Let the Circuits Roll

Flexible circuits fabricated by a roll-to-roll process that mimics newspaper printing may someday become ubiquitous in everyday electronic devices thanks to the Center for Advanced Microelectronics Manufacturing at Binghamton University

By Tom Gibson

Driving down North Street in Endicott, New York, the memories started to come back. This is the site of a massive former IBM complex that once cranked out mainframe computers and components that went in them, and I worked here the summer of 1976 between semesters at college. Many gray monolithic buildings remain from those days, and I tried to find the one I worked in. Industrial trappings accompany them, including pipelines for steam and compressed air snaking from one building to another and over streets. Enclosed elevated walkways that once carried employees from one building to another also spanned across roads.

But then I came upon a building that stood out, one that has been renovated in a darker shade of gray and green trim, gleaming with glass and steel. Endicott Interconnect Technologies (EIT) now occupies this structure, as IBM got mostly out of manufacturing here after its transformation in the 1990s and moved its remaining manufacturing operations elsewhere.

Nostalgia notwithstanding, why am I here? EIT has teamed with the Center for Advanced Microelectronics Manufacturing (CAMM) at nearby Binghamton University (BU) to develop a cutting-edge type of electronic manufacturing capability. A first-hand look awaits me.

I meet with Mark Poliks, director of research and development at EIT, who tells me, “There’s a whole new opportunity opening up where electronics could be printed or made in a roll-to-roll manufacturing environment, which could make a variety of sensors ubiquitous.” Many compare roll-to-roll (R2R) with the process for printing newspapers. “The idea is to make these devices in massive quantities, enabling you to do the same thing over and over again very quickly. Then you cut out the individual electronic circuits from the roll using a laser or stamp and die, and these go into electronic devices.”

Being conducive to mundane, lower-cost electronics, R2R electronics manufacturing will have applications in myriad areas, such as medical devices, the military, flexible displays, computers, telecommunications, and consumer products. You might see it in the form of a flexible readout on the arm of a soldier, on a Band-Aid or cereal box, or even in wallpaper with lighting in it.

Brings Partners Together
CAMM came about in 2005 when the United States Display Consortium, later renamed the Flex Tech Alliance, selected BU to spearhead development of next-generation R2R electronics manufacturing capabilities. BU had a reputation as a leader in electronics packaging and small-scale systems integration. A unique collaborative effort, CAMM brings together partners from government, industry, and academia. To carry out its mission, CAMM would acquire prototype tools and develop processes to produce low-volume test products.

Besides BU and EIT, CAMM’s founding partners include Cornell University, and corporate members include General Electric, Kodak, Samsung Electronics Company, Corning, Plastic Knowledge and Rohm and Haas. Additional partners and supporters include NASA; the Army Research Laboratory; the New York State Office of Science, Technology and Academic Research; and the Arizona State University Flexible Display Center.

CAMM is part of BU’s Small Scale Systems Integration and Packaging Center (SSSIPC), and principals include BU and Cornell professors of mechanical, materials, and industrial engineering and physics and chemistry. Bahgat Sammakia serves as director of CAMM and SSSIPC as well as a professor in the Department of Mechanical Engineering at BU. He received his B.S. in Mechanical Engineering from the University of Alexandria in Egypt and M.S. and doctoral degrees in mechanical engineering from the State University of New York at Buffalo. He started working at IBM in 1984 as an engineer in the thermal management area and stayed there until 1998, and his last position was manager of development for organic packaging in the Microelectronics Division.

Crossing the Susquehanna River and driving to the campus of BU, I again experience a flashback to days past. I remember older buildings made of brick and concrete populating the campus, and these remain. But a smattering of shiny modern structures now stand among them. Credit this to the emergence of a new engineering school and associated high-tech facilities.

I met with Mary Beth Curtin, associate director of CAMM and SSSIPC, at the Innovative Technologies Complex on campus. With green and gray halls to match BU’s colors, as well of those of EIT’s facility, and lots of high-tech equipment, this epitomized the trend. And adding to it, a new engineering building was under construction next door.

IBM and Binghamton-based Singer Link were instrumental in forming the Thomas J. Watson School of Engineering and Applied Science at BU nearly 30 years ago. The school is still taking shape and gaining prominence as it continues to grow.

Continuing the IBM Tradition
Meanwhile, EIT offered strong infrastructure for CAMM when it came time for BU’s engineering school to find a facility for it. Formerly the IBM Microelectronics Division, EIT took its present form in 2002. Today, nearly 1800 employees occupy 18 buildings in Endicott’s North Street Campus, formerly the IBM complex.

Poliks comments that breaking from IBM allowed EIT to diversify and do things they couldn’t have done otherwise. “We knew each other throughout the years; we were IBM. Between Endicott and Binghamton, it’s a day-to-day working relationship, almost like we were the same organization.” A former IBM employee, Poliks has two offices, one at EIT and another at BU. And it goes beyond BU and EIT, as some IBM fellows had gotten Cornell into electronic systems and packaging and started a packaging research center at that university. “There has always been a close relationship among the three entities.”

