Dear Josh,
You pose some interesting questions regarding adhesives and the measurement
of ionics. In my estimation, you are entering into uncharted territory. I
apologize in advance for the length of the reply.
1. You asked what standards and procedures are accepted for measuring
ionics. The two industry accepted practices are ionic contamination using an
ionic cleanliness tester, such as an Omegameter, using IPC-TM-650, method
2.3.26.1 and 2.3.26.2. These are a relatively gross measure of cleanliness
and may not be sensitive enough to measure the minute quantities of ionics
available. The other accepted method is IPC-TM-650, method 2.3.28, which is
for ion chromatography (IC) testing. IC is much more sensitive and the
extraction method more complete. Depending on the sensitivity of the IC
system and the skill of the operator, you can pick up ionic impurities in the
parts per billion range.
2. These two test methods are for general purposes, measuring contaminants
on whatever you dunk in the test solution. Therefore, they could be used for
flex circuits, rigid boards, adhesives for PCBs and adhesives for chip
attach. If I had only one test to choose, it would be IC. You get much more
information and a much clearer picture.
3. In terms of acceptable levels of contamination, you must specify what
residues you are looking for on the substrate. Halides, such as chloride and
bromide, can be highly corrosive and can lead to electrochemical failure in
the long term. Other residues, such as weak organic acids, are generally
benign. You must also take into account the source of the bad residues. An
example is from rosin fluxes. RMA flux has amine hydrochloride activators,
which can be corrosive, but the rosin has an encapsulating effect.
Therefore, rosin-fluxed boards can tolerate higher levels of chloride than
can a no-clean flux board, which typically does not have such an
encapsulating effect.
4. The question of how clean is clean, or how clean is clean enough, has
been asked repeatedly in our industry. We have guidelines for our own
laboratory based on the failure analysis we do with our customers. I think
you would see such guidelines vary from lab to lab, or consultant to
consultant. To my knowledge, there are no nationally recognized standards or
pass-fail criteria for ionic contamination of adhesives, and no pass-fail
criteria for ion chromatography. In many cases, you have to take into
account the specifics of the application, the sources of the residues, and
the materials involved. Since many of the pass-fail criteria of labs will be
experience-based, most (including me) will be somewhat reluctant to broadcast
them over the Net. If you (or anyone else) has an interest, I would be happy
to discuss it (317-457-8095).
5. There are different levels of tolerance towards various residues.
Chloride, as an example, needs to be 10 times lower for hybrid applications
than for PCB applications. There are a variety of reasons for this, but most
involve the materials of construction. Hybrids generally have silver
metallizations and so electromigrate easily. Chloride, as an agent to
initiate and propogate electromigration, needs to be held to lower levels to
avoid electrochemical failure. Due to the ceramic substrates, chlorides will
also contribute to leakage currents at much lower levels than you would see
for PCBs and their organic laminates.
6. The primary ions to cause trouble vary depending on application. The
halides are usually pointed to as culprits. Chloride and fluoride are more
electroactive than bromide, and so are a greater threat. In hybrid
applications, sodium can be a problem.
7. I think you will find that most adhesives contain a chloride element or
an acidic reaction byproduct. Some adhesives are impervious to extraction
when cured, but we have seen that such adhesives have poor SIR performance
when contacting energized circuits.
Hope this has been of some help. Please call me if you have any other
questions
Douglas Pauls
Director of R&D
Contamination Studies Laboratories
[log in to unmask]
(317) 457-8095
|
