I would recommend reading a text book on negative and positive resists
to understand all the aspects.

The negative and positive resists have different aspects that would make
one superior to another in different situations. For the great variety
of MEMS materials one might have advantages over the other.

0. Defect density issues. If you are exposing contacts, you want
positive resist since dirt in the field of the pattern on the mask won't
cause defects on the wafer, where as with negative resist every piece of
dirt on the mask will cause an unwanted contact. With positive resist
you are only interested in substrate and mask defects in the area of the
contacts with is a small percentage of your device, with negative resist
you would be concerned by defects in every non-Contact area which would
be a high percentage of your device.

If you were doing resist pillars, you would have an oppositive yield
strategy.

Though nowdays the objective is to have clean substrates and masks and
this type of negative-positive resist strategy for yield is generally
obsolete for semiconductors, but for your MEMS process it might be
vital.

1. Positive resists generally have higher resolution capability than a
negative resist.

2. Contacts are easier to image in positive resists.

3. Negative resists might have much better adhesion to some surfaces
than a positive resist would have and vice versa. You might not be able
to apply a HMDS vapor to some surfaces which helps positive resist
adhesion.

4. Negative resists can resist alkalies, which dissolve positive
resists.

5. Negative resists require the disposal of solvent developers.

6. Some substrates might not be able to withstand a solvent developer,
others might not withstand an alkali (positive resist) developer.

7. I don't think negative resists take plasma etching well, but I may be
out-of-date and there might be new negative resist chemistries that
resist plasma.

8. I would consider the temperature resistance of the resists for post
image processing.

9. Positive resists can be processed to have negative slopes.

10. The use of negative and positive resists together in a fab, can
result in one mask being used to make both a positive and negative
image. This can sometimes be useful.

11. Is there a negative equivalent to Fusion processing of a postive
resist?

12. On a reflecting surface, stray light causes polymerization in
negative resists and could result in linked resist in areas where it
isn't wanted. Positive resists would have thinning possibly in some
areas or lines pinched off. Depending on your design, one or the other
problem might be more tolerable. I think an array of small contacts
might be much easier in positive resist.

Also, positive resists can be dyed to suppress stray light, I don't know
if the equivalent exists in negative resists.

13. Wettability of a surface might be much different between one resist
and the other. Positive resist sometimes has real problems with
dewetting on some surfaces. I have seen problems with positive resist
being spun on high topology wafers, even though the usual tricks were
tried, such as pre-spinning with resist solvent. A negative might coat
such a surface just fine.

14. Positive resist technology has been driven by the semiconductor
industry, negative resists were dropped by the semiconductor industry
decades ago. Though there may be technological progress I don't know
off.

15. Positive resist wafer edge bead is easily removed with Edge bead
removal tools integrated into spin tools. It isn't an issue.

16. Are you doing contact, proximity, or projection exposure? I am not
sure how well negative resists work with contact exposure in terms of
sticking to the mask.

17. Is your substrate compatible with the resist stripper for the mask.
Positive resists use organic amines.

With all the varieties of substrate materials, films, and patterns that
occur in the MEMS world, I don't think either negative or positive
resist is always the better choice. The above is just a short list of
issues that readily come to mind.

What you need to look at is your specific application and consult your
resist vendor. Some resist vendors are not at all helpful and some
resist vendors are very helpful. Contact me for a brand recommendation.

Ed



-----Original Message-----
From: mems-talk-bounces@memsnet.org
[mailto:mems-talk-bounces@memsnet.org] On Behalf Of Paul Nguyen
Sent: Tuesday, October 27, 2009 10:24 AM
To: MEMS-talk@memsnet.org
Subject: [mems-talk] Positive vs. Negative Resists in MEMS processing


Dear all:

I would like to hear your experiences of using Positive vs. Negative
resists
in MEMS processing.   What are the PRO vs. CON?  I appreciate your
information.

Thanks and regards,
Paul