delivery process to expose new features.  Layers will be linked to the vehicles via peptides sensitive to proteases such as MMP-1, a collagen-degrading enzyme upregulated by fibroblasts as they migrate through the extracellular matrix.  In vivo, fibroblast migration is typically observed in tumor stroma and at sites of injury, making MMP-1 secretion a useful signal for targeting these areas.  To validate our targeting system, we have investigated the levels of MMP-1 expression by two model fibroblast cell lines.  Human and mouse fibroblasts were cultured on collagen-coated substrates, and migration was stimulated by scrape wounding and by treatment with Tumor Necrosis Factor-alpha (TNF-α).  Immunofluorescence staining for F-actin, nuclei, and MMP-1 was performed on these migrating cells.  Western blots of the conditioned media and cell lysates of these cells were also analyzed for MMP-1 expression.  Preliminary results with these techniques validate the efficacy of this migration model to upregulate MMP-1.  Having established the conditions that stimulate MMP-1 secretion, future work will focus on a fluorescence resonance energy transfer (FRET) system for demonstrating MMP-1-mediated peptide cleavage.  Once validation of the responsive nature of this peptide is achieved, it can be incorporated into the described hierarchical DNA delivery vehicle.  This project is supported by HHMI and NSF grant 0707583.


Effects of sorbitol on folding/unfolding and aggregation of a-chymotrypsinogen
Rebecca K. Pagels, Rebecca K. Brummitt, and  Christopher J. Roberts
Department of Chemical Engineering

Irreversible, non-native aggregation is a constant concern for the biopharmaceutical industry throughout the production process and storage as it can result in loss of viable product. a-Chymotrypsinogen (aCgn) is a well-studied model protein for non-native aggregation under acidic solution conditions where the aggregates remain soluble. The focus of this study is to assess the impact of a canonical stabilizing additive (sorbitol) on different stages of aggregation. The primary methods of experimentation were differential scanning calorimetry, size exclusion high performance liquid chromatography, circular dichroism, and fluorescence spectroscopy. These techniches helped determine the change in Gibbs free energy of unfolding (DGunf ), monomer loss kinetics, secondary structure, and tertiary structure, respectively, of the protein with increasing sorbitol concentration at fixed temperature. Changes in DGunf  with increasing sorbitol concentrations were qualitatively consistent with observations in the literature in that addition of sorbitol (slightly) stabilized the native monomer state over the unfolded state. Addition of sorbitol caused only minor changes to the rate of aggregation, with the exception of a notable increase in aggregation rate at intermediate sorbitol concentration. This anomaly cannot be explained by changes in the thermodynamics of unfolding, and the data is currently inconclusive concerning changes in the secondary and tertiary structure of the unfolded monomer. Future work will involve light scattering and seeding experiments in an attempt to separate the intrinsic aggregate nucleation and growth timescales. Support from the University of Delaware Howard Hughes Medical Institute Scholars Program (RKP) and from the National Science Foundation (RKB) is gratefully acknowledged.


Biodegradation and Feasibility of Biocomposite Panels for Roofing Applications
Matthew Parkhurst, Tim Strickland, Harry Shenton, and Daniel Cha
Department of Civil and Environmental Engineering 

We are living in age where people are attempting to find alternative ways to live other than depending on petroleum-based products. Infrastructure consumes the majority of petroleum. We have fabricated a biocomposite material made from all natural ingredients, including soybean oil resin and recycled paper. This biocomposite is seen as an alternative to the present day roof. Previous tests included basic strength and creep tests. Tests are currently being conducted to investigate two other properties of the material. One is to understand if this biocomposite can undergo biodegradation in both soil and water. Another we are testing is the feasibility of the real world usage of this material. This means if the biocomposite be exposed to various elements such as heat, moisture and sunlight, what will change and what will remain the same. The main thing we want to see is if the composite loses or retains its original strength when exposed to various elements, such as temperature, humidity, and sunlight. Tests are also being conducted on wood and asphalt shingles to see how the biocomposite compares to the behavior of these known materials. If the biocomposite compares similar or better than the known materials, the biocomposite is a feasible alternative. Full results are not available at this time. This project is funded by the Walter L. and David P. Hernson Scholarship Endowment.


