Index Home About Blog
Newsgroups: rec.pyrotechnics
Subject: Re: Cyclonite (RDX)/ (HMX)
Message-ID: <>
Date: 15 Apr 91 22:04:41 GMT

In article <>, (Grant Waldram [Atom Smash]) writes:
> Yes. Typo in 'The CRC' :  The correct empirical formula is C H N (NO )
> 							    3 6 3   2 3
> It is most easily produced by the nitration of Hexamethylene tetramine
> by concentrated nitric acid. When you simply pour the nitrated mixture

No, using acetic anhydride, ammonium nitrate and paraformaldehyde at
60 oC is easier. You need no nitric acid. Even the yield is better than
with the direct nitration giving 25..30 % yields. This process
(referred to as K-method ("K-Verfahren")) gives yields up to 60 %.
Moreover, this method is safe, there is essentially no possibility
of a cook-off reaction, since you don't have any nitric acid in the process.

> into ice water, the Hexogen precipitates out. This process is inherently
> dangerous, due to the fact that you are splitting off (by oxidation) three

Yes, pouring the nitrated mixture into ice and boiling the solution
after that is a very dangerous procedure. At this stage the instabile
byproducts decompose giving off lots of noxious fumes. Sometimes
the reaction reaches explosive violence ...

> Impact sensitivity : 0.74 kp m = 7.4 Nm    (what the hell is a 'kp m'?)

An old unit, kilopond meter, i.e. the energy required to lift 1 kg up to
1 meter. It corresponds to 9,81 Nm, i.e. the above is not quite correct.

> insensitive, and could be comfortably processed at normal room temps,
> and probably up to about 70C in solution. However, I would not want to

With the KA-method (WFNA, hexamine nitrate, AN, acetic anhydride) the
working temperature is 60 to 80 oC, optimum being at 68 oC (gives the
best yield).

> play with this shit, and would seek a very experienced chemists opinion
> on the characteristics if forced to play with it. Speaking of which,
> do we have any experienced chemists who would like to pass comment?

RDX and PETN are two of the easiest and safest made high explosives.
You really have to sqrew up intentionally to make a hexogen process to
explode, unless you are nitrating hexamine directly.

                               /          \        O
                             /     ArNO     \    //
                             \\        2    //--N+
                              \\          //     \
                                \\------//         O-

From: (Arno Hahma)
Newsgroups: rec.pyrotechnics
Subject: Re: RDX / HMX manufacture
Message-ID: <>
Date: 18 Apr 91 12:52:52 GMT

In article <> writes:

>In article <>, (Lawrence Curcio) writes:

>> I'd be interested in seeing a detailed rundown of the KA method from Arno,
>> including an isolation workup at the end (that was my problem). Once

The product is precipitated during the process - that is actually the problem.
It precipitates on everything that is in contact with the solution, often as
a thick crystal slurry or as plate-like crystals. The stirrer and the walls
of the vessel are covered with a thick layer of it and the mixing becomes
increasingly difficult. To prevent this, the stirrer should run along the walls
to loosen any crystals at once they are precipitated.

In any case, the product should be boiled in water after filtering and washing
it. A recrystallization from acetone is recommended to remove traces of
possibly unstable impurities.

>work-up to minimise the HMX in the product.  Why not use this method to
>produce the more powerful HMX at lower temperatures 40 oC (if my memory is

Yes, it is possible to get about 90 to 98 % pure HMX, if BF3 is used as a
catalyst, the temperature is maintained at 44 oC (+-1 oC) and the
concentrations of the reagents are held constant during the reaction.

>RDX without HNO3 would be interesting and I would like to read more on this.

This method is called E-method after Ebele. This is probably the safest method
for making RDX, since no nitric acid is involved. On the other hand, the method
requires a lot of acetic anhydride, which is quite expensive.

First the acetic anhydride is warmed up to 60..65 oC and 0,5 % of BF3 is
added. Then, AN and paraformaldehyde are added in right proportions maintaining
the temperature by cooling. As all the reagents have been added, the mixture
is cooled to 20 oC and the product is filtrated, washed, boiled in water to
remove instabile components. Yield 63..65 %, in laboratory scale up to 80 %
has been achieved.

The amounts of the reagents can be calculated according to the

3 CH2O + 3 NH4NO3 + 6 (CH3CO)2O -> (CH2N-NO2)3 + 12 CH3COOH

Note, that paraformaldehyde is (CH2O)n . H2O, the water content has to
be determined to calculate the necessary excess of acetic anhydride for
binding the excess water. Of course, an excess of the acetic anhydride
will not do any harm.

