-
- News
- Books
Featured Books
- smt007 Magazine
Latest Issues
Current IssueComing to Terms With AI
In this issue, we examine the profound effect artificial intelligence and machine learning are having on manufacturing and business processes. We follow technology, innovation, and money as automation becomes the new key indicator of growth in our industry.
Box Build
One trend is to add box build and final assembly to your product offering. In this issue, we explore the opportunities and risks of adding system assembly to your service portfolio.
IPC APEX EXPO 2024 Pre-show
This month’s issue devotes its pages to a comprehensive preview of the IPC APEX EXPO 2024 event. Whether your role is technical or business, if you're new-to-the-industry or seasoned veteran, you'll find value throughout this program.
- Articles
- Columns
Search Console
- Links
- Events
||| MENU - smt007 Magazine
WEB EXCLUSIVE: MEPTEC Converging Technologies Symposium Discusses Merging MEMS and Wafer-Level Packaging
May 11, 2004 |Estimated reading time: 8 minutes
By Jody Mahaffey
Santa Clara, Calif. — MEMS (microelectromechnical systems) has become one of the hottest topics discussed in the industry today. MEMS revenues are forecast to increase at a compound annual growth rate (CAGR) of 15 to 20 percent over the next few years, while units shipped are forecast to grow at a CAGR of 25 to 30 percent.
MEMS devices are being used today in high-volume applications such as accelerometers, gyroscopes, digital light processors (DLP) and printers. But not all MEMS are considered mainstream or fully commercialized, in large part due to the issues related to packaging MEMS devices. Many believe that wafer-level packaging (WLP) is the ideal technology to deal with those issues.
To discuss the issues related to the convergence of MEMS and WLP, MEPTEC, the Microelectronics Packaging and Test Engineering Council brought some of the most prestigious names in MEMS together for a one-day technical symposium on May 11 in Santa Clara, Calif. Some of the conference speakers offered their insights into the quickly changing world of MEMS and WLP.
Long-time industry consultant Bance Hom served as symposium chair for the MEPTEC conference. Hom believes that there are many common skill sets and cross-linking packaging technologies easily translated bilaterally into both arenas, so tying the common links together should enhance the progression and agendas for both technologies.
When packaging MEMS, the lines between the front and back ends of the industry blur. Many believe that a partnership between the front-end and back-end is the most successful route in merging MEMS and WLP, including Hom. "With packaging one of the biggest challenges for MEMS, it becomes a natural migration in the U.S. for packaging professionals to find new challenges in this emerging MEMS field," she says. "In my own personal experience and expertise, plating, which has traditionally been a back-end operation in IC packaging has migrated into the IC front-end (Damascene Process), and now is fully integrated into the MEMS front-end fulfilling purposes as functional device designs, sacrificial layers, etch resist and eutectic gold/tin layers for wafer bonding/lid seal operations.
"It also became a priority for MEMS technologists to have a fundamental understanding of packaging technology to design and sync up their devices to the outside world in a more cost-effective and reliable mode. Wafer-level packaging becomes a key enabling technology as packaging and fabs converge."
Whether MEMS companies add packaging personnel, or through partnerships with back-end companies, most experts agree that packaging MEMS at the wafer level is a front-end responsibility. John Heck, senior engineer, MEMS and packaging for Intel Corp. and session chair for the Enabling Technologies Session of the conference, believes merging MEMS and WLP should fall to the front-end because critical steps must be done in a clean room, and because handling MEMS wafers should be minimized with the fragile devices exposed.
Joe Brown, MEMS director at SUSS MicroTec, whose company presented in the Enabling Technologies Session, agrees with Heck, saying, "For the most part, MEMS have been front-end process-centered, and for this reason have suffered by lacking advanced packaging solutions. With packaging integrated by front-end processing (prior to CMOS), MEMS can take a big step toward meeting cost targets by utilizing novel processing techniques."
Anthony Flannery, director of development, MEMS Gyroscope for Invensense, spoke in the Assembly Processes Session of the conference. Flannery believes that wafer-level integration, the term he prefers over WLP, must be addressed by the MEMS foundry. According to Flannery, "The trend in the MEMS industry is moving towards a fabless model for most new companies. It is difficult for the new MEMS startup to tackle packaging. While it may be an enabling technology, investors no longer have the stomach to invest in the capital infrastructure necessary to bring it up in-house.
"The cost of development and implementation must be amortized over the company's specific product line. In contrast, the foundries can leverage their infrastructure and development costs over many customers and product lines. It just makes better financial sense," he adds."It is also doubtful that wafer-level integration will be successfully addressed at the packaging level," Flannery continues. "The infrastructure does not exist there. The level of precision and control is greater than what they or their infrastructure is used to dealing with. Perhaps most importantly, many times the wafer-level integration step is not the last. Additional MEMS processing must be done. That would mean it would have to go back to the MEMS foundry, which smacks of inefficiency and the possibility for errors and contamination."
