The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Modeling >> Semiconductor Devices >> Strain & mobility
https://designers-guide.org/forum/YaBB.pl?num=1178685786

Message started by mg777 on May 8^th, 2007, 9:43pm

Title: Strain & mobility
Post by mg777 on May 8^th, 2007, 9:43pm

In sub-90 nm CMOS processes, mobility is incerased by creating channel stress - tensile for n-channel and compressive for p-channel. It's interesting that electrons and holes need opposite types of stress. I was trying to figure out this behavior intuitively.

Electrons are in the conduction band and their mobility is limited by atomic scattering. So further apart the atoms less the scattering. Now why doesn't this same logic hold for holes? Because holes are in the valence band, more or less orbiting the silicon atoms. The valence band is formed because of Si-Si covalent bonds whereby valence electrons can tunnel from atom to atom, forming propagating wave functions in the periodic potential of the lattice. Ergo, the closer the atoms the easier the atom-hopping.

So, electrons and holes are indeed very different beasts.

M.G.Rajan
www.eecalc.com

Title: Re: Strain & mobility
Post by Croaker on May 17^th, 2007, 6:46am

Good explanation! :)

Title: Re: Strain & mobility
Post by krishnap on May 22^nd, 2007, 6:03am

Hi Rajan,
Good concept.
Usually how the tensile and compressive stress are formed?
By varying the doping concentration?

Title: Re: Strain & mobility
Post by Geoffrey_Coram on May 23^rd, 2007, 4:49am

krishnap wrote on May 22^nd, 2007, 6:03am:

Usually how the tensile and compressive stress are formed?
By varying the doping concentration?

No. The "STI stress effect" is a compressive stress caused by the (shallow trench isolation) oxide around the device. LOCOS (local oxidation of silicon) also causes it, because the oxide takes up more room than the silicon it was oxidized from, but most foundries switched from locos to sti before the device sizes got small enough that anyone noticed the effect.

Lattice mismatch is another way to get stress: grow your Si device on Ge or SiGe. I think there's also something in the literature about the gate oxide (or nitride?) applying a stress.

Title: Re: Strain & mobility
Post by krishnap on May 23^rd, 2007, 6:56am

Hi Geoffrey, Thanks for the explanation