By Claudia M. Caruana, associate editor
Encouraged by the positive experience of the pharmaceutical industry in speeding-up new drug discoveries using combinatorial chemistry, researchers here and abroad are starting to employ this technique to design better catalysts and create new materials more quickly. Already, new catalysts that possibly could be used for improved fuel cells (See below: Spotting Better Fuel-Cell Catalysts) are under consideration, and several industrial alliances have been formed to explore the possibilities of this technique.
Combinatorial chemistry uses a parallel approach to discover thousands of target materials and then produces "libraries" of these substances quickly, often in a few hours or days. The underlying principle of the combinatorial approach is to synthesize microscale quantities of compounds, which are placed on special grids, and test thousands of them quickly and reliably by a variety of methods such as laser and photoionization examination. Many biotechnology-based and other companies offer combinatorial chemistry systems that can produce a large library, plus provide rapid screening of candidates. The results also must be reproducible.
Dr. Ian Maxwell, technology manager for catalyst research and development at Shell Research & Development, Amsterdam, The Netherlands, says "combinatorial chemistry holds promise for improving the speed of innovation related to new and improved catalysts and novel materials. But, it should be more accurately be called, 'robotic synthesis and high speed screening.'"
In contrast, conventional discovery strategies usually are based on the time-consuming "one-sample-at-a-time" approach, which can take months or years to determine suitable candidate materials.
The potential in the CPI
Although the pharmaceutical industry is notoriously secretive about new drug discoveries, it is believed that several drug candidates developed using combinatorial chemistry are in early clinical trials. In addition, several major chemical companies, including Monsanto, DuPont, and Dow Chemical, according to Chemical and Engineering News, are using combinatorial chemistry to find new chemicals that might be useful as pesticides. And, the potential of this technology has not gone unnoticed by the chemical process industries (CPI), says Dr. Selim Senkan, a chemical engineering professor at the University of California at Los Angeles (UCLA), who is using this technique to optimize and develop better catalysts.
Senkan emphasizes that combinatorial chemistry, originally the bailiwick of chemists and biochemists, "is now ripe for chemical engineers to explore. These methods facilitate the rapid preparation and processing of large libraries of solid-state materials. Their use, plus the appropriate screening techniques, recently led to the discovery of materials with promising superconduction, magnetoresistivity, luminescence, and dielectric properties. But, equally important, combinatorial chemistry tools will help chemical engineers optimize catalysts, which has not been easy to do.
"Typically, catalyst discovery involves inefficient trial-and-error, because catalytic activity is difficult to screen. In contrast to superconductivity, magnetoresistivity, and dielectric properties, which can be visually observed, catalytic activity is difficult to detect. Combinational chemistry can help weed out deadends early on," he adds.
Senkan says the assessment of catalytic activity requires detection of a specific product molecule above a small catalytic site on a large library. The use of in situ infrared thermography and microprobe sampling mass spectrometry have been suggested by several researchers, "but while showing activity, the former provides no information on reaction products. The latter is difficult to implement because it requires the transport of minute gas samples from each library site to the detection system. This is the reason we developed the laser screening process," Senkan says.
Shell's Maxwell says the development of high-speed, positional resolution screening techniques is critical for the application to larger libraries with smaller site sizes. "Success in this area would, indeed, open up the possibility of exploiting combinatorial techniques as an efficient method for the discovery and optimization of solid-state catalysts," he adds.
"Because combinatorial discovery is capital intensive, continued innovations in hardware and software techniques will be necessary to drive down costs and facilitate its use in the chemicals and materials sectors," says Dr. John Hewes, program manager of the National Institute of Technology's (NIST) Advanced Technology Program (ATP) in Combinatorial Chemistry.
He emphasizes that new and better software will have to be developed that is suitable for creating better electronic databases and libraries. "Better library design will also require chemical synthesis expertise integrated with the use of design tools and statistics to reduce the number of samples and experiments. New tools will have to be developed to enable the input of this information into modeling engines, and the increased number of data that can be input into computational engines will increase their power in iterative cycles."
Adds Hewes: "Contrary to what many people might think, it simply is not a matter of using the technology developed for the pharmaceutical industry and transferring it to catalyst or new materials discovery. New technology, especially in sensors, needs to be developed for these applications."
