|French, Alfred - Al|
Submitted to: Tetrahedron Asymmetry
Publication Type: Peer reviewed journal
Publication Acceptance Date: 12/8/2004
Publication Date: 9/1/2005
Citation: French, A.D., Johnson, G.P., Kelterer, A., Csonka, G. Fluorinated cellobiose and maltose as stand-ins for energy surface calculations. Tetrahedron Asymmetry. 2005. v. 16(2). p. 577-586. Interpretive Summary: Because the performance characteristics of cotton fabric depend on the detailed molecular structures involved, and because improvements in the performance will involve alterations of the detailed molecular structures, it is vital to know the structures before and after chemical, physical and biological modifications. One method for determining structure is computerized molecular modeling, which is undergoing development so that it can eventually fulfill its promise. In this study, the effects of replacing the hydroxyl groups with fluorine atoms were explored. This replacement simplifies the modeling studies, so high-quality calculations could be used. It was quite predictive of structures for the cellulose that could be observed under ideal conditions, but not as predictive for maltose structures. This information is primarily of interest to scientists engaged in modeling as well as those making use of structural information on carbohydrate materials such as cotton and starch.
Technical Abstract: To better understand computational predictions of disaccharide conformations, Phi,Psi maps were constructed for two analogs in which all hydroxyl groups were replaced with fluorine atoms (F-cellobiose and F-maltose.) Hartree Fock and density functional quantum mechanics (QM) theory were used. These molecules do not permit hydrogen bonding but should give better steric representation than previous analogs that use hydrogen atoms instead of exo-cyclic groups. The preferred ring shape for fluorinated glucose depends on the level of QM theory, but during mapping over limited f,y space, the rings remained in the 4C1 form. Also, fluorine atoms are remote enough that they do not affect the torsional energies for the glycosidic bonds. F-cellobiose maps seemed predictive of the conformations in crystals, but F-maltose maps were less so. F-disaccharide maps take far less computer time than a comprehensive QM study of the hydroxylated carbohydrates but are still expensive. The QM, F-cellobiose map and a more quickly prepared hybrid map based on MM4 and QM for cellobiose were similar. However, the hybrid maltose map had many more experimental conformations within its 2 kcal/mol contour than did the QM F-maltose map. The apparent mean strength of an intra molecular, inter-residue hydrogen bond is about 3 kcal/mol, based on the energy for many of the hydrogen bonded maltose structures on the F-maltose map. The F-maltose map was similar to one for an analog of maltose in which all hydroxyl groups were replaced with hydrogen atoms.