Abstract
A model is presented to explain the effect of oxide bias on photodiode collection efficiency. Its attempts to distinguish the effects of charge density from those of recombination velocity at the Si–SiO2 interface. It has been tested by comparison with experimental oxide bias curves recorded on a p+nn+ photodiode with a thermally grown oxide at different stages of its degradation by excessive oxide bias and partial restoration by exposure to UV radiation. By fitting one parameter to the spectral responsivity and two to each oxide-bias curve, the ratio of the photocurrent as a function of oxide bias to the photocurrent with an oxide bias of −60 V could be reproduced with a maximum error ranging from 0.0033 for the undegraded diode to 0.019 when the response was lowest. The results indicate that both the interface charge density and trap density are affected in the degradation process. This study has also shown that oxide bias does not restore the collection efficiency strictly to unity.
© 1984 Optical Society of America
Full Article | PDF ArticleMore Like This
Robert L. Booker and Jon Geist
Appl. Opt. 23(12) 1940-1945 (1984)
J. Geist, A. J. D. Farmer, P. J. Martin, F. J. Wilkinson, and S. J. Collocott
Appl. Opt. 21(6) 1130-1135 (1982)
James L. Gardner and Frank J. Wilkinson
Appl. Opt. 24(10) 1531-1534 (1985)