Submitted to: Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: 5/12/2009
Publication Date: 5/16/2009
Citation: Schnupf, U., Momany, F.A. 2009. Why use DFT methods in the study of carbohydrates?. Meeting Abstract #129. Interpretive Summary:
Technical Abstract: The recent advances in density functional theory (DFT) and computer technology allow us to study systems with more than 100 atoms routinely. This makes it feasible to study large carbohydrate molecules via quantum mechanical methods, whereas in the past, studies of carbohydrates were restricted to empirical/semiempirical and Hartree Fock (HF) methods. The application of DFT methods for modeling carbohydrates generates a common question in the scientific community on how accurately energies, structures, and properties can be determined. Furthermore, what DFT methods and basis sets should be used in order to produce high quality results? Here we compare the results of DFT calculations with those of commonly used ab initio methods such as HF, MP2 and CCSD. In addition, we address the issue of selecting proper DFT methods and optimal basis sets for carbohydrates. Considering both the cost and accuracy, DFT methods using hybrid-functions such as B3LYP with at least a 6-31+G* basis set are the best choice for hydrogen-bond rich carbohydrate systems.