If Bis-nitrotriazoles weren’t enough for you…

A couple weeks ago, I talked about a patent published by the Klapötke group in which a series of bis(aminotriazole) salts were prepared and characterized.  It’s pretty neat stuff, and the molecules showed pretty solid energetic performance across the board.

Well, as luck would have it, another publication from the same group came out last week.  This piece is a followup on some chemistry from back in 2015 [1], wherein they prepared some triazole tetrazoles bearing nitro, azido, and amino ring substituents (compounds 5–8).  You know, for the severely carbon-averse out there.  A quick snapshot of this chemistry is reproduced below.  Compound 5 is afforded in ~64% overall yield in 5 steps.  The process chemists out there will appreciate this detail: all compounds are prepared by chromatography-free means.

Reproduced from [1]

Get your azoles, now with 33% more nitrogen!  — Reproduced from [1]

As one might expect, the sensitivities of the four parent compounds to external stimuli (impact, friction, ESD) ranged from “quite” to “extremely.”

The scope of the recent offering is to use these materials as energetic anions with various metallic and non-metallic cations [2].  Since the tetrazole proton is basically holding on for dear life — as evidenced by a 1H chemical shift of 16.19 ppm — preparing these salts was a matter of treating the parent compounds with a basic solution of the desired cation.  The goal here being use metal cations (a–c) to produce sensitive primary explosives, and nitrogen-rich cations (d–h) to produce less sensitive secondary explosives.

Get your nitrogen heterocycles, now with 33% more nitrogen!

Nitrogenous energetic materials 5–8 and corresponding metal (a–c) and nitrogen rich (d–h) cations — Reproduced from [2]

Silver’s an interesting choice; I’m generally content to not find myself in the same room as large quantities of any compound with both “azido” and “silver” in its name, but nonetheless, compound 6c exists, albeit only briefly.  The compound has impact and friction sensitivities below the minimum test threshold using the BAM method.  To quote the text:

The yield [of 6c] was not determined owing to the extremely high friction sensitivity of this compound.

The cesium analog, 6b, fares slightly better.  It holds together long enough to get some spectroscopy data, and the impact, friction, and ESD sensitivities of this compound were high, but measurable.  The text notes the ESD value is slightly less sensitive than lead azide, which is commonly employed in blasting caps.  But high sensitivity alone does not necessarily make for a good primary explosive — the material must be capable of initiating detonation in a secondary explosive.  To that end, a mass of RDX was loaded into a copper tube sealed at one end, and 50 mg of compound 6b was layered on top.  Firing with an electrical detonator resulted in the image below:

rdx det

Left to right: intact copper tube; copper tube after deflagration of 100 mg 6b; copper tube fragments after detonation of 300 mg RDX initiated by 50 mg 6b — Reproduced from [2].

So it does in fact kick off secondary explosives quite nicely — fragments of metal casings are often a welcome sign in energetic research.  It is also noted that compounds 6a–c all undergo a deflagration to detonation transition (DDT), although it’s not clear how this was determined.  Small samples of explosives generally do not detonate unless they are confined or there is a critical diameter present.  But I digress.

Now onward to the so-called “nitrogen rich” salts.  A total of 14 nitrogenous salts were prepared, however four were only isolable as the corresponding dihydrates: 5d, and 8-2e–g (compound 8 formed the corresponding dianion).  The remaining ten demonstrated a very high degree of insensitivity.  With the exception of 7e, all were less sensitive than RDX in BAM impact and friction, and ESD tests.  Additionally, all were within a 10% margin of RDX in both detonation velocity and detonation pressure according to EXPLO5 calculations.  Likewise, almost all burned cooler than RDX.

One major advantage of using nitrogen-based HEDMs is that they are non-oxidative.  That is, no oxidizing salts (say, perchlorate, nitrate), and minimal organic nitro-groups, so metal components of gun barrels won’t break down as quickly.  Since deflagrations/detonations are far from ideal from a stoichiometry standpoint, you can end up with significant amounts of NO and NO2 (and even HCl if your propellant is ammonium perchlorate).  Those gases get pretty warm in armament combustion chambers and can seriously damage barrel bores, which can lead to critical failure.  But with nitrogen heterocylces, you reduce the formation of nitrogen oxides and instead form mostly N2, and eliminate chlorine entirely.

So to that end, the authors reformulated two propellants, HN-1 and HN-2 (HN = “High nitrogen”), which are used in very large bore guns.  The compositions are mostly RDX, with some TAGzT, a relatively sensitive triaminoguanidinium salt of 5,5′-azotetrazole.  Substituting compound 5h, for TAGzT resulted in about a 10% boost in the specific energy of the formulation, with a modest increase in combustion temperature.

