Using sophisticated materials in advanced printed circuit boards (PCBs) can make the difference, but not without challenges! Increasing performance demands are driving the development of sophisticated materials to a new level. This article will discuss several of these performance drivers and how to best respond with material selection.
The rate of change in PCB materials technology is accelerating faster than ever before. It has come to the point where a person needs a database just to keep tabs on the countless material offerings available from the various material suppliers. The IPC-4101 and IPC/JPCA-4104 materials specifications utilize a specification sheet format to identify critical characteristics for each material type. This serves as an excellent tool to help categorize the new materials as they develop. The responsibility then falls on the material suppliers to submit the necessary data to IPC if a new material requiring a specification sheet has been created. The specification sheets can be used as a method to quickly rule out certain materials that will not support the design requirements.
It's not enough just to know what materials are available and their performance factors. It is also important to understand the manufacturing and reliability influence as they interact with each design and material set. For example, many materials with high-end electrical properties can only be used in double-sided applications unless combined with bonding materials from another material family or supplier. Polytetrafluoroethylene (PTFE) resin-based materials, to name one example, are not practical in a high-layer stack or thick board application due to the high Z-axis coefficient of thermal expansion (CTE) characteristics, not to mention the cost of the PTFE resin. However, a PTFE resin system may prove to be the best possible answer to achieve the required electrical performance in a critical power amplifier application with a low layer count.
It is important to keep in mind that the material performance attributes detailed in the IPC specification sheets are primarily lab tests on raw materials (resin systems, copper foil, laminate, pre-preg). This information may serve as a general barometer for material performance, but once these materials are incorporated into actual designs and influenced by processing, they may not respond favorably to PCB design and/or assembly stresses. In other words, materials that appear to be the most robust may actually prove otherwise once laminated, drilled, plated, final finish applied, etc.
For many years, dielectric constant (Dk) has been a significant material selection attribute to support controlled impedance requirements. Today, the fast growing market for wireless applications, router, server and high-end computer applications is driving the increasing need for low loss tangent (dissipation factor, Df) materials. The kicker is, these new material sets also must be manufacturable and reliable in high-layer stack of 20 to 50+ layers. These applications often demand thin dielectric capability in the 50-90 micron range, and dimensional stability for layer-to-layer registration and annular ring considerations in support of hole-to-land size relationships, as well as improved thermal robustness for lead-free assembly.
In an effort to support high-end electrical performance requirements, yet minimize material cost, a material combination approach might provide a solution. In order to benefit from the best of both worlds, a PCB application may need to utilize materials with high-end electrical performance properties. However, to reduce cost, the design could limit the use of these higher cost materials to only a few electrically critical dielectric openings. The balance of the PCB materials in the construction may only require the electrical properties of a standard FR-4 material. This approach can provide a successful solution, but be careful because failure to design a symmetrical material construction can result in problematic bow and twist results.
As with any construction, try to design the stack-up with mirror image selection of resin, glass and copper volume, and with the pattern density equally distributed on either side of the Z-axis centerline. One more word of caution: be sure your PCB supply base has UL recognition for the material combination that serves your needs, assuming your application requires UL. In the eyes of UL, even if each of the two materials has its own UL recognition, combining them creates a new material. If a PCB supplier needs to test the new material combination, the UL test program becomes more complex and time consuming than a homogeneous material construction.
When it comes to reliability and electrical performance, strong peel strength and low moisture absorption are always important. In the past, most materials with high performance electrical properties meant reduced peel strength and sometimes increased moisture absorption. Needless to say, these materials may not survive today's assembly rigors or the variation in electrical properties caused by moisture uptake. For the most part, the latest material offerings appear to be addressing these issues.
Many other performance drivers exist. Glass transition temperature (Tg) has been a critical parameter for many years in products with high layer stacks greater than 14 layers or thick construction greater than 2.5 mm. These applications generally use materials in the 150-200 C Tg range. The high Tg materials normally provide reduced Z-axis CTE, improving hole wall survival during the thermal excursions of assembly, rework and operation.
PCB and component CTE matching is also a concern. X/Y-axis CTE differences between the PCB and components can induce stresses resulting in solder joint cracks or even cracked components. The greater the component I/O count and footprint size, the greater the stress level is magnified by the X/Y CTE mismatch. Aside from cost and weight prohibitive solutions such as woven aramid or restraining metal cores (copper-invar-copper, copper-molybdenum-copper), very little development or improvement has taken place in the area of CTE matching materials.
Conductive Anodic Filament (CAF) growth is also a growing concern for potential field failures. During assembly and product life, resin cracks or resin-to-glass separation can occur. These flaws are then aggravated during product life by temperature and humidity exposure, creating the means for CAF growth. The mounting concern is triggered by decreasing dielectric separation between mechanically drilled hole walls. As holes are placed closer and closer together, the potential for the CAF issue increases. While material selection is a factor, the results also can vary significantly based on the PCB manufacturer's drill quality.
