Appendix C:
How ChemNMR Pro Works

---

Overview

---

ChemNMR Pro estimates chemical shifts for all hydrogen or carbon atoms for which additivity rules are available. Following a hierarchical list, it first identifies key substructures of a molecule. A substructure provides the base value for the estimated shift. For example, benzene would be identified as the key substructure of trinitrotoluene.

When a substructure happens to be a ring system not available in the data, ChemNMR approximates its base shift using embedded rings and, if necessary, it will even disassemble the ring into acyclic substructures.

ChemNMR views remaining parts of the molecule as substituents of a substructure. Each substituent adds to or subtracts from the base shift of the substructure to which it is attached. Additivity rules determine the increment of each contribution. If an increment for a substituent cannot be determined, ChemNMR uses embedded substituents-smaller structural units with the same neighboring atoms. Or, it will use increments of identical or embedded substituents of a corresponding substructure by assuming that the effects of the substituents are of the same magnitude.

ChemNMR provides a detailed protocol of the estimation process applied. It gives substructures as names, compound classes in most cases, substituents in form of a linear code, respectively.

The data set for the ¹H NMR Shift tool currently contains 700 base values and about 2000 increments. The ¹³C NMR Shift tool is based on 4000 parameters. It also implements models for ethylenes (cis/trans) and cyclohexanes (equatorial/axial).