ohnathan Faltermeier,
B.S.’91, M.S.’94, Ph.D.’97, can explain it, and, even better, he can actually
do some of the construction work.ohnathan Faltermeier,
B.S.’91, M.S.’94, Ph.D.’97, can explain it, and, even better, he can actually
do some of the construction work.
In fact, he’s been so well trained in the intricacies of today’s powerful computer chips by the University’s Center for Advanced Thin Film Technology (CAT) that he hadn’t yet actually received his Ph.D. before job offers from high-tech companies started coming his way.
Faltermeier now works
at IBM in Fishkill, developing processes to make the next generation of
memory chips. So does his wife and fellow Albany alum, Cheryl Faltermeier,
B.S.’91, M.S.’94, Ph.D.’97. Other CAT graduates have had similarly enviable
records in landing good jobs.
CAT Director Alain Kaloyeros is understandably proud of these graduates.
A key goal of the CAT, which was established in 1993, is to develop technologies
that will help strengthen New York and U.S. industries, particularly in
the fields of microelectronics and optoelectronics. And by a number of
measures, the CAT is succeeding.
But at the same time, the CAT has enhanced the University’s learning environment in countless ways, yielding tangible benefits for students like the Faltermeiers and employers like IBM. "Our students can hit the ground running when they’re hired by industry," says Kaloyeros.
"I worked on state-of-the-art equipment at the CAT, tools equivalent to what is in processing lines at semi-conductor manufacturing facilities," said Johnathan Faltermeier. "And the research I did was on the kinds of problems that the semiconductor industry is trying to solve in order to build the computer chips of the future."
IBM, Varian, Motorola, Texas Instruments, Intel and over 40 other companies large and small are CAT sponsors and partners, collaborating with CAT scientists or sup-porting CAT research projects. And that means that while Albany students are mastering advanced physics for their graduate work, they are also learning to deal with and meet the expectations of industry scientists working in the highly competitive semiconductor field.
When Faltermeier first arrived at the University almost ten years ago, he had no inkling that his future would be entwined with the revolution wrought by microprocessors. He had an interest in science, and the University’s Department of Physics, then as now, was noted for its strengths in materials physics.
But a few things happened that opened new doors for him and the University.
To build on its strengths in materials science, the University hired Alain Kaloyeros as an assistant professor in the Department of Physics in September 1988.
"I had him for
the first class he taught," recalls Faltermeier.
When he wasn’t teaching, Kaloyeros was pursuing his research interest in advanced electronic materials, also known as thin films, and his work began to attract support from industry. In 1991, Kaloyeros was named a Presidential Young Investigator by the National Science Foundation, an award that brought him federal and industrial matching funding of $100,000 a year for five years as well as recognition for his research. That same year, Faltermeier and his future wife, then Cheryl Wyetzner, received undergraduate degrees in physics and began doctoral studies at Albany.
Also about the same time, the Semiconductor Research Corporation started highlighting in its publications Kaloyeros’s work with copper as a material for interconnections within and between computer chips. Equipment vendors, chemical suppliers, analytical instrument makers and other companies which play essential roles in the construction of the modern-day computer chip started knocking on his door.
Industry support for the research being done by Kaloyeros and his colleagues grew, and in 1993, the University was designated by the state as a Center for Advanced Thin Film Technology. The designation carried with it state funding of $1 million a year for ten years, with a required industrial match of $1 million minimum, for a yearly budget of at least $2 million.
With the state and federal and industry funding, the CAT assembled state-of-the-art research facilities now valued at $30 million, recruited faculty with expertise in critical research areas, and supported increasing numbers of graduate students. In 1995-96, the CAT received $12.7 million in industry and government support.
Faltermeier, meanwhile, was in the thick of develop-ments in Kaloyeros’s laboratories in the time leading to the birth of the CAT and during its growth since.
"When I started as a graduate student, there were four or five students working in Professor Kaloyeros’s Advanced Materials Laboratory. Now there are 23 graduate students working in his group in the CAT," says Faltermeier. (More than 50 percent of those students were Albany undergraduates who elected to continue in the physics doctoral program because of the opportunities available at the CAT, Kaloyeros frequently points out.)
For both John and
Cheryl Faltermeier, the birth and growth of the CAT has meant long hours
of work—"most of us are here six or seven days a week," he said
shortly before he completed his doctoral work—but also a front-row seat
on what it takes to build ever-more-powerful computer chips. And, yes,
it’s been an excellent lesson in how a computer chip is made.
As almost everyone who has heard of Silicon Valley knows, silicon wafers are the basic building blocks of microprocessors or computer chips. Silicon is the substrate on which are packed the ever-smaller and ever-more numerous circuits and transistors that are at the heart of digital technology. All that circuitry is made through processes that deposit layers of thin films—copper interconnects are just one kind of thin film— and then shape or modify those layers through lithographic etching.
The first commercially available microprocessor, the Intel 4004 produced in 1971, had just 2,300 transistors. And ever since then, microprocessor manufacturers have sought ways to reduce the size of the components and pack more of them on a chip. Today, the most powerful chips have five million or more transistors and the semiconductor industry is looking for ways to make even more powerful chips.
The incredible increases in the computing power of chips have been spurred by significant developments in how they’re made. Advances in lithography, for example, dramatically reduced the size of the transistors that can be reliably fabricated on a chip. Manufacturing processes have come a long way from 25 years ago when silicon wafers were loaded by hand into red-hot furnaces, where they were exposed to various gases for specified amounts of time.
