Sciencemadness Discussion Board

Novichoks and Related 3rd/4th Gen. OP Agents

Sauron - 31-1-2007 at 21:36

This is pretty much a summary of what I have been able to find on this, there was a thread here about this 2.5 years ago, but that seems a bit stale to reopen.

Soviet development of these very interesting and potentially highly threatening compounds apparently began in the 60s and continued through the 80s and, with ever tightening secrecy especially after 1972. It is reported that some or all of these compounds can defeat MOPP gear, that some are supertoxic (many times more toxic than GA/GB/GD/VX), that they can'be detected by western military instrumentation, all of which if true are highly alarming.

The USG appears to be trying assiduously to conceal the existance or at least delay release of much details about Novichok (despite this now being a five decade old Soviet projecy.) Acronyms appear with some regularity to replace all references to the nickname Novichok in western official documents. Two examples are NTA ("Non Traditional Agents") and FGA "Fourth Generation Agents".

Novichok simply means "newcomers" in Russian, especially in the sense of Johnny-come-lately or the VN era US military slang for replacement grunts, "FNG" (Fucking New Guys").

Some clues have emerged as to structure of at least some of the Novichok class of agents.

Author Steven Hoenig in the very recent (11/06) Springer-Verlag "Compendium of Chemical Warfare Agents", q.v., is quoted on roguescience as identifying three supertoxic Novichoks:

CAS [26102-97-6]

CAS [26102-98-7]

CAS [26102-99-8]

I won't bother with Hoenig's nomenclature as it seems to be confused, but I did go draw these in ISIS/Draw 2.5

which indicates that these are all fundamentally more related to PCl3 or phosphorous acid. Note the absence of a C-P bond. P is bonded only to O and F.

The functional group on the right is a dihaloformaldoxime specifically the oxime of carbonyl chlorofluoride. The purpose of this group is to irreversibly bind to AChE and BuChE receptors and preclude reactivation by oxime therapies such as 2-PAM. The Russians have a long love affair with phosgene oximes as CW agents, going back to WWI. Carbonyl chlorofluoride is produced from Cl2, F2, and CO and reacted with hydroxylamine. However, I do not believe that the dihaloformaldoxyl group is introduced to POCl3 or a trialkyl phosphite in such a direct fashion.

The other alkoxy group is 2-chloroethyl from POCl3 and 2-chloroethanol (ethylene chlorohydrin).

The P-F moiety is probably introduced by fluorination of a P-Cl function as a final step, in a manner similar to the preparation of DFP (cf.Sanders book.)

When Ellison's "Handbook of Chemical & Biological Warfare Agents" appeared a very odd review of it showed up in The reviewer used the name of a recently deceased Russian general. He summarized some of the Novichok project history and that of the parallel "Dusty Agents" weaponization, named key researchers, cited journal publication prior to 1973, and gave two alleged Novichok structures. This review is now available in the MadHatter FTP site.

Later: The preparation of the dihaloformadoxime functionality attached to the P nucleus is done at low temperatures in liquid SO2 solvent by reacting POCl3, (RO)3P=O (a trialkyl phospate), (RO)3P (a trialkyl phosphite) or RP(O)Cl2 or RP(O)(OR)2 etc with chloropicrin, dichlorofluoronitromethane, dichloronitrosomethane, chlorofluoronitrosomethane, or other halogenated nitroalkanes. Possible substitutents alpha to the nitro or nitroso group are Cl, Br, F, or a stable pseudohalogen such as -CN. The entry point for these nitro/nitroso compounds is the corresponding substituted acetic acid, just as in the classical prep of nitromethane from chloroacetic acid.

So, dichlorofluoroacetic acid, as an example, is treated with sodium nitrite to obtain chlorofluoronitroacetic acid. This is then decarboxylated. The product is chlorofluoronitromethane. I do not have experimental details or physicochemical date as yet, but this ought to be in the open lit. The Novichoks are not so mysterious after all, but they constute a large and structurally varied group of agents (apparently mostly solids).

[Edited on 1-2-2007 by Sauron]

[Edited on 1-2-2007 by Sauron]

novichoks.jpg - 9kB

Sauron - 31-1-2007 at 22:37

Fellow students of OP agents will recognize the progression in the series above follows the observations of pioneers Schrader and Saunders, in that the A-232 alkoxy is secondary as opposed to the primary in A-230. We would expect therefore A-232 to be more toxic. A-234 on the other hand is more hindered and like soman (GD) may well be more persistant than the other two.

The review in, q.v., which some have called suspect, gives a different structure for the same CAS number as A-230, and a very different structure for another Novichok.

As that review was edited by our own @samosa maybe he can shed some light on this discrepency. Was it deliberate obfuscation by the "deceased Russian general"? Or a mistake based on faulty nomenclature? The structures in appear to be, on their face, unlikely to me.

(Later) On the E&W Forum at roguescience, Megalomania states flatly that the structures given below are incorrect and that fits in with my own view stated above.

[Edited on 1-2-2007 by Sauron]

[Edited on 1-2-2007 by Sauron]

novichok2.jpg - 7kB

Sauron - 1-2-2007 at 12:32

It appears that the only relatively exotic precursors to the Novichok class of OPAs are the dihalonitromethanes and dihalonitrosomethanes.

However, see p.164-165 of Sartori's "The War Gases"

Trichloronitrosomethane (b.p. 5 C) and trifluoronitrosomethane (b.p. -80 C) were both evaluated as CW agents a long time ago along with the rest of the chloropicrin family.

Therefore it seems very likely that the chlorofluoronitrosomethane precursor to the Novichek agents is to be found well described in the open literature.

Five minutes later after a Google: the energetics boys to the rescue. Chlorofluoronitrosomethane is an intermediate in the synthesis of an energetic plasticizer fluorodinitromethane (FDNM). From a DTIC report (Fluorochem Inc.) in 1990 (attached abstract):

"Nitration of 1,2-dichlorodifluoroethylene gave chlorofluoronitroacetic acid, which was converted to chlorofluoronitrosomethane with red fuming nitric acid and water."

[Edited on 1-2-2007 by Sauron]

Attachment: http___stinet.dtic.mil_oai_oai_&verb=getRecord&metadataPrefix=html&identifier=ADA219385.pdf (20kB)
This file has been downloaded 1468 times

Ozone - 1-2-2007 at 15:43

Very nasty! It looks like these, besides being lipophillic, will also penetrate most forms of rubber (latex and nitrile for sure, butyl, and maybe even, chloroprene). There appears to also be very real possibility that they can chew into the skin to adminster themselves (these would probably be very vessicating if you lived long enough to find out).

I agree that the Amazon structures look like hooey. The acetylenics would likely polymerize, and further, I see no special biologically disruptive reason for them to be there.

Interesting, I had not previously been aware of these agents.



Sauron - 1-2-2007 at 21:34

Certainly if the free dihaloformaldoxime formed, there would be vessication. But if someone were exposed to enough of this stuff to have to worry about vessication they would be dead anyway. Besides, the point of that functional group is to NOT hydrolyze, not even exchange with other oximes (reactivators) so I expect it to be extremely stable both at the P-O-N bond and at the receptor site. Therefore unlikely any free oxime can form.

Yes absolutely defeat of MOPP gear was and is a design goal. This is analogous to the GF agent "cyclosarin" which defeats at least some protective gear.

On that basis I would anticipate that a Novichok in which the 2-chloroethyloxy function was replaced by a cyclohexyloxy function or a 2-chlorocyclohexyloxy function, would have enhanced capabilities against protective gear. See sketch below. I have proposed an appropriate trivial name for it.

As to the erroneous structures, I remember well once in the 1970s a house organ of the ACS (C&EN) published a VX structure with the side chain -S-CH=CHN< and we all had a good laugh about it. That pi bond of course should be a sigma. So, there's precedent for just this sort of error.

[Edited on 2-2-2007 by Sauron]

cyclosaurin.jpg - 4kB

Ozone - 1-2-2007 at 21:43


The F-...Same game the medicinal Chemists are playing with drug stability! Selling these compounds as the oxons as also quite nasty, but will, no doubt effect the life of the active moeity--which could be "good" because, who wants to deploy a chemical agent when they cannot enter post-haste? I think that the O-Et-Cl chain on the other end is supect as well...

Good night,


Sauron - 1-2-2007 at 22:27

The SIPRI folks always proposed standardizing on GF because they argued that a CWA against which there was no defense would be a deterrant in the MAD sense. I think they were being overly optimistic, is not pollyannish.

Should we assume that just because the russkies engineered Novichoks to defeat western MOPP, that they did not develop their own protective and detection countermeasures against it? Not a safe assumption!

Of course there is much discussion and has been for a long time of CWAs as area denial weapons, and this is pretty consistent with the historical practice of various communist forces of not mapping their own minefields. Certainly the case in Cambodia where there are 20 million mines and the Khmer people mostly find them one at a time, the hard way, by treading on one.


I am having a hell of a time finding out anything about 1,2-dichloro-1,2-difluoroethylene [27156-03-2].

