The germanium crystal documented here comes from the estate of Professor Heinz Beneking. Professor Beneking was one of the first to carry out research and teaching in the field of semiconductor technology in Germany. He worked at the RWTH Aachen.
The germanium disk was certainly cut from a longer crystal pulled from a melt. This slice was then cut horizontally and vertically to produce rods of the same size. Some of the rods were broken out of the crystal. As will become clear, it must be germanium.
The crystal comes with the paper shown here. According to this, transistors should be made from it.
A comparison with a CR2032 button cell shows that the crystal disk is not very large.
Over time quite some dirt has accumulated in the gaps.
The dimensions of the rods are 0,6mm x 0,6mm x 1,8mm. The surface is surprisingly smooth. Perhaps the surface was etched after sawing. This would make sense, as impurities and dirt on the surface can have a negative effect on the properties of the later transistors.
The fifth edition of the “General Electric Transistor Manual” explains what is behind the term metlback. It is a lesser known method of creating an NPN transistor. In principle, the process works with both germanium and silicon, although it was usually used with germanium crystals.
First, n and p doping are introduced into a germanium crystal so that the n doping is dominant. The crystal is cut into rods, attached with one side to a heat sink and heated until the other side melts. If the melt is allowed to cool, it first recrystallizes slowly at the still solid area and then increasingly faster towards the end. The incorporation of n and p dopant atoms depends in part on the speed of crystallization. Gallium, which is usually used as a p-dopant, is hardly influenced by the speed of recrystallization. Antimony, which is usually used as an n-dopant, builds up more strongly in the crystal structure the faster the crystallization takes place. This physical effect ensures that a layer is formed on the non-melted rod that has a dominant p-doping, while the n-doping predominates again in the remaining area. This results in the desired NPN structure.
In the book “Getting Started With Transistors”, Lous E. Garner shows what a transistor with a so-called meltback junction usually looks like.
However, the term meltback is also used in another context. Robert Widlar is the author of a training document entitled “Transistors”, which was used in the “Department of Weapons Training Lowry Air Force Base Colorado”. The fifth chapter of this document describes another way to create NPN transistors. This involves pulling a germanium crystal from a melt containing n- and p-dopants, with the n-dopant (antimony) dominating. By increasing the temperature of the melt, the growth rate of the crystal can be reduced. The physical relationship described above then ensures that the p-dopant dominates in this area.
If the temperature of the melt is increased a little more during this process, part of the crystal that has already been produced melts again. As a result, the transition to the p-type semiconductor, which occurred when the temperature was first increased, melts again and an advantageous, more abrupt transition occurs. This process is also referred to as meltback. The germanium crystal handed down by Professor Beneking is certainly the variant in which the germanium rods still have to be melted. This is clear due to the paper which describes that transistors should be built by meltback.
https://www.richis-lab.de/BipolarA66.htm 
Home
Help
Search
Login
Register
