by James Montgomery, News Editor, Solid-State Technology

A technique pioneered for drug discovery has successfully been adopted for materials research, to speed up the process for finding and characterizing new structures. Early adopters of this method are realizing “powerful, sustainable competitive advantages” in terms of R&D speed and efficiency, according to Isy Goldwasser, president/COO of Symyx Technologies, speaking to a ConFab audience on Wednesday about “Lessons Learned from Other Industries.”

The process of “combinatorial chemistry”, which screens thousands of compounds for specific properties, has been used for years in drug discovery, but only in the past decade have other industries including materials and electronics looked to it to help them with R&D. Essentially, this method can help speed up the process for finding and characterizing new structures — e.g. which ones have specific magnetic, conducting, or optical properties — and increase materials discovery rate by factors of 1000 or more, Goldwasser explained.

Well-defined R&D projects go through three phases, he explained: synthesis, testing, and informatics, resulting in a step-change R&D execution (i.e., more commercial outcomes in significantly less time). High-throughput research, mainly a systems problem (only 25% of this method involves hardware), requires an informatics-driven culture and mindset, with sustained investment and specialization, he noted. In every case where the method has been applied, Symyx was told it was impossible to break R&D down into small-scale experiments and parallelize them. But it has always proved feasible.

R&D execution value drivers include increasing experimental throughput (by 10x-100x), reducing cost/experiment, integrating/automating complete R&D tasks and procedures, and enabling enterprise-wide access and reutilization of scientific information. Inputs must be intelligently selected, and analysis and evaluation must also be expertly structured.

This high-throughput research methodology has been applied in numerous areas by Symyx and partners, including workflows for process optimization, electrochemistry, gas phase heterogeneous catalysis, and pigments, Goldwasser showed. He also laid out a map for enterprise-wide efforts to deliver step-change R&D execution, including integrating, linking, and visualizing R&D data from the lab to plants, and forming solid alliance structures (e.g., Symyx with Dow Chemical and ExxonMobil) with collaborative, strategic R&D programs that transfer best practices and know-how. Small incremental R&D projects are done rapidly in parallel with data analysis to select the most promising candidates. The key is process sequence integration rather than material science, so that thousands of small-scale experiments can be performed quickly and at low cost/experiment.

Goldwasser described several case studies highlighting four areas of improvement: new product development, controlling product costs, building volume with existing yield assets, and improving chemical processes by analyzing existing feedstocks. He cited work with Dow to squeeze higher value products from existing operations, which resulted in the commercial launch of a new product family of plastomers and elastomers. The two firms also paired to push through what he termed a “rapid” R&D cycle for a new line of specialty olefin copolymers.

Providing more specific numbers to back the case for a high-throughput (HTR) R&D model, Goldwasser cited optimization of a portfolio of 100 formulation products (4-6 components per formulation), where HTR shortened the project length from an “impractical” three years to six months, and generated 1%/year reduction of the total portfolio’s $300M cost of goods sold.

Finally, targeting a commodity chemical process, Goldwasser noted that a partner took 10 years to see just incremental improvements in raising yields for a core product, without reaching commercial performance targets — but incorporating HTR by working with Symyx met technical targets after hundreds of thousands of experiments, and achieved more progress in the first two years than in previous 15 years. — J.M.

by James Montgomery, News Editor, Solid-State Technology

Three key approaches have helped Dow Chemical launch nearly 20 products over the last decade, rethink how effective R&D is done, and generate significant improvements in cycle times. These factors include a molecular architecture approach, a focus on “speed-based development,” and adoption of high-throughput technology, noted Kurt Swogger, VP of PP&C business development, in his Wednesday presentation at the ConFab in Las Vegas, NV.

Back in the mid-90s, Dow’s vision for success was not to do interesting science, Swogger explained, but rather to convert interesting science into good science — and, of course, make more money. Faced with a projected next cyclical bottom in 1999-2000, and normal product launch times approaching a decade or more, Dow wanted to shift its focus to higher-volume, higher-value specialty products (with the goal of making these products 20% of the company’s 1991 volume).

