The hysteresis observed in capacitance-voltage (C-V) measurements on metal-sputtered silicon nitride—silicon structures indicates that carriers are injected predominantly as holes rather than electrons. Shifts in the C-V characteristic after bias-temperature stress at 300°C support this finding. In dc conduction measurements on these samples a linear relation was found between the logarithm of current and the square root of field. The slope of this plot and the independence of current on bias polarity indicate a bulk-limited conduction mechanism of emission of carriers from traps in the silicon nitride.

Silicon nitride films (from 500 to 7500A in thickness) have been deposited on silicon and silicon dioxide by reactive sputtering of a silicon cathode in a N2 glow discharge. Both d-c and rf sputtering have been investigated. Physical and chemical properties of silicon nitride films prepared by sputtering were examined in reference to the process variables. The rate of deposition was aDDroximately proportional to the square root of rf power density. Film density increased with power density, and decreased with gas pressure. Optimum gas pressure was in the range 3-10 x 10 -3 Torr. Higher pressures resulted in less dense and electronically leaky films. Dielectric constant and etch rate in HF solution appeared to correlate well with film density. Residual pressure higher than I-2 x 10 -6 Torr was found to have a most detrimental effect. Argon and nitrogen gas mixtures resulted in excess silicon incorporation in the deposited nitride films.Silicon dioxide has for the past several years been the most important material to serve the multiple function of insulation, passivation, and diffusion masking. Because of its chemical nature and its structural characteristics, there are a number of shortcomings inherent in silicon dioxide. The most serious of these is its low resistance to diffusion of certain impurities and to field migration of ions, which causes instability and deterioration of surface characteristics of electronic devices, and, in particular, field effect devices. From the properties of crystalline silicon nitride one may extrapolate to the generally superior properties of amorphous silicon nitride compared to silicon dioxide, except for higher electronic conductivity.Precisely because of its desirable high resistance to the diffusion of atomic, molecular, or ionic species, silicon nitride cannot be grown on a silicon substrate by thermal nitridation, either in nitrogen or in anhydrous ammonia, in contrast to the thermal oxidation of silicon in oxygen and in steam. Experiments with thermal nitridation (1) and plasma anodization (2) failed to produce coherent silicon nitride films on silicon substrates. In addition, thermal nitridation or oxidation has a limitation on the scope of its use to silicon only. A deposition method thus seems to be a more logical approach.Sterling and Swann (3) and Doo et al. (4) have succeeded in the deposition of continuous amorphous silicon nitride films by means of vapor phase reaction of silane and anhydrous ammonia. At the same time the authors and Pennebaker (5) also succeeded in obtaining continuous silicon nitride films, properly identified as silicon nitride (dielectric constant 6.0-7.0; IR absorption band at 11.5-12~) by means of d-c and rf reactive sputtering, respectively. Recently, others have described various processes for depositing silicon nitride (6).Sputtering deposition of insulating-passivating films over semiconductors is a process which involves more difficulties than the sputtering of metal films. Some of these are: (i) most insulating films are chemical compo...

No abstract

No abstract

Multivariate statistical process control is the continuation and development of unitary statistical process control. Most multivariate statistical quality control charts are usually used (in manufacturing and service industries) to determine whether a process is performing as intended or if there are some unnatural causes of variation upon an overall statistics. Once the control chart detects out-of-control signals, one difficulty encountered with multivariate control charts is the interpretation of an out-of-control signal. That is, we have to determine whether one or more or a combination of variables is responsible for the abnormal signal. A novel approach for diagnosing the out-of-control signals in the multivariate process is described in this paper. The proposed methodology uses the optimized support vector machines (support vector machine classification based on genetic algorithm) to recognize set of subclasses of multivariate abnormal patters, identify the responsible variable(s) on the occurrence of abnormal pattern. Multiple sets of experiments are used to verify this model. The performance of the proposed approach demonstrates that this model can accurately classify the source(s) of out-of-control signal and even outperforms the conventional multivariate control scheme.