Abstract
This paper compares several popular accelerated test methods for projecting SiO/sub 2/ lifetime distribution or failure rate: constant-voltage and constant-current time-to-breakdown and charge-to-breakdown tests, ramp-voltage breakdown test, and ramp-current charge to-breakdown test. Charge trapping affects the electric field acceleration parameter for time-to-breakdown and the value of breakdown voltage. Practical considerations favor ramp breakdown testing for gate oxide defect characterization. The effective thinning model is used for defect characterization and the ramp-voltage breakdown test is shown to be superior to the ramp-current Q/sub BD/ test for extraction of the defect distribution. Measurement issues are also discussed.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call