Evaluating Rinsing Effectiveness in Spray-In-Air Cleaners
Umut Tosun, M.S.Ch.E., Jigar Patel, M.S.Ch.E.
Manassas, VA, USA
Electronic assemblies manufactured with WS (Water Soluble) or RMA solder pastes are cleaned following the reflow process. No Clean solder paste is likely to be cleaned when used within high reliability applications. WS solder paste is cleaned with DI-water or with an aqueous based cleaning agent. RMA and No Clean solder pastes are cleaned with a water based engineered cleaning agent. In many cases, the cleaning equipment selected will be a spray-in-air type, either batch or inline configuration.
To effectively clean post solder flux residues, the cleaning process requires chemical, thermal and mechanical energy. Once optimized, the result is an effective and efficient cleaning process. The chemical energy is derived from the engineered cleaning agent, thermal energy from the temperature of the cleaning agent and the mechanical energy from the pump system employed within the cleaning equipment, that is, spray pressure, spray bar configuration and nozzle selection.
In general, spray-in-air systems include wash, rinse and dry cycles. Engineered cleaning agents are critical to the process. In addition to the ability to solubilize the residues, they exhibit low surface tension averaging about 30 dynes/cm whereas pure DI-water has an average surface tension of 70 dynes/cm. Given the low standoff heights of components on present day PCBs, low surface tension is required to enable the cleaning agent to penetrate underneath components for contact and solubilize the residue.
The rinse cycle is critical to the process for the rinse water must penetrate the same low standoff components in order to "rinse" or remove the residue laden cleaning agent. However, rinsing uses pure DI-water. Given the surface tension of water, how effective is this process?
This study was designed to investigate the efficiency of the rinse cycle using both a spray-in-air batch and inline
cleaner. Glass slides of various surface area were affixed to the test vehicle surface at various standoff heights to emulate low standoff components. Rinse efficiency was evaluated through visual inspection, ionic contamination and localized extraction for ion chromatography. Key variables were plotted and analyzed using Minitab(R)software.
Through this study, the authors confirmed that DI-water can effectively rinse wash solution underneath low standoff areas with spray-in-air inline and batch cleaning processes. The surface tension of the rinse water can be reduced through contact with residual wash solution and through the mechanical energy developed by the rinse process, penetrate and rinse the wash solution from under low standoff components at standoff heights as low as 1 mil.
Furthermore, this study confirmed that Ionic Contamination and Ion Chromatography (localized extraction) analyses are valid methodologies for assessing rinse effectiveness as confirmed through baseline Trials 2, 3, 5, and 6 and results detailed in Tables 10, 11, 14, and 15. The majority of aqueous based cleaning agents contain ionic constituents. Thus, if wash solution is not completely rinsed, these ionic constituents can be detected.
As one would expect, the study confirmed that as wash solution becomes flux loaded (as measured by NVR), and standoff height decreases (less than 1 mil), rinsing becomes more difficult. This is evident from the Minitab(R) results (Figures 11 and 13). Even so, both the batch and inline cleaners can effectively rinse under low standoff components when optimized. Using the Ionic Contamination and Ion Chromatography cleanliness assessment methodologies, the rinse processes can be optimized as required.
The test results presented as a result of Ionic Contamination and Ion Chromatography analyses confirmed that the Hypotheses H1 and H2 are indeed true.
This study presented challenging scenarios to assess the rinsability question. Glass slides were used representing low standoff components; however, the undercomponent's surface area of the glass slides was large, similar to that of QFPs and BGAs. Given the large surface area, thorough undercomponents rinsing was achieved. Although smaller undercomponents surface areas were not considered in this study, they should prove to be less challenging to achieve complete rinsing.
Although this study did not consider rinsing substrates with soldered components for the reasons stated in the Methodology section, the authors theorized that components which trap a large amount of wash solution such as vented BGAs, open sockets and components having cavities and capacitor sleeves, could be more challenging to rinse completely. For substrates including these types of components, and cleaned in aqueous based batch and inline cleaning processes, rinse optimization, including additional rinse time and/or mechanical energy (more spray bars in rinse section of inline cleaners) will be required.
For cases where a board has some fresh wash solution (partial or incomplete rinsing) but it is completely dried, remaining wash solution may not be corrosive and will not cause failure. But this may not be applicable for all cleaning agents. Some cleaning agents do not dry residue free and remaining residue could be ionic.
Dried loaded wash solution is ionic in nature. If loaded dried wash solution is left on the board, it can cause failure. Therefore, effective rinsing is a critical part of the cleaning process.
Initially Published in the SMTA Proceedings