Nothing in Aldrich, nothing in Merck, nothing on Acros, nothing in Ullmann. I have not even been able to locate the bp. I assume it is a gas, but with a MW of 192 maybe not. If it is a gas then I know where to look for a supplier, I am striking out with the usual suspects for reagents.

Even Fluorochem, the folks whose report DTIC has on file, do not sell it.

Is this in the unobtainium department?

[Edited on 2-2-2007 by Sauron]

Sauron - 2-2-2007 at 10:16

Here are the two "new" (to the public) G-agents details of which were released in Ellison's Handbook and a few more in the review. The structures are from Ellison and probably reliable. The CAS numbers and some of the nomenclature are from the somewhat dubious Russian review and therefore may not be reliable.

[Edited on 2-2-2007 by Sauron]

GPGV.jpg - 11kB

Ozone - 2-2-2007 at 19:19

CycloSauron:D, perhaps you should change your screen-name!

With the possible exception of cycloSauron, your earlier structures looked far nastier, but still, you are considering G and V agents which are frighteningly toxic to begin with!

The big meany with the new ones is the ability to insinuate themselves though the typical polymers used to prevent intoxication.

As for Soviet MOPP gear being resistant...the only things I know that might be OK are rigid aramides, etc. which must be woven (much to rigid as solids); woven materials tend to leak. How do you think teflon would do with those fluoro-modified "business ends"? My guess is that the answer would be something like "it depends on how thick".

The G progression just looks like the old combi-chem approach, and may well be a "realistic" model to use in misdirection (they don't like to disclose the "family jewels").



Sauron - 2-2-2007 at 20:28

Exactly. All the way back to Saunders, Schrader et al (rivals of course) they were screening many many structural variations and looking for toxicity spikes in the SAR. More recently we have speeded this up with QSAR and parallel synth and so on but in the 40s and 50s it was wet lab and more wet lab and I am sure they made hundreds if not thousands of variants - many of which were of low or indifferent toxicity or simply did not meet other well established military requirements. Vapor pressure, persistance, MW, phase at STP, and so on.

GP and GV look like a couple from such series, but I have not seen anything to indicate they were ever weaponized or stockpiled. GA, GB, GD yes. VX yes. These two? Dunno.

GF as far as I know was not stockpiled by anyone simply because the stuff does defeat at least some protective gear.
I didn't show that structure but I thought it was clear: GF is same as GB but with cyclohexyloxy replacing the isopropyloxy. In the G-series, secondary alcohols > primary alcohols.

2-propyl GB should be regarded as baseline
pinacolyl GD (hindered so slower hydrolysis, more persistant)
cyclohexyl GF defeats some protective gear

You can get a pretty good clue about other likely candidates for the secondary alcohol from reading the CWC schedules. All the cycloalkyl alcohols from C3 to C7 are probably similar to GF, the most practicable being cyclopentyl but who knows, maybe cyclobutyl and cyclopropyl would have special advantages?

As for the alkyl side chain linked directly to P, straight chains out to something like C10 would still be potent though likely to be solids.

But GV and GP have neither an alkyl group nor a simple alkoxy group. They have one bit from GA and one bit from a precursor of VX. I'd be curious about those structures if F were replaced with CN and/or the O in the side chain were replaced with S. But they aren't. So I smell a rat. I think those are bogus. I am not however sure enough that I'd want to get up close and personal with them.

Ozone - 2-2-2007 at 21:29

My guess is that the toxicity of the GV/P agents would be sufficent to make deconvolution of your data difficult. That is, with enormous dilution comes propogated error at DF 50,000 +-.0001 becomes significant. the hard nucleophile makes for a stable derivative that is less likely to be mitigated, metabolically.

I'd think that the nitrile in a P-C bonded system would be less persistant; this persistance vs. acute toxicity could be tested, I suppose, by somehow putting a nitro group onto that sucker. If this was easy, I am sure it would be a killer (and we might, could, possible have gotten wind of it--unless it was really good:).

As for a standard compound, I would be inclined to try diisopropylfluorophosphate (which is actually used as the surrogate for GB analysis--why they use a surrogate on nearly equivalent toxicity is beyond me; it makes the blanks CW toxic). It is of prototypic chemistry and toxicity for G agents *and* is well characterized.

On second thought, GP/V look like "missing links" in the evolution of G to V agents which means an increase of acute toxicity of ~2 orders of magnitude; I can only speculate on P-O agents. In a P-O agent, CN may exert a much more considerable inductive effect (the entire thing will be an electron transport chain) which can remove electron density from the active site causing it to bond much more tightly (and, probably with greater speed and specificity).

Still, I am both petrified and in awe of the hell that can be wrought with such things,


Sauron - 2-2-2007 at 23:59

Putting on a -CN is in some ways more facile than putting on an -F.

Take GA (Tabun) as archetype.

You start with POCle. This is also the case for GP and GV and it's one of the options for Novichoks.

The proceduire is given in Saunders and also in a number of open source pubs so I am not spilling any new beans here, just emphasizing a point.

P(=O)Cl3 -> Me2NP(=O)Cl2 is usually first step and this is why dimethylamine and POCl3 are highly restricted/watched.

The second step is done at a single go. This means, really, that the product of the first step, and the reagent for second step, constitute a crude binary.

What is the reagent?

An alkali cyanide in ethanol.

How simple can you get? Oh, you probably want to have Pyr or TEA or another tertiary base for H-acceptor to drive the equilibrium to the right, in both steps.

If we take a page from GB binaries then just use NaCn an equimolar mix of ethanol and ethylamine.

GA is Me2NP9=O)(CN)OEt. By the way, decontamination with the usual mixture DC (hypochlorite) gives rise in case of GA or any other P-CN bond hydrolysis, to NCCl that is cyanogen chloride, a military agent in its own right. Combination choking agent and blood agent.

Schrader knew what he was doing.

Incidentally you don't have to be Linus Pauling to observe that the order of toxicity of the halogens goes F>Cl>Br>I

-CN is a pseudohalogen and probably fits in their right after F. Why? Toxicity parallels halogen electronegativity. Consider the consequences of those electronic effects on the P and resulting effects for the binding potency to the AChE receptor sites.

GA is nonpersistant but the military wants both persistant and nonpersistant agent options.

GA is similar to DFP in ease of prep but a lot more potent. It is a popular first agent for countries starting out in CW for those reasons. CWC can't really control the precursors for this one. It's a joke.

Chris The Great - 3-2-2007 at 03:17

GD or soman is also impossible to treat with oximes. It ages within minutes and binds irreversibly, while most agents take several hours (V agents) to days (Sarin). When the ester hydrolyses, then it's aged. Don't know why pinacolyl would age so much faster than any of the others.

It certainly is a fascinating subject, and a completely misunderstood one in terms of the toxic effects. A simple proof is that there are a certain breed of rats with acetylcholinesterase, which nerve agents inhibit. They are pretty fragile but with car they can live hundreds of days. Despite not having any AcCholinesterase to inhibit, nerve agents still kill them, although at roughly half the potency. Nerve agents also have different effects as sub-toxic levels and exhibit many weird things which are not explained except by a vast array of complex interactions with the nervous system.

Sauron - 3-2-2007 at 12:24

OK I worked out four different feedstocks and routes to a variety of the immediate precursors to the Novichok dihaloformaldoxime side chain (which is the only real novelty about these agents). I deleted all the previous nattering which was just me thinking out loud. Instead I now present the Isis/Draw sketch of this set of schemes which I believe is pretty close to what the Russians are up to.

[Edited on 4-2-2007 by Sauron]

precursors.jpg - 47kB

Sauron - 4-2-2007 at 09:26

The patent for FDNE with chlorofluoronitrosomethane as intermediate (see above.)

The patent describes how the starting material 1,2-dichlorodifluoroethylene can be prepared from commercially available 1,2difluorotetrachloroethylene, citing Locke et at, JACS 56, 1726 (1934) for details. It also states that the 1,2-dichlorodifluoroethylene is available commercially.

Chlorofluoronitrosomethane is described as an intensely blue liquid, stable at r.t. for a few days, but usually used directly in subsequent steps. I would expect this dihalogenated nitrosomethane to be a highly irritating substance, based on the known irritant properties of trichloronitrosomethane and tricfluoronitrosomethane (see Sartori, "The War Gases" in the forum library, pp.164-65.)

Apart from that, the two stage nitration/oxidation does not appear to be any more hazardous (or any less so!) than many of the nitrations proposed on this forum.

[Edited on 4-2-2007 by Sauron]

hbx53 - 4-2-2007 at 09:40

If you think this stuff is too much toxic then why don't you write their synthesis processes?

By synthesis processes, more people came to know how this stuff was made.

Besides Novichok, Dimethylmercury was also good and toxic stuff, if possible provide synthesis processes of this also.

Sauron - 4-2-2007 at 10:24

All my source material is in the open literature. Anyone could read them. Anyone with the skills could draw the same conclusions. This is a chemistry forum, and as far as I am concerned I am talking about interesting chemistry. I am not advocating making this stuff, and certainly not advocating using it.

