Summer Research Experience for Undergraduates (REU)

2004 Summer Abstracts

 

 

 

Alexander O'Day - Fabrication of a Discriminating Gas Sensor

 

Brian Knapp - Using Logical Effort to Improve Delay of Integrated Circuits

 

Carl Herickhoff - Growing Nickel Nanowires for Application in Nanosensors

 

Chisom Maduike - Visually Animating Gate Activity in a Processor

 

Christopher Barsi - Numerical Modeling of Self-Assembled Photonic Crystals

 

Elizabeth (Za) Freeman - Living Microorganisms Entrapped in Nano-Structured Latex Formulations Piezoelectrically Printed onto Bioelectronic Devices

 

James Foltz - PRGUI:  A Visual Tool for Displaying Placement and Routing

 

Jon Brackbill - Fixed Point N-Tone Sigma Delta Modulators for UWB Applications

 

Kathryn Kelchner - Fabrication and Characterization of Magnetic Nanowires

 

Luis Avila - TerminatorBot’s Locomotion Interface

 

Margo Smith - Controlled Neuron Growth By Patterned Protein Gradients

 

Mikhail Lisovich - Fixed Point Simulation of Error Correcting Algorithms

 

Shamekia Dixon - Language Extensions

 

Shane Wood - nanotechnology, materials science, and pollution detection

 

Timo Mechler - Biomimetic Attachment Surfaces using Microparticle Self-Assembly and Dry Etching

 

Viet Anh Nguyen - Large-Scale Self-Assembly of Carbon Nanotube

2004 Summer Program Staff

Professor Douglas Ernie

Meghan Drivdahl

Program Assistant

Professor Bethanie Stadler

2004 Summer Program Abstracts

 

 

Participant: Alexander O'Day

Program: ECE NNIN REU
Home Institution: Cornell University
Faculty Mentor: Stephen Campbell
Project: Fabrication of a Discriminating Gas Sensor

Gas sensors have been used for many years for equipment control and environmental monitoring.  A basic problem is that they typically are unable to discriminate between gasses that are chemically similar.  This generic problem is the source of considerable research, typically involving the creating of arrays of sensors that are slightly different and then developing an empirical signature for different gasses.  In this project the student has designed and fabricated a new type of gas sensor.  The gas must diffuse down a nano sized cavity to reach the sensing elements.  By spacing the elements along the cavity one can determine the diffusion coefficient and from that infer the size of the diffusing species. The student laid out the test chip for this project, and developed several critical processing steps needed to build the device.  Next the necessary materials were deposited and patterned.  Finally, we intend to package the device and test its response.

The logical effort delay model is a method of minimizing the delay of a path in a digital circuit by distributing an equal effort to every stage in the path.  Since the delay in a gate relies on its size and the size of its output gates, its delay can be reduced simply by resizing it and other gates in the path.  This method, in conjunction with a logic gate equivalent replacement method called resynthesis, can reduce the delay in a digital circuit dramatically.  The methods of logical effort and resynthesis can be automated in C++ to make the process quick and convenient.

With many possible applications on the horizon, in areas such as sensor design and magnetic recording media, considerable attention has been given to nanowires, their fabrication, and their characterization. The goal for this project was to grow magnetostrictive nanowires with the intention of eventually using them in sensors. Nickel nanowires were created by means of electrochemical deposition into porous alumina membranes. The effects of the configuration of the copper electrode contact were investigated as well. Samples were analyzed using X-ray diffraction spectroscopy, a scanning electron microscope, and a vibrating sample magnetometer. These methods verified that nanowires composed of Nickel were successfully made, and that they exhibited predictable magnetic properties. Investigation of the electrode contact revealed that the wire deposition distribution was dependent on how copper tape was applied to the disc and the disc’s depth of submergence in the electrolyte, and independent of the thickness of the sputtered copper. The design of acoustic sensors with high spatial resolution and high sensitivity will be possible with continued research on magnetostriction and nanotechnology.

Given a text file of gate activity generated from a verilog simulator for an open core embedded processor I had to write a JAVA program which would visually represent this data.  The set of  classes from JAVA known as Swing were used to create the GUI (Graphical User Interface) application.  The program initially reads in the gate activity file and stores the data in a tree structure due to its hierarchical nature.  The leaves of the tree are the gates.  The actual GUI program looks like a bar graph:  after the user specifies a level of the tree, bars representing all the nodes at this level are displayed.  The height of a bar, or node, is the total gates active under the sub tree where the node is the root node.  The user is also able to specify which time stamp of the file to view.

