- The images in this gallery are microscope photos showing structures in semiconductor crystals, mostly silicon, some germanium, and mixtures of germanium and silicon. The area shown is listed in each figure caption. Most crystals include dopants like phosphorus, arsenic, antimony, or gallium to influence the electronic properties. They were crystallized from a melt for scientific purposes to investigate the influence of melt flow on the crystallization process. The resulting crystals resemble metallic cylindrical rods. After growth, such crystals are cut into slices parallel to the growth direction, are ground, lapped, polished, and finally etched to reveal differences in composition or other structures in the crystal. The etchant removes more or less material depending on the composition and structure, so the result is a polished flat surface with a relief on it that shows that internal structure. However, the height differences are in the nanometer to micrometer range, so the etched samples are viewed and photographed with an optical microscope using reflected polarized light in a technique called Nomarski Differential Interference Contrast* (NDIC). Reflected light NDIC transforms the small height differences into interference colors. The order of the interference depends on the height difference, but is also adjustable with an optical device in the microscope, the Nomarski prism. For these images, only the first order interference patterns from black to white are used. The relationship of the different gray values can be altered and even inversed by the setting of the Nomarski prism, so for any given image there is an infinite range of possible tonal relationships even before development and printing the images!
Whereas the part of the crystal that is of interest to science is often rather boring with just a few structures, the final part of a crystal growth experiment, were impurities and dopants are concentrated, and which is often solidified rapidly and uncontrolled, can have very interesting and unpredictable patterns. In addition, experiments where not everything went according to plan are a rich source for images. These parts of the crystals were usually not of immediate interest for the scientific questions I was after, but I always found them really intriguing from an aesthetic point of view. In 1995 I started bringing in some film on the weekends and took pictures of these patterns with regard only to the tonal and geometric relationships and the textures they presented, with no scientific purpose behind it. Some of them are purely geometric, in others one can fantasize seeing landscapes, aerial images and other things, and occasionally the eye and brain are fooled into seeing three dimensional structures that at are not really there or even impossible in the sense of M.C. Escher drawings.
In reality, many of those patterns are due to variable flows in the melt adjacent to the solidifying crystal, and some are due to a process called “constitutional supercooling”, where the interface between crystal and melt breaks down into a pattern of rods or dendrites, depending on the concentration and temperature gradients, and the solidification speed. Other patterns are due to the growth of twinned crystals, to facetted growth, or polycrystalline growth. Since I made those samples, I do know what these patterns are and their basic formation mechanisms, but I think that background is not necessary at all to enjoy them just as they are either as abstracts or as representations of something else.
For those interested in the technical details, the images were taken with a Leitz Aristomet microscope and Leitz PL Fluotar 5x/0.12D, 10x/0.25D, 20x/0.45D, and Plan L 40x/0.60 lenses. The microscope has an attached “Variophot/Orthomat E” camera system for 35mm film with automatic exposure and reciprocity failure correction. The images were recorded on fine-grained 35mm film (Agfa APX 25 and Rolleipan 25), with exposure times of a few seconds to several minutes, developed to normal contrast (N) in Tetenal Neofin blue, and printed on Adox MCC 110 paper using the usual darkroom controls, developed in Moersch SE6. All prints are selenium toned.
*An explanation of DIC can be found online at: http://micro.magnet.fsu.edu/primer/techniques/dic/dicconfiguration.html
For the reflected light configuration, which is actually more simple than the transmitted light version explained above, see: http://www.microscopyu.com/articles/dic/reflecteddic.html