Portland, Ore. - Rudolf "Ruud" Tromp is undertaking groundbreaking work in the study of the growth of organic-semiconducting crystals, such as pentacene, which is being applied in flat-panel display technology.
According to Tromp, who is IBM Corp.'s molecular assemblies and devices manager, poor attention to the atomic-surface physics of these thin films leaves researchers with grain sizes restricted to 1 micron-making them a difficult material from which to build displays. However, Tromp recently demonstrated that by paying careful attention to how semiconductors grow, he could increase the grain size to as large as 100 microns-large enough to contain a pixel's transistor inside the grain's boundary.
"All these grain boundaries make the pentacene transistors' performance poorer than it could be," said Tromp. "But by studying the way they grow, and by properly preparing the surface, we have been able to increase the grain size to a tenth of a millimeter-a factor of 100-fold bigger than previously-big enough to hold a transistor. Not only will the performance of such transistors improve in such large grains, but they also will introduce the possibility of engineering organic thin-film nanostructures that could enable new types of devices and materials that are presently out of reach," said Tromp.
Tromp's work today follows naturally from pioneering research he initiated at IBM's T.J. Watson Research Center in 1983. Indeed, the work he began there in understanding the structure and growth of semiconductor surfaces and interfaces has been rewarded by the American Physical Society, which has presented him with its Davisson-Germer Prize in Atomic or Surface Physics.
The insights resulting from Tromp's basic work in materials science have been critical for many advanced areas of semiconductor technology.
In reflecting on his career, Tromp described how his IBM Research team is continuing to break ground in such areas as organic, molecular and quantum devices.
"During the first third of my career, I sought to understand the structure of semiconductor surfaces-questions like: 'Where are the atoms, what do the bonds look like and how can we understand them?' " said Tromp.
"Where are the atoms?" was the "really big question," in 1983 when he started, said Tromp, but by the late 1980s his group, along with many other labs, had established a basic understanding of the atomic structure of crystals on surfaces.
"Where the atoms were was more or less resolved by the mid-to late 1980s-we had found silicon surfaces to have very complex structures with large unit cells, but by then we knew exactly where the atoms were," said Tromp.
From there, Tromp became interested not only in where the atoms were, but also in how they got there-work that Tromp accomplished with the United States' first low-energy electron microscope. LEEMs are ideal instruments for observing the growth of films on surfaces. The microscope was invented by Ernst Baurer, who now has his own LEEM at the University of Arizona. IBM's LEEM design was taken from the first practical system that was put together by a group at the University of Clausthal in Germany.
The LEEM allowed Tromp to make live-action movies of crystal growth with a lateral resolution of 5 nanometers. By observing how crystals grow, Tromp pioneered the industry's current understanding of how atoms move on surfaces and how they attach themselves to substrates. The LEEMs have been constantly improved by IBM and are currently licensed to many other labs worldwide.
Tromp's understanding of semiconductor surface growth progressed as his work using the LEEM crossed over from basic science to breakthrough technologies. "We found that the fastest transistors were made from germanium-silicon alloy films, but the growth of these films was very tricky business," he said. "If you try to grow pure germanium on silicon, you find that you can't grow a flat film, because it tends to ball up the way water droplets do on a newly polished car."
At this point, Tromp began fishing around for solutions to the balling-up problem, which turned out to be the use of "surfactants." The process used an additional monolayer of metal that could enable the growth of atomically smooth and defect-free films. That resulted in the ability to grow germanium films on top of semiconductors, solar cells and even magnetoresistive disk heads.
"What we found was you can play a little trick: Just put a single atomic layer of metal film on top of a silicon surface-it can be arsenic, antimony, gallium or many other things-and you can grow very beautiful, smooth films that make very good transistors," said Tromp.
Today, according to Tromp, there are more than 1,000 papers on the different aspects of the surfactant phenomenon. Since then, Tromp has become interested in observing the growth of crystals at high temperatures. Crystal growth, until recently, was believed to occur far from equilibrium, but Tromp has shown that for semiconductor surfaces such as silicon and germanium, under certain conditions growth happens close to equilibrium. Tromp claims that analysis can be extended to the growth of quantum dots and quantum wires.
"The thermodynamics of crystal growth was written in the 1920s and 1930s, but when people tried to apply it to semiconductor growth in the 1990s, it didn't really work . . . it was thought that these high temperatures were far from equilibrium, but we demonstrated that not to be true. On the contrary, high-temperature semiconductor growth is very close to thermal equilibrium," said Tromp.
This discovery accelerated the burgeoning semiconductor industry, by enabling a detailed, quantitative analysis of the energy costs of atomic steps, the rate at which atoms diffuse over surfaces and how different materials interact during growth.
"You can use this theory for silicon growing on silicon, but you can also use it for understanding quantum-dot growth, which has applications for lasers and quantum computing, as well as for a lot of this nanoscale self-assembly stuff that we are working on," said Tromp. "For quantum dots, it should help us control the size distribution. Ideally, quantum dots should all have the same size and shape, but if you grow them without knowing very well what you are doing, you might end up with a very broad distribution-very small ones and very large ones. But if you understand the thermodynamics, then you can grow quantum-dot populations that have more desirable and consistent properties," said Tromp.
Tromp currently manages a molecular-devices and assembly group at IBM's T.J. Watson Research Center (Yorktown Heights, N.Y.), which is building new types of memory devices and transistors using molecular and organic systems-ultimately to develop capabilities that cannot be accomplished with silicon.
Tromp was the recipient of the 1981 Wayne B. Nottingham Prize, and received his PhD in physics in 1982. In 1984 he was elected to the Boehmische Physical Society. He received the IBM Outstanding Innovation and Technical Achievement Award in 1987, 1991 and 1992.
In 1993, he became an APS Fellow and in 1995 he received the Materials Research Society Medal. Tromp serves as a member of the U.S. Department of Energy's Basic Energy Sciences Advisory Committee, and is the author of about 200 scientific papers and the originator of six U.S. patents.
The APS also awarded the Irving Langmuir Prize in Chemical Physics to fellow research worker Phaedon Avouris, for his "pioneering contributions" to nanostructures and atomic-scale phenomena at surfaces.