Are you advocating censorship? If so you are facing a hostile audience.

Dimethylmercury is not a military agent, it is a byproduct of the pulp and paper industry. Do you happen to reside near a pulp mill?

Ozone - 4-2-2007 at 10:54


A recently came across an article in "Archives of Toxicology" wherein it was proved that the F- does not dissociate from GB when AChE binding takes place. What do you suppose is the fate of the F- when bonded as N-P(F)-O-C? Seems to me that the F binding constant whoudl be smaller with these molecules.

@ hbx53

There are other threads that are more appropriate for the discussion of Me2Hg such as:

Detritus then too, hmm?



[Edited on 4-2-2007 by Ozone]

hinz - 4-2-2007 at 11:09

Maybe there are more OTC-methods:

If Freon 11 (CCl3F) or even Freon 21 (HCCl2F) is available, it would be possible to oxidise them with oelum or chromsulfonic acid to carbonyl chlorofluoride.
I expect because the C-F bond is stronger than the C-Cl bond, that the Cl-Atoms will oxidise first. So the reaction should work like with carbon tetrachloride or thrichlormethane:

SO3 + CCl3F =heat==> SO2Cl2 + O=CFCl
CHCl2F + (O2 from air but slow) or chromsulfonic acid ==> O=CClF + H2O +Cl2

The produced carbonyl chlorofluoride could then condensed with hydroxylamine, but there I don't know whether in the second part of the reaction the Cl or the OH group will be eliminated.

O=CFCl + H2N-O-H ==> HO-C(FCl)-NH-OH ==> FClC=N-OH + H2O
or in bad case:
O=CFCl + H2N-O-H ==> HO-C(FCl)-NH-OH ==> HO-C(F)=N-OH + HCl

Probably you'll get a mix of both, but the OH Group might be rechlorinated by thionyl chloride. The whole condensation must carried out in a non nucleophilic solvent, as the esters of carbonyl chlorofluoride are easily produced. A good solvent might be dimethyl sulfoxide.

Ozone - 4-2-2007 at 11:23

mmm. DMSO can explode when mixed with acyl chlorides (which is what carbonyl chlorofluoride is). This particular acyl chloride might be weird, though, since the F- will be strongly inductive leading to a positive C-Cl polarization. Still, I think it will react leading to, at the very least a loss of product (and maybe the synthesis of some unwanted and very nasty side products). Nevermind the immediate reduction of the DMSO to DMS (which stinks like hell) that would ultimately result when your primary oxidant is exhausted (you would have to watch it like a hawk).

The acidity usually required for operations like these take place in anhydrous HF:SbF5, for example. I would be very surprised if it (the oxidation) took place in oleum. Do you have any reference material for your proposed transformation?



[Edited on 4-2-2007 by Ozone]

Sauron - 4-2-2007 at 11:33

The oxime of carbonyl chlorofluoride was my very first supposition but, in the absence of experimental results reported I see no reason to assume that such an oxime =N-OH will bond with P-Cl, or replace P-OR.

The Russians are definitely using dihalogenatednitromethanes and nitrosomethanes which do react with both types of P bond.

The Russians were the first to weaponize phosgene oxime and use it in warfare, and they have more experience with its manufacture than anyone else AFAIK. So if they could use carbonyl chlorofluoride they could. But they don't.

BTW you don't need to oxidize one-carbon Freons to get that stuff, all you need is carbon monoxide, clorine and fluorine or, ClF interhalogen.

Anyway is starting with one CFC refrigerant any more OTC than starting with another? No more no less.

The purpose of this whole exercise from my point of view was to explicate what the Russians were doing and not especially to come up with alternative (and unproven) routes. I interjected the Fluorochem prep of the precursor because it is proven and because at the time I had not yet worked out the likely Russian processes.

Trichloroacetic acid, acetonitrile and chlorinated acetonitriles, and chloropicrin (cheap agro fumigant) are all very OTC as well.

But thanks for the comment.

[Edited on 4-2-2007 by Sauron]

hinz - 4-2-2007 at 11:50

The oxidation with oelum should work with trichlorofluormethane as well as with carbon tetrachloride for phosgene, since the fluorine group is not changed, the reference for this is "the war gasses " p61. But instead of heating CCl4 I would preheat the oelum and bubble gaseous CClF3 through it and trap unreacted Cl3CF in a cold trap with dry ice/acetone.

For the second reaction I've no reference, the courageous home chemist would have to try it, but your're right, the condensation of hydroxylamie with formaldehyde has less side reactions, there is only one OH group and no more halogens that might eliminate, the best solvent has also to be searched. A bit tricky as it has to dissolve both carbonyl chlorofluoride and hydroxylamie and be non ncleophilic, too.

I wish a lot of fun to the chemist that has to work with fluorine, chlorine and carbon monoxide at the same time, futhermore I don't think the chlorine in trichloracetic acid is that simple replaced by flourine or even by NO2-, it's far too much steric hindered

[Edited on 4-2-2007 by hinz]

Sauron - 4-2-2007 at 12:33

No more so than in trichloronitromethane (chloropicrin) yet that is precisely what the Russians are starting with, there's over 1000 Russian publications on this topic (mostly pretending to be about pesticides of course) and the evidence is all there.

To get from trichloronitromethane to chlorofluoronitromethane they must be knockig off one Cl, which solves your steric hindrance problem, then exchanging one Cl for F, and there's the reagent.

We are told that they are also using acetonitrile, and that some of their formaldoximes contain a -CN instead of one of the halogens. I say it is reasonable to assume that these two facts are not unrelated, and therefore they are making a halocyanonitrosomethane or -nitromethane, and if so, they are dichlorinating acetonitrile, or dechlorinating trichloroacetonitrile to the dichloro form. Again, no steric hindrance issue. (Zinc dechlorinates hexachloroethane just fine, which is how smoke grenades work.) After that the same two steps apply.

It would be trivial to painstakingly repeat the explanation of the procedures for the reported Br containing variants. I believe the F/Cl and F/CN pairs on the dihaloformaldoxime chains are probably the major ones.

As to carbonyl chlorofluoride, have a peek at Brauer's book. Personally I would not care to work with carbonyl chlorofluoride under any circumstances, any more than phosgene. But these do in fact get used in labs and industrially (well, at least phosgene is). My objection is that they are insidious. You can get a lethal exposure long before symptoms appear.

The nitrosomethane is probably intensely irritating rather than insidious. I's guess it is a likely carcinogen, after all nitromethane is now a suspect so nitrosomethane ought to be more so. And at least it's a liquid, and therefore more manageable.

Bottom line: show me a publication that says that P-Cl and HO-N=CX2 condense to P-O-N=CX2 and I'll agree that this route is practicable. Or demonstrate it experimentally, say with PCl3 and acetone oxime, harmless enough. Until then it is conjecture. Maybe reasonable conjecture but conjecture the same. What I am describing is not conjecture. I have filled in a few blanks but the starting materials were stated and the immediate precursors reported. I just played connect the dots.

Sauron - 4-2-2007 at 22:29

Here's a 1989 US patent on conversion of Freon-112 (1,1,2,2,-tetrachloro-1,2-difluoroethane) to 1,2-dichloro-1,2-difluoroethylene by catalytic hydrogenation at elevated temperatures in gas phase.

And also the older JACS article on doing same with zinc in absolute methanol. Not as selective but simpler.


[Edited on 5-2-2007 by Sauron]

Attachment: 4876405[1].zip (409kB)
This file has been downloaded 955 times

Sauron - 5-2-2007 at 06:18

Something I ignored till now mostly because I needed to look up the unfamiliar nomenclature.

The "suspicious" review, which now is not looking suspicious at all but 24K gold, prominently referenced an article in the proceedings of a 1969 conference. The title referred to 1,3,2-dixaphospholanes and iminophosphites.

I was unable to obtain the article but I did learn that the phospholanes are 5-membered cyclic phosphorus heterocycles and therefore [1,3,2[-dioxaphospolanes look like this:

[Edited on 5-2-2007 by Sauron]

The review actually talks about substitutes 1,3,2 phospholanes as precursors to the Novichoks with the final step being the ring opening by heating.

Well, I suspect that the ring opening may be by HCl hydrolysis. Remember, the examples of final product we have been shown all had a 2-chloroethyloxy side chain. The phospholanes shown are essentially cyclic esters of PClx and ethylene glycol. Open that heterocyclopentyl ring up with HCl and what do you get?

In examples we have been shown, R=H or Me, R2=H or Me, R3=F. After ring opening there's an available P-OH or P-Cl to react with the dihalonitrosomethane to form the dihaloformaldoxime side chain.

In short, the Russians may not be using PCl3 or POCl3 at all if they can buy this phospholane.


[Edited on 5-2-2007 by Sauron]

pring2.jpg - 4kB

Sauron - 9-2-2007 at 21:05

I obtained the full (25 page) DTIC document reporting on Fluorochem's preparation of FDNE via CFNM.

