Millipore Testing – Verification

Millipore testing is a very popular method of measuring and verifying part cleanliness.  The procedure is described in the blog Millipore Testing.  Although the procedures for Millipore testing are usually very well written and precise, there is still occasionally reason to question the results.  So how do you make sure that the numbers are good?

Since Millipore testing may be employed using gravimetric or particle size and distribution as a criteria for cleanliness the procedures and methodology to verify test results may vary but, in general, the following will apply.

If the numbers are higher than expected, obviously, the excess must have come from someplace.  Here’s where to look  –

• Is the solvent being used to harvest particles pure and well filtered?  The purity of solvent may vary from batch to batch or the solvent filtration may not have been adequate.  This may be due to a compromised filter.  For example, using the wrong size filter or a filter that it not properly mounted or even missing.
• The equipment being used to apply and collect the wash solvent must be, at least, cleaner than the parts being tested.  Any new piece of equipment may need to be cleaned several times to achieve the required level of cleanliness for use in the testing procedure.
• In the case of gravimetric analysis, proper drying of the patch prior to weighing must be verified.  Check the oven temperature and drying time.

If the numbers are higher than expected, verification of the validity of the Millipore testing is relatively easy and logical.  The first step, of course, is to make sure that the procedure has been performed exactly as specified.  It is easy to miss the details especially if there is a change in personnel or equipment.  Once this has been determined, the next step is to go through the motions of testing but without a part.  Since the part is the intended source of the contaminant being measured, the result should show no contamination at all!  In reality, this probably won’t be the case, but, at the very least, the number should be significantly less than the specification allows.  If the number is not significantly less then the specification allows, it is wise to conduct another “dummy” test to make sure that the result is consistent.  If it is, it is possible that the specification or collection procedure is not appropriate or that the source of contamination is in the collection procedure.

If the numbers are significantly lower than expected (based on historical trend), there may also be reason to question their validity.  Either the parts are truly cleaner than they were previously (a good thing) or for some reason there are contaminants that are not being removed in the collection process.  Things to look for here are –

• Low spray pressure
• Inadequate collection from all surfaces of the part
• Improper demagnetization of the part allowing particles to held by magnetism

Lower numbers might be a bit more difficult to verify than higher numbers as discussed above.

If the “dummy” test produces a result that is significantly lower than that allowed by the specification the next step is to conduct a test on a part from a lot that has been proven to meet the specification.  Another way to proceed would be to test a single part repeatedly.  With each repetition, the measured level of contamination should trend lower.  This is because at least some of the contamination that was originally on the part should have  been removed by the previous testing procedure.