Reading time ( words)
Nolan Johnson meets with Isola’s Ed Kelley at DesignCon and takes a tour of the dynamics in play in the materials market. As Isola prepares to roll out an extremely low-loss, halogen-free product, Kelley says the company continues to work on how to make the product reproducible and available to customers around the world.
Nolan Johnson: What’s front of mind for you and Isola in the materials market?
Ed Kelley: There are so many areas I could cover, but I’ll start with 5G as something critical. It’s been talked about for a long time, and even with the 5G rollout, there’s still more to do. If you break 5G down into low-, mid-, and high-band, most of the rollout has been low- and mid-band. But the high-band applications get interesting for base material suppliers and others in the supply chain because, where there are very high operating frequencies, people are looking for extremely low loss materials. Every piece of the base material is critical—the resin system, reinforcement, and the copper foil. There’s a lot to consider.
We do a lot of work developing extremely low loss resin systems because a lot goes into them, not just polymers, but fillers and flame retardants as well. Low Dk fiberglass cloth has become common as a reinforcement. The next-generation low-Dk glasses are more available now than a year ago, though still not at full commercial scale, and we’re looking at alternative types of reinforcements to achieve certain characteristics. Much of the loss at higher frequencies is due to conductor loss, creating a need for copper foils with lower and lower surface profiles.
How do you ensure those low-profile coppers stick to your dielectric material and give you a reliable printed circuit board? It’s one thing to achieve an electrical performance target by itself. It’s not as easy to get the electrical performance and satisfy everything else the material has to do. In addition to meeting signal integrity requirements, the material also needs to meet thermal reliability and CAF resistance requirements in complex, high-layer-count boards. We also focus on PCB manufacturability. We want our products to be as easy to use in PCB manufacturing as possible so that our customers can achieve high yields. So we spend a lot of time up front developing resin systems and integrating them with glass and copper options to meet all these requirements.
Back to electrical performance, we also have to consider how materials are used by the PCB fabricator. It’s important to recognize the nuances of keeping the system as low loss as possible. When the fabricator takes our material, makes an innerlayer circuit, and runs it through oxide-alternative chemistries, that’s another area of attention for 5G signal integrity performance. If you roughen up the copper surface too much, you degrade the loss performance. We’re all working as hard as possible to make sure the end product achieves the goals of the OEMs.
Johnson: That’s a lot to balance those constraints. You’re already working iteratively on known solutions, so how do we develop materials that help open a more reliable high-speed 5G application?
Kelley: We’re about to release a new product called TerraGreen 400G. As the name implies, it’s extremely low loss and halogen-free. We’ve got our UL certification. We’re focused not just on the performance of the material itself, but whether we can manufacture it repeatably day in, day out. As the product grows, is the supply chain stable enough to support demand growth? We are wrapping up some of the final qualification testing, not just to meet internal requirements, but with suppliers; we want to make sure we have a robust supply of the materials that go into it. Our goal with the product, with the combination of resin, glass, and copper, is 25% lower loss than our best performing high-speed digital product currently, which is Tachyon 100G.
With the combination of resin, glass and copper, we’re hitting those signal integrity targets with this product. We’ve done the thermal reliability testing. The CAF resistance we’re seeing is excellent. In fact, with these very smooth copper foils for high frequency applications, we’ve seen cases of copper migration at the bond line between an etched laminate surface and the prepreg that’s bonded to it.
Traditionally with CAF, people think of migration along the resin-to-glass interface, which is the most common failure. With some of these very low-profile copper foils, the copper suppliers are determining how to get a better chemical bond vs. mechanical bond between the copper and dielectric. There are various adhesion promoters that can be used, but some may not be compatible with all the resin systems that companies like us are trying to develop.
If you use the wrong adhesion promoter, when you etch off the copper it may still be on the surface and interfere with the bond of the prepreg to the etched laminate. We’ve spent considerable time making sure we understand what’s driving that and what we need for the right adhesion promoter for our resin system. As we wrap up that work, we are seeing some very good results.
Even as we introduce this new product, we’re looking at what comes next. Our chemists are looking to synthesize some of our own polymers, which would still be used in combination with commercially available resins, but may offer us some unique attributes in our applications.
As we develop resin systems, we recognize there is a range of operating frequencies to consider, and you don’t always need the lowest loss material. When you don’t, you don’t want to pay for the loss performance you don’t need. We are offering a halogen-free resin system which is extremely low loss, but we could put it on E-glass, low Dk glass, next-generation low Dk glass, and different copper combinations. With that one resin system, we can target certain loss properties and combinations that are attractive, both for performance and cost. You don’t always need to use the most costly option.
To read this entire conversation, which appeared in the May 2022 issue of PCB007 Magazine, click here.