Thin High Phos Electroless Nickel for 5G


With 5G rollouts around the globe, ENIG continues to hold on to first place in PCB final finishes. You would think with data published on how electroless nickel increases insertion loss, ENIG use would be declining. But there are a few reasons why the prophesized demise of ENIG has been premature.

Figure-1rev.jpgFirst, 5G introduction is just in its infancy only scratching the surface of potential speed. For the masses, today's 5G is still relatively low speed in the 10-20 GHz range. In work done by Rogers Corp, ENIG insertion/signal loss is higher than copper. However, it is a relatively small increase at 0.3 to 0.5 dB/inch over copper even up to 40 GHz. (Figure 1:  Microstrip insertion loss & rolled copper). This increase in signal loss is due to the electroless nickel in the ENIG deposit. The amount of loss will change with different RF designs, higher transmission speeds, and length of transmission lines. A more thorough explanation of insertion/signal loss 1 is covered in several papers and blogs. However, for this discussion, the bottom line is ENIG continues to work for the immediate future mostly because electroless nickel is an excellent barrier metal and it provides corrosion protection to the base metal.

Electroless nickel has between 4-13% phosphorus co-deposited with the nickel. The amount of phosphorus impacts the properties of the deposit.  Most PCB ENIG processes utilize what is referred to as a mid phos of 6-9%. This was a compromise to improve chemical resistance to help eliminate Black Pad and still provide a wide window for good coverage and good solderability. Electroless nickel does not auto catalyze on copper and requires a cleaning and Pd activation process to get complete coverage of exposed copper. This fact becomes important when we discuss electroless nickel deposit thickness in a future blog, Electroless Nickel, How Low Can You Go.

When phosphorus gets to 10-13% it’s considered a high phos electroless nickel that yields some interesting properties.Figure-2rev.jpg

  • First as shown in figure 2 (Ron Parkinson2) the magnetic properties of nickel goes to zero @ ~10.5% phosphorus.  As current flows through a fero-magnetic material like nickel, it creates a magnetic field that will disrupt current flow increasing signal loss
  • Second high phos EN provides the highest level of corrosion resistance. For 5G, protection will be key due to increase mobility and provide a reliable human interface
  • Third due to the amorphous structure high phos nickel is the most ductile and has the lowest as-plated stress. Again to provide mobility will require thinner flexible PCBs where ductility & stress of a deposit will be critical

So a high phos EN has the potential to reduce signal loss, improve protection and provide a better deposit for flexible applications. All necessary attributes for advancing 5G technology. Rogers Corporation agreed to do testing to try and quantify high phosphorus EN impact on signal loss.

The test plan measured the impact of thin high phos EN on two test vehicles, microstrip vs grounded coplanar waveguide (GCPW) and with two types of copper, rolled (RC) and low profile (LPC). We have lots of charts & graphs including the originals from Rogers showing the impact of each factor on signal loss. But to understand the true impact we needed to compare each of the factors from the test.  In the first attempt, figure 3, we looked at plotting signal loss for each factor at 20, 50 & 100 GHz.  It was difficult to see trends in figure 3 as the impact on signal loss was lower than demonstrated in other ENIG signal loss tests. But it was apparent CGPW & LPC yield the highest amount of signal loss.

Figure-3-4rev.jpg

So just the data for GCPW and low profile copper was broken out in figure 4. This showed as expected that the copper control has increased signal loss with higher frequency. However, the impact of ENIG using thin high phos EN was estimated at 0.55 dB/inch at 20 GHz and almost stabilized at 0.8-0.9 dB/inch from 50 to 100 GHz. This seemed low and we wanted to compare these results to other final finishes.

For this comparison, we used published data by John Coonrod from Rogers Corporation1 but had to be careful to compare similar test conditions. To see the impact of thin high phos ENIG to a typical ENIG we created figure 5 where the copper control data was very similar and both data sets were obtained on a GCPW design with rolled copper. The figure shows the thin high phos EN reduced the ENIG signal loss by ~50%.  It was a little more difficult to compare a newer final finish like EPIG as the Cu controls did not match up as nicely. If we plot just the EPIG signal loss against thin high phos ENIG, the ENIG response is actually better than EPIG. So in figure 6 we had to plot both copper controls (solid & dashed blue lines) along with the thin high phos ENIG (solid orange) & EPIG (solid yellow) to show what’s really going on. Yes under these test conditions thin high phos ENIG has lower overall signal loss, but this is exactly that, a test.

Figure-4-5rev.jpg

The data that’s important for the actual product is the difference between the dashed lines vs the difference between the solid lines. That difference (yellow arrow vs orange arrow) shows EPIG does have a lower impact on signal loss especially as frequency goes above 50 GHz. So thin high phos ENIG can provide 50% lower signal loss vs typical ENIG but will have more signal loss than a nickel-free final finish like EPIG.

But before going to place new orders with your PCB supplier, as we always say, you have to look at the whole picture. Since we want to limit the length of each blog and make them for a specific subject there will be additional information on this subject area, Final Finish Selection for 5G Webinar, and our Blog post, Electroless Nickel How Low Can You Go.

References

  1. J. Coonrod, Rogers Corp. “Insertion Loss Differences Due to Plated Finish and Different Structures”, IPC 2019
  2. Ron Parkinson, “Properties and Applications of Electroless Nickel”

Author

Matt%20Sylvestre-150-2021.png

Matt Sylvestre is currently a Senior Chemist at Technic and has over 20 years of experience in chemical formulations.

Add new comment