August 12, 2020, 04:27:54 PM
Forum Rules: Read This Before Posting


Topic: Melting Point  (Read 210 times)

0 Members and 1 Guest are viewing this topic.

Offline A13235378

  • New Member
  • **
  • Posts: 3
  • Mole Snacks: +0/-0
Melting Point
« on: July 27, 2020, 12:05:03 PM »
How can the geometry of a molecule change its melting point in this case:





Offline Enthalpy

  • Chemist
  • Sr. Member
  • *
  • Posts: 3528
  • Mole Snacks: +291/-57
Re: Melting Point
« Reply #1 on: July 27, 2020, 12:12:06 PM »
Welcome, A13235378!

Boiling points are rather easy, melting points are badly complicated. I know that I don't have all explanations, and that books and authors neither.

The ease of efficient stacking matters.
The symmetry matters.
The ease of deformation too.

For sure, simple rules like "more branched melts more easily" are wrong. So are additive rules that check the branches on a molecule and claim to predict a melting point: they fail by much.

Offline Corribus

  • Chemist
  • Sr. Member
  • *
  • Posts: 2915
  • Mole Snacks: +451/-22
  • Gender: Male
  • A lover of spectroscopy and chocolate.
Re: Melting Point
« Reply #2 on: July 27, 2020, 02:09:49 PM »
From what I've read, symmetry seems to play a very important role in determining melting point, which higher symmetry often leading to higher melting points, all other things equal. Your drawing is consistent with this. There's an entropic reason for that. If you think in terms of freezing points, freezing requires crystallization more or less, with molecules lining up just so. If a molecule is more symmetric, there are more ways for molecules to line up in the same mutual orientations, which makes it easier to crystallize, which means it happens at higher temperatures (because it is more favorable to be in a solid phase). It's not a hard and fast rule, but it's usually a decent bet that if two molecules have the same chemical formula, the one with the higher symmetry will have the higher melting point. "More branched" is a poor-man's way of saying "lower symmetry", so this is why higher branching often have lower melting point... but if the molecular weight is small, "more branched" can also mean higher symmetry, as in the case here of your 2,2-dimethyl propane. So considering symmetry is better than considering branching.
What men are poets who can speak of Jupiter if he were like a man, but if he is an immense spinning sphere of methane and ammonia must be silent?  - Richard P. Feynman

Offline Enthalpy

  • Chemist
  • Sr. Member
  • *
  • Posts: 3528
  • Mole Snacks: +291/-57
Re: Melting Point
« Reply #3 on: July 28, 2020, 04:54:44 AM »
And it's one interesting case more where entropy doesn't reduce to dQ/T.

Branching has more effects than reducing the symmetry.
- It hinders efficient stacking of the molecules, thus reducing the attraction
- It might help some rotations at low temperature by making the most favourable conformation less favourable. Rotations, available to the liquid only and incompatible with the crystal, favour the liquid,  hence raise the melting point. But beware, this is personal theory substantiated by AM1 computations only.

Offline Corribus

  • Chemist
  • Sr. Member
  • *
  • Posts: 2915
  • Mole Snacks: +451/-22
  • Gender: Male
  • A lover of spectroscopy and chocolate.
Re: Melting Point
« Reply #4 on: July 28, 2020, 10:43:47 AM »
No disagreement from me, Enthalpy! I once spent a long time data mining to try to get a better handle on melting points. Built myself a giant spreadsheet. Very complicated. My memory only is that symmetry seemed to play a huge role, particularly for smaller molecules. In large molecules, it seems not as important because average symmetry is low anyway, in which case branching-dependant intermolecular interactions become more influential. This is a qualitative argument only.

What men are poets who can speak of Jupiter if he were like a man, but if he is an immense spinning sphere of methane and ammonia must be silent?  - Richard P. Feynman

Offline Enthalpy

  • Chemist
  • Sr. Member
  • *
  • Posts: 3528
  • Mole Snacks: +291/-57
Re: Melting Point
« Reply #5 on: July 29, 2020, 04:44:20 AM »
So did I, Corribus, and I got the same conclusion... Spreadsheet from the available experimental data, which isn't so abundant (most data is software scam). The result was: I don't understand, most authors understand even less, and Mankind needs someone to tackle this and succeed.

Melting points are probably too difficult for hand computations and additive rules, with which software predictions fail the big way. If it's a matter of proper stacking and of ease of deformations, even small molecules take a fast computer. But already hydrocarbons or alkanes up to C15 or C20 would mean much.

Sponsored Links