Journal of Chemical Education, Vol. 55, issue 5, pg. 322 (May 1978)

Computer Assisted Analysis of Nuclear Magnetic Resonance Spectra

In our introductory organic course the student is taught an integrated approach to solving problems in spectroscopy. This brings together infrared, mass spectroscopy, nuclear magnetic resonance and any applicable ultraviolet data. The student then assembles what structural data he can from each type of spectrum and proposes a structure. It seems that, for many students, interpretation of nmr spectra is more difficult to grasp tahtn is any of the others. The average student needs experience in interactive practice sessions to aid in grasping the analysis of nmr spectra. Unfortunately, if much class time is spent discussing the finer points of nmr spectroscopy, other areas may suffer. But without these interactive practice sessions students will spend an inordinate amount of time in analyzing unknown spectra.

Thus we have implemented an interactive computer program designed to assist the beginning student and as a quick review for advanced students. This program, capable of giving assistance and rapid feedback to a student on his performance, was written in the hope that it would serve to supplement classroom exercises on spectral identification. In additon, the student must become more intimately involved in the actual analysis than in a classroom discussion.

The first problem in designing such a program is to analyze the way in which one would normally interpret an nmr spectrum. For an unknown compound the obvious features such as corboxylic acids, aldehydes, aromatic or vinyl protons are the first signals to be observed. After analysis of any signals present in these regions, one would then preceed to analyze any signals appearing in these regions, one would then proceed to analyze any signals appearing in the alipatic regions, spin-spin splittings, and coupling constants. Then, by assembling these bits of information and integrating the various signals, deduce a structure consistent with the available information.

This approach was utilized in the program design. The student is lead from low field to high field signals. The program, written in Fortran for a Xerox Sigma 6 system, presents a series of questions about the presence of specific structural features. Thus the first question asked is "Do you see any acids?" to which a student responds "Yes", "No", or "Help". If the student responds either "Yes" or "No" he is told whether he had made the correct decision. "Help" brings information about the characteristic chemical shift of the functionality in question and any other applicable data. There are, at present, twelve practice spectra coded in the program. These spectra are photocopied from "The Aldrich Library of NMR Spectra".1 Thus the student has at hand a copy of an actual spectrum for analysis. On completion of the run, the student is reminded of the structural features he correctly assigned to the unknown and asked to propose a structure and name the compound.

Provision has been made for a student to request assistance with one of his own unknowns. This second mode of operation leads the student through the analysis scheme in the same systematic approach as with the practice spectra. Most of the common types of proton signals have been incorporated in the program. Initial student comment has been very favorable.


1Pouchert, Charles J., and Campbell, John R., "The Aldrich Library of NMR
Spectra", ALdrich Chemical Co., Milwaukee, 1974.
2Author to whom correspondence should be addressed.
Andrews University                              Keith G. Calkins
Berrien Springs, Michigan 49104                 Richard F. Daley2