From cmcg at stanford.edu  Sat Jan 23 15:15:31 2010
From: cmcg at stanford.edu (Chris McGuinness)
Date: Sat, 23 Jan 2010 15:15:31 -0800 (PST)
Subject: etch rate for small features vs. entire wafer
Message-ID: <1713521115.1366311264288531059.JavaMail.root@zm02.stanford.edu>

Has anyone noticed a significant difference in etch rate for small features (400nm) compared to a blank test wafer when using the jim-ox process in chamber B?  

I've compared this twice before using the SEM to measure small feature etch rate, and nanospec for whole wafer etch rate and gotten very similar etch rates for both.  I did the comparison again recently for LTO oxide (densified and undensified) and found: 3700A/min for two blank test wafers, and 2850A/min for 400nm features.  This is almost 25% different.

If anyone else has done this analysis I would be grateful to hear what you have found.

Thanks,
Chris