Reading time ( words)
While Harald Wack, Ph.D., ZESTRON, is encouraged by the numerous pathways of innovation that have developed to meet new cleaning challenges, he is discouraged by one in particular. Recently, Wack witnessed the promotion of Hansen solubility parameters as a means to determine the performance of cleaning agents and processes.
In recent years, the precision cleaning market has gained tremendous momentum, due to many factors that influence the need for cleaning. The move to lead-free solder pastes and the corresponding hotter reflow profiles, use of low standoff devices, and miniaturization have raised the bar for cleaning performance when reliability is paramount. As a result, equipment and chemical suppliers are driven to innovate like never before. The resulting aqueous-based cleaning processes have improved cleanliness levels and lowered overall cleaning costs while proving to be safe, environmentally friendly alternatives to traditional solvent-based cleaning agents. Now, virtually all processes are approximately 10% cleaning agent and the remaining 90% is DI-water. You may be asking, “What does this have to do with Hansen-Hildebrand solubility parameters?”
The answer is simple. While I am encouraged by the numerous pathways of innovation that have developed to meet the new challenges, I am discouraged by one in particular. Recently, I have witnessed the promotion of Hansen solubility parameters as a means to determine cleaning agent and process performance. In one recent case, a contract manufacturer incorporated this method into his decision process only to find that residues were virtually untouched by the cleaning agents. Does this mean I am criticizing Hansen-Hildebrand directly or that I don’t see the value in the method? Absolutely not.
As an organic chemist, I have frequently used Hansen parameters for solvent selection and I advocate this method. In fact, the work produced by Drs. Hansen and Hildebrand changed the way researchers of all kinds approach unanswered questions about solubility and solvent systems. To understand my frustration over the use of this method in aqueous cleaning process selections, I will provide some background information.
The selection of solvents and/or cleaning agents for particular situations involves many factors, including evaporation rate, density, solution viscosity, etc. Also, their effectiveness depends on their ability to dissolve one material while leaving other materials unaffected. The selection of solvents or solvent blends to satisfy such criteria was at the time — and in some respects remains — an art. However, in the 1960s, Drs. Hansen and Hildebrand began to define and quantify solubility using an organized system, which facilitated the accurate prediction of complex solubility behavior and provided explanations for such rules of thumb as “like dissolves like.” While detailed discussion of this topic is far beyond the scope of this column, for our purposes, it is important to understand that the method is based on well-controlled experiments. Most published solubility data are derived from fixed (10%) concentrations of the solute (residue) and room-temperature conditions. The utility in applications where the conditions are not similarly controlled and change over time can be problematic. In our industry, production results in a loaded wash solution with countless variables and demands upon cleaning agents that, at the very least, require one to fully understand the limitations of applying such methods. Otherwise, as in the previous case I mentioned, valuable time and energy may be lost in the academic pursuit of practical solutions.
Among the prominent problems with the use of this method in our industry is that we are, in most cases, dealing with aqueous systems where approximately 90% of the cleaning solution is water. Water, with its ionic character and propensity to form hydrogen bonds, is problematic. In fact, the presence of water in a solvent blend can dramatically alter the accuracy of solubility predictions. Hydrogen bonding alone can change the dipole of certain solvents to such an extent that the initial solubility prediction based on the Hansen parameters assigned to the pure solvent is dramatically different and the new hydrogen-bonded solvent resembles a different solvent altogether. Therefore, the initially assigned Hansen parameter and the resulting prediction are thrown out the window. Of course, there are methods that attempt to account for this, but the degree of hydrogen bonding varies greatly in complex systems such as a fully loaded (flux, organic acids, ionic species, etc.) wash solution of an in-line cleaner under production conditions. This alone makes all the effort to predict the performance of cleaning agents based on Hansen methods purely academic; on top of this, there is, unfortunately, a host of other issues. We start with temperature.
Increasing temperature in a wash solution also increases the disorder (entropy) of the solvent system, which has the effect of reducing the differences between the components’ solubility parameters. In other words, the initial solubility values that are assigned to each component matter less as the temperature increases. The initial work that was done to predict solvent system (cleaning agent blend/water) performance becomes less meaningful.
The practical application of Hansen methods also requires users to understand some fundamental properties of the solute (material to be dissolved), such as polarity. Unfortunately, flux residues are as unique as reflow profiles, due to polymerization differences and burning in at high temperatures. We know that process conditions that remove residues from reflowing paste with one profile may be unsuitable when the profile is changed. Furthermore, flux residues are not homogeneous. So, the method is inherently inaccurate unless you want to spend time analyzing each residue at the exact profile to create a cleaning agent for every situation. Why waste time when high-quality modern cleaning agents are broadly applicable and capable of cleaning even the toughest residues?
Harald Wack, Ph.D., is the president of ZESTRON and an SMT Editorial Advisory Board member. Dr. Wack has authored and published several scientific articles and has provided technical information for various publications. He received his doctoral degree in organic chemistry from Johns Hopkins University. He may be contacted at (703) 393-9880 or email@example.com. Dr. Wack recently wrote Go Greener in Your Cleaner, The Future of Cleaning OA Fluxes, What is Innovation in Chemistry?, Reduce the Cost of Cleaning Processes, and IPA-Water (75/25): Ineffective for Cleanliness Test and co-authored Cleaning No-clean Solder and Flux
SMT May, 2010