Premium
Relationship between protonproton NMR coupling constants and substituent electronegativities. V —Empirical substituent constants deduced from ethanes and propanes
Author(s) -
Altona Cornelis,
Ippel Johannes H.,
Hoekzema Aldert J. A. Westra,
Erkelens Cornelis,
Groesbeek Michel,
Donders Lambertus A.
Publication year - 1989
Publication title -
magnetic resonance in chemistry
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.483
H-Index - 72
eISSN - 1097-458X
pISSN - 0749-1581
DOI - 10.1002/mrc.1260270609
Subject(s) - electronegativity , chemistry , substituent , coupling constant , isopropyl , proton , computational chemistry , chemical shift , yield (engineering) , analytical chemistry (journal) , stereochemistry , thermodynamics , medicinal chemistry , organic chemistry , quantum mechanics , physics
The electronegativity dependence of the torsion angle‐independent term in the Karplus equation, i.e. of the ‘constant’ A in the Fourier expansion A + B cos ϕ + C cos 2ϕ +…, was investigated. Experimental proton‐proton coupling constants of substituted ethanes and isopropanes appeared to be suitable for this purpose. A data set was constructed which contained 70 couplings, newly measured or remeasured at 300 MHz, and 25 couplings taken from the literature. The accuracy of each data point is estimated as ≤0.02 Hz, with a few exceptions. The actual analysis was carried out on 93 data points, i.e. on J values of 55 mono‐ and of 38 1,1‐di‐substituted ethanes, including 22 isopropyl derivatives. A total of 55 chemical groups is represented in the set; some of these were taken together, leaving 50 distinct groups. Regression analysis of the present data versus standard electronegativities did not yield acceptable results. Instead, substituent parameters λ e , valid for 3 J (HH) in saturated HCCH fragments, were derived in a least‐squares procedure from the data set. The couplings from mono‐ and 1,1‐di‐substituted ethanes could be accounted for in a simple expression that contains an interaction term C 012 (λ 1 λ 2 ). The best equation obtained is\documentclass{article}\pagestyle{empty}\begin{document}$$ {}^3J{\rm (HH) = 7}{\rm .84 - 0}{\rm .59(}\lambda _{\rm 1} + \lambda _{\rm 2}) - 0.42(\lambda _{\rm 1} \lambda _{\rm 2}) $$\end{document}The parameters are valid for λ e values scaled according to the Huggins electronegativities: λ H = 0, λ OR = 1.40. The equation fits 84 experimental couplings with a root‐mean‐square deviation of 0.018 Hz and a maximum deviation of 0.06 Hz. Some exceptions occur. (i) CH 3 CH 3 , CH 3 CHCl 2 and CH 3 CHF 2 appear to follow a different, but correlated, regression; (ii) C(O)H, C(O)R, SO 2 Cl and POCl 2 groups require different λ e values according to the substitution pattern, i.e. mono‐ or 1,1‐di‐substitution. The striking difference between the new λ e substituent‐effect scale and other empirical electronegativity scales lies in the inverse correlation of λ e with increasing electronegativity of β‐substituents. The inverse relationship is not only found for α‐carbon atoms, but appears to represent a general phenomenon, also seen for substituted α‐hetero atoms (O, N, S).