August 2001

From Virginia Tech

Chemists use statistics to improve creativity

(Blacksburg, Va., August 29, 2001) -- Using combinatorial and in-parallel experimentation techniques accelerates creativity and allows researchers to make discoveries more quickly, says Timothy E. Long, professor of chemistry at Virginia Tech.

Jeff Labadie of Argonaut Technologies, Inc. of San Carlos, Calif., Long, and Warren Ford of Oklahoma State University co-organized a symposium on "Combinatorial and Highly Parallel Techniques for New Materials," to be presented at the 222nd national meeting of the American Chemical Society, Aug. 26-30 in Chicago.

"In creating and testing polymers, it is now possible to perform hundreds of experiments at once to determine the various characteristics of various combinations," explains Long. "The result is a lot of data -- so it is very important to do the right experiments," he says.

One way for a researcher to make sure she or he is doing the most productive experiments is to use a computerized statistical design of experiments program in combination with a parallel reactor such as the Argonaut Quest 205 -- to help design the experiments.

"We can use existing statistics software to design more efficient experiments. If you want to test reaction temperatures and reaction times for various component concentrations, the software can tell you which experiments to do in order to generate enough information to create predictive models. The philosophy is not to do only a lot of experiments, but to do the right ones," Long says.

The two-day symposium is being offered by the Division of Polymer Chemistry of the ACS on Tuesday, Aug. 28, from 1:30 to 5 p.m. and Wednesday, Aug. 29, from 8:30 a.m. to 5 p.m. at McCormick Place South Room S102B/C, Level 1. Tuesday afternoon, Labadie will preside at the section on Tutorials and Integrated Approaches.

Wednesday morning, Long will preside over the section on Synthesis and Surface Properties and, at 9 a.m. Wednesday, he will give a talk on one application, "Combining statistical design of experiments with in-parallel polymerization methodologies (Poly 444)." The paper will discusses work David T. Williamson, a Ph.D. student at Virginia Tech, and Long did using statistically-designed in-parallel experiments to elucidate the role of monomer concentration and temperature on living anionic polymerization of 1,3 cyclohexadiene. A statistical approach has enabled the discovery of reaction conditions that ensure the efficient production of novel elastomers (rubbery materials). The work was done in collaboration with Kraton‘ Polymers in Houston, TX.

Ford will preside over the section on Characterization and Bulk Properties on Wednesday afternoon.

"Doing things statistically and efficiently -- these are the tools of the future, the techniques we need to bring into the research lab and into the laboratory classroom," Long says.

Contact for further information:
Dr. Timothy E. Long
540-231-2480
telong@vt.edu












This article comes from Science Blog. Copyright © 2004
http://www.scienceblog.com/community

Archives 2001 E