To achieve this involved adopting several approaches. Swogger discussed the hierarchy of Dow’s Insite technology platform, a molecular architecture approach supported by “wisdom, luck, or outside Dow sources.” A key advance was the ability to model molecular kinetics, which could help define final properties as well as provide solid data for scaling up processes from the lab to large reactors. This multipronged Insite approach was used to launch 10 products from 1993-2001, and another eight products over the past five years, a range of materials and chemicals including plastomers/elastomers, adhesives, polymers, and fibers. For polyolefins property relationships are structured by building molecular models from many (500-600) subroutines. The goal depends on the properties needed. This is being expanded to epoxies, urethanes, and water-based products (cellulosics).

Another key to the Insight strategy is “speed-based development,” defined as a practical philosophy for quickly converting science and technology into value. This is a concept applicable to any product in any market, ranging from architecture to autos to semiconductors, Swogger noted. Think of it as a five-pronged star of “rights,” he explained — the right leadership overseeing the right people, on the right projects, with the right information, and right execution.

The right leadership means creating and selling a vision that delivers. Using the right people means, for example, a scientist/manager partnership to lead development, and using a flexible allocation of resources. Also, projects must be selected with an eye toward “net present value,” and link business strategy to individual results, involving marketing early in the process. The right information means using science, involving customers early and utilizing experts (internal, consultants, etc). Performance requirements should be used for application development, and should link to prior experience to help avoid mistakes, he said.

All these add up to execution, governed by Six-Sigma methodology, Swogger noted, adding that other keys include making and delivering commitments, emphasizing decision making, and using a parallel process.

Swogger also discussed “high-throughput” research, which he described as a system of hardware, informatics, workflows, and researchers. He explained how a “high-throughput” catalyst development has accelerated Dow’s discovery rate, citing work with Symyx as an example, which started with polyethylene catalyst development and then pursued polypropylene. This approach is now being transferred across the company for formulations, material science, and chemistry workflows, and even licenses for new devices.

Combining speed and molecular architecture, Dow has seen cycle times drop from years to months for new product lines and from 6-24 months to 2-3 months for product extensions. Further, 17 of 19 new products under these strategies have been deemed “successful,” with eight out of 10 successful product extensions instead of one out of three. — J.M.

Investments to equip fabs will slow to just 3% growth this year, down from a 25% surge in 2006. Fab construction investments also are seen in the low single digits (4%-5%) this year — but both areas are expected to pick up again in 2008, according to a new report from SEMI.

Memory fab capacity will expand at double the rate of the rest of the industry this year (33% vs. 16%), with the industry’s top expanders led by Samsung and Hynix, as well as Intel and Micron. SEMI’s report also notes that 85% of all fab spending this year will go to 300mm activities, to increase 300mm capacity by 50% this year.

More than 30 major fab construction projects are taking place this year, soaking up $8 billion in investments ($10 billion for total construction by the end of 2008, when most of them will be online). Another 30 fabs will start volume production this year, with 16 fabs ramping in 1Q08-3Q08.

Geographically, spending is spread out fairly evenly. Taiwan and Japan each represent about 20% of all spending on equipping fabs, SEMI notes, while the US and South Korea make up about 18% each. Excluding Japan, the Asia-Pacific region is expected to account for more than half (53%) of fab equipment spending in 2007, surpassing $20 billion.

Within that A-P number is a big push from Singapore, where companies including Chartered, Qimonda, and the Tech Semiconductor and Intel/Micron JVs, will push domestic spending from $1.8 billion worth of equipment installations this year to about $3 billion in 2008.

The reports, including “FabFutures” and “Fab Capacity Report” as well as other related reports, were acquired earlier this year by SEMI from Strategic Marketing Associates.