They flatly state the idea came from the Russian literature and cite two articles by same principal investigator whose name shows up consistently through two decades of Novichok-related publications.

Unfortunately, this report was written at a point when the Naval Research Lab had redirected Fluorochem's labors, and prior to their eventual success, so for details we still must rely on their patents.

They do say that commercial 1,2-dichloro-1,2-difluoroethylene and also same prepared by then by Locke's method, were contaminated by 10% 1,1-difluoro isomer which if nitrated would generate a highly toxic product.

So I suppose their success came only after preparing this substrate without the byproduct.

[Edited on 10-2-2007 by Sauron]

Attachment: FDNE.pdf (958kB)
This file has been downloaded 1302 times

Sauron - 11-2-2007 at 12:07

A treasure trove of information on the halogenated nitromethanes and nitrosomethanes is to be found in PATR 2700. I have extracted the two key sections and combind them in a pdf attached below.

Volume 3 p.40-41
Volume 8 p.89-104

Specifically, chlorofluoronitromethane is found there.

Chlorofluoronitrosomethane can be prepapred from that by mild reduction, but AFAIK either compound reacts in same fashion with OP compounds, so there is no particular need to do that reduction unless you like blue colored compounds.

We were told that these dihaloformaldoxines have the general structure

-O-N=C(X)-y where X and Y are F, Cl, Br, or a stable pseudohalogen like CN. In the examples shown to us X and Y are different but we cannot disallow the possibility of X=Y. Therefore the number of possible combinations are trivial to calculate, and most of these combinations are now documented here as the immediate precursor. The only exceptions are the ones containing -CN.

The preparative methods are diverse and the literature extends back to the 19th century. An examination of this ought to allow us to do a better job of filling in the remaining blanks.

Picatinny's interest of course is in energetics applications such as the longstanding interest in fluorodinitroethanol. The long exploited process for this came to a halt in 1970 when the Swedish plant making nitroform (TNMe) blew up.

Obviouly though a different US Army arsenal two states to the south of Dover NJ also has a keen interest in these dihalonitromethanes and nitrosomethanes.

Fascinating reading. Enjoy!

[Edited on 11-2-2007 by Sauron]

Attachment: ExtractfromPATR2700.pdf (1.7MB)
This file has been downloaded 1823 times

Sauron - 12-2-2007 at 10:06

The mining of PATR 2700 for preps, references etc on the halonitromethanes, nitrosomethanes and related compounds is proving to be interesting and productive.

Some of these are proving to be easy to prepare (or so it appears on paper).

I'll present a full report when I am done digging.

Sauron - 13-2-2007 at 04:00

Nitrosomethane itself is fascinating.

In monomeric form it is a blue gas.

But it is usually isolated as its dimer Me-N(O)=(O)N-Me which is a solid, which rearranges to the gas when heated above its melting point. Further heating produced trimeric formaldoxime.

At ordinary temperatures nitrosomethane slowly changes into the formaldoxime cyclo(-N(OH)-CH2-}3, a six membered heterocycle C3H9N3O3

Most preps are by UV irradiation of tert-butyl nitrite via a methyl radical process. Yields are variable but in general not high. tert-butyl nitrite isn't cheap but can be prepared from cheaper materials.

Nitrosomethane and its dimer can be halogenated by the elements into dihalofonitrosomethanes such as dichloronitrosomethane and dibromonitrosomethane. Difluoronitosomethane is not so easy.

Both cis and trans isomers are possible.

Nitrosomethanes are easily oxidixed to nitromethanes.

Nitromethanes may is some cases be reduced to nitrosomethanes but overreduction gives methylamines.

Freshly prepared formaldehyde oxime hydrochloride (formaldoxime monomer) can be reacted with nitrosyl chloride in cold Et2O to give chloronitrosomethane I am hoping that this compound can be fluorinated or that one of the chlorines on dichloronitrosomethane can be exchanged for F.That wiuld give the important chlorofluoronitrosomethane.

Nitrosyl chloride is made from sodium nitrite and conc HCl per Inorg.Syn. vol 4.

Formldehyde oxime monomer hcl from hydroxylamine hydrochloride and formalin soln. The oxime is extracted with ether after salting out with CaCl2.

No need for messing about with carbonyl chloride fluoride, the above procedures don't appear to be anything like that hazardous.

The halogenated nitrosomethanes are doubtless irritants and should be assumed to be toxic and carcinogenic but, normal precautions ought to suffice.

[Edited on 13-2-2007 by Sauron]

Sauron - 13-2-2007 at 06:01

Here's another interesting prep, this one for monohalonitromethanes starting from readily available diethyl malonate:

1. The malonic ester is nitrated in high yield to diethyl nitromalonate.

2. The nitromalonate is halogenated. In the literature example the reagent is perchloryl fluoride gas, but I am sure for Cl and Br much less exotic reagents will suffice.

3. The product is saponified and decarboxylated to the halonitromethane.

Fluoronitromethane is stable below 5 C and the best yields were only 32% but the F compound is a difficult one.

So, diethyl malonate could be used instead of socium chloroacetate to prepare unsubstituted nitromethane, and ordinary conc nitric acid used instead of sodium nitrite. Ought to be useful for those of you who have expressed difficulty in obtaining sodium nitrite. Details and references on request.

Sauron - 14-2-2007 at 22:58

I am well along with literature searching to flesh out the information gleaned from PATR 2700 ("Federoff's") and extend it to present.

That covers the halogenated nitromethanes and nitrosomethanes.

What remains are the halonitroacetonitriles and halonitrosoacetonitriles. Some haloacetonitriles are commercial products.

ISIS/Draw's nomenclature routine did not 'like' the tautomer of the nitrosoacetonitriles with C=N bond although this is probably the stabler form since it is conjugated to the nitrilo group. However it accepted the other tautomer with N=O bond and named it correctly as nitrosoacetonitrile.

Sauron - 15-2-2007 at 11:59

There's an obscure derivative of phosgene with one Cl replaced by -CN.

It is a liquid with b.p. 120 C and is prepared by dehydrating and chlorinating ethyl oxamate with phthaloyl chloride. See Sartori, "The War Gases" p.58 and ref 1.

(ethyl oxamate is the ethyl ester of oxamic acid, the half amide of oxalic acid.)


The oxime if this NC-C(=NOH)Cl would be very very close to

NC-C(=N-O)Cl the nitroso cpd or NC-C(-NO2)Cl the chlorocyanonitromethane

Another route might be via methyl or ethyl nitroacetate (commercial products or easily prepared from nitromethane).

nitroacetate to nitroacetamide to nitroacetonitrile (aka cyanonitromethane). Start with a halonitroacetate and you end with same structure as from ethyl oxamate.

We are told the Russians use acetonitrile as a starting point, and also chloropicrin, but I still do not find any evidence that this is true. It is true that the intermediates and precursors are substituted nitromethanes/nitrosomethanes and some can be regarded as substituted acetonitriles, but that does not necessarily mean getting to these directly.

The nitroacetic esters by the way are tools in peptide synthesis, my primary synthetic interest. Very interesting that they turn up in novichok chemistry as well.

Formyl fluoride is known, so is carbonyl fluoride. So doubtless cyanoformyl fluoride may exist (may even be know.) Same with cyanoformyl bromide although it might be less toxic and less stable.

Wonder why iodine is absent from the novichoks? Carbonyl iodide is known but extremely unstable with a -80 C temp of decomposition. The same holds for the oxime, very low stability. Also reduced toxicity.

Sauron - 16-2-2007 at 00:21

Diethyl oxalate (commercially available and cheap) is treated with the calculated amount of ammonium hydroxide in ethanol to give ethyl oxamate.


This is treated with phthaloyl chloride, which dehydrtaes the amide to nitrile, cleaves the ester and chlorinates the acyl group to produce cyanoformyl chloride, oily liquid, bp 120 C at normal pressure.


Hit that carbonyl with hydroxylamine hcl and what have you got?


which is awfully akin to the target compound


In fact at the right pH I's expect these to be tautomers.


hbx53 - 20-3-2007 at 06:15

How can we increase the persistence capacity of dimethylmercury?

One of Novichok product consists of Acetonitrile and ______ ? Can you tell the second one?

Sauron - 20-3-2007 at 21:57

I'm not inclined to increase persistence of dimethylmercury. I'd rather reduce it to about zero.


Newer Developments on Novichoks and Folient Project

Sauron - 3-7-2008 at 02:54

The Allen Reaction appears to be pertinent to the novichoks and "Folient" Project and was first described in JACS 79, 3071-3073 (1957). I will post this here shortly. Various web pages mis-cite this paper as JACS 79, 3071 (1975) and give no DOI or author.

The DOI of the correct paper is 10.1021/ja01569a022

The title is: :

"Phosphorus (V) Acid Esters of Oximes"

which ought to be enough to make anyone who has read the posts upthread sit up and take notice.

