Oxford ICP-CVD SOP.pdf - 第11页
Page 11 of 11 * For Staff: Process chec ks: o Using 1x1” “den se mask ” pattern, measure etch rate of SiO2 and SiN x – week ly o Using full- wafer “dense mask” pa ttern, m easure unif ormity of S iO2 and SiN x – monthl…
Page 9 of 11
Appendix B: Explanation of the cleaning recipe
6)
Plasma Steps:
a) There are two CF4/N2O steps designed to chemically etch the deposited SiO2/SiNx: a high-
pressure and a low-pressure step
b) The high-pressure step is designed to remove the bulk of the deposited material, using a
relatively low ion-bombardment strength
c) The low-pressure step is designed to remove any remaining deposited material, using a high
ion-bombardment strength
7) Pump/purge steps
a) There is a set of 50 pump/purge cycles after these plasma steps, designed to induce flaking of
any thick material that was only partially-attacked during the plasma steps.
b) If the most recent deposition was <300nm thick, the potential for flaking during your process is
small (this grows at >1um total thickness), and it is likely “safe” to skip these pump/purge steps
8) To skip the pump/purge steps, click “stop” during the cleaning process, assuming it is in the middle
of the pump/purge steps
9) Make sure to allow the tool to unload the process wafer into the loadlock before starting your run
Version history
Draft
Date
Author
Notes on changes
v.0.1
3/23/17
Eric Mills
V 0.2
5/24/17
Eric Mills
More images
V 0.3
6/4/17
Eric Mills
Appendix B
V 0.4
4/3/18
Eric Mills
Format Changes
Page 10 of 11
Page 11 of 11
* For Staff:
Process checks:
o Using 1x1” “dense mask” pattern, measure etch rate of SiO2 and SiNx – weekly
o Using full-wafer “dense mask” pattern, measure uniformity of SiO2 and SiNx –
monthly
Maintenance:
o Run 30 minute “Clean SF6+O2” recipe - weekly
Known issues (2/21/17)
O2 plasmas may not strike below 40mTorr; this seems specific to O2
though