Although this increase is broadly in line with industry as a whole, as indicated in the 22nd annual R&D Trends Forecast from ACC’s Arlington neighbor, the Industrial Research Institute, it isn’t exactly a sign of an industry rushing to invest in an activity that has long been at the core of its creative and competitive success.
If anything, there appears to be a rush to take R&D out of the country. More and more chemical companies are opening new facilities in emerging economies like China and India. Dow Chemical, Midland, Mich., for example, recently announced plans to invest more than $200 million in manufacturing and R&D facilities for its Dow Epoxy subsidiary in Shanghai, while DuPont, Wilmington, Del., last year opened a facility in Zhangjiang near Shanghai. This isn’t just a U.S. trend either. Earlier this year BASF, Ludwigshafen, Germany, whose U.S. operations are headquartered in Florham Park, N.J., opened its first nanotechnology research center in Asia, in Singapore, while Degussa, Düsseldorf, Germany, and Parsippany, N.J., set up R&D labs in Mumbai, India, in 2004.
The lower cost of operating R&D labs in these countries obviously is a factor behind this trend, although it also reflects other competitive issues such as the need to be closer to increasingly important markets. Regardless, these moves underscore a broader pressure facing R&D — the constant weighing of its cost against the rewards it can deliver.
“If you’re going to be competitive in the chemical industry,” says Kurt Swogger, Dow’s vice president for performance plastics and chemicals R&D, based in Freeport, Texas, “you have to have the ability to be much more productive with your research dollar.” One relatively new (at least to the chemical industry) way of enhancing R&D productivity is via high throughput experimentation, coupled with the combinatorial chemistry method pioneered in the pharmaceutical industry.
This so-called combi/HTE approach has already brought Dow commercial success — first with a range of polymers and elastomers called Versify and this year with a new olefin-based elastomer range known as Infuse. What these products have in common is that they are both produced with catalysts discovered using high throughput techniques of Symyx Technologies, Santa Clara, Calif.
Last year Dow entered into a $120-million, five-year alliance with Symyx after working with the company for several years on the development of the Versify catalysts. “We started with catalysts,” explains Swogger, “but now we’re looking at materials science, polymerization and other chemistry that can be used with high throughput techniques.”
Symyx’s customer list reads almost like a “Who’s Who” of the chemical industry. The company “currently has 15 licensed discoveries that are in development and nine emerging development candidates,” says Teresa Thuruthiyil, its vice president for investor relations. One of those near-term development materials is a refining catalyst for ExxonMobil, Plano, Texas, which has a five-year strategic alliance with Symyx worth $200 million. And in April, Celanese, Dallas, Texas, introduced a new catalyst to reduce the cost of manufacturing vinyl acetate that was discovered by Symyx in another research collaboration. “We have worked with Celanese for several years,” Thuruthiyil notes. “The research to identify the catalyst was done at Symyx by scientists in our laboratories using our proprietary equipment, while the development was done by Celanese at their facilities.”
Founded in 1994, Symyx was one of the first companies to adapt the techniques of combinatorial chemistry and high throughput experimentation, which even then were becoming established research tools in the pharmaceutical industry, to materials science. It has developed a range of proprietary tools and software that enable researchers to generate hundreds to thousands of materials at a time, and then to rapidly screen them for desired properties. Its Polyolefin Workflow tool, for example, was used in the catalyst developments cited above.
Combining research and equipment
Although as Thuruthiyil says, “the labs are at the heart of what we do,” approximately half of Symyx’s business is now in the supply of its tools and software to companies — including Dow — wanting to take high throughput techniques into their own R&D departments.
Some of the tools — though not the highest throughput systems, says Thuruthiyil — are also licensed to other contract research and equipment companies, such as Solvias, Basel, Switzerland, and Fort Lee, N.J., and Biotage, Uppsala, Sweden, and Charlottesville, Va. Solvias recently launched its first full-scale high throughput catalysis screening service for asymmetric homogeneous hydrogenations, based on specially adapted Symyx instrumentation, while Biotage offers the Endeavor catalyst screening system. Originally licensed to Argonaut Technologies, which was acquired by Biotage last year, Endeavor is a parallel multireactor synthesizer that, with eight small reactors, is basically a smaller version of Symyx’s Parallel Polymerization Reactor (PPR) system of 48 reactors with a throughput of nearly 100 reactions a day (Figure 1).
Figure 1. Compact screening unit includes 48 separate reactors.
This dual business model approach — part research company, part equipment supplier — also applies to the German company hte AG, Heidelberg, which was founded in 1999, backed by two key collaborations with BASF’s catalysts department and ChevronTexaco’s research and technology center. In the last two years the systems side of hte’s business has grown strongly, says Dirk Demuth, CEO of the eponymous “high throughput experimentation” company. “There are really two schools of thought at play,” he says. “There are companies who build on strong outsourcing activities, even in R&D, and companies that want to have the equipment in house.”
About 80% of hte’s work is still catalyst related but lately this has been broadened to include process formulations such as surface coatings. The company has long-term collaborative research projects with BASF, Chevron, San Ramon, Calif., and Albemarle, Baton Rouge, La. in the field of refinery catalysis. It also recently announced that it is to collaborate with the Shell Group company CRI Catalyst, Houston, Texas, on the discovery and development of new ethylbenzene dehydrogenation catalysts for styrene monomer production.
The hte equipment finds use in stage I and stage II catalyst-screening systems. The micro-scale stage I “single bead reactor” is akin to the combinatorial methods of drug development, in which uniform spherical 1-mm beads of material are sequentially impregnated to create compositional libraries, with up to 625 beads positioned on a reaction plate no bigger than a credit card (Figure 2). The catalysts are screened at ambient pressure and temperatures up to 400ºC.