Because of the complexity of today’s microprocessors which can be damaged by the tiniest of particles, modern manufacturing plants must be pristine and free of vibration— enormous "clean rooms." Machines whisk wafers between stages.
But what must be done to make the chips of tomorrow?
Through its research, Albany’s CAT aims to provide answers to that question.
"Our goal," says Kaloyeros, "is to provide next-generation products to the semiconductor industry on a timely and cost-competitive basis."
The Semiconductor Industry Association has identified "interconnects," the connections between transistors on a chip and between chips, as a key technology for the development of more powerful chips, and the CAT has developed nationally recognized expertise in this area. Kaloyeros’s early work with copper laid the foundation for that expertise.
Aluminum alloys are the current metal of choice for interconnects. But as all the components of chips have become smaller and smaller, the deposition of aluminum alloys into the small holes made for interconnects has become increasingly difficult and generated industry concerns about chip reliability, says Kaloyeros. For that and other reasons, industry wants to explore alternatives.
A tour of CAT laboratories in the University’s Physics Building reveals one important reason so many companies come to the CAT for help in exploring alternatives. The facilities are truly state-of-the-art.
The CAT now possesses three commercial cluster tools— the core components of modern chip manufacturing. Whereas once people brought a silicon wafer from station to station to produce a computer chip, those "stations" are now chambers in a cluster tool to which wafers are automatically moved.
Electron microscopes are the centerpieces of the $3-million CAT laboratory facility led by Tung-Sheng Kuan, a nationally recognized expert in high-resolution imaging and structural analysis of thin films and multilayered structures. The analysis he provides helps CAT researchers determine what is happening at virtually the atomic level when a thin film is grown on a surface. Gottlieb Oehrlein is the director of the plasma process-ing laboratory, where he leads efforts to develop advanced processes for patterning of electronic and optoelectronic materials in thin film form.
And in the basement of the Physics Building is the $4.5 million pilot semiconductor manufacturing line, sponsored by Varian Associates Inc., that will demonstrate new manufacturing processes.
The CAT is "unique in the U.S. in its full compatibility with the industry’s standard eight-inch wafer size," says Kaloyeros. "The extent and depth of expertise and associated infrastructure in the fabrication and characterization of thin films is unrivaled at a university and rare in the corporate world."
For one recent project, Kaloyeros worked with researchers from National Semiconductor Corp.; Schumacher, a chemical supplier; and MKS In-struments, an equipment manufacturer. The group investigated the use of chemical vapor deposition (CVD) as a new and possibly better approach for depositing aluminum interconnects on chips.
Kaloyeros is an expert in CVD, a process that uses gases of various chemical compounds to transport metallic or other individual components to a substrate where they react to form the target material. Currently, chip manufacturers rely primarily on physical vapor deposition (PVD) to place materials on chips. "PVD is a line-of-sight process, analogous to spray-painting a wall, while CVD is surface-activated and thus can get anywhere and has the potential to be more precise," explains Kaloyeros.
But developing a CVD process is complicated. What are the best precursor chemicals to use? What kind of equipment will most effectively deliver the chemicals to the wafer surface? Schumacher develops and supplies chemical compounds for CVD processes and MKS makes the equipment that delivers chemicals to the reaction chambers in cluster tools. Both companies have obvious expertise in CVD technologies and a strong interest in the development of new ones, and through the CAT, they can be involved in and collaborate on the vital development stages of new approaches.
The CVD process developed for aluminum shows promise, says Kaloyeros, but more work remains to be done. And he continues to explore the potential of copper as an alternative to aluminum for next-generation interconnects.
The very complexity of chips and of modern manufacturing processes means that change can be enormously costly and risky. The CAT, says Kaloyeros, provides "low-risk, high-payoff" opportunities to develop alternatives.
"The center (CAT) gives us access to equipment that we don’t have in our factories yet," says James Ryan of IBM Microelectronics. "We get early exposure to new technology (prior to commercialization) and the center facilitates interaction with vendors and end-users... The center is a source of highly skilled professionals."
Varian, a California company that sells semiconductor equipment to manufacturers, clearly views the CAT as a vital resource for its future.
Varian chose Albany as the site for its pilot manufacturing line because CAT researchers, in its view, had made the most progress in the technology of chemical vapor deposition, said Varian chairman Tracy O’Rourke. The Varian equipment can use both PVD and CVD processes, and CAT researchers are looking for ways to make the best use of both approaches.
While the CAT is best known for its expertise in inter-connect technology, its work has applications in a number of other fields and the CAT is expanding its research in those areas, says Richard Saburro, the CAT’s deputy director for business development.
An aluminum alloy interconnect is just one example of a thin film. Besides being the building blocks of computer chips, the atomically engineered materials known as thin films are also the building blocks of solar cells, photonic devices, lasers, high resolution displays and sensors. Thin films are also important in applications as diverse as medical prosthetics and aircraft engines.
In optoelectronics, for example, the CAT is working with a computer-screen manufacturer to develop blue light-emitting material to improve flat-panel displays, such as those used in laptop computers, says Saburro. That technology could also be important in the creation of flat-panel televisions of the future.
Since its inception in 1993, the CAT has established joint research and development activities with more than 50 companies, including over 30 New York State companies. These activities have generated over $32 million in funding and equipment, resulted in nine inventions and yielded 22 high-tech products, Saburro notes.
The CAT, says Kaloyeros, stands ready to develop and commercialize more "vital enabling technologies" for the 21st century. He says the CAT’s "ambitious and entrepreneurial faculty and staff culture" has been the key to success thus far and will enable it to meet the challenges ahead.