Time zero determination of FET reliability

Claim 1

The method for predicting the failure rate of metalinsulator-semiconductor devices caused by variations in manufacturing process parameters, comprising the steps of: A. applying for a relatively short period of time, a first current limited voltage across the insulator of a first portion of a group of substantially identical metal-insulatorsemiconductor devices, said voltage having a magnitude substantially equal to the saturation voltage of said semiconductor device; B. determining the average percentage of said first portion of said devices which exhibit metal-to-semiconductor voltage breakdown under said first voltage at the end of said short period of time; C. applying for an extended period of time a second voltage across the insulator of at least a second portion of said group of devices, said second voltage having a magnitude substantially lower than said first voltage; D. determining the average percentage of said second portion of said devices which exhibit metal-to-semiconductor voltage breakdown under said second voltage at the end of said extended period of time; E. correlating the percentage of devices exhibiting voltage breakdown in said short and extended periods of time to determine the specific relationship between said percentages; and F. predicting the average percentage of devices expected to exhibit voltage-to-semiconductor breakdown at said extended period of time for a different group of metal-insulatorsemiconductor devices manufactured under different manufacturing process parameters by determining tHe average percentage of at least a portion of said different group of devices exhibiting voltage breakdown after applying said first voltage for said relatively short period of time and determining the expected average percentage of devices to exhibit breakdown at said extended period of time by applying the specific relationship determined in step E to said average percentage of devices exhibiting voltage breakdown in said portion of said group of different devices.

Claim 2

The method for predicting the failure rate of metalinsulator-semiconductor devices caused by variations in manufacturing process parameters, comprising the steps of: A. applying for a relatively short period of time, a first current limited voltage across the insulator of a first portion of a group of substantially identical metal-insulatorsemiconductor devices, said voltage having a magnitude substantially equal to the saturation voltage of said semiconductor device; B. determining the average percentage of said first portion of said devices which exhibit metal-to-semiconductor voltage breakdown under said first voltage at the end of said short period of time; C. applying for an extended period of time a second voltage across the insulator of at least a second portion of said group of devices, said second voltage having a magnitude substantially lower than said first voltage; D. determining the average percentage of said second portion of said devices which exhibit metal-to-semiconductor voltage breakdown under said second voltage at the end of said extended period of time; E. correlating the percentage of devices exhibiting voltage breakdown in said short and extended periods of time to determine the specific relationship between said percentages; and F. predicting the average percentage of devices expected to exhibit voltage-to-semiconductor breakdown at said extended period of time for a different group of metal-insulatorsemiconductor devices manufactured under different manufacturing process parameters by determining tHe average percentage of at least a portion of said different group of devices exhibiting voltage breakdown after applying said first voltage for said relatively short period of time and determining the expected average percentage of devices to exhibit breakdown at said extended period of time by applying the specific relationship determined in step E to said average percentage of devices exhibiting voltage breakdown in said portion of said group of different devices.

Claim 8

The method of determining at the time of manufacture the future expected percentage of gate dielectric failures of insulated gagte field effect devices of the same type as those having a known percentage of gate dielectric failures as determined from an extended constant stress test performed on previously manufactured devices, comprising the steps of: A. electrically stressing for a short period of time the gate dielectric of a plurality of recently manufactured devices with a current limited voltage having a magnitude substantially equal to the saturation voltage capable of being sustained by the gate dielectric of said devices; B. determining the average percentage of said recently manufactured devices exhibiting gate dielectric breakdown after said short period of time; C. correlating the average percentage of devices exhibiting breakdown with said known percentage of gate dielectric failures to determine the specific relationship between said percentages; D. manufacturing an additional plurality of devices of the same type as previously manufactured; E. determining the average percentage of gate dielectric breakdowns of at least a portion of said additional plurality of devices after the application of the conditions of step A; and F. determining the expected future percentage of gate dielectric failures by applying said specific relationship determined in step C to the average percentage determined in step E.

Claim 9

The method of determining at the time of manufacture the future expected percentage of gate dielectric failures of insulated gagte field effect devices of the same type as those having a known percentage of gate dielectric failures as determined from an extended constant stress test performed on previously manufactured devices, comprising the steps of: A. electrically stressing for a short period of time the gate dielectric of a plurality of recently manufactured devices with a current limited voltage having a magnitude substantially equal to the saturation voltage capable of being sustained by the gate dielectric of said devices; B. determining the average percentage of said recently manufactured devices exhibiting gate dielectric breakdown after said short period of time; C. correlating the average percentage of devices exhibiting breakdown with said known percentage of gate dielectric failures to determine the specific relationship between said percentages; D. manufacturing an additional plurality of devices of the same type as previously manufactured; E. determining the average percentage of gate dielectric breakdowns of at least a portion of said additional plurality of devices after the application of the conditions of step A; and F. determining the expected future percentage of gate dielectric failures by applying said specific relationship determined in step C to the average percentage determined in step E.