The same concern applies to horizontal (Z-axis) dielectric separation. Many high layer stack applications are already using 50-75 micron dielectric separations to keep overall thickness and drill aspect ratios in a manageable range. Thin dielectrics are also used in high density interconnect (HDI) applications, in the 35-75 micron range. Be it mechanical or laser drilling, poor PCB process optimization in areas of lamination, drilling and hole wall preparation can often give the CAF problem an unwanted jump-start.
Some exciting changes are coming in the form of flattened glass. Flattened glass is rapidly gaining interest for use in thin dielectric applications for laser microvia. Although resin-coated copper can produce nicely shaped microvias, many were apprehensive from the start with regards to the use of a solid resin dielectric with no woven glass support. One alternative was to use standard woven glass, 106 or 1080 weave. Unfortunately, due to the weave porosity of standard woven glass, any given laser hole could hit solid resin, no glass, or up to two bundles of glass with minimal resin. This not only makes it difficult to optimize the laser performance, but it could result in undesirable microvia shapes that are more difficult to copper plate. The flattened glass will provide more evenly distributed glass and resin resulting in improved laser productivity and a more consistent microvia shape. The flattened glass should also serve as a reliability enhancement by helping to prevent resin crack propagation.
Another whole area of sophisticated materials will develop around embedded passives, optoelectronics and thermal transfer materials. In general, performance requirements-electrical, mechanical and thermal-required by various product applications are reaching the limits of the traditional material sets (such as FR-4 epoxy resin on woven E-glass, and variations thereof). Consumer applications are cost driven and for the most part will continue to find a way to design products using low cost materials. Typically, designers deem the performance properties of the high-end material necessary to achieve required performance qualities, but over time, somehow they manage to redesign the product back into standard FR-4 material.
On the other hand, it's inevitable that performance-driven applications requiring high-end performance properties will eventually hit the wall and move beyond the comfort zone of the historically proven low cost material sets. Many applications already have. In many situations, even cost driven consumer products require the latest material developments in an effort to address environmental issues including halogen-free, lead-free assembly or increasing component densities.
While the emerging requirements for halogen-free materials are still a moving target, some of these material sets do exist (or at least they fall near or below the 300 ppm range). Honeywell Advanced Circuits has a production-ready process for halogen-free PCBs. The bad news is the material is more expensive than FR-4, less efficient to process, and has actually lost ground on some electrical properties. Material suppliers are working to address these issues.
It is probable that the lead-free move will eventually be driven by contract manufacturers (CMs), as they are exposed to end product requirements from multiple original equipment manufacturers (OEMs). The lead-free movement has a little more potential to come to fruition than the halogen-free effort. If assembly temperatures rise as predicted over the next few years to support lead-free assembly, this could potentially create a large decline in the use of today's traditional FR-4 materials. Most of today's current formulations perform marginally at best against the increased assembly temperature needs. Assembly temperatures may rise as much as 20-30C over current Sn/Pb reflow and wave temperatures. Efforts are underway by material suppliers to reformulate resins and create new variants that can withstand lead-free assembly temperatures.
Whenever possible, it is important that material suppliers work diligently to keep the new and more thermally robust formulations classified in the FR-4 UL/ANSI grade (IPC-4101 refers to this as FR-4 NEMA grade). If successful, this will help to minimize the testing requirements imposed by Underwriters Laboratories (UL). However, if the IR resin scan results and/or the support materials (woven glass, fibers, fillers) should push the materials out of the FR-4 grade, the timeframe and cost to perform UL testing will be significantly increased for the material suppliers and PCB manufacturers. The PCB industry and UL have truckloads of performance data on the FR-4 grade materials. Departing from the FR-4 grade pushes the material test plan into uncharted territory. Extensive and more time-consuming testing is required as a result. Even if the material suppliers are successful at maintaining the FR-4 designation, the testing will still be a challenge due to the increased test temperatures to represent lead-free assembly. Once again, uncharted territory.
The need for improved electrical performance properties is significant. Furthermore, the need for increased assembly temperature survivability for lead-free assembly is changing the whole playing field. Historically, PCB development efforts were primarily centered around design attributes such as line width, spacing, land size, hole size, aspect ratio, etc. However, today's PCB development efforts now include a new dimension involving countless material options and supporting reliability testing. Whether the driver is environmental considerations, electrical performance or improved reliability, material suppliers are developing new and more sophisticated material solutions to support the needs. The key to timely materials development lies within strong partnerships between material suppliers, PCB manufacturers and PCB customers in the areas of design, assembly, test and reliability.
