<berndj>
azonenberg, i don't get why *high*-k dielectrics are good for gate insulation.  doesn't that increase the gate capacitance?
<azonenberg>
berndj: but it also decreases leakage
<azonenberg>
i think
<azonenberg>
and lets you make the dielectric thinner
<azonenberg>
by contrast you want low-K for ILD
<azonenberg>
so you dont have capacitive coupling between wires etc
<azonenberg>
or wire capacitance getting in the way as much
<azonenberg>
berndj: i think it also lets you make the gate smaller
<berndj>
ILD?
<azonenberg>
inter-layer dielectric
<berndj>
okay so it's more like it *happens* to have high-K, but that isn't the reason you use it for gate insulation: instead you're using it for the higher dielectric strength it also has?
<azonenberg>
They're the same thing
<azonenberg>
higher dielectric strength will increase capacitance
<berndj>
err, correlated, i'm sure, but not the same thing?
<azonenberg>
And doing a little bit of reading it seems that really thin gate dielectrics have troubles with leakage
<berndj>
SF6 for example wouldn't follow that correlation too well
<azonenberg>
So the way to compensate is to use a *thicker* dielectric
<azonenberg>
but use high-K to get the same capacitance
<berndj>
but you *don't* want capacitance in a gate, surely?
<azonenberg>
I think you do
<azonenberg>
the higher the capacitance per unit area, the more charge you can put across it and thus the lower your on resistance
<berndj>
this is getting confusing :(
<azonenberg>
Yeah
<azonenberg>
Transistors are something i have not studied much
<azonenberg>
my focus has been MEMS so far
<berndj>
what you said there re charge makes sense, but i thought it's the E field that inverts the polarity of the channel
<azonenberg>
no idea, again
<azonenberg>
this is something i've been meaning to study for a while