BLACKSBURG, Va., — Daniel Crawford’s research in theoretical and computational quantum chemistry has earned him a National Science Foundation CAREER Award designed to encourage young researchers.
"We apply models within quantum mechanics to compute properties of individual molecules to compare with or predict experimental data," said Crawford, whose award is for $435,000 over five years. Some molecular properties of interest include structure, thermodynamic data, and spectra.
"We can also compute properties that cannot be measured experimentally," Crawford said. In addition, he said, "there are some molecules you don’t want to create experimentally—particularly ones that are highly toxic or explosive, for example."
Crawford’s proposed research is the study of large chiral molecules, such as amino acids, which have an inherent left or right handedness. The properties of each hand can be quite different. One example is the drug thalidomide, which was banned by the FDA in the 1960s. One hand of thalidomide has a sedative effect and the other damages fetal tissue.
A major area of chemical research today, Crawford said, involves the analysis of natural products—for example, compounds isolated from marine species, which often have biological importance. In order to synthesize such molecules in the lab, organic chemists must know which "hand" they have isolated. "They often have to test the end product using methods such as circular dichroism (CD) spectroscopy, which shows a different spectrum for left and right handedness," Crawford said. "Then they can compare the result with the original natural product to see if they have the right configuration. So only once they have taken all that time to generate a synthetic version can they see if it’s the correct hand."
This is where Crawford’s proposed research comes in. "When we prepare a calculation, we have to tell the computer what the structure is, including its absolute configuration," he said. "The proposal is to compute rapidly the CD spectrum of all possible hands and compare those results to the natural-product spectrum. If the theoretical spectra are accurate enough, we should be able to identify positively which structure the experimentalists should target in the laboratory.
However, quantum computation is very complicated mathematically and requires vast computer resources such as memory, speed, and disk space. "When the molecule gets larger, the calculations can become dramatically more expensive to run," Crawford said. For example, high-level calculations for an amino acid such as valine might take about five days, but calculations for the valine-valine dimer (two valines together) take not 10 days, but more than two years.
"We have ideas for significantly improving the scaling of these models," Crawford said. "We’re breaking the molecule into pieces and dealing with it a fragment at a time and then putting it together at the end. With this approach, we can leave out electronic interactions on pieces of the molecule that are widely separated." If these reduced-scaling techniques are successful, Crawford envisions his work helping to speed up the synthesis of natural products that can be used for such purposes as medicine.
Crawford, an assistant professor, has received several awards for his research, including the New Faculty Award from the Camille and Henry Dreyfus Foundation and a Research Innovation Award from the Research Corporation. He joined the Virginia Tech faculty in August of 2000 following a postdoctoral appointment at the University of Texas.
PR CONTACT: Sally Harris, 540-231-6759, slharris@vt.edu