Absent a standard, fabs feel the ups and downs of surface roughness.

In these columns, I try to touch on a topic that gets frequent questions, either from colleagues, customers or OEMs. This month, surface roughness seems to be dominating. Surface roughness is a difficult topic because there is no standardized desired roughness level. Also, the desired roughness prior to solder mask is different from the desired roughness prior to surface finishing. It should be noted that surface profilometry, the tool used to measure surface roughness, is not available at fabricators.

One of the measurements we use for comparing different incoming copper surfaces prior to final finishing is RSAR, or roughness surface area ratio. It is calculated by:

(Measured surface roughness/Area of a flat plane of same xy dimensions) – 1

High surface roughness can contribute to conditions such as poor solderability, poor tarnish resistance or poor wire bonding. Normally, these defects would not appear unless another factor is working with the roughness. I typically advise that roughness alone will not cause defects, but given high roughness, everything else in the process must be perfect. Factors that promote defects when roughness is nonuniform or high are poor cleanliness, thin deposit and poor storage conditions.
Poor copper plating process control can contribute to roughness issues or nonuniformity in the surface. Electrolytic copper plating is a leveling chemistry: Additives in the bath promote a smooth, level-plated surface with a uniform grain structure. If the chemical constituents of copper plating bath or the process parameters are not controlled, the resultant deposit will have an undesirable roughness and nonuniform grain structure. This can be a result of poor chemical control resulting in low brightener levels or poor process control, such as excessive current density during plating. When properly controlled, the electrolytic copper can level very rough surfaces created by poor drilling or other issues upstream. This is not possible with final finish chemistry.

Alternate PCB surface finishes for preserving solderability are not leveling chemistries. They contour to the underlying copper. This may be obvious for thin deposits such as organic solderability preservatives (OSP) and immersion silver, but the same is true for finishes that use electroless nickel. Electroless nickel, though substantially thicker than its competitive surface finishes, is not a leveling chemistry.

Although every surface finish line has etch chemistry, it is said to create peaks and valleys or rolling hills depending on the etch type used. The truth is that microetch chemistry, though critical to surface finish performance, is not dramatically changing the roughness of the incoming copper. It works just as stated, altering the micro topography.

The overwhelming attribute for copper roughness is created just prior to solder mask. Solder mask adhesion steps promote high surface roughness to get the solder mask to “stick.” This can include chemical etching, pumice scrubbing, mechanical brushing or the dreaded belt sanding. Without a solder mask pretreatment (Figure 1), solder mask would not adhere to the metal surface, and solder dam lifting could result. This is especially true for smaller traces and features.

When line maintenance is not controlled – brush cleaning and replacement, attention to the grit and slurry conditions of the pumice – there will be significant variability in the surface (Figure 2). Normally, this shows itself as inclusions or deep scratches in the copper.

Properly maintain the solder mask preclean treatments. Also, be aware of the pretreatment used. Many fabricators have multiple solder mask preclean types, depending on the designs they are working with.

From a surface finish standpoint, smoother is better. This is easier to clean, to apply a uniform coating (whether plated or bonded), and ultimately will produce better functional performance. The downside is smoother copper surfaces can result in solder mask adhesion issues.

Some etching chemistry suppliers will discuss a desired roughness prior to solder mask application, but this is specific to their adhesion promoters. Setting a roughness target for those using pumice or brushing is not common. As a surface finish supplier, I offer a rule-of-thumb for desired roughness before final plating, but again, this is not a standard accepted number in the industry. A proper balance between significant roughness for good solder mask adhesion and a smooth surface for quality surface finishing can be a challenge. Some have instituted a light mechanical or pumice step post-solder mask, but prior to surface to help achieve this balance. PCD&FScratching the Surface

Absent a standard, fabs feel the ups and downs of surface roughness.

In these columns, I try to touch on a topic that gets frequent questions, either from colleagues, customers or OEMs. This month, surface roughness seems to be dominating. Surface roughness is a difficult topic because there is no standardized desired roughness level. Also, the desired roughness prior to solder mask is different from the desired roughness prior to surface finishing. It should be noted that surface profilometry, the tool used to measure surface roughness, is not available at fabricators.

One of the measurements we use for comparing different incoming copper surfaces prior to final finishing is RSAR, or roughness surface area ratio. It is calculated by:

(Measured surface roughness/Area of a flat plane of same xy dimensions) – 1

High surface roughness can contribute to conditions such as poor solderability, poor tarnish resistance or poor wire bonding. Normally, these defects would not appear unless another factor is working with the roughness. I typically advise that roughness alone will not cause defects, but given high roughness, everything else in the process must be perfect. Factors that promote defects when roughness is nonuniform or high are poor cleanliness, thin deposit and poor storage conditions.

Poor copper plating process control can contribute to roughness issues or nonuniformity in the surface. Electrolytic copper plating is a leveling chemistry: Additives in the bath promote a smooth, level-plated surface with a uniform grain structure. If the chemical constituents of copper plating bath or the process parameters are not controlled, the resultant deposit will have an undesirable roughness and nonuniform grain structure. This can be a result of poor chemical control resulting in low brightener levels or poor process control, such as excessive current density during plating. When properly controlled, the electrolytic copper can level very rough surfaces created by poor drilling or other issues upstream. This is not possible with final finish chemistry.

Alternate PCB surface finishes for preserving solderability are not leveling chemistries. They contour to the underlying copper. This may be obvious for thin deposits such as organic solderability preservatives (OSP) and immersion silver, but the same is true for finishes that use electroless nickel. Electroless nickel, though substantially thicker than its competitive surface finishes, is not a leveling chemistry.

Although every surface finish line has etch chemistry, it is said to create peaks and valleys or rolling hills depending on the etch type used. The truth is that microetch chemistry, though critical to surface finish performance, is not dramatically changing the roughness of the incoming copper. It works just as stated, altering the micro topography.

The overwhelming attribute for copper roughness is created just prior to solder mask. Solder mask adhesion steps promote high surface roughness to get the solder mask to “stick.” This can include chemical etching, pumice scrubbing, mechanical brushing or the dreaded belt sanding. Without a solder mask pretreatment (Figure 1), solder mask would not adhere to the metal surface, and solder dam lifting could result. This is especially true for smaller traces and features.

When line maintenance is not controlled – brush cleaning and replacement, attention to the grit and slurry conditions of the pumice – there will be significant variability in the surface (Figure 2). Normally, this shows itself as inclusions or deep scratches in the copper.

Properly maintain the solder mask preclean treatments. Also, be aware of the pretreatment used. Many fabricators have multiple solder mask preclean types, depending on the designs they are working with.

From a surface finish standpoint, smoother is better. This is easier to clean, to apply a uniform coating (whether plated or bonded), and ultimately will produce better functional performance. The downside is smoother copper surfaces can result in solder mask adhesion issues.

Some etching chemistry suppliers will discuss a desired roughness prior to solder mask application, but this is specific to their adhesion promoters. Setting a roughness target for those using pumice or brushing is not common. As a surface finish supplier, I offer a rule-of-thumb for desired roughness before final plating, but again, this is not a standard accepted number in the industry. A proper balance between significant roughness for good solder mask adhesion and a smooth surface for quality surface finishing can be a challenge. Some have instituted a light mechanical or pumice step post-solder mask, but prior to surface to help achieve this balance.

Lenora Toscano is final finish product manager at MacDermid (macdermid.com); This e-mail address is being protected from spambots. You need JavaScript enabled to view it .