Slicing method cuts germanium wafer costs
Researchers say that they can reduce manufacturing costs of compound semiconductor solar cells by slicing germanium substrates thinner than existing wire saws currently are able to.
Eberhard Bamberg and Dinesh Rakwal of the University of Utah have shown that wire electrical discharge machining (WEDM) wastes less material than conventional cutting methods.
In their in-press paper, published online by the Journal of Materials Processing Technology on September 4, Bamberg and Rikwal cut a cylindrical germanium crystal boule with 75 µm thick molybdenum wire.
The usual brass-coated steel wire saws are around 170 µm wide and waste material in the process of cutting "“ known as kerf - equivalent to a 200 µm slice from the boule.
The Utah team s technique, which applies 150 V to the wire that then arcs across to and melts the germanium, creates just 125 µm kerf.
“At the current standard wafer thickness of 300 µm you can produce up to 30 percent more wafers using our method,” Bamberg said.
The narrow kerf becomes much more important for thinner wafers "“ which Bamberg says can improve substrate yields still further.
“If you go down to 100 µm thick, you can make up to 57 percent more wafers,” he claimed.
Slicing germanium boules made by fellow Utah inhabitants Sylarus Technologies into 66 mm diameter wafers by WEDM took 15 hours on average. By contrast, standard wire saw cutting takes only six hours.
The wafers were measured to have surface roughness, Ra, of 0.73 µm after cleaning with a hydrofluoric acid-based etchant, and were crack-free.
Bamberg hopes to speed up the overall cutting rate and has patented a process that uses multiple electrically-charged wires to produce several wafers simultaneously.
Substrate-making startup Sylarus "“ which supplies germanium substrates for the kind of multi-junction material that until recently held the photovoltaic efficiency record "“ is now considering the commercial viability of Bamberg's WEDM approach. Whether or not it is taken up, he and Rikwal also suggest that this method holds potential for all high-value semiconductor crystals.
In their paper the Utah researchers propose using WEDM for: “rare semiconductor materials, such as germanium, gallium arsenide and other III-V and II-VI semiconductors where material cost is a major contributor to the overall cost of the wafers.”
