Temperature Dependence Cgs

[Bakez]

I have read in several texts the gate-source capacitance expressed in terms of Cox, W, and L.

However I am wondering what its temperature dependence could be? I assume that W and L do not change with temperature, so it must be Cox is the temperature dependent parameter, which I find written as:

εox/tox

So I assume that thickness does not change with temperature and it must be εox as the temperature dependent parameter. BUT then I read that with increasing temperature this dielectric property supposedly DECREASES with temperature, so Cgs sould also decrease? But I am sure it is the opposite

I thought these capacitances were supposed to have an exponential relationship with temperature (a professor once told me this)

So what is the deal?

 

[shar2013]

I have read in several texts the gate-source capacitance expressed in terms of Cox, W, and L.

However I am wondering what its temperature dependence could be? I assume that W and L do not change with temperature, so it must be Cox is the temperature dependent parameter, which I find written as:

εox/tox

So I assume that thickness does not change with temperature and it must be εox as the temperature dependent parameter. BUT then I read that with increasing temperature this dielectric property supposedly DECREASES with temperature, so Cgs sould also decrease? But I am sure it is the opposite

I thought these capacitances were supposed to have an exponential relationship with temperature (a professor once told me this)

So what is the deal?

We need to remember that the channel capacitance is induced by a voltage applied between the gate and the bulk (source if they are shorted) . Cgs is therefore dependent on the threshold voltage , which decrease with temperature.

Hope that helps.