When is a new material production ready?
The answer to this question can vary significantly based on multiple factors:
Design attributes (layer count, construction type, buried or blind via, aspect ratio, HDI, etc.).
Reliability requirements. (Each customer or application may have a different test regiment.)
Targeted PCB manufacturer's process capabilities.
For example, a PCB facility may have many years of experience and reams of reliability test data for building 12-layer constructions using an epoxy/PPO resin system on woven E-glass. But what happens when they build the same material set on an 18-layer design? Or a 12-layer design incorporating a 6+6 blind via construction? Or smaller holes with increased aspect ratio? Some level of process development and reliability testing may become necessary.
To complicate things even further, the best material in the world is only as good as the manufacturer's process. Each PCB manufacturer may have varying levels of expertise with any given material set. Table 1 demonstrates the effects of PCB manufacturing experience. Material sets A and B may have near identical performance properties if processed properly. However, if a customer dictates that both suppliers X and Y must use material set A, the customer may be getting substandard boards from supplier Y. This is why it's important to work with your PCB supplier to understand their material offerings and capabilities (see Table 1).
How Does A Designer Select Materials?
Unfortunately, there are no materials in the market today that serve as the perfect, one-size-fits-all material solution. What is available is an extensive toolbox of materials and processes that meet most requirements depending on your performance priorities. The material options are numerous. They are far too complicated to list in a simple table, as many complicating factors exist. Each PCB application has a specific set of performance needs (electrical, mechanical, thermal, reliability and UL) with a specific order of importance. The second level of complication is the fact that each PCB manufacturer has a material supplier base and process experience levels they have developed and optimized over time for each material set. It would be in the designer's best interest to take advantage of the strengths of the PCB manufacturers and not the weaknesses.
New sophisticated materials typically cost more and are often more difficult to process resulting in greater PCB cost. On the other hand, they may offer substantial benefits:
Add greater value and reduce the ultimate system cost
Improve overall system performance
Solve a system limitation
Reduce the number of PCBs in a system
Reduce PCB layer count or size
Improve PCB electrical, mechanical or reliability performance
PCB manufacturers are obligated to be proactive and work closely with material suppliers to develop and test improved material offerings. Honeywell Advanced Circuits has a significant team of resources focused on materials and process development, and how they interact with different designs. Many new materials have surfaced and more are sure to come in support of the pressing performance needs. The greatest hurdle continues to be the time element necessary to establish a reliable PCB fabrication process, assembly level reliability testing, UL recognition and a history of performance in the field.
When considering sophisticated materials, it would be wise for designers to work closely with PCB manufacturers that have a wide diversity of materials experience. The best thing a designer can do is to present the proposed design, while still early in the development stage, to the intended PCB suppliers for discussion around available material options. This way, they can evaluate the design, the performance requirements and the performance priorities, making it possible to accomplish educated material selection decisions. By working closely with your PCB suppliers, you can often avoid the potential for catastrophic production start up issues.
David Backen is product development engineering manager at Honeywell Advanced Circuits Inc. [www.Honeywell .com], where he has been employed for over 24 years with primary focus in engineering related functions. Backen's experience includes 10 years of managing the pre-production engineering and tooling functions before changing emphasis to materials development, PCB design and product development. He has managed materials and product development over the last 12 years. His e-mail address is dave.backen@Honeywell.com.
IPC-4101, Specification for Base Materials for Rigid and Multilayer Printed Boards
IPC/JPCA-4104, Specification for High Density Interconnect (HDI) and Microvia Materials
Table 1 - PCB Manufacturing Experience
Board Supplier X
Status: Production Ready
Supplier X has a fully optimized process for material A. Full UL recognition, reliability testing performed successfully up to 20 layer construction
Board Supplier Y
Status: Process Development
Supplier Y has not developed optimized lamination cure cycles or drilling parameters for material A. Full UL recognition will take 4 months, reliability testing to date yielding poor results due to rough hole wall (poor drilling parameters)
Status: Process Development
Supplier X has not developed optimized lamination cure cycles or drilling parameters for material B. Full UL recognition will take 4 months, reliability testing to date yielding poor results due to rough hole wall (poor drilling parameters)
Status: Production Ready
Supplier Y has a fully optimized process for material B. Full UL recognition, reliability testing performed successfully up to 20 layer construction
© 2001 CMP Media Inc.
6/1/01, Issue # 1806, page 12.
Reproduction and distribution of material appearing in PCD and on the www.pcdmag.com Web site is forbidden without written permission from the editor.
Contact information for Reprints.
NOTE: The articles presented here contain only the text originally published in Printed Circuit Design magazine. Any accompanying graphics and illustrations have not been recreated here. You may view the article in its entirety in each printed issue of PCD.