Automated research: Not just for pharma


Nanotech development is beginning to benefit from methods popular in the pharmaceuticals industry

By Paula Doe

Like many nanotech companies, Intematix Corp. has an innovative new material-in this case yellow phosphors that convert blue LED light to white. But potentially far more interesting than the material itself is the automated research technology the company developed to design it. The new technology affords the nano niche the same kind of high-speed mass screening that has revolutionized pharmaceuticals development.

Intematix is not alone in its use of automated research for nanotech development. Japan’s JSR Corp. is selling a specialty polymer for electronics applications that Symyx Technologies Inc. developed using its high-throughput synthesis and screening system. Symyx says it is increasingly working not just with giant chemical companies, but also with smaller companies across a wide range of technologies.

Dramatic production improvements

Meanwhile, Intematix has gotten plenty of attention for its growing phosphor business. It has started producing commercial volumes of the phosphors to meet the growing demand for low-power lighting in portable electronics displays and says sales have tripled during the past year.

The company developed the new phosphor to meet required specifications in only 14 months. It accomplished this by using combinatory synthesis to make tiny dots of thousands of candidate compounds, and massive parallel screening with automated tools to select the ones that worked best. In doing so it also avoided the prior method of using Ce-doped YAG yellow phosphors. Intematix co-founder Xiao-dong Xiang did the early groundbreaking work on combinatory synthesis with Peter Schultz, who was affiliated with Affymax (which spawned Affymetrix) and founded Symyx to apply the approach in materials science.

Symyx pioneered the automation, miniaturization, and parallel processing of thousands of experiments.
Click here to enlarge image

While the high-volume screening method has significantly influenced drug discovery over the past decade, it hasn’t yet had much impact on other applications. The process still starts with modeling-choosing materials thought most likely to have the desired characteristics. “But in materials, our knowledge is not enough to pinpoint the parameters exactly, so we have to experiment,” says Xiang, who is now Intematix’s chief scientist.

His company created a behemoth machine it calls a “discovery engine” to deposit and anneal thin films-applying multiple materials in overlapping continuous gradients of varying composition, for the physical equivalent of a phase diagram-with thousands of possible variations of composition on a square centimeter of substrate. Intematix developed an automated x-ray microprobe that can resolve crystal structure and composition finely enough to analyze each dot on the matrix to screen the variant compositions. A noncontact, microwave-based microprobe analyzes optical and electrical functionality. The screening tests generate a global image to show the results in patterns of color. Xiang explains the dramatic acceleration this affords: “The complete cycle for a phase diagram of one million pixels is one day, or a week for the entire analysis-for what could otherwise take years,” he notes. “But the real challenge is to select the right application,” Xiang adds. “The process still takes a lot of time and money.”

Enabling quick market response

Conveniently, many customers came requesting new phosphors for white LEDs.With IP particularly tangled, new entrants in Taiwan and Korea muscling in, and new applications requiring phosphors that can withstand higher temperatures and produce warmer light, this happens to be a market of significant volume with many opportunities. The market for white LEDs in cell-phone displays is now largely saturated, but backlights for mid-sized (7-inch) displays for portable DVD players and car navigation systems are currently driving demand, says Intematix CEO Magnus Ryde. He adds that low-power backlights for laptops are likely to be the next driver, and demand for solid-state lighting is also picking up as quality improves and costs decline. “We’re seeing quicker adoption than anticipated in solid-state lighting,” he says, “in part because China is pushing ahead really aggressively to cut energy costs.”

Of course even in the bulk chemical industry, where Symyx has been developing high-throughput synthesis and screening systems to speed research on catalysts for a decade, the first major commercial products have reached the market in just the past couple of years. Symyx discovered new catalysts that enabled the Dow Chemical Co. to commercialize its Versify plastomers and elastomers, and ExxonMobil to develop the more efficient catalyst it plans to put into production in refineries this year. In addition, Dow used Symyx tools to develop the catalysts enabling its new Infuse specialty polymers. Both Dow and ExxonMobil say the automated synthesis and screening of micro quantities of hundreds or thousands of candidate materials greatly accelerated their development of major products. Dow reportedly brought its new Versify products to market in four years, about half the usual time, and has said it would never have found these catalysts using conventional research methods. ExxonMobil says in its annual report that the high-throughput approach doubled its R&D output.

New benchtop systems are starting to make the synthesis and screening tools more modular and more affordable for smaller companies and smaller markets. Isy Goldwasser, Symyx’s president, says, “As we develop more modules, it’s increasingly bringing down the cost of entry for using the high-throughput approach.” But a user will still have to invest from $500,000 to several millions of dollars to build an automated system, he points out, so it only makes sense for developing materials with markets of at least tens of millions to hundreds of millions of dollars, or for companies that can use the tools to develop several different products at the same time. “In the future, though, the price may drop by twofold to fivefold,” suggests Goldwasser. “That would enable a whole new range of applications.”

Symyx has started to sell some systems to university labs and has supported startups applying its technology to new areas. Besides its catalysis tools, the company has been working on performance materials and has developed automated systems for testing small samples for such qualities as adhesion, strength, compatibility and stability. It spun out Visyx to develop sensor technologies to measure oil conditions in cars and trucks and invested $13 million in Intermolecular, a start-up by some former Applied Materials executives developing applications for the semiconductor industry.

Intematix, meanwhile, has also applied its system to the design of catalysts that use smaller amounts of costly metals for fuel-cell applications. Work for DOE that involved screening hundreds of catalysts for releasing hydrogen from chemical storage systems found three candidates that worked almost as well as ruthenium-for about 3% to 5% of the cost. Other research has developed alloys of 10% to 30% platinum to replace pure platinum catalysts in methane fuel cells, using a thin film coating on a nanostructure for efficient exposure of the catalyst. The company aims eventually to develop phase-change materials for nonvolatile memory.