The abstract states:

"Alkyl esters of trivalent phosphorus acids react with a-halogenated nitro compounds to give alkyl halide and oxime esters of the corresponding pentavalent phosphorus acids."

[Edited on 3-7-2008 by Sauron]

Attachment: ja01569a022.pdf (421kB)
This file has been downloaded 1223 times

Sauron - 3-7-2008 at 04:18

Way upthread we examined some purported Novichok structures reported by "Samosa"

Now we have more details on the reaction and precursors.

The starting esters are glycol esters of PF3 or PFCl2 acid halides of phosphorous acid H3PO3 The first structure shown below is precursor to Novichok A-230 and the second, to Novichok A-234.

These are reacted at subzero temperatures with dichlorofluoronitrosomethane in dry Et2O and on warming to O C or above rearrange to the open chain esters.

The following are precursors to the compounds described by Samosa.

[Edited on 3-7-2008 by Sauron]

noviprec.jpg - 40kB

ScienceSquirrel - 3-7-2008 at 05:01

Personally I wonder how much of the Novichoks program is the product of the fevered imaginations of intelligence operatives and defence contractors out to get their hands on a wodge of Uncle Sam's dollars.

Many years ago there was a compound called red mercury; what it was, even if it contained mercury at all was in doubt, what was not in doubt was that it was the hottest thing in defence since bullets.
It could;

a) render an aeroplane completely invisible to radar

b) 'focus' the neutron flux in a A bomb so you could make one the size of a hand grenade

and a lot more besides.

A lot of money and time was wasted on red mercury and I reckon there are a few old KGB men relaxing in their dachas on the Black Sea coast courtesy of the CIA ninnies.
However no verifiable sample of red mercury has ever been produced.

Sauron - 3-7-2008 at 06:17

Having been main mover in the debunking of "red mercury" on this forum, I am here to assure you that no parallel exists in regard to novichoks.

Red mercury was the invention of scam artists and confidence tricksters not intelligence services. They were trying to fleece the sorts of people with lots of cash and an interest in buying supposed nuclear materials for bomb making.

The evidence for the reality of the novichok Folient Project in former USSR is massive. Russian scientists have been arrested, imprisoned and prosecuted for whistle blowing about this. The main sources are emigre former scientists. Russian journal articles confirm much of what they are saying. Ans American officialdom, which was quick to debunk red mercury as a fraud, is silent on the novichoks. Several well known authorities on CW have included novichoks in their books. It is not mere Wikipedia claptrap.

See the attached document.

[Edited on 3-7-2008 by Sauron]

Attachment: novichoks2.doc (61kB)
This file has been downloaded 2380 times

Sauron - 3-7-2008 at 07:10

Here are the unstable intermediate of the reaction of gluorinated phospholane shown above and dichlorofluoronitrosomethane. The lower structure is the final novichok identified as A-240.

The same compound could be built from 2-chloroethanol, POCl3, KF, and phosgene oxime. The last is a substance the Russians have more experience with than anyone as they mass produced it as a chemical weapon in WWII.

A fluorinated derivative of chloropicrin would also work.

[Edited on 3-7-2008 by Sauron]

A240.jpg - 121kB

ScienceSquirrel - 3-7-2008 at 08:21

Perhaps it would be unsporting of me to point out that red mercury was at it's height during the late eighties - early nineties.
Neither Wikipedia nor this board existed at the time and most of the stuff written about it is in the dead tree library.
There was definitely a good healthy dose of scam about it but quite who was scamming whom never really became clear.
Anyway I remember plenty of other 'intelligence' stories. What about the stuff that was being pedalled in the run up to the invasion of Iraq?
If you believed the line that Bush and Blair were running Sadam Hussein had enough nasties to render the whole Mediteranean basin free of life.

Sauron - 3-7-2008 at 08:50

Use the search engine, o unsporting one. There was some damned fool who started a thread about red mercury and some similarly imaginary isotopes on this forum last year and citing pages from Wikipedia. THAT is what I was referring to. I take note that you are very new here and can't really be expected to have read everything that has gone before your arrival. That, however, does not mean that the forum only started when you got here.

Anyway this thread is about the chemistry of the novichoks and not about red mercury at all. If you don't wish to stay on topic kindly go elsewhere.

[Edited on 4-7-2008 by Sauron]

ScienceSquirrel - 3-7-2008 at 09:12

Originally posted by Sauron
Diethyl oxalate (commercially available and cheap) is treated with the calculated amount of ammonium hydroxide in ethanol to give ethyl oxamate.


This is treated with phthaloyl chloride, which dehydrtaes the amide to nitrile, cleaves the ester and chlorinates the acyl group to produce cyanoformyl chloride, oily liquid, bp 120 C at normal pressure.


Hit that carbonyl with hydroxylamine hcl and what have you got?


which is awfully akin to the target compound


In fact at the right pH I'd expect these to be tautomers.

Perhaps you could explain how these would be tautomers as they do not have the same number of atoms.

ScienceSquirrel - 3-7-2008 at 09:34

Also could you give a reference for the reaction of a compound of the type RCOCl reaction to form an oxime.

I would normally expect the formation of RCONHOH via elimination of HCl from the transition intermediate.

Sauron - 3-7-2008 at 09:38

It's immaterial.

Sauron - 3-7-2008 at 09:52

You misreprersent the structure

The compound in question is a derivative of phosgene, and perhaps you have heard of phosgene oxime? Or the oxime of carbonyl chloride fluoride?

-CN is a psuedohalogen.

Actually, phthaloyl chloride will not dehydrate ethyl oxamate that way. The old paper is wrong. This was proven when the compound ClC(=O)-CN was prepared by other means and identified by modern instrumental methods. Its physical constants were nothing like the material obtsined from ethyl oxamate and phthaloyl chloride. In fact as you might expect, the bp of the mixed carbonyl was intermediate between those of phosgene and carbonyl cyanide.

Also the mixed nitrile is not stable and disassiciates to a mixture of phosgene and carbonyl cyanide, the symmetrical compounds, in a few days.

I did a seperate thread on this errata in the old literature. It will be fun one of these days to prepare ethyl oxamate and react it with phthaloyl chloride, and see just what that product really is. But it is not Cl-C(=O)-CN.

ScienceSquirrel - 3-7-2008 at 09:58

I would say that it is very far from immaterial.

Some of your chemistry is fine but some of it would have left my PhD supervisor with his jaw hanging open!

Sauron - 3-7-2008 at 10:03

You resurrected a post how old? to argue over a trivial point and then attack me personally? You haven't earned the right. Not by a long shot.

The point is immaterial because the novichok compounds (real or not) are reported to be prepared NOT from the oximes but from the alpha-halonitromethanes and nitrosomethanes, and the appropriate trivalent phosphorus compound.

The discussion in the early part of this thread a year and some months gone is therefore, rendered moot. Redacting it is a waste of time.

You are just doing your best to derail this thread. You will not succeed.

[Edited on 4-7-2008 by Sauron]

ScienceSquirrel - 3-7-2008 at 10:58

I have heard of phosgene oxime but you do not make it by reacting phosgene with hydroxylamine.

Reaction of ammonia or an amine produces the corresponding urea derivative,



I would expect hydroxylamine to form


Do you have to earn the 'right' to question someone's chemistry round here?


Sauron - 3-7-2008 at 16:32

The first few posts in this thread were written eighteen months ago and concerne a mention in Sartori's THE WAR GASES (in the forum library), q.v.) of the reaction between ethyl oxamate and phthaloyl chloride to produce the mixed Cl-C((=O)-CN.

Subsequently I learned that Sartori's cited paper was in error. The assymetrical carbonyl has been prepared by reaction of phosgene and HCN under autogenous pressure, and characterized by instrumental analysis unavailable to the author Sartori quoted. His high boiling product therefore cannot be the assymetrical carbonyl, which actually has the expected bp between phosgene and carbonyl cyanide, the symmetrical analogs.

As the editing time on this forum is limited I cannot go back and correct anything in those year and a half old posts. I only mentioned that old paper because it appeared to be pertinent to Samosa's document which indicated that the formaldoxime moiety in novichoks could include a pseudohalogen such as -CN on the carbon.

As it turns out those novichoks are prepared not from the phosgene oximes but from the halogenated nitromethanes or nitrosomethanes, analogs of chloropicrin, or chloropicrin itself.

Chloropicrin etc are unpleasant irritants and lachrymators but, phosgene oxime is deadly, and at the least will blind very quickly.

These phosphorus (V) acid oxime esters therefore arise indirectly by the Allen reaction and not via the oxime, although oximes have been used to prove structure in certain relatively nontoxic cases.

So, thanks for pointing out that phosgene oxime is not prepared from phosgene and hydroxylamine. But the matter of the oximes as far as this thread is concerned is now moot and your correction is 18 months too late.

On a personal note it is IMO very poor form to go round on an amateur chemistry forum brandishing your sois-dissant docrorate like a club with wish to bash all lesser chemists. There are many Ph.D. chemists on this forum, and I have not seen any of them being so crass as to establish an elite class.