The self-assembly of colloidal spheres provides a cheap and fast method of fabricating photonic crystals with a photonic band gap.  In this project, three-dimensional self-assembled photonic crystals were modeled with commercial software that utilized the finite-difference time-domain (FDTD) computational algorithm.  Monolayers of both opal and inverse opal photonic crystals were designed and found to have near-field energy densities concentrated near the dielectric of higher refractive index in each crystal.  In addition, a 10-layer opal was modeled and reflectance was measured.  It was shown to have a main reflectance peak at l=1550nm, the wavelength used for optical communications.  Moreover, secondary oscillations were found on either side of the main reflectance peak, a result of the finite number of layers in the propagation direction.  The reflectance peak and near field images of the photonic crystals were in agreement with previous findings and support characteristics of photonic band gap materials.

Using piezoelectric deposition, living microorganisms in a latex formulation can be printed onto bioelectric devices for mercury (Hg⁺²) detection. When Escherichia coli HB101containing the mer-lux plasmid pRB28 is entrapped in a nano-porous latex ink it survives piezoelectric printing, drying, rehydrating, and being induced with Hg⁺² to luminescence through Lux synthesis.  The reactivity of E. coli mer-lux when printed in dot arrays with two piezo tips (opening: 25µm and 50µm) was examined by inducing with 1 to 10,000 nM Hg⁺².  The necessity of forming nano-pores in the dried latex ink and the ability of E. coli to survive piezoelectric deposition were investigated.  Reactive microbial inks have many potential applications such as detecting mercury in the environment, in fish, or in metal recovery.

Current two-dimensional (2D) Field Programmable Gate Arrays (FPGA's) suffer from large delays of global interconnects and this is even more stringent as circuit sizes (and therefore FPGA's) are continuously increasing.  One possible solution to address this problem is the development of the three-dimensional (3D) FPGA's. 3D FPGA's can offer smaller average net delay compared to the 2D case and therefore they allow implementation of larger circuits while attaining the same performance.  In order to fully analyze 3D architectures one needs appropriate CAD tools.  This work is on developing a Placement and Routing Graphical User Interface (PRGUI), which can be used to visually display the placement and routing of circuits on 3D FPGA's. Features of PRGUI include the ability to view the entire three-dimensional circuit or every layer individually.  PRGUI has many filtering options for displaying the netlist and timing critical paths. These features, coupled with its abilities to zoom and rotate make PRGUI a very powerful program for displaying placement and routing on 3D FPGA's.

Participant: Jon Brackbill

Program: ECE REU
Home Institution: Ohio Northern University
Faculty Mentor: Professor Gerald E. Sobelman

Graduate Student: Kai-Chuan Chang 
Project: Fixed Point N-Tone Sigma Delta Modulators for UWB Applications

Wireless networking is a rapidly growing technology and current hardware is proving to be a bottleneck for networked devices.  Ultra wideband solutions aim to solve that bottleneck by providing high speed low powered devices.  Current sigma-delta filtering techniques are not sufficient for these new UWB applications as they mostly designed for low frequency applications.  N-Tone sigma-delta modulators are ideal for this type of application since the UWB signal has spectral gaps in the signal.  This allows for noise shaping out of the signal band for removal by a comb filter.  This project implements the N-Tone sigma-delta modulator for floating point and some limited fixed-point models as well.  The floating point model was able to reconstruct the discrete input signal with the accuracy of the amplitude increasing as the filter order increased.  The period and the power spectrum were correctly reconstructed by the filters.

The aim of nanotechnology research is to construct and control things on the very small scaleó even at the molecular level.  Nanowires are some of the simplest nano-scale devices, wires with diamters on the nanometer scale, yet they are scientifically interesting.  The study of nanowires involves determining methods for their fabrication and characterization, and how to control and measure their physical, electronic and material properties.  Nanowires made of magnetic materials allows for potential applications in high-density memory storage and magnetic field sensing.  One of the most complicated measurements of a magnetic nanowire is the magnetoresistance (MR), which measures how the electrical resistance changes with applied magnetic field.  This project was involved creating, handling, and studying cobalt nanowires, including experimenting with a method to measure their MR. This involved placing single nanowires on electrode arrays and measuring the voltage/current relationship using an electrical probe station.

This study involves the designing and implementing of a PDA program that imitates the real-time movements of a two limb robot.  The TerminatorBot (a.k.a. CRAWLER Scout) is specifically designed for “corebored inspection in search-and-rescue missions.”  Corebored inspection is the process of retrieving visual data via a small camera from a void.  The executions of [six] small motors control the limbs of the TerminatorBot which reflect the angles of each arm.  Tests have already been done to calculate the angles of different locomotive gaits.  The objective in designing the interface program is to have the user visualize and control the robots gait movements without looking at the actual robot.  To implement this design, TerminatorBot’s pictures are drawn based on the angles calculated from the locomotive gaits.  The PDA program will communicate with the TerminatorBot with a serial cable (RS232) connected to it.   The major benefit of the robot can be measured in human casualties.  Preliminary testing of the TerminatorBot shows that it is more efficient to have three people control the TerminatorBot than 10 people searching for victims in a void or rubble after a disaster.

