Predicting radiation response from process parameters: Verification of a physically based predictive model

Abstract

We evaluate the hole trapping rcsponsc of twcntytwo oxidcs subjected to twenty-two dikrcnt sets of processing parametcrs. Thc oxides were prepared in tbrec diffcrcnt €acilities, the Harris Semiconductor-Intersil Palm Bay facility, thc foriner Naval Research and Dcvclopinenl Laboratory (NRAD) 4" facility, aitd the new SPAWAR 6" fabrication facility in San Diego, Califoruia. In twenty of the tneiity-two cascs, oxidc hole trapping is almost coinplctcly determined by thc highest processing tcnipraturc aud is in reasonablc agrecment with a recently proposcd physically based prcdictivc model. We have also evaluated Si/SiOl interface trap (Dit) geucration in a subset of four vcry simply processcd oxides ntiliaed in the hole trapping study. Thc D, rcsults are also in rcasoiiable agreement with thc recently proposed modcl.Our results indicalc that it is possible lo make reasonably accurate predictions o€ radiatiou response from processing paranietcrs and that such predictious can be nude with our current understanding of radiation damagc phenomena. (It should be emphasizcd that the current level of undcrstanding is not yct complete. This work docs -not demonstrate that prccise predictious iuvolving all imaginable process paranicters are possible.)

Radiation-induced leakage currents: atomic scale mechanisms

“…Recent experimental studies have shown that the oxygen vacancy E′ hole trap precursor has an activation energy of about 1.5 eV [18]- [22]. Several more recent theoretical studies have provided calculated values in good agreement with experiment [23], [24].…”

“…As mentioned previously, positively charged E′ centers dominate hole trapping in (thicker) oxides irradiated to moderate total dose levels. Recent experimental work [18]- [22] indicates that the oxygen vacancy E′ precursor has an enthalpy of creation of approximately 1.5 eV. Although these experimental results have been somewhat controversial on theoretical grounds [29], [30], at least two careful theoretical studies by independent groups have recently generated results in agreement with the (1.5 eV) experimental value [25].…”

“…In view of the 1.5-eV activation energy of the oxygen vacancy precursor [18]- [22], it is not likely that it could be simply and entirely responsible for these quite high total dose numbers. It is thus likely that they are generated differently for the hole trap centers and also quite likely that the sorts of processing steps that yield an oxide hard with respect to hole trapping will not necessarily yield oxides hard for RILC phenomena.…”

“…It is well established that these centers are generated in oxides by both ionizing radiation [7]- [13] and high electric fields [14]- [16]. Most ESR E′ studies have dealt with the positively charged variant (O Si Si O ) in which an electrically neutral silicon dangling bond orbital is closely coupled to a positively charged silicon [7]- [12], [17]- [22]. This defect state is created when an oxygen vacancy site (O Si-Si O ) captures a hole.…”

Properties of latent interface-trap buildup in irradiated metal–oxide–semiconductor transistors determined by switched bias isothermal annealing experiments

“…15 This assumption is in contrast to the vast majority of work on radiation effects 1 and is still to be verified. 16 Finally, within the context of the P b approach, and not necessarily in direct connection with the LITB, we would like to stress the importance of taking into account the concurrent running of interface-trap formation and passivation processes. 5 This fact has often been neglected in analysis of the postirradiation response of MOSFETs.…”