So, are you done, or do you have anything else to add about red mercury, old posts, or the price of tea in China? Because I for one would prefer to return to the thread topic.

[Edited on 4-7-2008 by Sauron]

Sauron - 3-7-2008 at 17:12

Squirrel is correct. Phosgene oxime is not prepared, as I assumed, from hydroxylamine and phosgene.

It is prepared by reduction of trichloronitrosomethane with H2S or aluminum amalgam.

Or, from silver or mercury fulminate,

I never looked it up before, because it is a violent vesicant and I had no interest in its preparation, and still don't.

Sauron - 4-7-2008 at 21:46

The novichok reaction is the "Allen Reaction" between two moles of a trialkyl posphite and one molea alpha-trihalonitroalkane, the most familiar such compound being trichloronitromethane, (chloropicrin) Cl3CNO2. The much less familiar nitroso compound Cl3CNO also works. There are some constraints on the choice of substituents.

1. One of these must be chlorine, this is what attacks phosphorus initially.

2. The others may be Cl, Br, F or a psuedohalogen e.g. -CN. These are the substituents that remain in the oxime ester side chain. These may be identical or different.

The product is an exime ester of a dialkyl, dialkoxy, alkyl halo, or alkoxyhalophosphate.

See graphical scheme upthread and below.

The preparation of the chloropicrin and analogs is well known.

The most convenient prep of trichloronitrosomethane is one starting from trichloromethyl sulfenyl chloride, also known as perchloromethyl mercaptan, the familiar product of chlorination of CS2 with dry chlorine in diffuse light.

This is oxidized to trichloromethylsulfonyl chloride according to the procedure of Prandl and coworker in Ber., 62 1752, (1929). A technical grade of this may be available from Eastman Kodak.

The sulfonyl chloride is reduced with sodium in ethanol by the technique of Prandle and coworker in Ber. 65 (1932). The product is sodium trichloromethylsulffinate.

The procedure described in same peper for preparing the trichloronitrosomethane, a deep blue liquid, from the sodium salt of the trichloromethylsulfinic acid is simple but causes some thermal decomposition and also produces a product not free from water.

The same salt can be reacted in the cold with nitrosyl chloride in liquid phase, under autogenous pressure, the reaction being complete when the mixture is allowed to warm to O C. This procedure is detailed in a paper by Suttcliffe in JACS 79 3071 (1975). The anhydrous trichloromethylnitrosomethane is well suited for use directly in the Allen reaction.

Allen made no reference to any special precautions being necessary for the preparation of the simple oxime esters he characterized.

I would not assume them to be nontoxic, however.

It is clear now that the basic chemical principles behind the novichoks have been known to the United States for at least thirty three years. It is worth noting that Allen was a chemist employed by FMC Corporation and that corporation was a major contractor involved in the VX program.

At least three reported novichok structures have been described in detail, those being A-240, A-242, and A-244. Are these compounds AChE inhibitors possessing toxicities of military potential? It would be simple enough for someone in an appropriate facility to find out. But they aren't talking, one way or the other. Neither confirmation nor denial.

The questions are hanging there.

[Edited on 5-7-2008 by Sauron]

[Edited on 5-7-2008 by Sauron]

Cl3CNO.jpg - 75kB

Sauron - 5-7-2008 at 00:15

Two equivalents of trialkyl phosphite or dialkyl phosphorofluoridite such as the phospholanes shown upthread, are reacted with one equivalent of a tri-a-substituted nitrosoalkane or nitroalkane in ether at a low temperature.

[Edited on 5-7-2008 by Sauron]

allen.jpg - 112kB

Sauron - 5-7-2008 at 01:19

Now we are ready to fully understand the reaction of alpha-trihalonitroalkane or -nitrosoalkane with the fluorinated phospholane examples referred to in the Samosa document and explained further in the paper posted upthread.

In the phospholane two of the P-O bonds are bridged by an ethylene to form the 5-membered heterocyclic ring.

Nucleophilic attack by the Cl from the nitro compound cleaves one R-O bond, opening the ring. The P-O bond become pi. The Cl rather than exiting as an alkyl halide, remains on the beta position of the ethoxy chain. The second equivalent of fluorinated phospholane is likewise oxidized to pentalaent P but without any bond cleavage.

So the reaction products are one equivalent of the oxime ester (novichok) and one equivalent of phosphonofluoridate.

Slightly simplified reaction graphic (intermediate omitted):

[Edited on 5-7-2008 by Sauron]

allen2.jpg - 69kB

Sauron - 5-7-2008 at 07:42

As the alpha-substituents on the nitro or nitrosoalkanes have to be Cl plus any combination of Cl, Br, F, and -CN or another psuedohalogen, there are a limited number of possibilities.

Longer nitroalkanes than nitromethane are shown in the oxime ester literature but none in the five or six publicly disclosed novichok structures.

The other two substituents on P (besides the pi bonded O and the oxime ester) can be F, R = alkyl or OR = alkoxy. I would not rule out S containing variations although none of the dissidents or emigres have mentioned any such.

Sauron - 5-7-2008 at 22:24

In regard to the oxidation step of the rxn scheme above for trichloronitrosomethane, the oxidation reagent choices are rather unsurprising.

Rathke (one of the pioneering investigators of CS2 chlorination) reported a very slow two phase oxidation of Cl3CSCl with nitric acid.

Later workers found that carrying out the reaction in glacial acetic acid as mutual solvent allowed efficient production of trichloromethanesulfonyl chloride. JACS 63, 1764 (1941)

Oxidation with 30% H2O2 in GAA gave a much better yield (78%) of product of better quality.

Peracetic acid required forcing conditions which were IMO rather hazardous.

The urea-H2O2 adduct was also used succesfully.

Another reagent that has been applied is calcium hypochlorite.

The trichloromethanesulfonyl chloride is a solid. It has been found to be a useful and succesful chlorinating agent.

See attached paper.

If preparation of Cl3CSCl is deemed undesirable, Kolbe teaches that the sulfonyl chloride can be prepared directly from the same precursor (CS2) by chlorination with wet, rather than dry, chlorine. Ann., 54, 145 (1845)

If CCl4 is available it is reported to react under mild conditions with sodium dithionite (Na2S2O4) to give trichloromethylsulfinyl chloride. The preparation of the sodium salt from this ought to be straightforward.

Zhang, Kirchmeier, and Shreeve, Inorg.Chem. 31 492-494 (1992)

Huang, Huang, and Chen, Huaxue Xuebao (1982) 42 1114 (C.A. 102, 78313 (1985))

but since we can't buy CCl4 and prepare it from CS2, this is less helpful than it looks. For mixed halonitroalkanes however it is highly significant.

[Edited on 6-7-2008 by Sauron]

Attachment: jo01067a600.pdf (128kB)
This file has been downloaded 794 times

Sauron - 6-7-2008 at 05:33

The Prandtl paper with preparation of sodium trichloromethanesulfinate is attached. Most of the paper deals with reactions of phosgene oxime and derivatives, which can be ignored. The procedure for the salt is paragraph 1 of the experimental section. The method of O.W. Loew is employed. (Z.Chem., 12, 82 (1869) Trimethylsulfonic chloride, recrystallized from ethanol, is taken up in methanol and reduced with H2S. The solution is neutralized with anhydrius sodium carbonate using Congo Red paper to indicate end point.

The procedure immediately following for conversion of the salt to trichloronitrosomethane using sodium nitrite and potassium nitrate and hot 20% HNO3 is facile but the product is contaminated with water after distillation and some loss occurs due to thermal decomposition. Therefore the Suttcliffe procedure referenced above is superior. The dry sodium salt of allowed to react sans solvent with condensed nitrosyl chloride under autogenous pressure, reaction is complete when the vessel warms to O C or above. The trichloronitrosomethane is anhydrous and yield is improved.

[Edited on 6-7-2008 by Sauron]

Attachment: Ber1932.pdf (454kB)
This file has been downloaded 1337 times

Sauron - 6-7-2008 at 18:11

Hmmm. It appears that bromotrichloromethane (a heavy liquid) is commercially available and cheap, and not likely to be prohibited here (unlike CCl4.)

In this case a one step facile prep of sodium trichloromethanesulfinate is available.

The reaction between Cl3BrC and Na2S2O4 takes place in MeCN/H2O to form the sulfinate with sodium carbonate. See the Inorg.Chem article cited above and attached below.

[Edited on 7-7-2008 by Sauron]

Attachment: ic00029a028.pdf (419kB)
This file has been downloaded 927 times

Sauron - 6-7-2008 at 18:53

Note that the Na2S2O4 is sodium dithionite and NOT dithionate which is Na2S2O6. Sodium dithionate is also often called sodium hydrosulfite, which of course - it isn't.

This certainly looks simpler and less costly than starting with CS2 and working with the nasty trichloromethyl sulfenyl chloride (perchloromethyl mercaptan).

The observant reader will also take note that the same process can be applied to other perhalomethanes and so on, including Cl2FBr if you can get it or make it. This opens up potential routes to the mixed halogen nitrosomethanes and homologs.

