SiC operation at high temperatures

[ZeroFunGame]

TL;DR Summary

What is fundamentally preventing SiC BJTs and JFETs devices from operating at 1000C?

SiC BJT can operate at 500C and SiC JFET has shown to operate at 800C, due in part to their lack of a dielectric. What is fundamentally preventing SiC BJTs and JFETs devices from operating at 1000C?

 

[berkeman]

Links please. Thanks.

 

[ZeroFunGame]

BJT @500C: https://iopscience.iop.org/article/10.1088/1361-6641/aa59a7

 

[mfb]

The same as for other semiconductors, and the publication you linked discusses this:

The advantage of a wide energy bandgap is that the intrinsic concentration is much lower, and therefore the intrinsic temperature is much higher. The intrinsic temperature can be calculated as the temperature at which the intrinsic concentration is equal to the lowest doping concentration in the semiconductor device

This temperature is not a hard upper limit but the behavior of the transistor will change in that range.

You also have to design all mechanical parts to withstand these temperatures, but with a custom design this shouldn't be what limits the temperature.

 

[ZeroFunGame]

Thanks mfb! What is preventing the continual doping to increase the intrinsic temperature then? Seems like the hotter the environment, the higher the intrinsic concentration, then just increase the doping to offset reaching the intrinsic temperature?

 

[mfb]

Strong doping comes with its own disadvantages but I don't remember what exactly now.

 

[ZeroFunGame]

I guess it would just become more and more conducting, which leads to further leakage and higher junction temperatures and power dissipation reducing device reliability. Just an initial guess.

Thinking about this some more, i suspect that the ohmic/shottky contacts between metal and SiC will degrade as well, but looking for further feedback from the community regarding good resources for this topic! Thanks!

 

[Vagn]

I guess it would just become more and more conducting, which leads to further leakage and higher junction temperatures and power dissipation reducing device reliability. Just an initial guess.

Thinking about this some more, i suspect that the ohmic/shottky contacts between metal and SiC will degrade as well, but looking for further feedback from the community regarding good resources for this topic! Thanks!

At higher doping concentrations you typically get reduced mobilities due to ionized impurity scattering.

 

[ZeroFunGame]

At higher doping concentrations you typically get reduced mobilities due to ionized impurity scattering.

What's the minimum mobility needed for digital computation?

 

[ZeroFunGame]

At higher doping concentrations you typically get reduced mobilities due to ionized impurity scattering.

Also, would there be an appropriate ratio of doping concentration vs intrinsic carrier concentration that would be appropriate for IC applications?