Wireless Signal Propagation
Nikhil Paul, Steven Bohacek, and Hweechul Shin
Department of Electrical/Computer Engineering

Wireless communication is an integral part of everyday life. In order to understand the performance of any wireless protocol, one needs to develop models of wireless signal propagation. Specifically propagation has a significant impact on the performance of protocols due to the variation induced by it in terms of link data rate and error rate. The objective of propagation simulation is to model the channel gain (and other channel characteristics). Two models that we considered are interference in propagation and spatial variation in propagation, and how they both affect packet transmission probability. To measure the variation in propagation, we used a robotic linear actuator that changed its position and measured the spatial variation and propagation in both, inside and outside environments. For the measurements of the interference, we used a novel in vitro lab setup. Specifically, in order to control every aspect of the environment, “wireless” transmissions were restricted to wires, splitters and attenuators. The research is funded by National Science Foundation.



The Effects of Secondary Polymers Protein Adsorption on Cation Exchange Media
Matthew Petroff and Abraham M. Lenhoff
Department of Chemical Engineering

The addition of a secondary polymer to the base matrix of an ion-exchange stationary phase can greatly improve its chromatographic performance.  This work explores the performance of two such stationary phases, Toyopearl GigaCap S-650 M (Tosoh Bioscience) and Bakerbond XWP 500 PolyCSX-35 (Mallinckrodt), through determination of the equilibrium adsorption isotherms for the model protein lysozyme.  Isotherms were generated through batch adsorption of protein with known amounts of particles and data were fitted both to a colloidal isotherm model and the Langmuir model.  With the phase ratios of these particles yet to be determined, the parameters generated by the fits are only good for qualitative comparison.  The data demonstrate that the GigaCap particles have a higher static capacity and stronger particle-protein interactions than the PolyCSX particles. This work was made possible by the National Science Foundation EPSCoR Grant No. EPS-0447610.





Macroscopic Alignment of Electrospun Gelatin and Collagen Nanofibers via the Hall Effect
for use in Tissue Engineering and Environmental Monitoring
Glenn Puzio§, Kristin Sisson†, Young Shin Kim†, D. Bruce Chase‡, John F. Rabolt*,†
§Department of Chemical Engineering, †Department of Materials Science and Engineering, and ‡DuPont CR&D, Wilmington, Delaware 19880

The research presented here explores electrospun, highly macroscopically aligned nonwoven mats of gelatin and collagen nano-scale fibers.  The electrospinning process provides micro-scale pores and high surface to volume ratios, yielding mechanical integrity for cell migration and tissue growth.  Prior research concluded that macroscopic alignment occurred when positively charged PEO (polyethylene oxide) was electrospun onto two negatively charged aluminum plates separated by a finger’s width gap.  Use of the Hall Effect, which disperses the negative charges to the sharp edges of these plates, allowed for the alignment of the positively charged nanofibers.  Using this technique, for the first time gelatin and collagen nanofibers were also highly macroscopically aligned.  Obtaining aligned nanofibers provides many opportunities including: fabrication of tissue engineered blood vessels, heart and muscle tissue, and other forms, e.g. filters for viruses, bacteria, etc., where natural alignment would provide anisotropic properties. Using gelatin and collagen provides a cell-friendly scaffold for cell growth and most importantly cell migration.  Determination of macroscopic alignment was accomplished through use of the University of Delaware’s SEM (scanning electron microscope).  Microscopic alignment was determined using their polarized FTIR (Fourier transform infrared) spectrophotometer, as well as both the University of Delaware and DuPont CR&D’s polarized FT-Raman spectrometer.  Funding was provided in part by NSF (National Science Foundation) and EPSCoR-DE (Experimental Program to Stimulate Competitive Research – Delaware).


