Retool your strategic thinking skills with the MEMS Education Series

Karen Lightman, Managing Director, MEMS Industry Group

MEMS is hot – it’s hotter than ever. Market research firms expect the total market for MEMS devices to exceed US $8 billion by 2012. Demand for MEMS devices is rapidly creating new business opportunities for both entrepreneurs and established companies alike. As a diverse industry across numerous markets, MEMS poses unique problems for developing and executing profitable business models.

So, how are you going to get the knowledge needed to compete in this increasingly expanding and growing field?

Here’s an idea: clear your calendar for October 14, 2009. Due to the great success of our July 13, 2009 MEMS Education Series short course — Insider’s Guide to Business Strategy for the MEMS Industry – MIG is holding another session on Wednesday, October 14 at the Carnegie Mellon University campus in Silicon Valley.

MIG has compiled years of research, market analysis, publications, survey data, interviews, case studies and lessons learned from the school of hard-knocks to create the MEMS Education Series. The course will be taught by two fantastic instructors, Dr. Jim Knutti and Dr. Alissa Fitzgerald (check out the website for their impressive bios). This is the second time they’ve teamed up to teach a course in MIG’s MEMS Education Series and the reviews have been stellar!

Strategy is the name of the game at this year’s short course. Jim and Alissa will go over MEMS-specific product development, research and development decisions, business models, business and operations implementation, and financial & funding considerations. Plus, they will draw heavily on primary research MIG has compiled in the form of interviews with seasoned MEMS companies. From these interviews they will hash out what has brought companies success in MEMS, how the big players are reinventing themselves during the downturn, as well as key lessons that can be learned from the failures.

Should you attend? You definitely should if you’re a business development professional, marketing manager, or product manager or if you are looking to enter the MEMS industry. This course also makes solid sense for entrepreneurs, intrapreneurs, venture capitalists, or engineering students just starting out in MEMS or considering entering the industry.

I look forward to seeing you on October 14 at the Insider’s Guide to Business Strategy for the MEMS Industry!

SEMI MEMS Technical Committee

Tom Morrow, Vice President, Global Expositions & Marketing, SEMI

The MEMS industry is changing from a one-process-per-product technology to a one-process-many-product one.  Critical to this transformation are industry standards that reduce costs and spur meaningful innovation.  The SEMI MEMS Technical Committee has been meeting this challenge since 2003, developing several important standards that have yielded significant benefits to developers, manufacturers and suppliers.

The charter of the MEMS Technical Committee is to develop standards for MEMS devices that cannot be handled by existing semiconductor technical committees. Current topics include Wafer Bonding Alignment Targets; Step-Height Measurements of Thin, Reflecting Films using an Optical Interferometer; and Ultra High Purity Microscale Fluidic Systems for Use in Scalable Process Environments. Recently SEMI announced the latest in the series of MEMS standards, MS8 Guide to Evaluating Hermeticity of MEMS Packages, available to users on-line immediately.  The hermeticity guide is the first to address MEMS packaging and further efforts to standardize the methods used to evaluate MEMS packaging are being evaluated.

One of the goals of the new MS8 guide was to gather and summarize all the hermeticity aspects of MEMS. As more data and MEMS-specific details are gathered, sections of MS8 will be removed, revised, or expanded upon leading to publication of a specification or test method standard.

One topic that the committee is now looking at more closely involves the permeability of sealing materials.  For example, assuming one has inspected the seal for integrity, the bond quality is good at both interfaces and there are no direct leak paths between the external and internal environments, the next concern would be the actual permeation of gases through the seal. Permeation is dependent on both the seal geometries and material properties. The minimum allowable seal width for a given lifetime expectancy of the enclosed device depends on both the seal thickness and the permeability of the material itself. The Committee has prioritized this area because there is little published data on permeability of seal materials used for MEMS.

Another issue of concern is the validity of the current test methods utilized for leak testing and residual gas analysis. Current standardized test methods were developed for packages with larger cavity volumes and have proven to be not reliable or not applicable for smaller volume packages, typically <0.02 cc. The published guide discusses the relevant standard test methods as well as more recently introduced methods, providing an objective opinion on the advantages and disadvantages for each.

Ideally, as the production volume of commercially available hermetic-packaged MEMS increases, a method that supports high-throughput non-destructive hermetic QC of the MEMS devices will emerge. Perhaps one of the methods discussed in the MS8 guide will become a self-contained standard, but further investigation and input from the MEMS industry is required to develop consensus.

SEMI has published seven other standards related to MEMS and is looking forward to publishing more as needed by the industry. Participating in the standards development process provides industry benefits, provides company benefits, and helps advance the knowledge and enjoyment of the individual participants.  Please join us in helping advance the industry in proven and productive ways.

For more information on SEMI MEMS Standards Committee activities, please contact me at or Susan Turner at

Chipworks Looks Inside the Freescale HARMEMS Process

By St.J. Dixon-Warren, Manager, Process Engineering, Chipworks

The Freescale MMA6222AEG Accelerometer is targeted for automotive applications and is implemented in a relatively large 13.0 mm x 7.6 mm x 3.3 mm thick 20-LEAD SOIC package. The MMA6222AEG is fabricated with Freescale’s HARMEMS process using a thick SOI layer to form the mechanical layer of the MEMS structure. Chipworks has just completed a full MEMS Process Review analysis of this market leading technology.


Freescale MMA6222AEG Package

Cross-sectional analysis of the package reveals two chips: a MEMS chip with a hermetic cap and a separate ASIC, both mounted upside down beneath the package lead frame. Unlike most MEMS device we analyze, the MEMS chip is encapsulated in silicone gel, likely to reduce packaging stress on the MEMS die.


Freescale MMA6222AEG Package Cross Section

Detailed examination shows the MEMS layer, sandwiched between the MEMS die and the MEMS cap. The thick SOI layer provides increased stiffness and greater mass for the moving mechanical element, plus increased electrical capacitance. This should give increased sensitivity compared to Freescale’s standard surface micromachined process, which we found in the Freescale MMA7455L, where the mechanical poly 2 layer was only ~3 µm thick.


Freescale MMA6222AEG MEMS Structure Cross Section

The HARMEMS SOI process uses a deposited polysilicon layer, with air bridges, to form the electrical interconnects for the MEMS die. The SOI process does not permit the use of a buried poly for the electrical interconnects. The poly air bridges would have been formed on a sacrificial layer (likely oxide) after the deep reactive ion etch (DRIE) used to form the MEMS structures. A timed etch is then used to release the MEMS structures.


Freescale MMA6222AEG MEMS Structure Tilt View

The use of single crystal SOI should allow better control of the DRIE process, thus giving better consistency in the mechanical properties of the device; however, the SOI process limits the kinds of anchors that can be used. Essentially, released beams need to be narrow, while anchored structures need to be wide enough to ensure that enough oxide remains to provide a connection to the substrate. It is worth noting that, Freescale’s European competitors, Bosch, ST Microelectronics and SensorDynamics, have chosen to develop a thick polysilicon process which allows for greater flexibility in the formation of interconnect and in the formation of anchors.