Cleaving Semiconductors for (Blue) Lasers Diodes

The module is based on the paper of Kuramta et al: in FUJITSU Sci. Tech. J. 342 (1998). p.191. The authors provide data about cleavage planes in semiconductors.

The preferred cleavage planes of a semiconductor are not as clear-cut as it seems. As everybody know who just once dropped a {100} Si wafer knows, the fracture plans are the {110} planes; a large part of the literature, however, including the paper given above, insists that it should be {111}.

Be that as it may, here we take the planes given in the article.

From the Ref. given above
Comparison always to GaN

GaN as a semiconductor technology material only comes as a thin layer on a substrate other than GaN since there simply are no usable GaN single crystals

As we know, if we grow thin layers with different lattices, we have to watch out for misfit dislocations. It is important to look for substrates with a lattice constant as similar as possible to that of the thin layer to be grown. The table above shows the lattice mismatch of prospective substrates to GaN and thus gives a guideline.

If we want to make a Laser diode form the thin film, we have a few more requirements besides "just" avoiding misfit dislocations as best as we can:

The substrate should have a high electrical and thermal conductivity. The first property would make it easier to supply the large current densities we need to operate a Laser diode, the second to remove efficiently the heat generated during operation.

The whole stack of substrate and layers should cleave nicely on a well-defined and very flat plane because the two relevant surfaces obtained by cleavage will serve as the mirrors of the Faby-Perot resonator we need for a Laser. Now look at the possible cleavage relations:

Cleavage of semiconductors with substrates

From the Ref. given above

Summing up: There is no ideal substrate - you have to find the optimal compromise once more if you wan to make the blue Laser diode.

With frame

9.2.2 Laser Diodes