Conduction Studies in Silicon Nitride: Dark Currents and Photocurrents

Abstract

Dark conduction and photoconduction phenomena in amorphous, stoichiometric Si 3N4 films incorporated into metal-silicon nitride-silicon (MNS) structures are studied in detail. Results demonstrating the importance of hole conduction in Si 3N. thin films are presented. The importance of trapped space charge injected from the contacts is shown in an analysis of the effects of Si 3N4 thickness and the choice of gate metal on conduction. Contact effects that are apparent in the current characteristics of thin Si 3N. layers become dominated by bulk conduction mechanisms as the silicon nitride film increases in thickness. Optical interference effects in photoconduction are used to determine the origin of photocarriers.

Space‐Charge‐Limited Currents at High Electric Fields

“…The calculations are carried out as described above. The expression for the field c(L) in terms of E, Q, coincides with (12). Only higher-order terms, omitted in (12), differ.…”

“…We see a strong dependence of the current on the film thickness at fixed E. This is explained by the influence of space charge, but the calculation of SCLC is made numerically in [12].…”

“…To determine the trap density, N,, we need x and P. We assume the static relative dielectric constant to be equal to 7.5 [12]. P can be determined from the slope of 1 n J -@ for maximum L, when the last term of the exponent in (18) is small.…”

“…Returning to the original symbols we obtain where we retained the leading term -E3I2 in the pre-exponential polynomial. Note that the term -B is absent in this polynomial, and the largest of the omitted terms is that -E'" in accordance with (12). The Q value in (14) depends, in general, on the injecting contact.…”

Location of Trapped Charge in Aluminum-Implanted SiO2

“…The x value decreased by approximately 12.5 nm when photoinjection from the Al-Si0 2 interface with 4.5-eV light was used to fill traps; this result was independent of the number of trapped charges per unit area in the range from 1 x 1011 to 1 X 10 12 electrons/ern", These two experimental observations are consistent with each other and imply that photodetrapping can explain the anomaly mentioned in item 2. This photodetrapping was observed to be influenced by the local fields [3,4,19] and optical interference patterns [19] and/or the light energy in the Si0 2 layer since values for x measured by the photo I-V technique varied somewhat under photoinjection conditions (injecting interface and light wavelengths used).…”