So assuming the answer to MEMS packaging is WLP and that this is a front-end responsibility, the expert speakers were asked what they believed were the most significant technical and non-technical issues facing the integration of MEMS devices and wafer-level packaging, which possibly are holding them back from being fully commercialized. Cost and a lack of standardization (which affects cost) were common answers regarding non-technical issues. Brown of SUSS MicroTec says, "The most significant non-technical issue is cost. This can, in many ways, relate to other issues, such as standards, but it usually comes back to cost. Most MEMS today are commodity products used in cost-sensitive markets, such as automotive and consumer electronics. In many ways, cost adders to provide wafer-level packaging must be targeted at pennies per die."
With respect to commercialization, Brown adds, "Fully commercialized would infer the 800-pound gorillas have accepted this path for technology. Here, standards and other key elements for cost reduction are needed to fully realize the potential of MEMS and WLP."
Heck of Intel says, "I believe the lack of standard, commoditized wafer-level packaging materials is the biggest problem to address. Every MEMS company is forced to develop their own proprietary solution for packaging MEMS at the wafer level, since there are no standardized processes."
"Cost clearly is one of the most significant issues," agrees David Fork, principal scientist with the Palo Alto Research Center. Fork presented in the Assembly Processes Session. "Wafer-level packaging has the potential to lower cost in comparison to conventional packages, but this will require a large application in volume production in order to achieve the economy of scale possible with batch fabrication."
Cost is not the only issue to be considered. Flannery of Invensense believes that available infrastructure in the areas of equipment and training is a critical issue for the convergence of MEMS and WLP. "Because of the (justifiable) reluctance of MEMS foundries to spend on anything that looks like capital investment, there is a shortage of places MEMS companies can go to get wafer-level packaging services, and when they do exist, there usually is a lack of breadth or understanding on the part of the foundry. All the technical pieces exist out there (bonders, plating setups, wafer aligners, etc.), but it is up to the foundry to put it together. Right now, that has been lagging."
As far as technical issues are concerned, the answers are wide-ranging. Brown of SUSS says, "The most significant technical issue is to be able to provide "0" level packaging with hermetic sealing, maintaining vacuum < 1 x 10 -3 or controlled environments with 99.999 percent accuracy for 10 years, having electrical feed through, with high yield of >95 percent."
Fork of Palo Alto Research Center believes that the protection of fragile MEMS components at the dicing step is a critical technical issue for all MEMS devices and that wafer-level packaging may be just the solution.
Marc Madou, Ph.D., professor of biomedical engineering and integrated nanosystems research at UC Irvine gave the keynote address at the conference. Madou says there is too much variability in MEMS devices to come up with one overall strategy.
Flannery agrees. "I don't see any major technical issues. Many different bonding technologies have been demonstrated in the last several years. They all have their own challenges and strengths, but I don't believe there are any showstoppers," he says.Along with the potential problems and pitfalls, there are some real advantages to integrating MEMS and WLP. Most speakers agreed that cost can be one of the biggest advantages. Heck of Intel feels that parallel packaging of all devices on the wafer at once will reduce packaging cost dramatically. In addition, sealing delicate MEMS devices in the cleanroom will increase yield significantly, ultimately reducing cost.
Madou explains, "It is obvious that if we succeed to make more of the back-end processes front-end, we will drive costs down. We will also save space/volume and materials."
Most of the speakers at the May 11 conference discussed new technology that their companies have developed to help solve some of the issues relating to the integration of MEMS and WLP. But they all agree that even as new solutions are developed, many issues remain to be addressed before MEMS can become fully commercialized."They (MEMS) will become mainstream when they get into the typical conventional computer industry, which will be in communication — RF MEMS," says Heck.
Madou says, "Only consumer applications really make MEMS worthwhile in the long run."
"If we can figure out ways to lower cost, there would be more MEMS applications and larger sales volumes, bringing it closer to becoming fully commercialized," Fork weighs in. "There are many more novel processes and approaches than the current infrastructures can support. Only some will be adopted. Sorting this out will take time and a mixture of success and failure."
Ultimately, the goal of the MEPTEC conference was to address some of these issues and begin discussing some of the remaining ones so that MEMS and WLP can help each other get to a fully commercialized stage. Symposium Chair Bance Hom believes that, "Having mainstream IC packaging technologists and MEMS technologists together to share and learn common issues/challenges is important as both worlds evolve up this commercialization learning curve."
For more information on other MEPTEC technical symposiums, visit www.meptec.org.
Jody Mahaffey is with JDM Resources.