[discodermolide]

If you have the vicinal dichloride then it does not matter as the molecule is symmetrical. The worst thing that could happen is di-substitution!

[curiouscat]

Norbornadiene is offered at 120$/kg in 100kg amount on one site,

Which site?

[Enthalpy]

If you have the vicinal dichloride then it does not matter as the molecule is symmetrical. The worst thing that could happen is di-substitution!

But would the subsequent step, photoisomerization to quadricyclane, work if the new double bond is towards the tertiary carbon? I believe to understand the photo step as a 2+2 cycloaddition.

[Enthalpy]

Norbornadiene is offered at 120$/kg in 100kg amount on one site,

Which site?

Maya Biotech.
Such non-binding figures are little reliable, but the stone-old Chemdat CD from Merck wants 36€/0.25L, so a >100kg supplier could indeed bring the price to that area.

[discodermolide]

If you have the vicinal dichloride then it does not matter as the molecule is symmetrical. The worst thing that could happen is di-substitution!

But would the subsequent step, photoisomerization to quadricyclane, work if the new double bond is towards the tertiary carbon? I believe to understand the photo step as a 2+2 cycloaddition.

I shall try the SMILES to see if I can explain myself better:
Start with this compound:

then make this one

finally eliminate HCl:

then do the 2+2 to give:

The only problem I can see is the formation of

as a side product, which may happen but might sterically be less likely and I believe you could control the conditions to avoid much of this forming.

[curiouscat]

Norbornadiene is offered at 120$/kg in 100kg amount on one site,

Which site?

Maya Biotech.

I had a look at Maya Biotech & somehow I'm skeptical that they have ever come within miles of Norbornadiene.

[pgk]

UV Absorption spectra of norbornadiene can be found in:

http://cdn.intechopen.com/pdfs-wm/34620.pdf

See also, the attached file.

[Enthalpy]

pgk: thanks! They give spectra with the catalysts and sensitizers they use to bring some absorption in the visible band. Early in the paper, they state "edge below 300nm" for unaided norbornadiene, which is accessible to excimer lamps, adn then with better efficiency, but which one for which wavelength? I've also found indications in
http://www.uspkhim.ru/php/getFT.phtml?jrnid=rc&paperid=745&year_id=2002
but the language takes me some time to guess where is what information. Nice: they detail for several subtituant groups, including dimethylamine, which coud represent azetidine.

Curiouscat: I won't risk any kind of opinion on the topic and gladly follow your opinion. 400€/0.5L at Sigma-Aldrich, does it make 120$/kg elsewhere credible in 100kg?

Discodermolide: I had imagined from Zaitsev (misunderstood?) that the HCl elimination from

would rather put the double bond at

where the photo step looks compromised. Unless some trick exists?
Two azetidyls in a byproduct would raise the boiling point enough to be separated if any worrying. Or let it polymerize for removal maybe.

[discodermolide]

That is a very unlikely elimination as it forms a bridgehead C=C. These are not favoured energetically.

[Enthalpy]

Fine!

Then, this is a very serious fuel candidate. Not the top-magic in the list, but an excellent one maybe accessible to classical production. Far above the Boctane that tempts designers presently.

Cyclopentadiene is cheap, dichloroethylene too. Azetidine costs a bit but is obtained from thermal elimination of NH₃ from cheap 1,3-diaminopropane with HCl, so home production could be cheap.

If the reaction steps can be made the big way at a reasonable speed (I hope chlorine helps Diels-Alder), just scale to tons.

I'll re-check a heat of formation and performance. Then, the melting point, boiling point, stability... must be experimented.

[Enthalpy]

Got it (hopefully). The new double bond created by HCl elimination won't go to the bridgehead position as per Bredt's rule. Having all neighbours in the same plane is much more important at a double bond than being more substituted or being conjugated.

The path through dichloronorbornene being clearly the better one, I've appended an updated synthesis scheme.

----------

Wanting a single azetidyl on norbornene, how to avoid a second one...

The opinion of AM1 is that the second condensation is almost as favourable as the first one, just 12kJ less good, without any abvious collision. Though, this is for the reaction's products - no idea about the intermediates, the approach conditions and so on.

More promising: Mpbpvp estimates atmospheric boiling points as
+191°C Dichloronorbornene
+240°C Azetidylchloronorbornene
+284°C Diazetidylnorbornene
and azetidine boils at +53°C.

If the amine-chloro condensation can happen in gas phase (can it?) the reactor would just contain an excess of dichloronorbornene slightly above its boiling point, introduce azetidine slowly, and let azetidylchloro- rain down to collect it so it doesn't react further. If necessary (49K contrast in boiling points), the reactor's bottom could be a packed distillation column, to better separate azetidylchloro- from dichloro-. If any diazetidyl- forms, later distillation can separate it.

[discodermolide]

The only thing about gas phase reactions would be the reverse Diels-Ader reaction if the residence time is too long.
Perhaps a continuous flow system would be better to reduce the contact time with the temperature zone.
I think this may require some experimentation to get right, but conditions could very probably be found.

[Enthalpy]

Thanks Discodermolide! Again things that only experience tells...

----------

If aminating in liquid phase instead of gaseous, could some alkaline metal be first dissolved/reacted in the azetidine, as is done with ammonia and aryls, to kick the reaction? ...or would the reaction kick back maybe...?

----------

I've hand-estimated the heat of formation of the azetidyl-, as well as dimethylamino- and bis(dimethylamino)-quadricyclane. At +417, +329 and +356 kJ/mol (298K liquids), they would all bring 5s more than RG-1, just one second less than the flammable parent quadricyclane, so the flash point, melting point and others can decide - mixes being often better.

All aziridines seem toxic, alas. The bis(dimethylamino) would start from trichloroethylene, at least in my imagination. Dimethylamines takes a bit more oxygen.

Marc Schaefer, aka Enthalpy

[discodermolide]

A dissolving metal system may well just remove MCl instead of doing the substitution. Better would be a base such as sodium hydride to deprotonate the N-H.
Aziridines, I think you can forget due to the toxicity. The dimethylamino substitution should be ok. In either case the synthesis would be the same as previously discussed.

[Enthalpy]

It is my nebulous understanding that Na or K react more rapidly with Cl, and this helps the amine+chloride condensation in KNH₂ + ArCl -> ArNH₂ + KCl (conducted in liquid NH₃).

So I just hoped to mimick that with KNC₃H₆ with a swift reaction at room temperature.

OK, I definitely give up aziridines, as you confirm the impression. And all three-membered aminated cycles and polycycles, alas.

Having azetidyl-, methylamino- and bis(dimethylamino)- quadricyclane as candidates gives some hope that one or a mix will survive the many conditions for a fuel.