Live Cell Imaging: Quantifying Expression of Fusion Proteins in Sacchromyces Cerevisiae
Lindsay Schmiedel, Carissa Young, David Raden, Anne Skaja Robinson
Department of Chemical Engineering

Improvements in confocal light microscopy combined with genetically encoded fluorescent tags have allowed continuous monitoring of protein dynamics in living cells. To model biological systems, it is necessary to monitor, analyze, and quantify proteins. This research specifically pertains to proteins BiP and Sec63 in the Endoplasmic Reticulum (ER) of yeast, Sacchromyces Cerevisiae. The objectives of my research are to i) molecular engineer a plasmid that incorporates a retention sequence to properly localize the proteins of interest, ii) create specific fusion proteins by the insertion of fluorescent markers, Venus and enhanced Green Fluorescent Protein (eGFP) at the C-terminus of the proteins of interest, iii) determine the photostability of fluorescent markers, enhanced Yellow Fluorescent Protein (eYFP) and Cerulean, and iv) quantify the extent of photobleaching and recovery of fusion proteins in the ER. Experimental techniques completed are polymerase chain reaction, restriction digest, ligation, E. coli transformation, and homologous recombination including yeast transformation. Confocal light microscopy and time series analysis during photobleaching have been used to obtain live cell imaging in vivo and examine the physiological effects of fluorescent variants in Sacchromyces Cerevisiae. This project was funded by the Science and Engineering Scholars Program.


Utilization of a Poly(ethylene glycol)-Amine Condensing Agent for DNA Complexation
Josh A. Selekman, Peter G. Millili, Millicent O. Sullivan
Department of Chemical Engineering

Current synthetic DNA delivery systems utilized in gene therapeutic applications are plagued by cytotoxicity issues as well as poor delivery efficiencies.  To combat this, we are investigating the utilization of an α-Amino-ω-hydroxy Terminated Poly(ethylene glycol)  (PEG) polymer as a dual condensing and stabilizing agent.  Our hypothesis is that this PEGylated small cation will allow DNA complexation (via its amine) while simultaneously preventing DNA complex aggregation (via its PEG).  pH, buffer, time, and charge ratio (PEG amines:DNA phosphates) were systematically explored as formulation design parameters.  First, agarose gel electrophoresis was used to analyze numerous charge ratios, revealing that minimal DNA complexation was occurring via this condensation chemistry.  Dynamic light scattering experiments corroborated this finding.  Conversely, zeta potential experiments showed evidence of interaction between the PEG and the DNA, with the surface potential increasing from -27.49 ± 0.98 mV with no functionalized PEG to -19.68 ± 0.79 mV with the addition of PEG.  These results suggest that PEG-DNA interactions are occurring, however the condensation chemistry employed is insufficient to induce the formation of complexes that can scatter light.  Current and future work is focused on identifying strategies to enhance DNA compaction with this PEG-amine-based condensation approach.  A small molecule polycation, spermidine trihydrochloride, has been added to the formulation, resulting in particle sizes of 116 ± 2 nm, regardless of the amount of PEG used.  Future work will focus on a PEG functionalized with additional amines, which is expected to enhance interactions with DNA and promote particle formation. This research was funded by the Howard Hughes Medical Institute Undergraduate Research Program.


An experimental and mathematical study of M. oryzae spore germination and dispersal
in the presence of host and non-host volatiles
Kyle Stern and John A. Pelesko
Department of Mathematical Sciences

Each year, the fungus M. oryzae destroys enough of the world’s rice, barley, and wheat crops to feed more than sixty million people.  In this project we investigate whether or not there are volatiles in host plants that cause M. oryzae spores to react.  If true, these volatiles may cause the fungus to spread rapidly.  The first part of this project focuses on germ tube growth.  Spores and volatiles were strategically placed near each other in order to determine the angles the spores’ germ tubes made with the volatile.  The angles were measured using computer software and the data collated in a rose plot, which revealed the distribution of angles. The second part of this project focuses on spore dispersal in a controlled setting.  After placing either volatiles or actual leaves in a dish with the fungus and allowing ten days for the spores to be released from their stalks, the distance the spores traveled was measured using computer software.  Initial results from the germ tube experiment reveal that the spores tend to germinate in random orientations.  Data from our second experiment suggests that spores utilize an active dispersal process and that host volatiles may change the vigor with which spores disperse.  The results show that limonene, a volatile of the rice plant, is one such volatile that triggers vigorous active dispersal. Funded by Howard Hughes Medical Institute.


