Recombination at the silicon nitride silicon interface

A contactless technique to study the characteristic quantities of the SiNx/Si interface is of great interest. To investigate this interface, the influence of the light-induced excess charge carrier concentration (Delta n(b)) on the differential interface recombination velocity (S-diff) at the SiNx/Si interface is studied using frequency resolved microwave conductivity measurements. S-diff. is determined for SiNx/Si systems with different doping types and doping concentrations of the Si substrate. The excess carrier concentration is measured independently from the determination of S-diff by means of the method of the short circuit variation. Very good passivation of the Si surface by the SiNx film was observed with surface recombination velocities smaller than 10 cm/s. The experimentally obtained values of S-diff as a function of Delta n(b) show good agreement with calculations based on the Shockley-Read-Hall formalism. Capacity-voltage measurements enabled the determination of the interface defect density and the positive fixed charge density in the SiNx(Q(f) = 2 x 10(12) cm(-2)). The dependence of S-diff on the excess carrier concentration is caused by the large positive fixed charge density and the high ratio of the charged-to-neutral acceptor capture cross sections (sigma(ca)/sigma(na)). The comparison between the model and the experimental results yields acceptor capture cross sections of sigma(ca)=5 x 10(-15) cm(2), sigma(na)=5 x 10(-21) cm(2) for n-Si and sigma(ca)=1 x 10(-14) cm(2), sigma(na)=1 x 10(-20) cm(2) for p-Si. The influence of the donor capture cross sections on S-diff can be neglected. Additionally, the experimentally determined interface defect density could be verified with model calculations. (C) 1997 American Vacuum Society.