[irvingfox]

I assume Glycerol's BP is so high due to its ability to form intermolecular h-bonds..

BUT, since the hydroxyl groups are all in such close proximity with one another, wouldn't they simply H-bond intramoleculary, limiting their ability to H-bond with other glycerol molecules?

Based on its high BP I can tell my prediction is off - but can anybody explain why I am incorrect?

[phth]

Please note that even molecules with the same functional groups, but heavier, are different!

We can look at it form a weight-BP graph. Methanol (64), HOCH2CH2OH (197 C), glycerol (290), 1,2,3,4-tetrahydroxybutane (330), and xylitol (347).  The point is the weight is going up by CHOH for each chemical.  So we can see that the graph is not linear, and it is decreasing with increasing molecular weight.  If the only force present was weight, then the graph would be linear which it is not.  The R₂ value of the natural logrithmic trendline is 0.98 using approximate boiling point values.  We can see that ehtylene glycol can form a 5 membered ring within itself.  H-O-C-C-O.  Glycerin can form a 6-membered ring and a 5-membered ring through h bonding (H-O-C-C-C-O).  So the favored conformation on average is 1 6-membered ring.  The conformation of the 6-memebred ring makes the β-hydroxyl stick out, or the less favored conformation (lower temperature) a 5-memebred ring with a terminal hydroxyl bonding intra or intermolecularly.  The opposing dihydroxy dipoles (on average this is the conformation) crease weaker intermolecular hydrogen bonds because it pulls the lone pairs on the free hydroxyl closer to it. 


As we go up to 1,2,3,4-tetrahydroxybutane, it can form  2 6-memebred rings bonding only intramolecularly. The boiling point increases at a slower rate (2nd derivative).  As we go up further, the molecules can form helices, so stabilization occurs through a different mechanism.  If you look at the logrithmic regression, it deviates from the fit more as we go up in mass.  This is directly because of additional forces: e.g. the helical forces, macrocyclic ring stabilization.... 

[Enthalpy]

Intramolecular H-bonds form in gases as there is no available partner, but H-bonds are directional and the molecule generally doesn't offer the most favourable atom positions (despite deforming itself for that), so H-bonds favour liquids over gases.

Phth, why do you seek small fixed numbers of molecules associated in a liquid? I had imagined that molecules form vast agglomerates that reorganize as heat shakens them, permitting the liquid to flow. At least theories for permittivity and viscosity suggest that.