[Enthalpy]

Hi dear friends!

I'd like to show some sensible rocket fuels that improve over the RG-1 "kerosene", which is approximately a Diesel fuel freed of all constituents volatile or unstable at heat, since it flows in the jacket to cool the engine. The US RP-1 is a tad lighter and more energetic.

Rocket fuels must be (1) stable (2) even at heat (3) healthy (4) not easily flammable if possible (5) cheap and abundent (6) efficient. RG-1 has all that, so improving isn't easy. For instance nitromethane and propyne aren't stable enough, alkenes polymerize at heat, boranes are toxic and expensive. If a fuel is easily flammable, we prefer methane or hydrogen. Also, burnt hydrogen brings more heat per mass unit than carbon, so only strained cycles are useful.

----------

How much an improvement? A 2% better Isp (ejection speed / g) makes a difference every time the rocket has accelerated by one ejection speed, or more due to inert masses. To a low-Earth orbit (9500m/s cost) it acts >3 times to carry ~10% more payload, or 4M$ over a 40M$ launch. 200t of fuel must then cost <<20$/kg - but 40t that gain 5% at the second stage must cost <<50$/kg. To the geosynchronous orbit it carries ~14% more, and landed on the Moon (Leo+5800m/s) it's nearly 20% more, where the last stages are tiny.

----------

To begin, here are banal commercial compounds that outperform RG-1. Amazing for rocket designers; future posts could me more exciting to chemists. The appended images compare (expansion from 245bar to 0.8bar) and depict the molecules; log in to see them.

Cis-pinane is strained by the trees, saturated by paper factories and sold as a byproduct. Much more available than RG-1, a bit better. Flammable under +55°C alas, but alkylation, cyclopropanation or cyclobutanation instead of hydrogenation would improve it.

Branched alkanes differ by the liquid range. The heavier products of an alkylation unit are good; farnesane, phytane would stay liquid on Mars but must be mass-produced first.

Branched amines are easier to make and better than branched alkanes because N₂ produces heat; tertiary amines improve the liquid range. Pmdeta and Pmdpta are mass products, and I've put there how I imagine to synthesize Detmdpta, hopefully liquid on Mars:
http://www.chemicalforums.com/index.php?topic=56069.msg272080#msg272080

JP-10 (saturated dimer of cyclopentadiene, good to burn with air) illustrates the cycles being less efficient if unstrained.

Boctane (di-cyclobutyl) is considered as an RG-1 upgrade
http://www.chemicalforums.com/index.php?topic=50579.0
but banal amines are safer, cheaper and almost as efficient - yes, I'll describe strained amines also.

Marc Schaefer, aka Enthalpy

[Enthalpy]

Here are some ring amines, efficient and hopefully possible to mass-produce. Their performances are close, hence the ease of synthesis, flash point and melting point shall decide. Mixes of several compounds and their isomers are welcome in a fuel, especially if they depress the melting point. The unmentioned cyclopropylmethyl- fall between the dicyclopropyl- and dimethyl-.

In the appended property list (log in to see it), the ring amines beat Boctane by 1-3s, and some feasible compound or blend will hopefully improve the flash point. Estimates of the melting points and densities are too inaccurate to mention, but I've put boiling points from the Mpbpvp software, and flash points extrapolated linearly from azetidine through the atmospheric boiling point. I've hand-evaluated the heats of formation carefully.

The compounds and reactants suggest some synthesis routes. Cyclopropylamine is available in tons for few $/kg if I believe Alibaba, and closing azetidine rings was done in the 19th century. Maybe imines can dimerize at UV like ethylene does as a route to 1,3-diazetidines; 1,2-dimethyldiazetidine (supposedly unhealthy) synthesis is on page 28 of "Improved liquid propellants, hydrazine derivatives" (with a wrong heat of formation, to my opinion)
http://www.dtic.mil/cgi-bin/GetTRDoc?Location=U2&doc=GetTRDoc.pdf&AD=AD0093924
and did I see elsewhere the 1,3- synthesized from dihalomethane and a substituted amine?

Diazaspiroheptanes were synthesized by Burkhard and Carreira:
"A concise and scalable synthesis of a 2,6-diazaspiro[3.3]heptane building blocks"
starting with a halogenated equivalent of pentaerythritol.

