Maintaining OSP Coating Integrity During the Cleaning Process

Estimated reading time: 6 minutes

Editor's Note: This article originally appeared in the September 2012 issue of SMT Magazine. During the PCB manufacturing process, exposed copper traces must be treated with an anti-corrosion coating. Traditionally, it is coated with solder by hot air solder leveling (HASL). The HASL finish prevents oxidation of the underlying copper and therefore guarantees a solderable surface. Although this is a sound approach, organic solderability preservatives (OSPs) have become prevalent in the electronics industry as an alternative to the HASL process. Even though this is a reliable surface finish method, care must be taken during the SMT process to minimize surface thickness degradation, particularly due to reflow and cleaning.  Although OSP film integrity can be affected by various handling and processing procedures, this study focused on the potential removal of OSP through the use of a chemically assisted cleaning process. Many equipment options are available for cleaning PCBs; however, for purposes of this study, a spray-in-air inline cleaning system was selected. Additionally, two different cleaning agents varying in alkalinity values were used. Variables monitored for their effect on the OSP film thickness included concentration and temperature of cleaning agent and conveyor belt speed, i.e., time of exposure to cleaning agent. Finally, OSP coupons identified for this test were sourced from two different manufacturers. Film thickness was measured prior to processing to confirm a baseline thickness. The coupons were reflowed using a lead-free profile. Subsequently, the film thickness was rechecked prior to cleaning to confirm any OSP change resulting from reflow as well as to re-define the baseline film thickness prior to cleaning. The results of this study detail the effect of chemically assisted cleaning on OSP film thickness and quantify the impact of key variables.   Introduction In general, thin film coatings are used to increase solderability as well as act as an oxidation inhibitor for the copper leads. While OSP is viewed as a viable alternative to the HASL process, thinning of its protective coating can occur at various stages throughout the SMT process and particularly through reflow and cleaning. Today, a typical cleaning process involves the use of water-based organic chemicals in a spray-in-air system. This study focused first on determining the effect of reflow and secondly, the effect of the cleaning agent on OSP thinning, irrespective of solder pastes. Compatibility of the cleaning process (cleaning agent + process parameters) with the OSP layer will determine flexibility of the cleaning process in terms of effectively removing various types of flux residues.    A design of experiment (DOE) was developed to explore the main factors affecting an OSP layer, including reflow and cleaning. Additionally, this study focused on optimizing the cleaning process in order to maintain OSP integrity. Conclusions were drawn based on changes in the OSP film thickness.      Design of Experiment   The main goal of this study was to determine the effect of the cleaning agent on OSP thinning. Two cleaning agents were selected and are identified as follows: Cleaning Agent 1: Traditional surfactant based - Several surfactant molecules are required to bond to one residue molecule.   Cleaning Agent 2: Dynamic surfactant based - Branch-like structure requiring fewer active ingredient molecules to bond to multiple residue molecules. Also, it should be noted that Cleaning Agent 2 includes inhibitors with the same functional group as the OSP coating. Thus, the authors hypothesized that the effect on OSP layer due to the cleaning agent itself would be minimal. Test vehicles used were high-temperature OSP-coated copper coupons and were sourced from two different vendors. The DOE developed included 10 variables. Given the interest in determining the most influential variable, the authors decided to use a full factorial matrix design. Therefore, a total of 36 tests were conducted (Table 1). Table 1: Test variables. Test evaluation methodology and technique:

• OSP thickness measurements were taken using a Filmetrics F-42OSP measurement system [1]. Film thickness was determined by taking three measurements from each coupon at various locations and averaging the values for use in the thickness analysis.
• The OSP layer thickness was measured before and after reflow to determine the effect of reflow. Since all OSP coupons were reflowed for this DOE, post-reflow film thickness was used as the baseline for analysis following the cleaning process.
• Test results were analyzed using Minitab Statistical Software generating a quantitative comparison of all variables with main effects and interaction plots (Figures 1-4). 