[In November, NIST's ATP program is sponsoring a one-day workshop in Atlanta, GA, focusing on combinatorial chemistry for catalyst and new material development. For information, contact NIST at http://www.nist.gov/fallmeeting or (301) 975-5416.]
But, not everyone contacted by CEP editors believed that the use of combinatorial chemistry is a panacea for new catalyst development and new materials discovery. One consultant acknowledged that while combinatorial chemistry can be advantageous for faster, more systematic drug discovery, he did not believe this would be the case with catalyst and materials. "A catalyst library is smaller than a biological library. The biological library, however, is easier to generate. We also are looking for a lot of different things, for example, how increasing the use of one catalyst in a combination changes the effectiveness of another. Combinatorial chemistry might be an effective tool is certain situations, but it is not a substitute for a good catalyst person who knows which combinations to try and which not to try."
Strategic alliances, joint venture
Nevertheless, increased interest is evident from recent business linkups. Early this year, for example, Santa Clara, CA-based Symyx Technologies, a start-up, privately held company that has created combinatorial material libraries for the chemical and electronics industries, signed collaboration agreements with Bayer AG, Leverkusen, Germany, Celanese, Dallas, TX, and Ciba Speciality Chemicals, Basel, Switzerland. With them, Symyx will use combinatorial chemistry to help discover and develop new catalysts, polymers, electronics materials, and new pigments.
Last year, Symyx began collaborating with Hoechst AG, Frankfurt, on long-term catalyst research. Symyx engineers have also announced that they have discovered a new blue-white phosphor after screening more than 25,000 luminescent materials at a time using combinatorial methodologies. This phosphor has a novel luminescence mechanism that may have potential in products such as flat-panel displays, fluorescent lights, and computer screens.
At press time, DSM Research, the research and development organization of DSM, NV, Geleen, The Netherlands, and Cambridge Combinatorial, Cambridge, U.K., have formed what is believed to be the first European joint venture focusing on the use of combinatorial chemistry to discover new catalysts and optimize already existing ones. This project is expected to continue until 2000.
Drs. Thomas E. Mallouk, a chemistry professor at Penn State and Eugene S. Smotkin, a chemical and environmental engineering professor at IIT, knew that methanol presents a special problem for fuel cells because its oxidation in the cell poisons the catalytic electrode surface. "The platinum catalysts that work so well in hydrogen fuels cells are basically useless for methanol," Smotkin says. "They do not adsorb water, which is needed to oxidize away the carbon monoxide that builds up on the platinum surface. That is why platinum alloys containing elements that bind the oxygen atoms in water make much better catalysts."
Although they had a good idea of which elements to mix together, making and testing these alloys was a time-consuming process that increased if additional elements were added. "To do a reasonable job at testing four elements, you would have to look at hundreds of catalysts," Mallouk says. "This could not be done in a reasonable time by making them serially."
Instead, the group devised a method for making and testing hundreds of catalysts simultaneously. Using an ink-jet printer, they printed dots of metal salt mixtures onto a large carbon electrode. Each dot, about the size of a lower case "o" in this article, contained a slightly different mixture of five elements: platinum, ruthenium, osmium, iridium, and rhodium. All the dots were converted into alloy catalysts by a solution process similar to that used to make bulk fuel-cell catalysts. To establish the catalytic activity of each dot, they converted the electrical current to an optical signal.
"The good catalysts lit up, similar to a litmus test," Mallouk says. "Wherever methanol is oxidized on the array electrode it generates acid. A fluorescent acid-base indicator in the solution pinpointed the most active dots, where the concentration of acid was highest."
Using this method, they were able to determine quickly which might be best catalysts. These catalytic compositions were made in larger quantities at ICET and later tested for suitability in methanol/air fuel cells by Smotkin and his coworkers. Finally, a quaternary alloy containing platinum, ruthenium, osmium, and iridium was selected, Mallouk notes. The performance of the quaternary alloy is between 40 and 100% better than the best binary alloy catalyst for this application, he adds.
Their research was supported by the Office of Naval Research, the Army Research Office, and the Defense Advanced Research Projects Agency.
© Copyright 1998 American Institute of Chemical Engineers. All rights reserved. Copying and downloading permitted with restrictions.