Titanium Silicide - why it takes 2-steps anneal?

[balutan]

For Titanium silicide, why does it require 2-step anneal instead of 1? I know the 1st anneal is at ~600 degree C to form the silicide and the 2nd at ~800 degree C to further lower the resistivity. But why can't it be done in just the 2nd step? I was told it has something to do with the weird nature of titanium. Can some1 tell me why? thanks.

 

[balutan]

NVM, I finally found out why.

 

[melone]

Well............ You can't tease us like that!

 

[nbucska]

Looks like he can....

<nbucska@pc33peripherals.com> omit 33 Use subj: ENG-TIPS
Plesae read FAQ240-1032

 

[balutan]

There are two TiSi2: C54 (lower resistivity) and C49 (higher resistivity). In order to form C54 TiSi2, it has to go thru the intermediate C49 TiSi2 phase. C54 forms at higher temperature (800 degree C) while C49 forms at lower temperature (600 degree C).

SO WHY 2-STEP ANNEAL??
For one, using one-step anneal might introduce Bridging.
Bridging here refers to the formation of Titanium Silicide between the gate and the source/drain. During the anneal process, Silicon atoms will diffuse thru the TiSi2 layer to form new TiSi2 layer. This silicide growth is isotropic. The lateral growth is something that we do not want. Combination of high temperature and long anneal time will promote growth of the TiSi2. Eventually the TiSi2 at the gate will make contact with the TiSi2 at the source/drain, causing a short. With the 2-step anneal, First anneal is done at lower temperature which diffusion constant is lower. When enough C49 is formed, the unreacted Titanium is etched away and thus no more Titanium to form bridging during the second anneal. Bridging can be avoided.

Hope you all like my explaination.