Injection and doping dependence of SEM and scanning light spot diffusion length measurements in silicon power rectifiers

SEM electron beam-induced current (EBIC) and light spot scanner photocurrent (PC) versus distance characteristics recorded from silicon power rectifiers were observed to vary consistently with beam power over the range 0.1 μW to 1 mW. To interpret these results, existing theory has been extended to incorporate a minority carrier lifetime which varies with position, doping and injection level. A simple procedure was devised to take surface recombination into account; under these conditions good agreement was obtained between theory and experiment. In the diffused end regions the results were best explained by a lifetime which varied with doping at N_d^{-( 1 2)}. The variation in shape with increasing beam power could be understood on the basis of an injection dependent surface recombination velocity. In the base, it was necessary to consider injection dependent effects both in the bulk and at the surface to explain the observed characteristics. We conclude that EBIC/PC techniques, while excellent for measuring surface recombination velocity and low level lifetime, are less satisfactory at higher levels. The wide range of injection levels produced by an electron probe or laser light spot makes the analysis rather uncertain. The techniques do, however, provide a good method for assessing lifetime in power devices and, suitably interpreted, can provide useful information about the way lifetime varies with injection level. Taking surface recombination into account is absolutely essential if valid results are to be obtained.