[Enthalpy]

Syntin contains a methyl group, maybe to ease its synthesis. Without the methyl, ter(cyclopropane) gains half a second specific impulse and quater(cyclopropane) loses half a second, so their mix should keep the performance and improve the flash point. It just needs to oligomerize (OK, condensate) the commercially available cyclopropane...

Boctane is also a dimer from cyclobutane, preferably with some trimer mixed. And most fuels beyond that are tiny fused cycles, to be joined for a good liquid range - picture attached.

Again, I don't care much about isomers, and mixes are good. If some isomers have a better melting point, they can be separated and put aside for the special uses.

"This would be nice" is always the easy part, isn't it? But a "that way" should follow.

[Enthalpy]

The halogenation of spiropentane and cyclopropane is difficult, with yields of 1/3 reported by Houben-Weyl and by
Applequist, Fanta and Henrikson in "The chlorination of spiropentane".
Some reasons:

The C-H bond is strong in cyclopropane and spiropentane, 7kJ stronger than in methane according to Neumann, stronger than H-H and H-Cl.

Consistently, the reactions see a significant barrier. Nist publishes an activation energy for spiropentane and atomic oxygen to spiropentyl, thanks
http://kinetics.nist.gov/kinetics/ReactionSearch?r0=157404&r1=17778802&r2=0&r3=0&r4=0&p0=3352576&p1=83321226&p2=0&p3=0&p4=0&expandResults=true
from which I "deduced" other activation energies in the appended table - shifted according to the bond energies to hydrogen, which shouldn't be done.

The ring-opening addition to cyclopropane is exothermal. If this were an indication for the ease of reaction:
H₂O₂ -325kJ, Cl₂ -212kJ, HBrO -187kJ approx, BrCl -184kJ, Br₂ -125kJ, HCl -93kJ, H₂O -66kJ
and ring-opened dichloro uses to be as abundent as chlorocyclo in the products.

Cyclopropane absorbs light at 174nm from Xe₂ lamps. 193nm light from less good ArF lamps acts less at the target species, and could even harm bigger cyclopropanes, while 222nm from KrCl seems impossible.

These difficulties don't apply to cyclobutane, for which I still hope Xe2 light and water achieve the oligomers
http://www.chemicalforums.com/index.php?topic=81721.msg297374#msg297374
among other potential methods
http://www.chemicalforums.com/index.php?topic=50579.msg296973#msg296973

----------

In the photochlorination process with least drawbacks I saw, at least on paper (whatever meaningful this is), the species in the reactor are by decreasing abundance:

• 1. Spiropentane (or cyclopropane), for instance at atmospheric pressure - hopefully transparent.
• Gaseous HCl kept around 0,05bar. Absorbs 63% of ArF's 193nm in 100mm.
• The produced chloride, to be removed continuously since it absorbs light as much as HCl does. It condenses 50K earlier than the hydrocarbon.
• By-produced H₂, to be kept lower than HCl.
• An optional bit of Br₂ to quench chain reactions.
• Possibly some hydrocarbon dimer, and some self-destroying Cl₂ that doesn't hurt in small amount.
• H°, Cl° and spiropentyl radicals.

The reaction shall proceed this way:

• 193nm light splits efficiently HCl into H° and Cl°.
• H° and Cl° abstract each an H from spiropentane to make HCl and H₂. The activation energy would be estimated to 16kJ and 20kJ, and the reaction endothermal by 10kJ and 13kJ approximately, but the radicals are emitted excited.
• The spiropentyls react with HCl exothermally to make spiropentane back and the product chlorospiropentane.
  If some bis-spiropentyl appears, fine! Strong concentrated light should favour it.

At best every second radical creates a chlorospiropentane, I take a third - but one 6.4eV photon makes two radicals, so it needs 14MJ of light to get 1kg of spiropentyl. The 15% power-efficient lamp uses then 25kWh/kg, and alone the electricity costs 2.5€/kg.