[Edited on 7-7-2008 by Sauron]

onestep.jpg - 78kB

Errata and Corrections

Sauron - 7-7-2008 at 18:14

A few errors to correct:

All references to A-240 and A-244 should read A230 and A-234.

There is also a reported A-232 Novichok intermediate in structure between those.

Some sources state that Novichok 5 and Novichok 7, said to be the most potent of the series and 5-8X more potent than VX, are the binary forms of two of the structures shown. This has not been confirmed.

In the graphical reaction scheme from trichloromethanesulfenyl chloride to trichloronitrosomethane, the oxidative step to give trichloromethanesulfonyl chloride can be performed by either nitric acid in glacial acetic acid, or, 30% H2O2 in GAA, the second method being preferred.

The reduction of that product to trichloromethylsulfinic acid and conversion to the sodium salt is classically done with H2S followed by sodium carbonate. The graphic shows, incorrectly, sodium in ethanol. NaOEt would react with alpha chlorines.
Graphic below shows two more fluorodioxaphospholanes along with A-232, one of the reported unitary Novichoks..

It is worth noting that there is no C-P bond in this molecule. This compound therefore is related to the fluorophosphates (e.g., Saunders' DFP) and is NOT within the scope of the Chemical Weapons Convention.

The secondary alkoxy side chain, not being Me, Et, Pr, or 2-Pr, is also outside of the CWC "envelope" while still following the rules set down by Saunders and Schrader for optimizing toxicity.

The P-O-N linkage is also outside of CWC's scope.

The requisite fluorodichloronitromethane or -nitrosomethane is also not CWC regulated (though chloropicrin itself is.)

The oxime ester does appear to be intended to frustrate NATO oxime-based OP therapies.

At first glance it does strain credulity to postulate that a DFP analog could match, exceed or even approach the toxicity of VX, as the two are several orders of magnitude apart. However, this should not be a matter of mere surmise but of experimental determination and the synthesis is not a difficult one.

This would need to be done in an appropriate facility and is therefore way outside of our status.

I would not at all be surprised if these compounds as shown could reach the toxicity of the G-series, as well as the versatility of engineered persistence/nonpersistance, defeat of MOPP gear via cycloalkoxy moiety, etc. I do not know enough to intelligently comment about detection technology.

But if the oxime esters do in fact thwart known therapies then these could in fact be very dangerous military agents.

[Edited on 8-7-2008 by Sauron]

A232.jpg - 119kB

Sauron - 8-7-2008 at 08:53

The Russians call the reaction between phosphites and alpha-trihalonitro/nitrosoalkanes to form oxime esters of phospahtes the Allen Reaction, although I have not seen any Western reference that indicates this to be an accepted name reaction. The JACS paper by J.Forrest Allen is posted abovethread.

This reaction was an extension of reactions of phosphites with a-halocarbonyl compounds to give vinyl esters of phospahtes in a manner entirely analogous to the prep of oxime esters.

Two earlier JACS articles by Allen and coworkers describing these reactions are attached, I combined them into a single pdf.

This allows a deeper understanding of the reaction central to novichok chemistry, to the extent that we know it.

See also that the employement of a dioxaphospholane as the phosphite was not lost on Allen. Ethyl ethylene phosphite is specifically mentioned.

[Edited on 8-7-2008 by Sauron]

Attachment: Allen.pdf (1.1MB)
This file has been downloaded 1180 times

Sauron - 27-9-2008 at 04:24

It is apparent that the precursors for the required cyclic phosphite esters are simple common aliphatic glycols, namely

ethylene glycol
1,2-propanediol (propylene glycol)
meso-2,3-butanediol (2,3-BDO)

The latter two for the alleged components for Foliant 5 and Foliant 7 binary agent.

The preparation of these as the chloro compounds from the glycol and PCl3 equimolar in DCM is described in Lucas et al, JACS 72 5491-5497 (1950). I will post this shortly.

Anyone skilled in the art will recognize numerous ways to convert the chloro compound to fluoro. Another option is to use PCl2F with the glycol in first place. See Saunders.

The proper pronounciation of Novichok is "noweeshock" snd the term is not so universally recognized in defense circles. The code name Folient (or Foliant) however gets an immediate heightening of attention. I have been assured, and reassured, by people who are in a position to know, that the United States and NATO are equipped with Folient-impermeable protective gear, although I am told that this gear is incredibly costly and difficult to manufacture. The transition to this apparel etc took place around the time of the first Golf War. I conclude therefore, that these newcomer agents are very very real, and that their threat has been taken most sriously by the West for about the last 20 years.

That the former Soviet Union succesfully hoodwinked and neutered the Chemical Warfare Convention is now very clear. The protocols of CWC do not encompass the precursors of the Folient agents nor the agents themselves in either unitary or binary form. The phosphorus containing precursors, and the agents, contain no carbon-phosphorus bond. The dissident and defector accounts are IMO accurate although necessarily partial.

It has beena pleasure to tease out the chemistry of these newcomer agents.

The correct name for these compounds is 2-halo-1,3,2-dioxaphospholane (for the ethylene glycol case) while the others are 4-methyl and 4,5-dimethyl substituted.

PF3 and PCl2F are pains to prepare and store, so so it is probably better to prepare the 2-chloro compounds then perform a halogen exchange. This might be as simple as NaF in CHCl3 or CCl4. Again, see Saunders for analogous work on dialkyl fluorophospates.

[Edited on 28-9-2008 by Sauron]

Attachment: lucas.pdf (946kB)
This file has been downloaded 1061 times

Sauron - 27-9-2008 at 20:57

Backgrounder on the 2-chloro-1,3,2-dioxaphospholanes:

Here is the pertinent section from Volume 1 of the invaluable "Chemistry of Hetrocyclic Compounds" series (thanks, kmno4!)

Despite the jawbreaker name from Ring Index, these are just cyclic chlorophosphonite esters of simple 1,2-diols (glycols).

The trick therefore is to obtain PCl3.

[Edited on 28-9-2008 by Sauron]

Attachment: Pages from 1.pdf (711kB)
This file has been downloaded 1239 times

Sauron - 27-9-2008 at 21:01

The foregoing would be pretty useless without the references from its chapter, I have not bothered to cull this down to just the ones for that short section.

Sorry for double posting but the two files were a bit large to be rolled into one so I am attaching them seperately.

[Edited on 28-9-2008 by Sauron]

Attachment: refs.pdf (1021kB)
This file has been downloaded 1736 times

Sauron - 28-9-2008 at 09:38

The trivial name for 2-chloro-1,3,2-dioxaphospholane is ethylene glycol phosphorochloridite. A googling of this name produced a Chinese paper on the use of this reagent in phosphorylating sugars. The preparation from EG and PCl3 in 72% yield is described, CHCl3 or Et2O used as solvent. Reactants were in equimolar proportions. The procedure appears to be a simplification of Lucas, q.v. Therefore I recommend employing Lucas precautions against side reaction leading to linear esters, and also his admonition that the EG ester is very readily hydrolyzed and requires manipulation under anhydrous conditions.

The 4-methyl and 4,5-dimethyl substituted glycol phosphorochloridites are reportedly much less readily hydrolyzed. These are from 1,2-propanediol and meso-2,3-butanediol respectively.

[Edited on 29-9-2008 by Sauron]

Attachment: 150215-171-03-0100-p4.pdf (89kB)
This file has been downloaded 1231 times

Sauron - 28-9-2008 at 23:56

2,3-butanediol is actually quite interesting. It can exist in three stereoisomers, 2R,3R, 2S,3S, and the meso form 2R 3S. The 2R,3R form exists in nature and can be produced by fermentation.

Commercial 2,3-BDO is a mixture of the racemate and meso forms in approximately equal proportions (25:25:50 %). The racemic fraction can be isolated in 97% purity by esterifying with Ac2O then holding at 4 C till the diacetate crystallizes out. The mother liquor is enriched in meso form. The literature suggests that meso-enriched mixtures can be enriched in the racemate by distillation to about 70% racemate, this sounds like the meso form can be removed relatively cleanly. The fractional crystallization can then be repeated.

The Lucas procedure calls for the meso form only, and while Aldrich does sell this it's only in 10 ml bottles and I shudder to contemplate the price.

Even the technical (racemate + meso) 2,3-BDO is expensive, like $300 a liter. Compare this with the dirt cheap propylene glycol (1,2-propanediol) and I am forced to wonder why the Russians would give themselves a headache like this?

Well not such a headache as it turns out. Meso-2,3BDO can be prepared unequivocally from trans-2,3-butene oxide (epoxybutane) while the cis-isomer gives only the dl form (racemate). Quite pure trans-2-butene is a commercially available gas and Org.Syn has the procedure for preparing the oxide, while JACS has the procedure for converting tat to the meso-2,3-butylenediol.

I have now obtained the paper from Cytobiology on purification of dl-2,3-butanediol from mixtures with mes-2,3-butanediol. The former is a useful cytoptotectant. However their procedure is not so useful in producing high purity meso-diol.