Figure 2. This microbead reaction plate serves for stage I catalyst screening.
For stage II work — basically optimizing the process efficacies of catalysts that come through the screening process — hte has developed 16- and 48-reactor systems, designed using CFD to ensure that they behave as ideal plug-flow reactors. Systems are available for reaction pressures up to 150 bar and operating temperatures to 700ºC; hte can configure test units to accommodate its clients’ chemistry as required.
Another research company that has recently branched out into selling HT systems is Avantium Technologies, Amsterdam, the Netherlands, which this year opened a branch office in Davenport, Iowa, “to better service our fast expanding customer base in the U.S.,” according to Guus Scheefhals, chief business officer.
“In the past our business was focused on providing high-value-added research services,” notes Tom van Aken, Avantium’s CEO. However, he explains a change was necessary: “While we’ve been successful in growing our services business [Avantium this year entered into a strategic collaboration with BP to apply HT techniques to R&D related to purified terephthalic acid (PTA)], it became clear to us that we couldn’t access part of the market by selling services only.”
Avantium increased in-house HT capabilities at its Delft facility last year with the installation of its “Polyolefin Platform” — an eight-reactor system with integrated handling of catalysts, co-catalysts and additives under inert atmosphere — which has already attracted “three leading polymer companies” to commit to research projects. At the same time, however, it was launching its first HT tool onto the open market — the Crystal16 system and software for crystallization studies. And it has followed this up in 2006 with the Block96 system, designed for HT experiments on a 1- to 2-ml scale in up to 96 parallel high-pressure batch reactors.
Demuth certainly isn’t alone in seeing a surge of interest in high throughput (HT). “In general, the concept of high throughput experimentation is becoming more widely accepted,” agrees van Aken. Likewise, Duncan Akporiaye, chief scientist with independent research foundation SINTEF, Oslo, Norway, and Houston, Texas, echoes the view. “There was a period when it was rather difficult to try and sell the concept [of HT],” he says, “but in the last couple of years the interest has suddenly taken off. Now most of the major companies are involved in some way.”
SINTEF started developing high throughput systems in 1995 and has had a strategic alliance with process technology licensor UOP, Des Plaines, Ill., since 1997. The first result was a more-environmentally-friendly catalysts system for paraffin isomerization, which now is in regular service at Big West Oil’s refinery in Salt Lake City, Utah. “In this study we were able to look at over 500 formulations in five weeks; using traditional means would have required about a year for preparation and two and a half for testing,” notes Jennifer Holmgren, UOP’s director of exploratory and fundamental research.
Figure 3. This 48-array gravimetric unit is designed for parallel processing applications.
Working with SINTEF, UOP has now put together a procedure, called CombiVision, that uses HT at every stage of development — from material synthesis (of zeolites, for example), through material modification (such as metal loading and ion exchange) and characterization, and finally the reaction assay module — based on SINTEF’s CombiChem parallel technology (Figure 3). “Tying it all together,” says Holmgren, “is an informatics system that includes statistical experiment design, automation, data analysis, visualization and modeling tools.”
Developments elsewhere in Europe also testify to the heightened interest. Chemspeed Technologies, Augst, Germany, and Monmouth Junction, N.J., for example, has worked with BASF on the development of “MiniPlant” reactor technology for HT polymerization studies; flow chemistry specialist Syrris. Royston, U.K., last month launched its new Atlas automated lab synthesis system in the U.S. at the American Chemical Society’s exposition in San Francisco.
Meanwhile, the European Union-funded TopCombi project is now in full swing after its launch last year. A €23-million, five-year project, TopCombi — “Towards Optimized Chemical Processes and New Materials by Combinatorial Science” — is a collaboration between 22 industrial and academic partners from 11 European countries, aimed at developing and applying HT techniques to the discovery of more environmentally friendly catalytic processes. Maria Raimondi, who is with one of the TopCombi partners, Insight Faraday, a U.K.-government HT initiative, says one of the priorities of the program is to develop a central library that will provide both academic and commercial HT research labs with easy access to high quality commercial catalyst samples from producers such as Johnson Matthey, London, U.K., and West Chester, Pa.
Another TopCombi partner, InforSense, London, is contributing that all-important aspect of all high throughput techniques, the “informatics” or IT infrastructure required to handle the vast amounts of reaction data generated by the systems, apply statistical techniques to the design of experiments, and provide the level of automated control necessary for parallel experimentation. (Other companies that have informatics tools on the market include Accelrys, San Diego, Calif., which was set up in 2001 from a group of five scientific software companies involved in everything from molecular modeling to the management and mining of experimental data.)
As Dow’s Swogger says: “The amount of information generated off one of these [HT] systems is awesome. What people don’t realize about HT is that it’s not just a bunch of automated equipment that does hundreds of experiments. If you can’t handle the data you’re in trouble. That’s where we look at the benefits. High throughput is expensive, so you always have to make sure the problems you work on are the ones that are going to pay you back.”
HT is viewed as expensive, agrees Mike Fasolka, director of the Combinatorial Methods Center at the National Institute of Standards and Technology, Washington, D.C. “Many companies believe that getting into HT is too costly for them and, depending on how they go about it, this could be true,” he says. “At NIST our aim is to develop HT techniques which don’t require an expensive laboratory infrastructure.”
Expensive or not, high throughput experimentation certainly looks well placed to boost R&D productivity across the chemical industry.