Self‐association of thiols. The baggy fit problem in weak complexes

Abstract

Abstract-The PMR technique has been used to obtain thermodynamic data for hydrogen bonding of alkanethiols (RSH) in 1:l dimers in carbon tetrachloride. At ca. 303°K these are (R, C,HI1, 1.3 2, 0.7 & 0.3, 15 f 1. Alkanethiol self-association is weak, and although an exact expression [Eqn. (5)] reproduces spectral data precisely, the fit is sufficiently 'loose' or 'baggy' so that values of K, AH" and AS' are uncertain. The methodology of the treatment of self-association data and their errors is examined and Deranleau's useful approach is extended. The impossibility of obtaining reliable data for very weak (< 10 %) or very strong (>90 %) associations by techniques equivalent to ours is emphasized. The possibility of cyclic thiol dimers is discussed. It is suggested that the PMR method cannot give trustworthy selfassociation data for aryl or arylalkylthiols because of the relatively large anisotropy effects introduced into the dilution shift.IO4KDESPITE its importance, little is known about the quantitative aspects of hydrogen bonding at a sulfur site. Certainly, many qualitative observations and a few equilibrium constants ( K ) have been r e c~r d e d ,~.~, * but there is only one complete set of thermodynamic data, i.e. for hydrogen ~ulfide.~ Although infrared (IR) techniques have been applied to this problem, the extraction of K values from the observations appears to be difficult or ~n c e r t a i n .~,~ to * Proton magnetic resonance (PMR) spectroscopy, which is the approach we used, is not without its p r o b l e m~.~~~J~J l Nevertheless, we obtained complete thermodynamic data, viz. K , AH" and AS', for the self-association of five alkanethiols in carbon tetrachloride. 2RSW + (RSH),(1) In recent years, there has been increased concern over the theory and methodology of studying association phenomena.ll to l4 Advances, especially computer processing of data, have rendered some older measurements, controversies, experimental strategies, etc., irrelevant or obsolete. Perhaps even more important to this work is the suggestion that conventionally obtained 'association' constants of weak complexes by nonthermodynamic measurements probably include 'sociation' constants, which involve those 'complexes' over and above those one would estimate from probabilities for a random mixture.12 For this reason, we cannot rely on most of the previously published equilibrium data for selfassociation of t h i o l~.~,~,~J~ In fact , our present analysis provides a good example of the vulnerable state of the art in the area of weak hydrogen bonding, as studied by PMR. EXPERIMENTALThiols were carefully distilled and the middle cuts were stored over Linde 4A molecular sieves; checks by VPC and IR and PMR spectroscopy established their purities (>99.5 %).' The solvent was Eastman Spectrograde carbon tetrachloride. SolutionsAll flasks and NMR tubes were intensively dried and flushed with prepurified nitrogen before use. The thiols were weighed in volumetric flasks and made up to volume at 25" with carbon tetrachloride containing -1 % t...