One of the potential areas for improvement in the nanowire growth process is the Silicon substrate. Even though the reaction happens in a mostly inert atmosphere, there is still the possibility for the Silicon substrate to undergo unwanted side reactions such as oxidation, especially if there are any impurities. Working with my research mentor, we hypothesized that the use of a Stainless Steel substrate would largely eliminate side reactions, so that the Cadmium Sulfide would interact only with the gold, and the product would be as pure as possible.
Chris made numerous attempts to use the Gold Colloid, but he could not find a way to get the Colloid to adhere to the Stainless Steel. In all of these attempts, there was no wire growth on the substrate because essentially the Gold Colloid would not "stick to" the steel.
sputter-coated the substrate with an extremely thin layer of gold. Sputter-coating is a method that utilizes a brief high-energy plasma to send a pulse of ions from a sputtering target to a substrate. If you look at the picture to the left, we used Gold as a target and the Stainless Steel as the substrate. We applied the plasma for 20 seconds to get a thin coat of gold on the steel. Even after just 20 seconds, there was enough gold deposition that the substrate appeared gold. Now that we had a thin layer of gold that had adhered to the substrate, it was our hope that by heating up this substrate, the gold would anneal and its structure would transform from a sheet into something that resembled small nano-sized dots. It was my job to vary conditions such as temperature, pressure and annealing time and analyze the results. Unfortunately, after numerous attempts, we did not see any significant deposition of CdS onto the substrate.
Based on these results and the appearance of the substrate when viewed under a light microscope, we hypothesized that the layer of gold that was sputter-coated onto the substrate was so thick that it was remaining as a sheet and not forming the bubbles that would be required for wire growth. If this hypothesis was correct, we figured that a shorter sputter-coat time may lead to a more suitable substrate. We sputter-coated the substrate again, cutting the time down to 5 seconds. This time it did not appear gold, but we were confident that a very thin layer had deposited after looking at the substrate under the microscope. Again, I spent a few days varying the temperature, pressure, and annealing time.
Although we were discouraged by these results, we took them to the Scanning Electron Microscope (SEM) for further analysis. The Scanning Electron Microscope uses a beam of electrons rather than a beam of light to scan the surface of a material. It is an invaluable tool for researching nano-sized materials, as it gives you a quick picture of the size and shape of extremely small structures.
Below are 3 images of my attempts to grow nanowires on the Stainless Steel substrate at 140 torr. The images get more zoomed in from left to right, as you can see in the scale bars.
Stainless Steel Substrates (140 torr) zoomed in to 50 µm, 10
µm, and 500 nm scales (click for more detail)
The results of studying the sample in the SEM were interesting and somewhat surprising. From far away (as in the left image) it looked like the gold film just cracked apart, but we were able to find numerous areas (as in the center image) where it looked like the gold was bubbling up, and when we zoomed in even farther (right image) we actually saw what looked like nano-sized gold beads that seem like exactly what we would have wanted to grow nanowires. In the right-most image, the small circles are very similar in shape and size to what would be found in the prepared Gold Colloid. The samples grown in different pressure environments showed very similar results. Below are three images of the substrate at 250 torr.
Stainless Steel Substrates (250 torr) at 20 µm, 5 µm, and 500 nm scales (click for more detail)
At the higher pressure the results are very similar and again we were surprised to see regions of high densities of nano-sized gold particles.
After discussing the results with Chris, we were able to conclude that even though we were successful in getting a thin layer of gold to deposit onto the Stainless Steel and heat it up to a point where it would anneal, there was something that was preventing the CdS from depositing onto the Gold and forming wires. The most likely explanation is that the binding energy between Gold and Steel is actually greater than the binding energy between Gold and Silicon. In order for wires to grow, there must be some degree of attraction between the gold and the substrate, but not too much, because the CdS needs to be able to attach to the Gold and then replace the Gold as it grows into a wire. In the case of Stainless Steel, this was not able to happen because the Gold was just too attracted to the Steel.
While we were not able to grow any wires, we learned a lot about the process and the interplay between Gold, Steel and Cadmium Sulfide.