May 30, 2003 – The co-founder of a nanotechnology startup that prides itself on its modeling capabilities predicts the industry will have commercialized design software tools like those used for microsystems, but not anytime soon.
He and other nanotech experts said the growth of nanoscale design tools will require a critical mass of people working on common products and acceptance by the field’s do-it-yourself software developers.
“Up to now, there’s been no real demand for it,” said Marvin Cohen, a physicist at University of California, Berkeley, and co-founder of California-based Nanomix Inc. “Right now, everyone develops their own.”
Cohen has been doing that for four decades at Berkeley.
Along with colleague and Nanomix co-founder Alex Zettl, Cohen developed a way to build carbon nanotube structures that act as chemical sensors. They also work on hydrogen storage systems for fuel cells. Their primary strength has been using theoretical modeling to predict the behavior of nanostructures.
Cohen devised a model that describes how materials are put together on the atomic level. The model, based on theories he has been developing since the 1960s, is used worldwide to calculate the properties of metals, superconductors and semiconductors.
He said the benefit is someone can theoretically figure out how things work at the nanoscale, saving time and money over trial-and-error approaches before taking a nanodevice into the laboratory. Nanomix is using Cohen’s proprietary techniques to virtually synthesize materials and predict electronic, physical and chemical properties of carbon nanotubes with a high degree of accuracy.
“To model how a nanotube will react to a given atom, whether it’s anthrax or carbon dioxide or hydrogen, you can use these techniques,” he said. There’s no reason to believe there won’t be as many applications of these approaches … as there are in the MEMS area.”
The field is broad, but Cohen can see ultimate demand for standard tools that all nano companies will face, such as a commercial software platform for testing and explaining properties of materials or particles.
Tim Harper, chief executive of nanotech business research firm Madrid-based CMP Cientifica, said there will be great software opportunities for nano, but it will require a critical mass of people focusing on one particular area — and that type of standardization doesn’t exist. For now, he said, many researchers take a cobbled approach: using general computer modeling and simulation suites and writing programs for their specific needs. “It ain’t pretty,” Harper said.
“We come across a lot of companies that say, ‘We’re in the Valley and in the software business. We’re going to develop software for nanotech.’ We say, ‘What particular flavor of nano?’” he said. “It’s like writing software applications for electricity. First you have to figure out if you’re writing software for a microprocessor or a toaster.”
One startup hoping to break into the space is California-based Junius Tech Inc., a nanoscale design and consulting services firm with technology and expertise from Stanford University and NASA. Today, the privately funded firm leans more toward providing services to outside firms, but co-founder Jonathan Woo said it is developing simulation software products.
The broadness of the technology is a risk to early players, Woo said, but Junius hedges its bets by offering customized consulting while it makes standardized tools for modeling and simulating nanomaterials for several applications, including electronic, optical, energy and biotech.
“It’s an evolution,” he said. “Think about any new technology that comes in — you have to grow with the industry.”
For now, however, most of the work is centered on research labs. The National Science Foundation-funded Network for Computational Nanotechnology at Purdue University provides access to simulation tools and services, develops new theories and approaches for simulation and links computational and experimental scientists to solve problems.
Rice University received a supercomputer from IBM to calculate interactions of subatomic particles in carbon nanotubes. IBM provided the $1 million system under a Shared University Research Award, which it gives to universities in fields of mutual interest. The partnership will allow IBM to develop more sophisticated software and systems applicable for nanoscale science and technology.
Cohen said most research funding is funneled into government and academic labs, and much of the work coming out is promising. Still, he cautioned against getting too excited about a rapidly rising market for software tools.
“I think it’s generally agreed that nanoscience is very healthy … (but) the commercial aspects of the field are not developed at this point,” he said. “The small companies, you hear a lot about them, but often the main product is hype. It’s just going to take a while.”