Comments welcome, of course and as usual!
Marc Schaefer, aka Enthalpy

[Enthalpy]

Here are some hydrocarbons that have been considered as RG-1 improvements - list and molecules appended. Many have drawbacks, but they belong to rocketry fantasy, help comparisons and show general trends.

Only Syntin has flown, on Soyouz three decades ago, is abandoned allegedly because of cost (synthesis on Wiki), but permitted one cosmonaut more per flight. In the existing tanks and pumps, it replaced RG-1 at identical volume. Notice that four hydrogen atoms less lose 5s over plain cyclopropane, the big cost of a safer liquid.

Soctane seems abandoned, but the easier Boctane is fashionable.
http://www.chemicalforums.com/index.php?topic=50579.0
Some mixed "trimer" would improve the flash point. One could also dissolve cyclobutane and cubane if meaningful. Worry: I've suggested here cyclic amines that outperform it, and some must be easier to synthesize.

With 4-Ladderane we approach the performance of simple cyclic amines - nitrogen brings so much. It just remains to mass-produce... If lucky, the laser pulses or the metal vapour I described there will help:
http://www.chemicalforums.com/index.php?topic=77307.msg283262#msg283262
http://www.chemicalforums.com/index.php?topic=72951.0

The hexyl-5-ladderane, is made with a carboxylic end by a bacterium, so Wiki's article about ladderanes raised big hopes
http://en.wikipedia.org/wiki/Ladderane
but the unstrained tail botches the advantage. Consider it only if its cost or liquid range are magic, or if a cheap method isolates the strained part.

The volatile Housane (Bicyclo[2,1,0]pentane) and its civilized "dimer" are difficult to synthesize, and easier molecules outperform them.

Quadricyclane has a decent synthesis path
http://www.orgsyn.org/demo.aspx?prep=cv6p0962
so sizeable amounts were characterized for rocketry:
"Chemical, Physical and Hazards Properties of Quadricyclane"
www.dtic.mil/dtic/tr/fulltext/u2/a345589.pdf
but a cyclopropyl-aza variant would be less flammable and keep the performance
http://www.chemicalforums.com/index.php?topic=79658.0

Methane is often considered the successor of RG-1 because it's cheap and its hydrogen gains 9s. Several adapted engines tested it with success. It leaves the engines cleaner, important if they're reused in the future. Though, it's very flammable and needs bigger tanks and pumps, so no designer has chosen it up to now. I claim that cyclopropane and spiropentane are better: more efficient, denser, easy to produce. The density would even allow a better chamber pressure not included in the comparison. More importantly, di-spiropentyl (and di-bicyclopentyl) bring methane's performance without the flammability.

Bicyclo[1,1,1]pentane is nicely stable and would have some synthesis paths, often through propellane
http://www.orgsyn.org/demo.aspx?prep=v75p0098
so it has been considered as a propellant
http://www.dtic.mil/dtic/tr/fulltext/u2/a267508.pdf
as well as its seducing "dimer", to blend with the "trimer" as needed. With a good process, it's useable.

Cyclopropane is a mass product, with only advantages over methane. Its dimer is volatile, ts trimer is Syntin.

Tricyclo[3.1.0.0(2,4)]hexane is more difficult to synthesize; maybe my laser pulses or metal vapour help.

Spiropentane is made since the 19th century just from pentaerythritol, bromine and zinc or sodium
http://de.wikipedia.org/wiki/Spiropentan (German or Polish as you prefer)
where bromine and the metal could be recycled on the site. Spiropentane is liquid at room temperature, rather dense, nicely stable, and it outperforms methane. Its "dimer", helped by some trimer, is not easily flammable, and equals methane through the gained pressure. Could a pentahalo-neopentane reacted with metal make the dimer directly?

Cheap ethylene gains 2s more if a designer accepts to cool the engine's walls with oxygen, which isn't trivial. But heavier alkenes are bad, even tetravinylmethane, alkynes too unstable, allenes probably too.

cubane is still a fantasy, stable and so efficient. For solid engines, or dissolved, or tuned to be liquid.

