April’s Tech Talk discussed the topic of preparing a copper surface for dry film lamination with brushed or jetted pumice. A variation of a jet “pumice” machine was first offered by Ishii Hyoki (Japan). It is also referred to as “jet brushing.” A slurry of aluminum oxide (also called alumina, Al2O3) is sprayed on the board in a conveyorized spray module that looks very much like a jet pumice machine. Other suppliers of brush pumice machines then offered modifications of their pumice brush machines to accommodate aluminum oxide. Since aluminum oxide is less abrasive than pumice (the particles “wear round”), one can use plastic nozzles. The particles last longer than pumice, i.e., the operation creates much less sludge or fines. Slurry replenishment is minimal and there is less sludge waste disposal. The working Al2O3 slurry settles readily and % solids determination is quickly accomplished by sedimentation in a graduated cylinder.

We had become aware of a problem at a PWB fabrication site where aluminum oxide scrubbing was used to prepare an electroless copper surface for dry film lamination and subsequent copper electroplating. Varying degrees of underplating had been reported. We suspected that the age of the aluminum oxide may be a factor and that underplating was related to the fact that aluminum oxide particles become a less effective surface preparation media when used too long.

Enlarge this picture Fig 1 Effect of Age on the Shape of Aluminum Oxide Particles

We set up a test to correlate throughput of boards through an aluminum oxide jet scrubber, with surface roughness Ra of the electroless copper after surface preparation, and with the degree of resist lifting observed after exposing a resist pattern on such a surface to the plating bath environment.

The test panels represented changes in the surface preparation process over the course of two months’ production: the first panel had been cleaned with fresh aluminum oxide shortly after the jet nozzles had been replaced. The next three panels represented one week’s use of the aluminum oxide particles, one month, and two months respectively. The monthly throughput of the jet spray machine was roughly equivalent to 60,000 panels, corresponding to about 2,700 square meters of surface area.

Resist lifting is evident by the lighter color of the lifted resist edge, the degree of which can be measured under the microscope. Resist lifting on panels after one week and one month throughput was similar to that on panels processed with fresh alumina. But an increase in lifting was evident on panels processed with two month old alumina. We had also noticed a difference in Ra values on the first panel used in the study depending on the profilometer scan direction. Months later, when we initiated collection of panels showing the effect of age of the media, we again observed this directionality of the Ra results. We have shown the average from the profilometer scan direction that gave low Ra values and the direction that gave high Ra values (see Table 1). A large number of scans were made on each panel to enable us to test for statistically significant differences. Comparison of the 1 week panel and the 2 month panel showed that the Ra values were statistically different. Thus, after two months there was a real drop in the roughness of the surface as measured by contact profilometry. The corresponding resist lifting results showed little difference between the one week old and the one month old samples, but did show noticeably more lifting on the surface prepared with the two month old aluminum oxide (see Table 1). These findings are in good agreement with production results. In production a different dry film resist with slightly better adhesion was in use, and incidents of underplating were reported when aluminum oxide, and nozzles, were not changed within three months.

Samples of the aluminum oxide media were obtained at the time each panel was processed. Scanning Electron Microscope photos show the sharp edges of the fresh particles. By the end of two months the particles had gradually worn. The edges were no longer sharp, but had become more rounded and smooth (see Figure 1).

On a Personal Note

Advertisement

In the interest of truth and full disclosure, I find it necessary to replace my fifteen year old picture with a current one. The green shirt remains for continuity’s sake.