This is a perfect example of one of the fundamental challenges of energetics research, which I mentioned in passing in my previous post: you can’t have it all.  You can always cram more energy density into a molecule — just stick more nitro groups on it, or replace a triazole with a tetrazole.  But increasing the energy density comes at a price, almost without exception: your molecule burns hotter, or is more sensitive, or both.  And the authors acknowledge this, stating that while the combustion temperature increases, it is still well below the critical temperature for gun propellants.


  1. Dalton Trans., 2015, 44, 17054, DOI: 10.1039/c5dt03044g
  2. Euro. J. Inorg. Chem., 2016, DOI: 10.1002/ejic.201600108

Bigger booms, through chemistry

The Klapötke group at LMU is marching relentlessly onward with their quest to find new and interesting ways to stick as many nitrogen atoms onto one molecule in as close proximity as (barely) possible for long enough to get NMR data.

You may remember the Klapötke group from Derek’s post over at ItP in the “Azidoazide Azide” issue of Things I Won’t Work With.  This is the group that would look at pentazole and think “Gee, I wonder if we could replace that proton with an azide…”  I’ve always thought this kind of work was pretty cool; most of these crazy nitrogen heterocycles are practically useless but they serve the important purpose of giving us a better understanding of the nature of chemical bonds at the margin of what is possible.

Klapötke et al is back with a published patent application that showed up on my scanner.  This time, they’ve taken a step back from the realm of the ridiculous and have prepared a reasonable looking energetic active ingredient: 3,3′-dinitro-5,5′-bis-triazole-1,1′-diol (and a couple bis salts thereof).

And that structure looks not at all unreasonable.  Sure, electron deficient triazoles aren’t the most stable, but that hydroxyl contributes some electron density back to the ring system.  Oxygen balance looks good.  Slightly under-oxidized, actually, which as a rule gives you a bit of stability back.

But enough with speculation, let’s take a look at the thermal and sensitivity data provided in the text.  In energetics, RDX is commonly used as a benchmark: it has good (not great) explosive performance, and it reasonably insensitive to impact, friction, and electrostatic discharge.  Interestingly, the application does not present characterization data on the parent diol, but instead offers three salts: dihydroxylammonium (MAD-X1), diguanidinium (MAD-X2), and di-triaminoguanidinium (MAD-X3).

And the lead compound, MAD-X1, outperforms RDX across the board: better sensitivity in all three metrics, high detonation velocity (9.3 km/s to RDX’s 8.7), greater crystal density, higher thermal decomposition onset, larger heat of formation, and lower detonation temperature.  As anyone who works in the field knows, it’s really hard to have it all; you can always increase you explosive performance… at the expense of sensitivity.  And vice versa.  But, as far as performance metrics go, MAD-X1 seems to pretty handily have a leg up on the competition.

Even the synthesis is pretty straightforward and uses decidedly non-exotic reagents.  First, oxalic acid is condensed with aminoguanidinium bicarbonate in concentrated HCl, then worked up under basic conditions, affording 3,3′-diamino-5,5′-bis-(1H-1,2,4-triazole) (“DABT”).  DABT is then oxidized to the bis-nitro derivative as the corresponding dihydrate, which is fantastic from a energetics processing standpoint.  Treatment with potassium peroxymonosulfate affords the anhydrous diol, which reacts subsequently with an ethanolic solution of hydroxylamine, which yields MAD-X1 in 44% overall yield over four steps.

synthesis of MADx1

While not as concise as the two-step Bachmann process, which yields RDX from hexamethylenetetramine in 57% overall yield on an industrial scale, Klapötke’s preparation of MAD-X1 appears scalable.  Namely, it dispenses with the wildly exothermic nitrolysis process used to make nitramines — if you’ve ever had the pleasure of performing such a reaction you’ll know it’s incredibly easy to end up with a runaway reaction and a resultant yield rapidly approaching zero.  Do that on a large scale, and you’ll have a pilot plant rapidly approaching low earth orbit.

Overall, I’m pretty impressed with this compound’s prep and apparent utility.  My main criticism is: how’s that alkoxide salt going to hold up in an environment where metals are present?  Namely, in a casing or shell.  If the the use of picric acid has taught us anything, it’s that acidic energetics tend to not play well with metals.  I’d love to see some followup formulation work addressing this issue.