Bio-based Composite Resins
Gregory M. Stirparo, Richard P. Wool, and Alejandrina Campanella
Department of Chemical Engineering

Alternatives are being sought to the traditional petroleum-based resins.  This is in response to the recent increase in petroleum prices, as well as the need for a more environmentally friendly product.  The traditional resins also commonly include styrene, which is something we hope to replace.  The petroleum based components of the resins will be replaced by Acrylated Epoxidized Soybean Oil (AESO) and the styrene will be replaced by Methacrylated Fatty Acids (MFAs), in particular methacrylated lauric acid.  The advantages of using the AESO and MFAs are they can be acquired at low cost, and are less harmful to the environment.  The goal is to use the AESO and MFAs to replace the petroleum-based components and the styrene while maintaining mechanical properties (Loss and Storage Modulus) on par with the existing resins, in addition to comparable viscosities.  To examine the mechanical properties the sample resins were cured, polished and cut the examined using the Dynamic Materials Analyzer (DMA).  The DMA provides a graph of the loss and storage modulus and aids us in determining the glass-transition temperature.  What has been found is that the petroleum-based components and styrene cannot be completely removed without compromising the mechanical properties, or the viscosity.  Rather a combination of all four components is likely to be needed to achieve the desired results.  Future work will be done to find the mixture containing the least amount of the unwanted materials with comparable mechanical properties.  Support Provided by the Science and Engineering Scholars Program.


Induction Heat Phase Transformation in the L10 FePt Magnetic Nanoparticle System
for FCT-ordered Self-assembled Arrays
Andrew J. Sweriduk, Michael J. Bonder, and George C. Hadjipanayis
Department of Physics and Astronomy

Iron-platinum nanoparticles hold great potential in the field of magnetic data storage for moving past current data storage density limits.  When heated, FePt nanoparticles with an atomic ratio of approximately 1:1 undergo a phase change from face-centered cubic (FCC) to face-centered tetragonal (FCT), which has a high anisotropy and could subsequently serve as a stable recording medium.  In this study, FePt nanoparticles were synthesized via the reduction of iron chloride and platinum chloride by sodium borohydride in a Y-tube junction, with the final sample suspended in oleic acid.  To attempt to achieve the desired phase transformation, the samples were subjected to radiofrequency (RF) induction heating using various coil configurations and over a range of power inputs and durations.   Both the as-made and heated samples were studied using a transmission electron microscope (TEM) to obtain images and determine their atomic composition.  Data on the crystalline diffraction of the samples (by which transformation could potentially be observed) was obtained via TEM and X-ray diffraction (XRD).  In addition, the magnetic response of the samples was measured via vibrating sample (VSM) and SQUID magnetometers.  Synthesis of FePt nanoparticles with an average size of 5 nm diameter and an atomic composition of 55% Pt to 45% Fe was achieved.  X-ray diffraction data showed some evidence for transformation to the FCT phase after RF induction.  Samples heated conventionally showed an increased coercivity when compared to the as-made samples, while samples subjected to RF induction showed no appreciable coercivity.  Further studies are necessary to explain this discrepancy. This work is funded by the National Science Foundation.