Nitrogen would ease some syntheses: azacubane, ladderanes, housane, diasteranes and more exotic ones
http://saposjoint.net/Forum/viewtopic.php?f=66&t=2372&start=20 (take with mistrust)
but the toxicity of aziridine versus azetidine suggests that nitrogen shouldn't be too strained, and the harmlessness of new candidates be tested early. Cyclopropanone might help
http://www.chemicalforums.com/index.php?topic=78226.0

Marc Schaefer, aka Enthalpy

[phth]

What I can add to the conversation: amines have an anomalous pKa values i.e. 2° (by about 4 times greater)>3°>1°.  We can do marcus theory calculations with bond enthalpy to draw qualitative conclusions using bond enthalpies and ΔG°; ~1kcal/mol uncertainty.

[Enthalpy]

Thanks phth!

[phth]

Marcus theory works well for homolytic cleavage in the gas phase because of how unstable ions are, and thus their transition states.  But it becomes a bad model in the liquid phase because of competing transition states may be lower or higher depending on the conditions.

[Enthalpy]

As they flow in a jacket to cool the chamber, heated fuels shall leave no solid decomposition products, or a runaway hot spot would destroy the wall. Some  present designs even wish a clean jacket after flight to reuse the stage.

The appended two-sheet xls (cut the txt extension) and result image estimate a temperature where 1000 moles of a fuel leave 10g decomposition products over 300s. The sources are in the xls; some fuels are deduced from much faster rates, which is inaccurate and pessimistic.

The fuels in green lines fly successfully. Most others refute their reputation.

Bicyclo[1.1.0]butane seems too fragile, but oxygen instead might cool a future engine. Hvap=427kJ/kg for cyclobutane and Hf=+217kJ/mol for gaseous bicyclobutane at 298K according to
http://webbook.nist.gov/cgi/cbook.cgi?ID=C157335&Mask=1
so Hf=+194kJ/mol for the liquid, and then bicyclobutane is 14s better than Rg-1, 3s better than ethylene, just 1s behind cubane, fabulous. Do you know if it may detonate or polymerize as a liquid without air?

Marc Schaefer, aka Enthalpy

[Enthalpy]

60m/s I_sp improvement to 3300m/s is little, but a rocket flight accumulates it 4×. The inert mass, a fixed cost, also amplifies the relative gain at each stage.

An existing stage carries more of a denser fuel that accepts as much oxygen. The thrust increases proportionally, so the acceleration is kept. Pumping would take the same power for the same volume; a little bit is wasted as the fuel pump achieves unneeded pressure.

The example Zenit-3SLB launcher gains much because it reaches demanding orbits from Baikonur at difficult 51°N and because its dry stages are too heavy.
https://en.wikipedia.org/wiki/Zenit_(rocket_family)
The third's 3220kg (from the user's guide) leave 1600kg payload on geosynchronous orbit, the second's 8367kg leave 4940kg on sun-synchronous orbit.

Rg-1 "kerosene" shall become
N-cyclopropyl-N'-methyl-2,6-diazaspiroheptane C₉H₁₆N₂ CN1CC2(C1)CN(C2)C3CC3
http://www.chemicalforums.com/index.php?topic=86972.msg314306#msg314306
with hand-estimated ΔH_f=+308kJ/mol and ρ=972kg/m³ for the liquid at 298K.

The engines use more Rg-1 than optimum, maybe to dump some in the nozzle for cooling, so the same volume ratio would favour the amine too much. I've added only 60m/s to Rg-1's exhaust speed. The appended spreadsheet (remove the .txt) supposes amine at all stages.

    Rg-1  Amine  Orbit
--------------------------------------
kg  1600  2321   Geosynchronous Gso
kg  3600  4585   1500m/s short of Gso
kg  4940  6516   Sun-synchronous 800km
--------------------------------------

At 50M$ a launch, the improvement is worth 15M$. Spending 5M$ in 120t fuel justifies 20$/kg production cost, or 3M$ for 30t in the upper stages 50$/kg. Easier orbits, or a launcher with lighter tanks, would spend less on a fuel. Already Pmdeta brings 1/4 the advantage at no overcost.

Marc Schaefer, aka Enthalpy

[Enthalpy]

Finally, I've found a measured decomposition speed for the Rocket Propellant N°1 or Rp-1, called "kerosene" but nearer to Diesel oil, at Nist of course
http://ws680.nist.gov/publication/get_pdf.cfm?pub_id=901027
but still not for hydrazine, Mmh nor Udmh despite they were investigated.

