Undergraduate Jenny Hu Studies Nanowires
|
September 30, 2004 -- Devices are being scaled downward into smaller, more efficient devices. One method of scaling semiconductor devices involves improving the design. Another method is to create a new family of devices using nanowires to enable nanoelectronics, a research topic still in its infancy.
Multiple nanowires of different materials can be connected together to perform more complicated functions. However, for the nanowires to be useful, they need to be able to connect to other devices.
Ideally, an ohmic contact with minimal resistance is desired. With lower contact resistance, less voltage is dropped across the contact, so the same amount of current can be passed using less voltage, which requires less power.
I became involved in my project first by shadowing a graduate student, who showed me his daily routine. He taught me how to use the equipment and the details of the photolithography process for making contacts to intrinsic InP nanowires, which were being grown in the lab. After I had learned the process, I continued it on my own and refined portions of the method.
Throughout my research I encountered many, mostly technical, problems. In the lift-off process to create the contact, I could not get the metal to lift off when it was supposed to. Instead, I ended up with a full layer of metal, signifying a failed attempt. I could not figure out the reason for the problem, so I tested the process by varying the conditions and iterating again. I eventually discovered it was due to the excessively high temperature of the hard bake, which had caused changes in the properties of the photoresist.
Another problem was the inconsistency of the results, which appeared to differ each time. Unfortunately, I have not found a solution yet. Possibilities include excessive strain on and frailty of the wires. Both of these problems can lead to a breakdown.
There were also many problems involving the equipment, such as failures or other difficulties. These issues were solved easily by asking for help from graduate students who were more familiar with the equipment.
The contact used was a Ti-Pd-Au structure, with Ti used for strong adhesion to the surface of the substrate, Pd used based on its reactivity with InP, and Au for protection. The contact was annealed at temperatures between 300 to 400 oC for varying times.
Upon inspection of the current in the -1 to 1 V range, there was a definite improvement in the quality of the contact compared with the contact prior to annealing. Very recently, an extremely linear contact was obtained, though it must be examined further to confirm the results.
Throughout the coming school year, I hope to continue research in this area. I intend to study the contact to p-type InP wires as well as core shell structures. The p-type wires are more difficult to make contacts to, due to a pinned Fermi level, which results in a large Schottky barrier height to overcome. I would also like to compare the contacts of other materials to that of the Ti-Pd-Au structure. These are only a few of the many interesting aspects I would like to explore.