I’ve been thinking about homebrewing a scanning tunneling microscope for a couple of years now, since finding out that it was possible. However, I did not have the engineering expertise to actually pull it off. Conceptually, the device is very simple. Basically, a sharp conductive probe is passed over a conductive sample, a voltage is established between the probe and the sample, and the tunneling current is constantly monitored. Using an STM, one can image a conductive surface down to the atomic level.
There are several challenges for the average hacker. The most daunting is that because nanotechnology is such a young field, one cannot just Google for this stuff and find usable plans for STM construction. I have found several potentially useful websites; since none of them describe workable projects with reasonable detail, I’ll be using them for ideas only. However, I intend to document this project as well as I can, for the benefit of future hackers of nanotechnology.
Besides the low availability of information, the project is difficult because of the level of precision required to build it. STM’s are used to visualize individual atoms, and this means that the probe must be extremely fine and be scanned very accurately across the sample. Furthermore, the whole device must be protected from vibrations during operation. The electronics must also be designed to minimize noise.
Mechanically, the STM will consist of a coarse adjustment mechanism responsible for positioning the probe in close proximity to the sample, as well as a fine control mechanism to actually scan the sample. The coarse adjustment will be done using a finely threaded screw attached to the scanning head. The assembly would be moved up and down using a tapped disc. The fine control mechanism will be entirely piezoelectric, as is customary for STM design. Instead of using highly expensive sticks of crystal, one of my information sources specifies a modified unimorph disc. Unimorph discs are commonly available in piezoelectric buzzers, and are easily modified for multi-poled XYZ motion.
The probe itself also requires attention. To achieve the high resolution typically associated with STM’s, the tip must be a single atom across. Probes are often made of platinum/iridium or tungsten for their strength, conductivity, and nonreactivity. However, copper may be usable for low-budget scanning. To generate the fine tip, I intend to pull a piece of copper wire taught, then make a diagonal cut. As the cut goes through the wire, the tension should start stretching the metal at the cut, making it thinner.
Finally, there needs to be a way for the STM to communicate with a computer. The easy way out would be to use a microcontroller kit. The computer would send positioning and measurement commands to the microcontroller, while the microcontroller does the necessary mathematics and drives the STM.
Most of this will probably have to wait another semester, because I cannot hack and study for exams at the same time. But it can’t hurt to put together some ideas in the interim, right?