Silicon

This is the top of a small diameter silicon crystal produced in research. The small cutoff circle is where the seed crystal was attached, where that seed was lowered into a pool of molten silicon and this crystal slowly pulled out, sticking to the solidified material by surface tension and then solidifying itself.

This is a mistake from a university research laboratory- someone was trying to melt some high purity silicon into a crucible, probably graphite. The melter they were using employed a vacuum/argon electric arc melting chamber, with a vacuum-suction casting attachment below where a trap door opens below the main melting chamber and your molten material is sucked down into the crucible (hopefully).

For whatever reason, it didn't work in this case. You can see the crucible shape in the really rough, unhappy looking bottom, where the solidification occurred too rapidly. However, on the top is a beautiful crystalline display, showing both the purity of the material and the tendency of silicon to crystallize, even when solidification occurs rapidly, which it likely did in this case.

Here is another silicon boule for research, much smaller than would be actually sliced into wafers for chip fabrication. This sample was produced either by the Czochralski pulling method or possibly the less-common floating zone method, though the end result is basically the same.

The distribution of silicon crystals such as this one for research in the second half of the 20th century is quite interesting, and strangely incestuous in a way. For example, General Electric might have received a government grant to grow crystals such as these as a pilot for making into electronic devices. General Electric would then send a box of crystals such as this one to MIT, contracting a professor there to determine the dissolved hydrogen content and check the homogeneity. The professor would then save the pieces, eventually giving them to a colleague of his at a different university, where this new professor would saw thin wafer-thin slices off and use them to test a chemical vapor deposition (CVD) process. A graduate student would end up with the unused crystals when they took over the CVD project from the professor, still a couple pieces such as this sample, The graduate student might then spend a year or two working on their Ph.D thesis at a US DOE National Lab, taking the crystals with them, where they would be abandoned and finally used as extremely pure evaporation material for a electron beam evaporation partner project with General Electric, coming full circle. I have not read anything about this phenomenon, nor heard it acknowledged widely in the research community; rather, this is something I have discovered myself in my pursuit of saving element samples from the landfill, and I think it is both fascinating and quite entertaining.

Hello,

I was enjoying looking at your periodic table today when i read you comments about the square silicon boule you have. This crystal was not pulled using the Czochralski method like many of the other boules you own. I believe that it was made using the Stephanov process. In this process, a graphite shaper is inserted into the pool of liquid Si, and almost any shape can be produced. It is also possible that another technique called the edge-defined film-fed growth(EFG) process could have been used. The process is similar to the Stephanov method, however capillary feeding through the shaper is used to feed material to the growing crystal. Also, you mentioned that you believed that all boules start square. In the Czochralski process, a single crystal seed is introduced into a liquid silicon pool, and is the slowly lifted and rotated while the pool is counter-rotated. This rotation is what causes the resulting crystal to be round. In the other two methods the shaper defines the dimensions of the boule and no rotation occurs. I hope that this information is useful to you.