For the comparison, I take the temperature that decomposes 10g from 136kg of Rp1 in 300s, or +316°C. It tells that the small cyclo- and spiro- are stable enough, azetidine probably too. Data about the commonly used hydrazines would validate more candidates.

Until we cool the engine walls with the oxygen to leave the engines clean for reuse and enable better fuels like ethylene, bicyclobutane, spirohex-1-ene...

Marc Schaefer, aka Enthalpy

[Enthalpy]

SpaceX and Blue Origin develop rocket engines burning methane, so methane will likely become the next irrational worldwide fashion. Time to tell my discordant opinion and compare methane with other fuels.

I have already explaned that Robert Zubrin's "Marc Direct" scenario, making methane on Mars from atmospheric CO₂ and imported hydrogen isn't optimum
http://saposjoint.net/Forum/viewtopic.php?f=21&t=1953#p36418
because burning directly the hydrogen avoids the chemical plant, needs only twice as much hydrogen, while Fischer-Tropsch demands a complicated and heavy recycling plant to waste less than half of the hydrogen. That said, I like the in-situ production of oxygen, as the hot electrolysis of CO₂ in a ceramic is simple and proven.

The Mars argument for methane drops, and the claimed efficiency doesn't stand neither against good fuels. Here I compare fuels for a launcher putting 21.5t in Leo.

• Methane gets a "full-flow staged combustion" pumping scheme to achieve 300bar in the main chamber, huge unfair advantage
  http://www.scienceforums.net/topic/81051-staged-combustion-rocket-engines/
  I admit temporarily that methane doesn't soot in its fuel-rich pre-chamber.
• The competitors have a simple staged combustion with one oxygen-rich pre-chamber. I've determined individually the achievable pressure by extrapolation from the old RD-170: same 500°C and 535bar, same 74% 79% 88% efficiency at the pumps, turbine and injectors, same power margin.
• The exit pressure is uniformly 3.6kPa at the second stage (refining this would spread the performance more) and 0.7bar at the first one.
• I take the needed delta-V to Low-Earth-Orbit as 9800m/s equivalent in vacuum. Spread as 5800 and 4000m/s.
• The dry mass is 70kg per ton of propellants for oxygen and Rg-1 (Zenit is worse) and I scale it proportionally to the propellants' volume.
  Everything: the tanks, interstages and engines, which is sensible, but also the equipment.
• I had to hand-estimate some heats of formation (should be fine) and densities (less sure).

All competitors make a lighter and smaller launcher than methane, even without the full-flow, much because their smaller volume saves inert mass.

• Cyclopropane is available industrially.
• Spiropentane is very stable too and may have methods for mass-production
  http://www.chemicalforums.com/index.php?topic=79757.msg314570#msg314570
  it's even more efficient and it evaporates less quickly.
• Diazetidylcyclopropane seems easy to mass-produce
  http://www.chemicalforums.com/index.php?topic=86972.msg314020#msg314020
  and C+N=11 shall make it safer against fire. I've estimated 988kg/m³. Unless drawbacks appear, diazetidylcyclopropane can be the best choice.
• Cyclobutene is too fragile, so oxygen shall cool the engine walls. This keeps the jacket clean for reuse: last argument lost for methane. It enables also the simpler expander pumping cycle for more fuels
  http://www.scienceforums.net/topic/82346-expansion-cycke-rocket-engines/?p=797705
• Bicyclobutane has plausible production paths and it improves over cyclobutene. I've take the same 841kg/m³. I ignore whether these two decompose catastrophically.
• Ladderdienes would improve further, but I lack data.
• Dispirooctadiene is prospective and its safety unknown. C=8 makes it a quite flammable liquid, which trispirodecatriene would improve.
  http://www.chemicalforums.com/index.php?topic=91735.msg327577#msg327577
  I've take arbitrary 880kg/m³.

In a pressure-fed rocket stage, these propellants would rank nearly in the same order.

CompareMethane.pdf here under is a renamed xls file.

Marc Schaefer, aka Enthalpy

[wildfyr]

Do you have any connections to try to get such calculations into the hands of the engineers at these companies? SpaceX especially seems like it has a culture where it will take a good idea no matter where it comes from.

