Computational Chemical Software to Aid in the Structural Determination of Organic Molecules by Means of Nuclear Magnetic Resonance

点击翻译按钮获取中文摘要

(1) Program CONFLEX for Conformational Sparch of Flexible Molecules The program was completed within the year 1992 and has been deposited in the Japan Chemistry Program Exchange for public use. The algorithm of CONFLEX is unique in that the Program not only effects systematic rotations to the rotatable bonds, but also simulates elementary processes occurring in the thermal conformational interconversions. A remarkable aspect of this algorighm is its ability of confining the search to within the low-energy region and still capable of finding new domains in the conformational space by crossing over energy barriers. One can arbitrarily control the extent of search depending on the nature of computer resource and the purpose of search. Careful tests revealed that CONFLEX is superior in its performace than any other existing programs. Thereupon the program has been applied to a number of pending problems including the origin of regiospecific O-methylation reaction of Erythromycins, the caus … More e of failure in taxane synthesis and other problems involving macrocyclic conformation and macrocyclization reactions.(2) Program for Universal Molecular Mechanics Calculations Within the year 1992, we have been able to complete that part of molecular mechanics force field program which takes care of the careful evaluation of solvent effects on intramolecular steric interactions. The original idea of this part rests on Clark Still, Columbia University, but we have modified the original formulations to a considerable extent. Technically speaking, we use the first and second analytical derivatives of molecular volume, not the molecular surface as other investigators does, as the key variables. This part is general and can be implemented in other force field programs.(3) Program 3JHH2 for Calculation of Vicinal Coupling Constants The present version of program 3JHH2 has been critically analyzed, together with other known algorithmus for computing coupling constants, for its performance over a variety of organic molecules. As it turned out, all of the presently known methods have unexpectedly low precision in reproducing and predicting coupling constants. Unexpectedly, the greatest source of error is the experimental precision. It is concluded that a new approach is necessary in the use of Karplus equation. Based on the conclusion we will start a new project beginning from the year 1993. Less