Test protocol:

• The OSP coupons were reflowed using a typical lead-free profile prior to cleaning to simulate a worst-case scenario. The peak temperature achieved was 245°F.
• All coupons were cleaned in a spray-in-air inline cleaner.
• Each trial within the DOE matrix was conducted using two coupons from each vendor for test repeatability of results.

The following cleaning equipment process parameters were maintained constant throughout the DOE (Table 2).

Table 2: Cleaning process. Results and Analysis The effect of reflow on OSP layer:

• The OSP-coated coupons from the two different vendors yielded dramatically different results with regard to film thinning following reflow. The OSP layer was thinned by 3.67% on Vendor A coupons and 47.47% on Vendor B coupons. Thus the OSP vendor must be carefully considered and their product evaluated in order to minimize the OSP thinning after reflow.
• Since all the coupons were reflowed for this DOE, the OSP thickness post-reflow was used as the baseline for judging the impact of the cleaning process post-reflow.

Vendor A Coupon Analysis Figure 1: Main effects plot; Vendor A coupon. For this OSP type, the main effects plot clearly indicated that cleaning agent selection had a major impact on film reduction. Dynamic surfactant is clearly the preferred cleaning agent.  Additionally, the OSP layer was negatively impacted as the concentration and temperature were increased and the belt speed was decreased.  Best results were achieved with the following process settings (Figure 1):

• Concentration: 10%.
• Temperature: 120°F.
• Belt Speed: 3 ft./min. (1 min. exposure time).

Figure 2: Interaction plot; Vendor A coupon. The interaction plot (Figure 2) reinforces that the dynamic surfactant is the preferred cleaning agent. The concentration of the dynamic surfactant has minimal impact on the OSP coating. This is critical since it allows for flexibility in selecting the concentration level based on the difficulty of solubilizing the flux residues.  Additionally, this plot demonstrates that lower temperatures and faster belt speeds (shorter wash time) reduce the impact on the OSP layer. However, even the most aggressive parameters, i.e., 15% concentration, 150°F wash temperature and 1ft./min. belt speed, resulted in a significantly lower percentage of OSP reduction with the dynamic surfactant as compared to the traditional surfactant cleaner.  Vendor B Coupon Analysis Once again, the main effects plot indicates that cleaning agent selection had a major impact on film thickness reduction and the dynamic surfactant was clearly the preferred cleaning agent. With regard to the remaining variables, there was no significant impact on the OSP layer (Figures 3 and 4).  However, it should be noted that with this coupon, the film integrity was greatly affected by the reflow process itself.   Figure 3: Main effects plot; Vendor B coupon.Figure 4: Interaction plot; Vendor B coupon. Conclusion As demonstrated through this DOE, not all OSP finishes are the same and film thickness can be greatly affected by reflow and cleaning processes. Vendor A's OSP-coated coupons resulted in far less film thinning across all parameters. With regard to OSP thinning resulting from the cleaning agent, the authors' original hypothesis proved true. Cleaning agent type proved to be most critical and clearly the dynamic surfactant based cleaning agent was preferred. Thus, when properly selected, the cleaning agent can minimize the effect on OSP stability while offering a wide process window with respect to concentration, temperature and exposure time. This is a critical consideration for the electronics manufacturer. Reference: 1. "Optical Reflectivity as a Nondestructive Measurement Technique for OSP Coating Thickness on Production PCBs," CircuiTree.com, March, 2008. Umut Tosun, M.S. Chemical Engineering, is the application technology manager at ZESTRON America. Questions and comments can be addressed to u.tosun@zestronusa.com. Naveen Ravindran, M.S. Chemical Engineering, is an application engineer at ZESTRON America. Questions and comments can be addressed to n.ravindran@zestronusa.com. Michael McCutchen, M.A. Chemistry, is the vice president of ZESTRON Americas and South Asia. Questions and comments can be addressed to m.mccutchen@zestronusa.com.