[Enthalpy]

Direct connections to the companies, no... But many people read the forums. When I put ideas on a forum to mitigate the catastrophe in Fukushima, they arrived to the right persons. And from time to time, I spread a CD with my inventions.

SpaceX has advanced a lot in their methane engine, I don't imagine that they change the fuel now, except if they are completely stopped by a problem, like soot in the fuel-rich prechamber - but even then, I suppose they would give up the full-flow and continue with the oxygen-rich prechamber alone, or add a second pumping scheme for the methane. Anyway, the comparison I make here is for tanks a bit heavy, in which case bulky methane is a drawback, but SpaceX has announced very light tanks, of graphite fibre if I remember well.

The last message is more an incentive to think twice, for other companies prone to copy whatever SpaceX does.

[Enthalpy]

Cheap allene (propadiene) oligomerizes easily to precursors of spiroalkanes.

Methenylcyclobutanes can be cyclopropanated. Maybe the final alkane is cheap enough. I didn't find if a 2+2 cycloaddition provides spiroheptanes; light with decent efficiency would be clean and affordable.

As heats of formation, I've seen only spirohexane, dispirooctanes and [4]rotane in
 The Chemistry of Cyclobutanes, by Rappoport and Liebman
which suggests that spirohexanes bring no more strain energy than the cyclopropane plus cyclobutane, surprisingly. I wish more measurements were available. So to estimate spiroheptanes, I added the rings' strain energy, not forgetting the heat lost in branchings. This is pessimistic.

5s over RG-1 equals hydrocarbons more difficult to produce, or more flammable, or unstable at heat. The listed C₈ is flammable and the C₁₀ marginal, but limited proportions can stay in a mix to lower the melting point. One could also cyclopropanate one methylidene and cyclobutanate the other, or fully cyclopropanate the trimer and cyclobutanate the dimer. The density is unknown but probably excellent.

I've found a decomposition speed for hot spirohexane, no idea if it's the main path
https://kinetics.nist.gov/kinetics/rpSearch?cas=157459
Spirohexane would then bear excellent +317°C under the criterion here, 1000 moles of a fuel leaving 10g decomposition products over 300s.

Marc Schaefer, aka Enthalpy

[Enthalpy]

A hypothetic synthesis for ladderanes:
  chemicalforums and next

==========

Some (= all) launcher engines still cool the walls with the fuel instead of the oxygen. Their fuel must create no gum, polymer nor bubbles at heat. This is interpreted as "remove all multiple bonds, all gaseous and solid solutes".

For such engines, I computed ladderanes, often obtained from more efficient ladderenes. A better way exists (drawing): end the ladderdienes like housane with C₃ rings. My estimated heats of formation claim that at the same number of carbons, which nearly means at same flash point, the housane-ended ladderanes are as efficient as ladderdienes, 2-3s better than ladderanes.

Again, a C₁₀ compound is about as reluctant to flame as turpentine, so a mix centred near C₁₂ is excellent.

==========

The synthesis could try ladderdienes cyclopropanations or (drawing) C₃ ring closures by halogens and metals, which work for spiropentane and bicyclobutane. The halogens include Cl, Br, I and the metals Zn, Mg, Li, Na, K usually, as gases, liquids, amalgams. Zn seems easier to recycle, but Li has electrolytes and electrodes for accumulators, and Na too since a recent Chinese car sells with an Na battery.

An adjusted acetylene amount can centre the produced mix near C₁₂ to heal the marginal flash point of C₁₀.

Marc Schaefer, aka Enthalpy

[Enthalpy]

(Pseudo-) Rotanes are energetic, some are flame-safe, or heat-resistant, or freeze uneasily, or look easy to mass-produce. Maybe a mix of them makes a good rocket fuel?

They all lack badly data. The few heats of formation look very wrong. As I have nothing about spiroheptane and spirohexane isn't credible, I can't even extrapolate additively. I've no Mp neither, just the very symmetrical C₁₂H₁₆ is solid.

Software isn't too wrong at estimating Bp. I subtracted 120K for rough Fp, please doubt.

Thoughts about synthesis and production may follow.

Marc Schaefer, aka Enthalpy