The test for cleanliness is challenged by numerous subjective assessments.

Solubility is a funny thing, but on the other hand easily explained. A liquid can solubilize based on its physical properties such as polarity, intermolecular interactions, temperature and other key features.

Recently, I took part in a presentation that made the case that an extraction with IPA-water at 176°C would provide superior ion chromatography results when injected (compared to room temperature). Thankfully, I have spent years in an organic chemistry lab. Despite the fact that our resources were limited at the time, we were able to find spare parts and assemble a number of chromatography instruments to further our research on asymmetric catalysis. I noticed the chromatography columns were sensitive to temperature and solvents. In other words, the extracted products were never injected onto the column at temperatures higher than room temperature. We had to make sure the solubility of our organic molecules was 100% to obtain accurate data on reaction result and yield (i.e., enantiomeric excess). Why is this important, you might wonder?

Numerous companies are investigating viable alternatives to IPA and water (currently in use for ROSE test, IPC-TM-650 methods 2.3.25.1 and 2.3.28, and MIL-P-28809 section 4.8.3). One might state that heating IPA and water temporarily increases the solubility of the contamination. However, lowering the temperature will make this reaction reversible. During a chromatography experiment, the respective results from heated extraction, therefore, will not be any better than ionic contamination measurements acquired at room temperature.

While people like stability and are generally resistant to change, we have to continue to acknowledge that our analytical assessments today are far from satisfactory or correct. Not only did we author numerous articles to raise public awareness, but we also initiated various internal projects to further the possibilities of developing a new extraction method that would not only dissolve today’s flux residues, but also produce numeric data far superior to frequently used ion chromatography and ionic contamination. (Please understand that we do support the latter methods and use them daily in our labs. My main point simply is why not look for something better?)

Related to this area of improvement, the test for cleanliness via ion chromatography is also challenged by other subjective assessments. During each and every analysis, the extracted liquid is injected onto the various columns to detect ions and cations. This results in quantitative numbers, which still require a correlation to “cleanliness.” Here subjectivity comes into play. The values are, in most cases, correlated to the “personal experiences” made by each of the analytical labs performing this test. My question is: Who is to judge whether a certain sodium (Na+), potassium (K+) or weak organic acid (ROO-) means clean or not clean? Further, I do not think there is a true relationship/correlation possible between long-term reliability and a certain ion level. Assumptions can be made, but users have to be aware that each assembly is so different in nature that one can hardly find a numerical value applicable to each user and industry.

For example, consider a highly complex populated board compared to a less populated board with widely spaced components. The risk of electrochemical migration or corrosion might be given in the first case, but the same level of ions may not harm the second board. This problem is compounded by the fact that analytical labs are hard-pressed to relate the environmental exposure/application to the individual assemblies. An assembly might not fail in a dry environment, but would fail once used under humid conditions. One can continue similar examples for some time.

Imagine for a moment the conundrum presented to the industry. While we are searching to find a better solvent than IPA/water, we also have to honestly ask what impartial (non-vendor based) method will be used to employ it, and finally, how best to find a numerical value for the ultimate cleanliness testing method that is representative and can be applied by most industry users. I assume that to overcome this dilemma, research will eventually lead to a combination of test methods that will provide the best overall estimate of the risk associated with your assembly.

Research is currently underway to provide viable, general replacement solutions that can and should be evaluated and compared to the present method. 

Harald Wack, Ph.D., is president of Zestron (zestron.com); h.wack@zestronusa.com. His column appears regularly.