>The direct method has the advantage of only producing RDX.

There is another method for pure RDX, with about 80 % yield. The product is
purer than that made with the direct nitration. This method, called W- or
Wolfram-method uses the potassium salt of 1,3,5-triazacyclohexanetrisulphonic
acid as a raw material. This is nitrated under anhydrous conditions using 99+%
nitric acid and 20 % SO3 dissolved in it. The raw material may be prepared by
condensing potassium sulphamate and formaldehyde.

>I have only produced RDX in 35% yields compared to the 70% claimed.

I think the 70 % is for the process with ammonium nitrate involved, i.e.
a direct nitration with ammonium nitrate added to the mixture
- called K-method after the inventor Knoeffler. In this method, also producing
pure hexogen,
1 part of hexamine per 8,6 parts of nitric acid, in which the calculated amount
of ammonium nitrate has been dissolved, is added maintaining the temperature
below 15 oC. After all the hexamine has been added, the temperature is raised
to 80 oC, whereby the reaction with AN takes place. The slurry is filtrated,
washed with dilute sodium carbonate and recrystallized.

An improved method, worked out indepently in the USA and Germany, is called
a Bachmann-process or KA-method. These are essentially the same process, with
slight differences in the temperatures and the proportions of the reagents.
The KA-method gives a somewhat better yield and uses less acetic anhydride.

Hexamine dinitrate is used as the raw material. It is prepared mixing a
concentrated solution of hexamine in water with 60..70 % nitric acid, molar
ratio 1:2. The temperature is kept below 15 oC and finally the solution is
cooled as low as possible to crystallize out the dinitrate. Yields of over 95 %
are obtainable. The dinitrate is dried at room temperature or slightly above.

The process itself is simple. First, 350 parts of acetic anhydride is heated
to 40..50 oC. Into this, 24 parts of ammonium dinitrate (a solution of
AN in nitric acid, 1:1 molar ratio) and 22 parts of hexamine dinitrate is added
in portions maintaining the temperature in the given limits. The proportions
of the reactants have to be 12:11 all the time. Preferably, the addition
is made continuosly in the right proportions. After this, 270 parts of acetic
anhydride is added in batch following by 5 parts of ammonium dinitrate. After
this, the whole batch is heated to 60..68 oC and maintained there for 0,5 hours
. Finally, it is cooled to 20 oC, the product is filtrated, washed with
sodium carbonate and recrystallized, if necessary. The yield can be improved,
if BF3 is added as a catalyst, for example as BF3-acetic acid complex.
This also aids to prevent any run-off reactions, since the catalyst
accelerates the reaction and the materials do not accumulate in the solution.
The yield without BF3 is about 75..80 % calculated on hexamine used.

Note, that the temperatures are of great importance to maintain as given.
If the temperature is too high, oxidation occurs and the yield is strongly
reduced. Also, instabile compounds are formed and these may be present in the
filtrate. If the temperature is too low, a run-off reaction will occur, as
the materials react too slowly. The reactants build up in the solution
without reacting at once. Then, as the reaction starts due to slow temperature
rise, too much heat is generated and no cooling is enough. The mixture
begins to boil and strong oxidation reactions produce even more heat.
Finally, if the batch is large enough, even an explosion may result.
Of course, the same might result, if the cooling is ineffective or the
reagents are added too fast.

The reactions produce great amounts of heat, so an effective cooling system
will be necessary. Immersing a beaker in cold water is not enough, unless very
small amounts are made. A cooling coil and an effective stirring of the
solution is required. Otherwise the process will be very slow, as you'll have
to wait for a long time before adding the next proportion of the reactants.

The KA method is probably the most effective and economical. It is not as safe
as the E-method, there is more chance for a cook-off to happen.

>I have a few ideas on increasing my yield the next time I get around to making
>some more.  I would like to test the RDX in a composite rocket propellant one

>        CH2NO3
>        |
>        |
>O3NH2C--C--CH2NO3     PETN  (Potentially Explosive Treat Nicely)
>        |
>        |
>        CH2NO3

B-). The same applies to hexogen as well.

For more details about the processes see T. Urbanski's "Chemistry and
Technology of Explosives", vol. 3.


                               /          \        O
                             /     ArNO     \    //
                             \\        2    //--N+
                              \\          //     \
                                \\______//         O-

Index Home About Blog