Refinement of Catalytic Microkinetic Models Using Singular Methods Coupled with the Design of New Experiments
Zachary Ulissi, Vinay Prasad, Dionisios Vlachos
Department of Chemical Engineering

Identifying useful kinetic models for catalytic systems is a long standing problem which is often done qualitatively by fitting arbitrary parameters to simplified models.  Modeling catalyzed chemical reactions using large sets of elementary reactions allows for more accurate and physically meaningful simulations, but requires large amounts of kinetic information for each forward and backward reaction step.  An important example of this is ammonia decomposition on Ruthenium surfaces, an important reaction with a known reaction model but only loosely identified kinetic parameters.  In order to understand which reactions (and corresponding parameters) are most important, realistic transient reactor models were first used with reasonable guesses for each kinetic parameter.  Computational Singular Perturbation (CSP), a mathematical analysis technique  previously used outside the realm of catalysis, was then applied to extract which reactions and species (and indirectly kinetic parameters) were active throughout the reaction.  This information was then be used to identify new experiments that will yield the most useful information.  This cyclic process of revised simulations and new experiments can then be used to successively refine the necessary parameters.  The developed method is general in nature, and can be directly applied to other catalytic systems. This work was sponsored by the Northeastern Chemical Association.

Microburst Synthesis of cPEG Coated Magnetic Nanoparticles with Highly Uniform Particle Size Distributions
Elisabeth C. Van Wie, M. J. Bonder and George C. Hadjipanayis
Department of Physics and Astronomy

Magnetic nanoparticles are an attractive alternative in medicine today for cancer detection and drug delivery due to their ability to be functionalized, controllable size, composition, and magnetic properties.  In order for these particles to perform optimally while suspended in a medium, they must be aqueous and have a narrow particle size distribution.  In this project, magnetic Fe nanoparticles were prepared using a borohydride reduction of FeCl2 in a frequency modulated flow reactor.  Reagents are passed through a cross T junction and collected in a beaker.  Particle size distributions were determined as a function of frequency from bright field transmission electron microscope (TEM) images as a function of synthesis parameters.  The smallest particles (~8 nm) were produced at a frequency of 10 Hz.  Particle size was also varied by altering reagent concentrations.  The coercivity of the particles increases from 524 Oe to 733 Oe as a function of frequency. 



A Study of Word Usage in Java Program Identifiers
Meilani Williams, Emily Hill, Lori Pollock, and K. Vijay-Shanker
Department of Computer and Information Sciences

In today’s largest and complex software, a piece of code will need to be read and understood by many software developers. To communicate their thought processes in code, developers use meaningful identifier names.  Thus, identifiers capture what a developer intends to accomplish with a portion of code.  We have found that natural language clues in identifiers can improve automated software tools to increase developer program comprehension and facilitate software maintenance tasks.  In this poster, we present a study of word usage in Java program identifiers.  We present our case study methodology, examples of word usage,  observations from the study, and how these observations can improve automated software tools. This research study was sponsored by the Computer Research Association for Women Distributed Mentor Program (CRA-W DMP).

Synthesis and Characterization of Poly(e-caprolactone) and Polyethylene Glycol Copolymers
for Use in Vocal Fold Tissue Engineering
Meghan Woods1, Sarah Grieshaber2, Xinqiao Jia2
1Department of Chemistry and Biochemistry, 2Deparment of Materials Science and Engineering, University of Delaware
  and Department of Materials Science

The ability of vocal folds to withstand drastic changes in froscopy and gel permeation chromatography (GPC).  Next, the phase transitions of each polymer at various concentrations in water were analyzed from 20-70 ºC.  It was found that with a sufficient PCL/PEG ratio and molecular weight, the copolymers transition from a solution at room temperature to a gel near body temperature (39-50 ºC) and back to solution at higher temperatures (55-60 ºC).  These copolymers are being investigated as the building blocks for the synthesis of elastin mimetic hybrid polymers.  This research was funded by HHMI. 

Links: Summer 2007 Undergraduate Research Symposium, Symposium Abstracts from other Colleges and Departments,
Undergraduate Research Summer Enrichment Program,  Unversity of Delaware Undergraduate Research Program, Howard Hughes Undergraduate Program.
Created  23 July 2007. Last up dated 15 August 2007 by Hal White
Copyright 2007, University of Delaware