Some attempts I set aside:

• Added Br₂, more abundant than HCl, would absorb little light and react with the spiropentyl to yield bromide - but only if H° and Cl° didn't react first with Br₂, which is easier than with spiropentane.
• H₂O₂ (towards alcohol), Cl₂+Br₂ and BrCl (towards bromide), HBrO (towards bromide) supposedly open the cycle like Cl₂ does.

Also, photohalogenation papers are often older and use little light at Hg wavelengths to trigger long chain reactions. Present exciplex lamps can provide a strong photon per molecule, which creates excited radicals, so direct bromination or quenched chlorination could be tried again.

Marc Schaefer, aka Enthalpy

[Enthalpy]

Dendralenes
https://en.wikipedia.org/wiki/Dendralene
offer alternatives to the oligomerization, possibly more practical, with cyclopropanation or cyclobutanation carried out on the polyene backbone as sketched below.

Several paths exist to dendralenes, those cited here come from Wiki. The multiple cyclopropanation to "ivyanes" was carried out through Simmons-Smith by Sherburn et al
https://cen.acs.org/articles/89/i6/Meet-Ivyanes.html
and I vaguely imagine the cyclobutanation made optically with ethylene.

These molecules have no isomers and the paths here give a fixed number of rings, but ivyanes are liquid up to 6 rings, so the C₁₂ compounds must balance the flash point, melting point, viscosity. To lower the melting point, it must be possible to mix both C₁₂, or mix several dendralenes before making the rings, or mix with 1,2-dicyclobutylcyclobutane
http://www.chemicalforums.com/index.php?topic=50579.msg233917#msg233917
http://www.chemicalforums.com/index.php?topic=50579.msg296973#msg296973

At least methyl-triscyclopropane was used as "syntin". Removing the methyl improves the performance. The gem- backbone may reduce the viscosity and must burn just as nicely. The cited synthesis paths look a bit lengthy, maybe then can be optimized.

Marc Schaefer, aka Enthalpy

[Enthalpy]

Tetracyclopropyl C₁₂H₁₈ C1CC1C2(CC2)C3(CC3)C4CC4 would be less flammable than Syntin and more efficient without the methyl. Expansion from 245bar to 0.8bar computed by Propep:

m/s     +s    O2:100    Bp °C   kJ/mol
-----------------------------------------------------------
3360    +5      250     +198     +232    Tetracyclopropyl
3354    +4      249     +291     +364    Hexacyclopropyl
3313   REF      283     +197             Rg-1 as reference
-----------------------------------------------------------
m/s     +s    O2:100    Bp °C   kJ/mol

Maybe a synthesis in the sketch, if the rings survive the coupling. More reactive geminal dihalo let me expect mostly even ring numbers from metal coupling, but odd ones are good too, so steps 2 and 3 can happen at once.

Other coupling mediums are better known, but gas phase could let the dihalotetracyclopropyl rain down from the reaction phase, as its vapour pressure is >30* smaller than the dihalodicyclopropyl.

Some hexacyclopropyl is welcome in the tetracyclopropyl, especially if it makes a eutectic. The melting points are essentially unpredictable: symmetric and somewhat heavy hydrocarbons, but with easy rotations.

As a hypothetic alternative, light would break C-X bonds. Bromine and iodine would make no secondary reactions around room temperature.

Recycling on site the halogen and metal would reduce costs and risks. Gaseous potassium is a known reactant, others may be good
http://www.chemicalforums.com/index.php?topic=72951.0
and seemingly zinc would separate from iodine by heat, cheaper than electricity. Or take the same metal that makes the carbene to recycle in one operation.

Marc Schaefer, aka Enthalpy

[Enthalpy]

"Bromoform can be prepared by [...] the electrolysis of potassium bromide in ethanol", from
 https://en.wikipedia.org/wiki/Bromoform
that would close nicely the loop for tetracyclopropyl synthesis. Potassium must still be reduced.