High Temperature Operation of Silicon Carbide Schottky Diodes with Recoverable Avalanche Breakdown

Abstract

4H-SiC diodes with nickel silicide (Ni2Si) and molybdenum (Mo) Schottky contacts have been fabricated and characterised at temperature up to 400°C. Room temperature boron implantation has been used to form a single zone junction termination extension. Both Ni2Si and Mo diodes revealed unchanging ideality factors and barrier heights (1.45 and 1.3 eV, respectively) at temperatures up to 400°C. Soft recoverable breakdowns were observed both in Ni2Si and Mo Schottky diodes at voltages above 1450 V and 3400 V depending on the epitaxial structure used. These values are about 76% and 94% of the ideal avalanche breakdown voltages. The Ni2Si diodes revealed positive temperature coefficients of breakdown voltage at temperature up to 240°C.

Reverse bias-dependence of schottky barrier height on silicon carbide: influence of the temperature and donor concentration

“…The existing of many different curvatures predicts that the variation of the SBH with bias voltage is not the same for all diodes. Moreover, the coincidence of the experimental and calculated curves for the Ti(Ni)/4H-SiC ref 6 SD predicts no important variation in the barrier height with the bias voltage and the effective mass (m * = 0.2m) obtained (see Table 2) in this case is equal to the theoretical value. For the reasons discussed above, we assume in the following section reverse bias-dependence of Schottky barrier height and we use the previously extraction method discussed in the section 2.…”

“…As can be seen from Fig. 10, the experimental barrier height increases linearly with increasing temperature and the slopes are higher for the low bias than for the high bias; the slope is about 1.49x10 -3 V/K at -500 V and 1.15x10 -3 V/K at -1200 V for Ti(Ni)/4H-SiC ref 6 . This increase in barrier height with increasing temperature is the same behavior at the forward bias, where the thermionic model is used.…”

“…Table 2 and the experimental data for several diodes. It can be seen from this figure that the experimental ln(I)-V plots and simulated ln(I)-V plots do not coincide over the entire bias ranges except in the case of the Ti(Ni)/4H-SiC ref 6 SD, which exhibits an almost total coincidence. The difference in the nature of the experimental ln (I)-V plots and the difference in the effective mass for the same material (4H-SiC) supports the dependence of b  on the bias voltage.…”

“…While Fig. 6b shows the barrier height as a function of the reverse bias for Ni/4H-SiC SDs at various doping concentrations (reverse I-V data are from several works [5], [6], [21], and [22]). From these figures, we can see two important observations: First, we can see that the barrier height tends to increase with an increase in the donor concentration, while, several authors [28][29][30][31] observed that the barrier height decreases with increasing donor concentration in the case of the forward bias, where the thermionic model is used.…”

“…However, due to the high electric fields normally encountered in SiC devices, the reverse leakage current of Schottky diodes can be significantly enhanced prior to junction breakdown due to the tunneling mechanism [3]. The high electric fields encountered in SiC and other wide band gap materials lead to significant tunneling through the Schottky barrier thus increasing the leakage current by many orders of magnitude over that predicted by simple thermionic emission (TE) calculations [3][4][5][6][7][8][9][10]. However, the tunneling leakage current, which calculated with the extracted SBH from forward I-V data, is discrepancy of the experimental data [3], [4], [9].…”