In the first two columns in this series, the author presented two critical areas of the PCB fabrication process thought to contribute to the mouse bite and pitting defects seen in productionat a fabrication facility. In those first two parts, photoresist lamination and exposure parameters were investigated as to the possible root cause of the defects.
Since the original assessment determined correctly that the mouse bites and pits had different origins, the troubleshooting effort focused on the potential for air or gas bubbles in the acid copper plating tanks (Part 2).
While corrections (with positive results) were made in the acid copper plating tank filtration, the issue of pitting down to the laminate or electroless copper was not completely eliminated. As stated previously, the developing operation was scrutinized as a possible contributing factor.
Developing of Negative-Acting Resist
Unexposed resist is selectively removed through the chemical and mechanical actions of a sprayed carbonate solution (a.k.a., the developing solution). The addition of a foam control agent may be required. “Time to clean” and “break point” (i.e., the percentage of useful development chamber length traveled by the board to achieve the visible removal of unexposed resist) are characteristic data that is generated in the development process.
A water rinse removes residual unexposed, partially polymerized resist and developer solution. The remaining exposed resist is “fixed” by stopping the development action through rapid removal of the developer solution in the water rinse. This is a critical and often overlooked step. Any residual developer solution that remains and is not adequately rinsed away will potentially lead to sidewall damage as well as contaminate the copper surface.
This in turn will lead to pitting and other plating anomalies. One idea worth exploring (to ensure development is stopped) is to rinse after develop with hard or acidified water. This action turns the binder molecules within the resist to insoluble entities. A final drying step removes residual moisture to harden the exposed resist for better survival in etching or plating solutions. An example of what can happen when resist residue is not effectively removed is shown in Figure 1.
Solving the Problem
At this point the troubleshooting team moved to the developing area in order to ascertain what was required to solve the residue problem shown in Figure 1. The team looked at several plausible areas including:
• Developer concentration
• Break point
• Rinsing time and rinsing effectiveness
• Spray pressure in developer
In Table 1, an extensive list of developer variables is shown along with the respective process effects. However, the four areas listed above showed the most promise as to the root cause of the pitting defect.
One of the first areas studied was the actual spray pressure within the developer chamber. It was recommended by the resist supplier that the minimum spray pressure be 25-35 psi. During the routine check, the spray pressure was reading less than 20 psi.
Upon further examination by the engineering team, it was discovered that some of the spray nozzles were plugged with resist residues and others were badly worn from extended use. These two conditions reduced the effective spray pressure.
A second area that needed attention was the seemingly poor ability to rinse away the solubilized unexposed resist during development. It was also discovered that while the fabricator was using city water for post-development rinsing, that incoming water was too soft to ensure effective removal of resist binder salts. A word to the wise: Use of soft water for rinsing after developing is not as effective as hard water. It was decided that to improve the efficiency of post-developer rinsing, the engineers added 200 ppm of calcium carbonate to the rinsing solution. This made a marked difference in the quality of the rinsing and the subsequent plating quality.
In Figure 2, one can see the remaining copper that was not etched away. This illustrates that resist residue remained in the areas that should no longer have resist. The remaining residues—most likely from incomplete development—acted to prevent complete removal of the copper during the develop, etch, and strip (DES) process.
Based on the changes discussed above, the fabricator solved the mouse bite and pitting issues. However, one can see from this case study that there were several factors that contributed to the defects. From this study it is quite clear that most defects are not caused by just one process issue. It is usually a combination of factors that must be vetted properly. From there, the experienced troubleshooter can make corrections to the processes and eliminate loss of yield.
“Pits and Mouse Bite Issues, Part 2,” PCB007 Magazine, January 2018, p.90.
Michael Carano is VP of technology and business development for RBP Chemical Technology. To reach Carano, or read past columns, click here.
This column originally appeared in the March 2018 issue of PCB007 Magazine, click here.