29

I understand that for cyclooctatetraene (COT) to "escape" the horrible prospect of becoming an anti-aromatic molecule, it must adopt a non-planar conformation.

enter image description here

The widely touted conformation of COT is said to be "tub" shaped. My sources (random links Google threw up) also concur on this particular "tub" conformation.

But why isn't it "chair" shaped instead?

Going along the same line of cyclohexane (non-aromatic), wouldn't a "chair" conformation reduce (even if it's a little bit) the steric repulsions due to crowding of hydrogens atoms on one side/face of the molecule?

Or even if "avoiding steric issues" isn't really the big idea here, then I could re-phrase my question as:

What stops COT from assuming a "chair" shaped structure?

The "chair" structure that I have in mind for COT is something like this:

enter image description here

(I don't have access to a decent enough organic chemistry sketcher to have this drawn any better, sorry.)

Melanie Shebel
  • 6,704
  • 10
  • 45
  • 86
paracetamol
  • 18,649
  • 29
  • 106
  • 169
  • 3
    Well, what you've drawn are either alternating cis and trans double bonds. If you do the tub shape you can have nice cis double bond geometry while if you do the "chair" you need to go towards trans, so in the end very twisted cis bonds. – DSVA Jul 10 '17 at 09:29
  • 2
    Interestingly, I can’t find any references to cis-trans-cis-trans-COT. I found some claiming the synthesis of mono-trans-COT derivatives, though. Naively, I assume cis-trans-cis-trans-COT should exist. However, the two are, of course, different compounds (and mono-trans-COTs tend to equilibrate back to all-cis-COT after being synthesised by irradiation). – Jan Jul 10 '17 at 10:30
  • @Jan If it c-t-c-t COT does exist...then know why people tend to stick with "tubs" rather than "chairs"? [Google images threw up tubs only :/] – paracetamol Jul 10 '17 at 10:35
  • 1
    The all-cis-form seems to be a stable diastereomer. I actually also found a few crystal structures confirming the synthesis of all-cis from precursors that I would have expected cis-trans-cis-trans to derive from. Well, I don’t know O:) – Jan Jul 10 '17 at 10:38
  • 1
    No, you should assume all COTs are all-cis-COTs unless otherwise specified ;) – Jan Jul 10 '17 at 10:45
  • @Mithoron no it sn't you cannot build the 3D structure of the "chair" confomer using planar cis double bonds. you would need to "twist" the double bond (which would be extremly high in energy) or go right to the cis,trans,cis,trans configuration. See my post below. – DSVA Jul 10 '17 at 12:18

1 Answers1

23

It's all about the 3D structure of double bonds.

If we look at the tub form we see that all dihedral C-C=C-C angles are 0°. the C=C-C angles are 125°, also pretty close to the optimal value. Everything works out fine, there's basically no strain on the whole molecule. enter image description here

I tried to build the chair form of all cis, but every optimization ended up in either the all cis tube form or the cis-trans-cis-trans chair form. The best structure that looks like the all cis chair form I ended up with is this:

enter image description here

First of all, it's pretty flat, so antiaromaticity would again play a role and the dihedral angles are also ~33° for the double bonds in front and in the back, which is energetically quite bad and unstable.

The cis-trans-cis-trans compound is definitely no monstrosity. I'm working a lot with trans-cyclooctenes and I don't see why this shouldn't be possible if a trans,trans-cycloocta-1,5-diene is possible.

enter image description here

Here we've got 0° for the dihedral angles for the cis bonds and 125° for the trans. Yes, this is far away from the optimal value of 180° but quite similar to what you expect from trans-cyclooctenes (around 135°). I feel like it would be quite reactive and unstable but it could definitely exist. There's just no good way to produce it I guess.

So overall: all angles in the tub form are much closer to the optimal values than in either chair form and the all cis form is quite planar.

Melanie Shebel
  • 6,704
  • 10
  • 45
  • 86
DSVA
  • 5,772
  • 2
  • 27
  • 42
  • Mind a spin-off question? You mentioned: "... it's pretty flat, so antiaromaticity would again play a role ...". I was under the impression that once an annulene with 4n π electrons adopts a non-planar conformation, it totally loses its antiaromatic status. But what you said suggests that antiaromaticity is not so sharply defined with regard to conformation (i.e- antiaromaticity of the annulene depends on the degree of "flatness" of the molecule...rather then simply considering whether the molecule is flat or not). Was this a slip-of-tongue? Or did you really intend that? – paracetamol Jul 10 '17 at 14:29
  • 1
    Yes that was intended. The Hückel rules, which I guess you are taking the "it has to be flat" rule from, are just very simple basic rules but they tend to be quite inaccurate in some cases. (Anti)Aromaticity depends on the conjugated pi-system, That's why you would normally say it has to be flat. However, if you go out of plane the overlap will be worse but it's still there, you'll still have the aromatic or antiaromatic character, but it's not as strong. – DSVA Jul 10 '17 at 15:35
  • 1
    There is no imaginary mode in the 'flat' case? – Martin - マーチン Jul 10 '17 at 17:22
  • @Martin-マーチン there are most certainly several imaginary modes in that structure. I wasn't able to optimize to a stable structure that had chair conformation and only cis double bonds. This one occured during optimization and was just taken to show that the bond angles are terrible for such a confomer. – DSVA Jul 10 '17 at 18:33
  • 2
    0 degrees? is that a typo? – NH. Jul 10 '17 at 21:51
  • 2
    @NH. no, why do you think so? Keep in mind it's the dihedral angle. 0° basically means planar and cis. – DSVA Jul 10 '17 at 21:55
  • @DSVA : What is an "optimal value" for an angle? IMHO, an optimal structure of a molecule corresponds to a local minimum on the potential energy hypersurface (PES). The PES maps structural arrangements of a set of atoms to the total energy of the molecule. Experimental evidence in favor of one structure may indicate that the structure corresponds to a local minimum whereas others aren't. However, in this particular case the electronic Pi-System is composed of four distinct bonds (unlike in, e.g., benzene) which could be stabilized by $\pi$-electron interaction through space. – TheFox Oct 10 '17 at 11:34
  • @TheFox the optimal structure for the whole molecule, yes that's the minimum on the PES. But that doesn't need to be true for the bondlengths/angles in fragments of this molecule. If we look at a trans double bond in a simple alkyl chain we see that the energy minima is at the planar arrangement. Now if we force the double bond into non planar conformation, for example by making an 8-membered ring, then it is what we call "strained" and higher in energy since it's not the optimal configuration for this part of the molecule. – DSVA Oct 12 '17 at 07:25
  • @DSVA It may well be that the "strainedness" is one portion of the energy difference. However, it is disputable to call the reason for domination of the tube over the chair. Your argument simply reduces the electronic structure to a background force that is invariant to the actual molecule under consideration. As $\pi$ and especially $\sigma$ electrons play a crucial role for the structure of a compound, I find it doubtful to exclude their role in an argument. – TheFox Oct 12 '17 at 08:01
  • @TheFox I really don't know what to say. I suggest you read some literature about the concept of strain and how it is applied to our understanding of organic molecules, since this doesn't exclude sigma and pi bonds at all. In fact that's the whole point. If you twist this molecule into a chair like confomer you are losing essentially two double bonds due to twist between the carbons. – DSVA Oct 12 '17 at 15:53