[Enthalpy]

The report that compares (with too elementary methods) hypergolic ignition:
http://www.dtic.mil/cgi-bin/GetTRDoc?Location=U2&doc=GetTRDoc.pdf&AD=AD0093924
it's the same report that tries dimethyldiazetidine. The link has worked this time as well.

Textbooks about propellants are full of mistakes, misunderstandings and false assumptions. I recommend anyone interested in the subject to download and run
Propep http://www.dark.dk/download/
Rpa http://software.lpre.de/downloads.htm (more recent, nice, easier to use, but no three propellants, hence no aluminium)
because propellant combinations can't be compared with hand computations.

Any oxygen atom is detrimental to an efficient rocket fuel, even if it's linked to a nitrogen.

I prefer not to discuss aluminium here, to focus on azetidine, diazetidine, diazaspiroheptane and their derivatives.
Data about safety, liquid range, heat of formation, ease of synthesis...?

[AndersHoveland]

HOCH2CH(R)CH2NH{C6H4}OCH3 is refluxed with (C3H3N2)2CO, using CH3CN solvent for two hours,
forming (C3H3N2)C(=O)OCH2CH(R)CH2NH{C6H4}OCH3. This was then heated to 150degC for 2 hours under vacuum pressure (0.05 mbar) to form R{C3H4N}NH{C6H4}OCH3. (C3H3N2) is an imidazole ring (nitrogens in 1- and 3- positions) which is connected on one of its nitrogen atoms.--{C3H4N}-- is a 1,3-substituted azetidine ring. (C3H3N2)2CO is also known as 1,1'-Carbonyldiimidazole. The variable R group ends in the 3-position on the azetidine ring. --{C6H4}OCH3 is a para-methoxyphenyl protecting group. This can usually be removed by hydrolyzing off the ester methoxy group with 20%HCl, then oxidizing. The --CH2NH{C6H4}OH turns into
--CH2N={C6H4}=O, and this resulting imine easily hydrolyzes off with more acid, leaving --CH2NH2 and benzoquinone  
O={C6H4}=O. R. M. de Figueiredo, R. Fröhlich, M. Christmann, J. Org. Chem., 2006, 71, 4147-4154

I had a thought about reacting glycerol with limited bromoethane, then distilling off the glycerol mono-ethyl ether.
66% would be 1- or 3- substituted, while 33% would be desirable 2-substituted. Then react the distillate with
acetyl chloride, putting acetyl esters on the remaining hydroxyl groups. The 1,3-acetyl,2-ethoxy propane could react with acetamide CH3CONH2 (the reaction would take days) to give low yields of 1N-acetyl,3-ethoxy-azetidine.
The azetidine derivitive could be distilled out, since it would be the most volatile of any of the reaction products. I would think 10% yield could be reasonably expected. based on the initial glycerol.
The ethoxy-ester could be hydrolyzed to 1N-acetyl,3-hydroxyl-azetidine, then the acetyl hydrolyzed off in conc NH4OH solution, leaving plain 3-hydroxy-azetidine. Any comments or opinions about this idea are welcome.

[AndersHoveland]

"A novel and efficient method for the synthesis of 1,2-diazetidines"
Wei Miaoa, Weiliang Xua, Zhiliu Zhanga, Rujian Ma, Shu-Hui Chena, Ge Li

Summary : A novel and efficient method has been developed for the preparation of racemic or optically pure 1,2-diazetidine from 1-(1-hydroxy-propan-2-yl)hydrazine-1,2-dicarboxylate under very mild conditions with excellent yield.

Basically,  HOCH2CH(R)--N(CO2H)--NH(CO2H) either undergoes a Mitsunobu reaction, or reacts with methanesulfonyl chloride CH3SO2Cl. I think the later option is very similar to the Mitsunobu reaction, because
I found an image for EtOCONHN(SO2CH3)CO2Et (there are not any N--O bonds in the structure),
which can be seen here:  http://journals.iucr.org/e/issues/2007/04/00/hg2206/hg2206contents.gif

Then the intermediate is reacted with a base.
Apparently this reaction produces the diazetidine derivitive (R)C2H3N2(CO2H) in 85-96% yield.

This article: http://www.ncbi.nlm.nih.gov/pubmed/8103389 suggests that 1,2-diazetidine-1,2-di-N-oxides exist, although I am not sure how stable they are. This has an oxygen atom on each nitrogen atom.

One would therfore wonder whether 1-hydroxy-2-hydrazino-cyclobutane could similarly undergo a Mitsunobu reaction to form  2,3-diazobicyclo[2.1,1]hexane. (with a diazo bridge connected to the 1- and 3- positions of the cyclobutane ring, forming a caged molecule.

[Enthalpy]

But why a 2,3-diazo in 2,3-diazobicyclo[2.1,1]hexane? And why 1,2-diazetidine?

These N-N compounds are expected to have more heat of formation, but
- hydrazines are already good propellants. Users want to get rid of them because they're carcinogens.
- more recently, nitramines are considered carcinogens.

So aren't all these N-N compounds unhealthy? Would we switch from molecules meanwhile known to be deleterious to ones whose detrimental effects will only be known later?

The idea behind 1,3-diazetidine derivatives and 1,4-diazobicyclo[2.1,1]hexane was to get an energetic compound without any N-N bond.

They would also be tertiary amines, which have a lower melting point and better hypergolic ignition.

As well, if a safe propellant needs dangerous reactants, the hazard has only moved.

[AndersHoveland]

Yes, you are quite correct, nearly all N--N bonded compounds (including double bonded) seem to be toxic.

Starting with 1,3-diamino cyclobutane...

http://upload.wikimedia.org/wikipedia/commons/c/cf/SulfonamideSynthesisfromSO2.png
http://www.orgsyn.org/orgsyn/orgsyn/prepContent.asp?prep=cv6p0078

(I previously lost my entire post when this site logged me out    )