Silicon Cross-Section Sample Preparation (Cleaving)

Abstract I have found the best way to cleave [001]-oriented silicon wafers is rather different compared to GaAs or InP. The problem is that Si prefers to cleave on (111) planes rather than (110) and so one gets an angled face with the cleave, which is usually rather uneven and often doesn't run straight. This is even worse when cleaving close to an existing edge, which attracts the crack front as it propagates. Good for making low-angle cleaved specimens, but a problem otherwise. It is possible to make Si cleave on (110) by cleaving the wafer without any support.