[Enthalpy]

Hi everybody and everyone!

Quadricyclane is an energetic hydrocarbon that some people would like to use as guess what - but I checked a few prices for norbornadiene and it seems expensive.

The synthesis would be, if I grasped it properly:
Cyclopentadiene from refineries and acetylene

make norbornadiene by Diels-Alder

and by photoisomerization quadricyclane

(the appended images may be clearer)

----------

Norbornadiene is offered at 120$/kg in 100kg amount on one site, but only 36€/0,25L by Merck a decade ago. Is that a realistic price in tons amounts?

What is expensive there? Is the reaction slow, dangerous...? Could it accelerate with a pair of chlorines or bromines at proper locations? On the cyclopentadiene or (hopefully not) on the acetylene?

----------

The photoisomerization looks rather affordable with a mercury lamp and a sensitizer, but its quantum efficiency is like 1%... As we have a choice of UV lamps now, powerful and at varied wavelengths, I vaguely hope that the proper one without a sensitizer would be more efficient.

Would you know a UV absorption spectrum for norbornadiene?

Thank you!

[discodermolide]

The trouble with these Diels-Alder reactions is that they are thermally carried out.
Cyclopentadiene itself is a dimer and has to be cracked to get the monomer. This involved heating the dimer, which undergoes a retro-Diels-Alder reaction to give the monomer which is then trapped with the new dienophile. Then you have to purify the entire mess.
Now I don't know if they are crystalline so it may require distillation.
All in all quite a tricky process.
Halogen substitution will probably slow the entire process down requiring higher temperatures for reaction.
Quadricyclane is a [2+2] cycloaddition, again quite reversible so difficult to isolate and keep intact. I don't know the best quantum yield that has been achieved.
These and other problems may well contribute to the high prices of these compounds.

[Enthalpy]

Thanks Disco! What I have read so far:

The dimer of cyclopentadiene is broken by heat, the monomer separated by distillation can be stored for a limited duration without having to combine it first.

Something like electrophile groups help Diels-Alder. No idea if this applies here with acetylene, nor if the groups must be on acetylene or on cyclopentadiene. But if electron-rich groups are better, just fine!

Under mercury UV and a sensitizer, the equilibrium between norbornadiene and quadricyclane is like 95% alkane versus 5% diene.

And meanwhile I've checked that 172nm from a xenon lamp fits all alkenes with separated double bonds. Heavy alkenes would accept also 193nm from an ArF lamp, but Xe lamps are better. Excitation of the pi bond happens essentially with every absorbed photon, but the story isn't finished then.

[discodermolide]

Normally dienes are electron rich, dienophiles electron poor helping the reaction move along and making the reverse reaction less likely.
In your system you just have normal electron demand and you have several competing reactions: re-dimerisation of cyclopentadiene (a Diels-Alder reaction) and reaction with acetylene, also a Diels-Alder reaction as well as the thermal reversal of both. So there is nothing to stop the equlibrium forming, unless you remove one of the constituents. Norbornadiene can be nicely coordinated to transition metals such as Nickel.
I'm not a photochemical expert but 95% cyclised, even with that strain is a bit strange?

[Enthalpy]

O yes, competition of dimerization once reacting with acetylene - thanks for re-explaining.

Well, if somebody seriously wants to burn quadricyclane in an engine, it's a matter of 100t/month, so a simple setup can be taylored to produce it. Separating the much bigger cyclopentadiene dimer from the products and reinjecting it in the cracking unit looks accessible, doesn't it?

Meanwhile I've seen that usual Diels-Alder put a pair of -COOMe on the acetylene, so this must be where the electron-withdrawing group belongs. Conversely, would amines (dimethyl-) on the cyclopentadiene do the trick?

Reaction yield of the photoisomerization: I have obviously no opinion about it...
US Patent 6,635,152 shows 80% to ~100% depending on the sensitizer and Nbd's purity
Orgsyn stopped at 80%
I vaguely imagine that the photosensitizer acts on the alkene far better than on the alkane, and that changing from the Hg lamp to a Xe lamp without sensitizer may squander this advantage.

[discodermolide]

Yes that's correct, making the dienophile electron deficient works well.
Amines on the dienophile also work well.
I would not really want to combust quadricyclane in my car engine. My car is so old it just manages to run with me in it never mind 100t of quadricyclane.

