Boiling point variation with structure

As with alkanes, the boiling points and melting points of alkenes decrease with increasing molecular weight, but show some variations that depend on the shape of the molecule. Alkenes with the same molecular formula are isomers of one another if the position and the stereochemistry of the double bond differ. For example, there are four different acyclic structures that can be drawn for butene (C4H8). They have different b.p. and m.p. as follows. 

Branched-chain alkanes do not exhibit the same smooth gradation of physical properties as do the continuous-chain alkanes. Usually there is too great a variation in molecular structure for regularities to be apparent. Nevertheless, in any one set of isomeric hydrocarbons, volatility increases with increased branching. This can be seen from the data in Table 4-2, which lists the physical properties of the five hexane isomers. The most striking feature of the data is the 19° difference between the boiling points of hexane and 2,2-dimethylbutane. 

The linear correlation between the refractive indices and the boiling points was found in 27 series of organic compounds with a wide range of structural variation, as summarized in Table 1.1. So long as the structures of a series of molecules resemble each other, these linear correlations have fair to excellent correlation coefficients [9]. Except for hydrogen-bonded and apparently rod-shaped molecules, the main coherent forces of the liquids come from the electronic polarizability. 

Branching in alkanes, which provides structural variation and alters the effects of intermolecular interactions, does not lend itself to the same kind of analysis. However, in general, branching (in compounds with the same molecular weight) lowers the boiling point. 

A plot (Figure 115) of the Apiezon retention index versus boiling point, where known, shows considerably more uniformity despite the variations in structure represented. From such a plot the retention index can be predicted from the boiling point with an average error of 1%, the poorest case being that of tetravinyl tin, where the error is 6%. 

---