Participant: Margo Smith

Program: ECE NNIN REU

Home Institution: Chapman University

Faculty Mentors: Prof. David Pui, Dr. Christopher Stipe, Dr. Seong-Chan Kim

Project: Controlled Neuron Growth By Patterned Protein Gradients

Neuron growth is controlled by the growth cone, located at the end of the axon. Extra-cellular components, such as the presence of various proteins, control the directional growth of these neurons; this allows for the possibility to manipulate extra-cellular factors to aide in the re-growth of damaged neurons. The objective of this project is to create protein gradients using an electrospray system, ultimately producing a system to target the growth of axons to specified endpoints. The electrospray system is first studied using polystyrene latex (PSL) particles as a substitute to the growth-effecting proteins. The electrosprayer deposits particles uniformly on a substrate of silicon, and the ultimate surface concentration of particles is controlled by moving the substrate under the electrospray tip at a constant velocity or various accelerations. Neuron cells from extracted dorsal root ganglia (DRG) of fetal chickens are deposited on the substrate, and the growth patterns are observed under varying conditions, including several protein gradient slopes and the presence of different proteins.

Participant: Mikhail Lisovich

Program: ECE REU

Home Institution: Penn State-Schreyer Honors College

Faculty Mentors: Professor Jaekyun Moon and Jihoon Park

Project: Fixed Point Simulation of Error Correcting Algorithms

The trend in the magnetic recording industry has always been toward miniaturization, with data density growing at more than 100 % per year. The resulting increase in complexity along with bit error rate requirements of less 10^-14 have prompted the development of numerous error correcting algorithms. One such algorithm being developed by Dr. Moon’s research group combines an adaptive equalizer, a Viterbi detector and parity check algorithm on a versatile, FPGA-contained system. This algorithm is currently implemented using MatLab scripting and floating point arithmetic, which is impractical to realize with an FPGA’s limited resources. Consequently, my work involves conducting fixed point simulation for various parameters of the program in order to determine quantization scheme which would achieve the optimum balance between memory consumption and performance reduction.

Participant: Shamekia Dixon

Program: ECE REU

Home Institution: Alabama Agricultural and Mechanical University

Faculty Mentor: Eric Van Wyk, Dept. of Computer Science and Engineering

Project: Language Extensions

Writing code is a hassle, especially when programs become lengthy and almost impossible to read.   Glancing at the code for only a few seconds can become tedious to understand. Another problem associated with writing code is the unnecessary declaration of various methods and statements to retrieve only the desired data from a file.  My project at the University of Minnesota required the use of SQL (Standard Query Language) in Java. The function of SQL was to retrieve and abstract the various tables and files from the Java database. Retrieving data can become complicated when SQL errors are undetectable by the Java compiler. The need for a compiler to detect errors in both languages is clear. It would be easier if a programmer could extract only desired features from a language instead of purchasing or retrieving an entire programming language for the use of a few of its features.

Participant: Shane Wood

Program: ECE RET

School: Irondale High School

Research Topics: nanotechnology, materials science, and pollution detection technology

Participant: Timo Mechler

Program: ECE REU

Home Institution: Luther College

Faculty Mentor: Babak Ziaie

Graduate Student: Tingrui Pan

Project: Biomimetic Attachment Surfaces using Microparticle Self-Assembly and Dry Etching

This research project is aimed at creating a nanofabricated biomimetic attachment device. Adhesive surfaces have been commonly found in nature and these successful evolution products present a strong promise for future micro and nano scale applications in a wide variety of biological and medical fields (e.g., lab-on-chip system). Long-range self-organized 2-D nanopatterns on a selective silicon substrate were achieved for this attachment surface by using current nanosphere self-assembly monolayer (SAM) techniques. The resulting patterns were transferred into silicon substrates through reactive ion etching (RIE). Following the removal of the silica nanospheres, one type of silicone elastomer, Polydimethylsiloxane (PDMS), was then applied onto the silicon substrate to replicate the selective nanopatterns formed by the dry etching process. The final attachment device was carefully released from its silicon mold by peeling it off from the edge. Modeling and testing were proposed to characterize this attachment surface.

Participant: Viet Anh Nguyen

Program: ECE REU

Home Institution: Louisiana Tech. University

Faculty Mentor: Heiko Jacobs

Project: Large-Scale Self-Assembly of Carbon Nanotube

Carbon nanotubes offer great promise as molecular wires because they exhibit high electrical conductivity and chemical stability. However, constructing nanotube-based electronic devices requires a controlled means of assembling the tubes. With the self-assembly monolayer method, we have been success in creating a new way to align millions of nanotubes precisely in place. In our process, individual polar molecular marks captivate and align single-walled Carbon Nanotubes along predestined lines without external force, enabling any single-walled Carbon Nanotubes-based structure to be assembled simply by using polar molecular patterns with the required shapes. This method will allow wafer-scale fabrication of millions of carbon-nanotube circuits with single-nanotube precision, and may enable nanotube-based devices, such as transistors and sensor arrays, to be produced industrially