In describing EIT’s engineering work, Poliks reveals, “Our R&D group started as a group of eight, and now it’s a group of about 40.” They have several hundred engineers on staff spread throughout the company, some working in applications, where they interface with customers in their design requirements, and others in manufacturing engineering, R&D, quality, liability, and procurement. They use electrical, mechanical, chemical, industrial, and materials engineers.

As its core work, EIT produces electronic interconnect solutions, including the fabrication and assembly of printed circuit boards, integrated circuits, and semiconductor packaging. As Poliks puts it, “We do everything but the chip.” They manufacture some products and license other patented technologies. Contrary to their work with R2R electronics, their core electronics see use in lower-quantity, high-reliability applications such as corporate servers and government supercomputers. So in pioneering R2R electronics manufacturing, EIT finds itself drawing upon its knowledge of high-reliability components and fabrication to develop machinery for mass production of flexible circuits.

In its development work, CAMM uses standard production machines, some vestiges of IBM, modified for their use. Starting on a tour of the R2R line, Chris Chase, research engineer at CAMM, showed me a Kraemer Koating machine for roll-to-roll processing. A machine like this can handle rolls of film 24 inches wide and 5 mils (125 microns) thick, and they have handled down to 1.2 mils on occasion. The plastic film, commonly PET or polyethylene napthalate (PEN) on rolls, goes through several deposition stages, including metal and photoresist. Then they etch it off to form the circuitry. Chase likens it to silver halide on photographic film.

Chase has an AAS degree in chemical engineering technology from nearby Broome Community College, and his past work experience suits him ideally for CAMM. It includes high-speed web coating of silver halide emulsions on various plastic and metal rolls of material, photolithography, PVD (physical vapor deposition), electron microscopy, and process engineering and development. As a BU employee, he works with CAMM member companies and their scientists and engineers on their projects, guides several PhD students in CAMM, and keeps the complex tools up and running. The latter involves interacting with tool vendors and their engineers.

Moving on to another machine, Chase says, “This is like a glorified car wash.” The film passes through a chemical developing solution, then sprays of de-ionized water to rinse it.

Next up, a cleanroom, which we viewed from outside through a window. A photolithography machine stood inside. Chase tells me they can get 2.5-micron-wide lines in a circuit and can draw nearly any shape. Poliks interjects, “This is the only printed wiring facility certainly in North America and probably in the world that can make patterns down to two to four microns on areas extremely large.” Chase then shows me an ECD vision inspection tool that uses lasers to reflect off contamination on the film.

“This is the GVE roll-to-roll sputter tool for thin metal deposition. You can etch the copper in a chemical bath, and you can also plate up a much thicker layer of copper on existing traces using a plating tool,” Chase explains as we take in another machine. The tool uses a two-stage mechanical pumping system as well as four turbomolecular pumps to achieve the desired pressure, often referred to as vacuum.”

In reflecting on his work, Chase reveals, “I get to work with many different companies and experience areas of research most people never see. Seeing the new daily applications that come from our members is amazing. Several of our larger tools are unique in the world, and it’s an honor to be able to work with them. There is so much potential to the facility.”

Some History Behind It
In describing the history of R2R electronics, Poliks reveals, “Roll-to-roll processing probably evolved out of printing. But companies like IBM and other packaging companies in the United States and beyond studied that kind of technology and made the first flexible circuits.” IBM did it in EIT’s building 10 or 12 years ago. Poliks notes that printed circuit boards have long used photolithography and photoresists, in additive and subtractive processes with etching or plating.

Several years ago, Motorola undertook an effort to run electronic inks through a printing process on substrates and build up layers. They added conductor and semiconductor materials to make what they hoped were the first printed radio frequency identification (RFID) devices. The materials didn’t work adequately in the RF range, but the processes are advancing and may lead to things like having an RFID on a tube of toothpaste.

As Poliks explains, “We’re learning from that and saying ‘how can we take more and more processes currently restricted to silicon wafers and apply them?’” You can’t heat typical plastic film substrates to 1000 degrees like you can a silicon wafer. “With those limitations, you apply thin films of silicon, silicon oxides and nitrides, and metal to actually build up thin, flexible transistors, which make the backbone for pretty much anything electronic.”

While the R2R work mostly involves testing prototypes, Poliks reveals that a couple of products have already come out of it. For example, “We make a flexible circuit that goes into a medical device. This goes on the end of a catheter, and it’s a point-source ultrasound emitter actually being used in diagnostic medicine now. The circuits in this device, which are as small as 14 micron lines in space, are made in the CAMM facility.”

On another exciting front, CAMM and EIT are looking at developing flexible photovoltaics. While not as good as those made from silicon, they would be cheaper to make. They envision putting them on tents along with flexible LED lights, with the PV panels powering the lights. The Center for Autonomous Solar Power (CASP) at BU, another part of SSSIPC, is already researching and developing large-area, flexible, lightweight solar cells for aerospace, consumer, and industrial markets.

In taking an objective view, Poliks knows R2R electronics technology stands “pretty much at the beginning.” He sees it as an exciting opportunity — “of course it is” – and it may not be that long before we start seeing a stream of flexible circuits in myriad everyday devices. When that happens, I’ll look forward even more to returning to Endicott and Binghamton to reminisce and catch up on the latest technology coming out of the old IBM complex.