I vaguely remember that using a sensitizer reduces the energy of intersystem crossover thus reducing the energy required for the photochemical reaction, i.e. the alkene and suppressing side reactions?

[Enthalpy]

Some people did consider running cars with the allegedly stable quadricyclane, brrmmm!
Others want to store sunlight in a norbornadiene, modified or catalyzed to absorb at visible wevelength, and let the quadricyclane release heat when needed.

But the engines I thought of are more thirsty. For instance the pretty-cute RD-170
http://www.lpre.de/energomash/RD-170/index.htm
whose four main chambers swallow each 433kg/s (1 bathtub) of already hot oxygen to burn 165kg/s of kerosene in D=380mm~L - because, to lift ten railway engines, this is what it takes.

To water such thirsty boys, one better has beverages decently easy to produce.

-----

Amines at the acetylene are just as good to burn as at the cyclopentadiene, and may lower the freezing point, fine.

[Enthalpy]

Billions of blue boiled and barbecued barnacles! Dichloro-, dibromo-, diiodo-, bis(dimethylamino)- acetylene have considerable drawbacks. But I still prefer a heavier, less flammable molecule to plain quadricyclane.

Would the appended synthesis of the N-cyclopropyl-7-aza-quadricyclane make any sense? Starting from butadiene or furane instead of cyclopropadiene dimer. This one keeps the interesting performance of quadricyclane.

Though, I'd like C+N~12. Is there a means to replace the -COOMe by an amine (without dihaloacetylene), instead of making a cyclopropylaza?

[discodermolide]

The first step probably won't work. You get better substitution patterns with the classical Paal-Knorr synthesis
http://en.wikipedia.org/wiki/Paal–Knorr_synthesis

The rest should be OK.

[Enthalpy]

Thank you Disco!

[Enthalpy]

The flash point of cyclopropylazaquadricyclane may be good, or maybe not, with C+N=10; estimation software is inaccurate even for quadricyclane's boiling point. With one carbon more, azetidylquadricyclane would raise the flash point and nearly keep the performance - provided it can be mass-produced.

As haloacetylene has drawbacks, maybe 1,2-di- or trichloroethylene can provide a halonorbornene, where a double bond would be restored and azetidine be coupled? I imagine chlorine helps the Diels-Alder step.

Though, I fear annoyances in both cases (drawings appended):
- From dichloronorbornene, abstracting HCl would create the double bond towards the tertiary carbon, according to Zaitsev;
- From trichloronorbornene, abstraction of Cl₂ seems forbidden because the chlorine aren't anticoplanar.
- (Or maybe the attempt makes no sense for other reasons...)

What do you think? Thanks!

[discodermolide]

I think the reaction of the vinyl chloride with the azetidine is unlikely. I would probably try and introduce this group at the start of the synthesis.

[Enthalpy]

Thank you!

So instead, first dichloroazetidylethylene, then Diels-Alder to he substituted norbornene?

What about the CL₂ elimination: can it proceed despite the C that hold the Cl can't rotate?

[discodermolide]

Do the displacement before the HCl elimination.
You could try a Pd catalysed coupling of the vinyl chloride with the azetidine. Usually this works well for aromatic halides (Buchwald-Hartwig). In this case it may well have a chance. See for example
A new reactivity pattern for vinyl bromides: cine-substitution via palladium catalysed C-N coupling/Michael addition reactions
M. C. Willis, J. Chauhan, W. G. Whittinham, Org. Biomol. Chem., 2005, 3, 3094-3095.
and
(t-Bu)2PNP(i-BuNCH2CH2)3N: New Efficient Ligand for Palladium-Catalyzed C-N Couplings of Aryl and Heteroaryl Bromides and Chlorides and for Vinyl Bromides at Room Temperature
Ch. V. Reddy, J. V. Kingston, J. G. Verkade, J. Org. Chem., 2008, 73, 3047-3062.

[Enthalpy]

Thank you!

Found the papers' abstracts, so apparently routes exist.

You wrote "HCl elimination": do you consider it could create a double bond at the proper location? I thought the double bond would appear towards the tertiary carbon where I don"t need it, that's why I considered the trichloro.