Google “homemade transistor” and you’ll find very little information concerning the actual making of a transistor from scratch. Making a point contact transistor is certainly doable, and there is some information available on the web on specifically how to approach the problem. However, start talking about any other kind of transistor and about all you’ll find is reasons why you cannot or should not attempt to do so. Mostly this is due to the complex equipment requirements (high vacuum, high temperatures) and extremely toxic chemicals used to make modern devices.
Try googling “homemade semiconductor” instead and the first thing you find (April 12, 2008) is Nyle Steiner’s report on making negative resistance oscillators. His two sites, home.earthlink.net/~lenyr/ and www.sparkbangbuzz.com, are filling with interesting and reproducible experiments. As soon as I get or build a power supply capable of the 100 to 150 volts required, I plan on attempting to reproduce his flame triode experiment.
At several points in the past I’ve looked at what was required to make a transistor from scratch, for reasons that can probably be summed up as “it is an interesting problem” and “it simply is not done, so it is a really interesting problem.” I’ve always stopped looking about the time I get to the part of the process that requires extremely toxic gases in a partial vacuum at high temperature.
Recently, Jeri Ellsworth has reported developing a home chip lab consisting, essentially, of a tube furnace capable of reaching 1000°C, a source of nitrogen gas, and brushed-on doping solutions. Reportedly, an article for Make is being written, which I am really looking forward to reading.
Lately, I’ve been researching and experimenting with a different approach to making semiconducting devices at home, one based on the production of semiconducting thin films deposited from liquid solutions. Chemical bath/solution deposition, liquid phase deposition via spin coating or dipping, and successive ionic layer adsorption and reaction (SILAR, which, in spite of its incredibly long name, is really just a matter of dipping a substrate into one solution followed by dipping it into another, repeating as necessary) all appear to be practical methods by which a workable semiconducting thin film can be made by the hobbyist.
The Amateur Scientist column in the June 1970 edition of Scientific American discusses making a variety of thin films, and gives a recipe for making a cadmium sulfide-based thin film transistor using chemical bath disposition. Cadmium, however, is a relatively toxic material and notably bad for the environment.
Zinc-based compounds, on the other hand, appear to have relatively low toxicities, and the production of zinc oxide, in particular, is very easy. Put a drop of an aqueous or alcoholic solution containing a zinc salt on a slide, heat it at 400°C for an hour, and the material remaining on the slide will be mostly zinc oxide. While the production of a good zinc oxide thin film involves a bit more chemistry and processing, it is not horribly complex nor does it require particularly expensive equipment. Zinc oxide fumes can cause “metal fever,” however, so adequate ventilation is mandatory.
There is quite a bit of useful information concerning the making of zinc oxide based thin films and transistors that is hidden around the web. Bits and pieces of useful information can be found on-line in various open-access journals and papers. Some of the most useful information that I’ve found so far is found in a couple of theses and dissertations found on the ScholarsArchive at the Oregon State University. In particular, the dissertations of D. Hong and B. Norris, and theses of M. Grover and D. Heineck contain a wealth of practical information.
I’ve purchased a copy of the book Transparent Conductive Zinc Oxide, Basics and Applications in Thin Film Solar Cells and it has a lot of good information about zinc oxide films in general, but its coverage of disposition methods is mostly limited to techniques such as sputtering, chemical vapor disposition and pulsed laser disposition. None of these methods is particularly practical for the casual hobbyist.
The go-to book for sol-gel information appears to be the Handbook of Sol-gel Science and Technology: Sol-gel. However, at $1400 for a used copy it is pretty much unobtainable, save for the little bits that are viewable via Google Books.
I’ve ordered a used copy of Chemical Solution Deposition Of Semiconductor Films, as it looks to contain a lot of very practical information.
Google Scholar provides links to a large number of published papers that look fascinating, but, for the most part, are also practically unobtainable. Some of the papers are open access, which is really wonderful, but the remainder cost between 20 or 30 dollars apiece, which is pretty steep for a paper that may or may not contain information that is useful to this project.
I do need to look into getting a library card at one of the local university libraries. That should provide some access to many of the otherwise overly expensive books and papers, and allow me to more carefully target the books and papers I do purchase in the future.
To date, I’ve been able to demonstrate the UV photoconductive sensitivity of a zinc oxide thin (or thick) film, and to demonstrate a change in the conductivity of a zinc tin oxide thin film in response to a voltage applied above the film from an insulated gate. The effect, however, based on a voltage change at the gate of 30 volts, is a changed of 0.07uA on top of a leakage current of about 6.53uA, or about 1 part in 93. Given that the gate insulator is a round #0 slide cover that is approximately 0.1mm thick, which is about 1000 times thicker than a useful gate insulator would be expected to be, I’m actually surprised that I can detect the effect at all.