Abstract
One basic assumption underlying all the deep level transient spectroscopy derivations is the existence of a single dominant emission process during the transient. It is demonstrated in this paper that the use of conventional procedure could result in significant errors in all the calculated defect parameters when the concerned level has comparable emission rates for both carriers in the detected range. A simple verification method and an improved scheme are proposed to determine the accurate defect parameters. By using this scheme, it is confirmed that EL2 in GaAs is an electron trap with a dominant electron emission rate. A level, designated as E3, in Si shows similar electron and hole emission rates in detected range. It is proven for this level that the conventional Arrhenius plot will lead to large errors in all the calculated defect properties, while the improved scheme provides the accurate information.