Deep Level Impurities (fwd)

Jim McVittie mcvittie at cis.Stanford.EDU
Wed Nov 3 13:22:21 PST 2004


THis is the reply I got from Baylor regarding deep level metal impurities 
such as Ni.


---------- Forwarded message ----------
Date: Mon, 20 Sep 2004 19:14:19 -1000
From: Baylor Triplett <baylortriplett at>
To: Jim McVittie <mcvittie at cis.Stanford.EDU>
Subject: Re: Deep Level Impurities

     Deep levels are a problem...if they get into the wafer or more 
precisely    the channel. But, since the channel has gotten very shallow 
and CMOS processing has gotten very good....these impurities don't 
matter nearly as much as they used to. Just make sure the BACK side of 
the wafer is protected by oxide or nitride (preferably the latter) and 
keep the contamination away from wafers undergoing high temperature 
processing like RTA anneals or dopant activation. This means cleaning 
the wafers before and after the sputtering step.
      By the way, Ni is a known problem if it gets into the wafer...but 
for Stanford yields...who cares whether the yield of research devices is 
85% or 95% 
      As far as Mn goes, I don't know whether it has a deep level or not 
but since it has d-band magnetism, I assume it does and thus is likely 
to be a problem like Cu, Ni, and Fe. Titanium is the only d band 
materials (with 0 d electrons) completely free of this problem.
Jim McVittie wrote:

>Hi Baylor,
>We are planing on letting Ni and Co into our new sputtering tool. In 
>addition, I just got a request to deposit Al with a few % Mn. According to 
>Sze's book, Ni, Co and Mn all deep levels very near the mid-gap of Si. We 
>know that Ni and Co silcides are commonly used with CMOS. Do you know why 
>their deep levels are not a problem in CMOS? And does Mn behave the same 
>as Ni and Co.
>	Thanks,  Jim

More information about the specmat mailing list