| |
    
|
by Rodney R. Gayle
Sean Seutter, a Minnesota native, completed his undergraduate work and his Master's program at the University of Minnesota and is now working on his Ph.D. here. His career began in electrical engineering, but he has since shifted his focus to material science, in which he is currently doing research.
Seutter's senior honors project led him to continue his electrical engineering studies. His honors project, given to him by his adviser, Dr. William Peria, was to create a scanning tunneling microscope (STM) that would allow the user to see the surface characteristics of a semiconductor as it was being assembled without exposing its surface to oxygen. Although his first attempt failed, Dr. Phil Cohen, an electrical engineering professor, offered him a position on a team that would design and create a working model of his senior project.
The Molecular Beam Epitaxy (MBE) lab, a branch of the electrical engineering department, deals with semiconductor fabrication. MBE was developed early in the '70s to grow high-purity epitaxial layers of compound semiconductors. Because of its ability to control surface thickness and composition, this method has been important in the development of technologically advanced electronic and optoelectronic devices.
The principle of molecular beam epitaxy is heating and evaporating two sources of gallium (Ga) and arsenic (As) in a high vacuum on a crystalline substrate that comes directly from an industrial manufacturer.
The initial substrate given to the lab is very rough, so a large amount of GaAs is needed to smooth it, the buffer layer, out. Next, a layer of tin (Sn) is placed on top of the buffer layer. Seutter and his colleagues are testing the effects of adding Sn to the buffer layer, a newly developed concept, on the formation of each layer.
In order to prevent impurities and oxidation from occurring in the crystalline structure, a high vacuum is needed for growth. Growth of the layers can be controlled by two methods. Placing the shutter in front of the sources prevents them from being evaporated onto the substrate.
Reflection high energy diffraction (RHEED) allows control of the surface structure, growth mode, and growth rates. In this process, electrons are emitted across the surface of a monolayer at a shallow angle, then reflect from the surface and strike a phosphor screen, forming a pattern which indicates surface structure. The more intense the electron beam is on the phosphor screen, the smoother the surface is. The sample is then sent to the STM, where the photo multiplier gives the signal to the computer to monitor real-time growth.
The STM at the University of Minnesota was custom built in the MBE lab. There are relatively few such machines around the world, so each machine is unique to its environment. The STM can produce very clear pictures of individual Ga and As molecules by isolating vibration from the machine using two sets of springs and dampers.
The STM works by placing a tungsten wire 1 nm from the surface of the sample. As the wire passes by the surface, it acquires an electrical charge by coming very close to the electron cloud of the sample. When a voltage is applied between the wire and the sample, the actual surface structure can be determined by measuring the potential difference across these two surfaces.
The STM represents a large step forward in semiconductor fabrication. "It helps us answer the question of how surfaces grow," said Seutter. "It helps us ask questions such as: Where is the gallium going to go as it grows? What type of barriers are at the step edges? How does tin change the surface morphology?"
After completing his undergraduate work at the University of India in Prune, Nikhil Sarpotdar came to the University of Minnesota to earn his Master's degree, which he is currently pursuing.
Sarpotdar is working with micro-chip circuitry using very large scale integration (VLSI), which involves placing an ever-increasing number of transistors on a microchip. As the technology has advanced in the past two decades, the number of transistors on a microchip has increased from a couple thousand to a couple million. With this advancement, more and more products today have converged on the idea of application specific integrated circuits (ASICs), which can be used for various designs (e.g., a computer chip for a car's anti-lock braking system) specific to that product.
For his Master's project, Sarpotdar is working with eight other engineering students to develop new digital signal processing (DSP) methodologies. Digital signal processing incorporates the theoretical and practical aspects of representing information-bearing signals in a digital format. It also involves using computers or special-purpose digital hardware to either extract the information or to transform the signals.
Sarpotdar and his colleagues seek to improve DSP by increasing the speed with which the circuit operates and reducing its power consumption. One proposed idea was to use a pipeline method, which "would take a group of digits at a time instead of one at a time. This would increase the speed of the machine but would compromise the area," said Sarpotdar.
If the chip is successful, future improvements may include making a microprocessor that will increase the rate of algorithmic exchange. This method will eventually be faster and cheaper.
After completing her undergraduate degree at Worcester Polytechnic Institute, Andrea Goring came to the University of Minnesota in 1993 to complete her Master's degree in signal processing.
In the fall of 1995, Goring switched the direction of her studies to biomedical engineering. She began working on a Ph.D. program in nuclear magnetic resonance (NMR), which operates on the ability to decipher the chemical make-up of a compound based on its elemental bond strengths.
Her goal was to build an integrated NMR sensor by perfecting a chip with micro pick-up coils. These micro coils are fabricated in a very sterile environment to prevent contamination, but even in an artificially controlled room, malfunctioning prototypes are still made. In some instances, the coil can meet the most rigid specifications; yet, when connected to a power source, the coil does not work.
For Goring, the next step involves finishing the micro coil and being able to characterize it. This will probably involve making the coil meet certain specifications and maintain a steady state equilibrium while in operation. Achieving the optimum signal-to-noise ratio, which has been a problem in the past, is necessary to increase the system's efficiency. Diagnostic tests must then be run in order to compare theoretical values to experimental values.
The future of these NMR coils could very well be in use as an invasive NMR catheter. However, a project of this magnitude will need much more research before it can be realized. One complication involves the orientation of the coil itself. Once inside the body, the coil must be oriented a certain way, and a method of accurately sending the coil through the body has not been developed yet. The impact that this coil will have as an invasive NMR catheter is not quite clear, though the prognosis is hopeful.