I also obtained another Lucas paper from JACS in vol 58 which details the preparation of pure meso-2,3-butanediol starting from mixtures of cis and trans-2-butene, readily obtained by dehydration of n-butanol with H2SO4. The liquified 2-butenes are converted to their chlorohydrins, then to their oxides, and fractionated. The epoxides are hydrolyzed with perchloric acid. trans-2-butene oxide gives exclusively meso-2,3-butanediol.

This procedure can be modified according to the method in Org.Syn. in which trans-2-butene oxide is prepared from trans-2-butene by means of 40% peracetic acid. However, those authors used a spinning band column to purify their epoxide, and who has one sitting around?

Trans-2-butene is available commercially in cylinders. Having the pure (99%) alkene would obviate the fractionation of the opoxides.

In any case these papers amply demonstrate that practical meso-2,3-butanediol can be prepared with a reasonable amount of effort.

Here is the JACS paper, which is the best of the three IMO.

[Edited on 29-9-2008 by Sauron]

Attachment: lucas2.pdf (730kB)
This file has been downloaded 994 times

Sauron - 29-9-2008 at 08:49

I see no particular need to post the Cytobiology paper unless requested.

As the only part of the Org.Syn. procedure that is pertinent is Note 2 I will simply quote it in entirety.

"2. trans-2-Butene oxide was prepared by appropriate modification of the procedure in Org. Synth., Coll. Vol. 4, 860 (1963). A 2-l., four-necked, round-bottomed flask fitted with a mechanical stirrer, a 1-l. dropping funnel, an acetone–dry ice condenser, and a thermometer is charged with 1 l. of 1,1,2,2-tetrachloroethane. The condenser is packed with dry ice and acetone, and the flask is cooled in a methanol-ice bath to −15°. trans-2-Butene (153 g., 2.73 moles) (Phillips Petroleum Company, 99%) is distilled into the flask from a tared, chilled trap. Six hundred milliliters of 40% peracetic acid (FMC Corporation), to which has been added 30 g. sodium acetate to neutralize the sulfuric acid present, is added to the stirred solution from the dropping funnel over a period of 2 hours. The mixture is stirred at −15° for another hour, then allowed to warm to room temperature. The mixture is poured into 1 l. of ice-cold water. The organic layer is separated, washed first with 10% sodium carbonate solution, then with water, dried over magnesium sulfate, and filtered. Distillation of the filtrate through a 75-cm. spinning-band column gives 133 g. (68%) of trans-2-butene oxide as a colorless oil, b.p. 52.5–55°."

It is worth noting that the same product from Lucas using a conventional fractionation and starting from a mixture of cis- and trans-2-butene oxides, gave the same b.p. Therefore I suggest the spinning band distillation is a luxury and not a necessity.

Panache - 5-10-2008 at 23:30

Can i just say sauron i have thoroughly enjoyed following your investigation into these compounds and i appreciate you taking the time to compose and post your musings, research, conclusions and the like.


Sauron - 6-10-2008 at 01:27

Thanks. I should qualify my remarks a little.

I am satisfied that Folient (or Foliant) is/was a real Russian/Soviet program and that is has been known to the West for several decades, and that current MOPP gear was introduced to defeat it.

As to the structures, I cannot say with any certainty that the published structures, which are only three of what is supposedly a very large class of compounds, are correct. The only way to find out would be to make them and see if you die! Unless you work at Edgewood or Porton Down etc. In which case you probably know already one way or the other.

I have discussed the matter with a friend who formerly was one of the world's leading manufacturers of protective apparel includinf for CW and he described to me the sea-change in MOPP (military organphosphorus protective) gear that took place around the time of the first Gulf War (Desert Storm) and why.

The new clothing is complex, a total departure from previous technologies and very very expensive. He says it was developed specifically because of the Novichoks.

Controversy is likely to continue over any possible connection between the so called Gulf War Syndrome and undisclosed Russian transfers of CW agents to Saddam, specifically the vaguely defined Russian variant of VX. It is generally agreed that Saddam did NOT possess any Novichok agents.

Polverone - 3-12-2008 at 16:46

It's not illegal to discuss chemical weapons or share literature references, at least in the United States, where both I and the web host for this site are located. I draw the line at discussing weaponization, but that is just part of the larger forum policy against discussing weapons construction. Of course I also discourage people who are lazy or trolling and want to be spoon fed information, like the recently banned member NovichokVX.

Edit: In case it is not already clear, I do not object to the chemical weapons discussions that have been conducted by Sauron, Ritter, Samosa, and others over the years here. It remains an acceptable topic for discussion in the future too if it is reasonably academic.

[Edited on 12-3-2008 by Polverone]

Polverone - 4-12-2008 at 09:52

I don't want any further off-topic discussion in this thread. Additional off-topic posts are subject to deletion.

Edit: it is not apparent now, but there was a long discussion in this thread about whether or not it's wrong to talk about chemical weapons in public. The off topic material has been removed.

[Edited on 12-4-2008 by Polverone]

DDTea - 7-10-2009 at 15:33

I've been avoiding these forums for a long time after some scares, but I'm really impressed with the level of discussion taking place--especially in my particular area of interest ;-).

Let me digress a little bit: one thing that I have been looking into is not so much the chemistry of these compounds, but their actual metabolism in the body. Questions such as: what is the mechanism of skin penetration? What is the mechanism and rate of binding to AChE and other enzymes and neurotransmitters? Maybe a year ago, I wrote a report for my physical organic chemistry course on a Structure Activity Relationship for the reaction between carbamates and acetylcholine. Unfortunately, that has been lost to a dead computer...

In any case, returning to "NTAs," as they are now called: at this point, it might be useful to compile a formal review of the public-domain literature regarding these agents. Are there any forum dwellers who are fluent in Russian, perchance?

Speaking honestly: after completing my degree in Chemistry, I initially looked down on these forums as "not true science." However, with the discussion taking place, it would be great if we could take the initiative to put more things into prepublication. I may be terribly mistaken, but I don't know if such a comprehensive review of the Novichoks has been compiled yet--it might even be worthy of journal publication; or at least something to add to your portfolio (let me tell you: the guys in Naval Research and Army Research are thirsty for people with any kind of background or interest in this field--don't be deterred or afraid of big brother. He just might be on your side, so long as you're not aiding insidious types).

Sorry that I have nothing to add to this thread at the moment. I need to get back into the swing of things; I'll probably be spending more time using the University databases to pull up some useful documents in the mean time.

Sauron--good job on the posts! Would it be possible for you to compile a list of your sources for others to glean over?

Cheers all.

simply RED - 8-3-2010 at 23:53

According to docking and dynamic/kinetic analysis - this shoud be the most toxic of the series.
Extremely fast aging - "to perfection" - Enzyme-O-P(O)(OH)2.

The same toxicity should have the analogue where Sulphur-CH2- is substituted with -N=C(CH3)- .

Higher stability but less toxicity should have the sulphur substituted with oxygen and the fluorine substituted with -CH3.

[Edited on 9-3-2010 by simply RED]

Nov.JPG - 12kB

[Edited on 9-3-2010 by simply RED]

[Edited on 9-3-2010 by simply RED]

Sauron - 9-3-2010 at 06:22

I lost most of my interest in this series once I ascertained that the US and NATO has been well aware of them for c.25 years and tota;;y redesigned MOPP gear and detection systems to deal with these by the time of Desert Shield, 20 years ago.

In short it is all old news, and can only be regarded as novel in comparison with preWWII G agents and early 1950s V agents'

[Edited on 9-3-2010 by Sauron]

simply RED - 10-3-2010 at 23:44

Of course they are old news!

They are just substituted VX.
The "drawback" of VX is that it contains P-O-R part, where R is Et or -CH2-CH(CH3)2 (Russian variety). This makes the enzyme inhibited by it regeneratable with 2-PAM and such compounds, because it does not hydrolise (fast) further (aging) to P-OH. Aged enzyme (Enzyme-O-P(O)(OH)2 or Enzyme-O-P(O)(OH)(CH3) is unregeneratable by any means.

If the -OR part in VX is : Pinacolyl alcohol, -O-CH=CCl2 (as in DDVP), -O-N=CCl2, etc - the enzyme ages much quickly and can not be regenerated. This is the theory of the latest generation OPs like Novichoks. Also the H-O-N=CCl2 or H-O-N=CClF is very toxic and reactive on its own(a battlefield toxin) and is released inside the synaptic cleft! This theory is published widely in the 1970s.

See attachment. Quantum/MM structural models may be wrong but most of the article is good.

Attachment: Aging.pdf (505kB)
This file has been downloaded 794 times

Antidote or desactivator for these final generation OPs could be medium mass organic, polymer or biopolymer molecules designed to bind and desactivate the OPs. Designed as key-keyhole to them - hydrolisisng the "tail" part, that is recognized by the acetylcholinesterase and the receptors. Such antidotes could easily be enginnered with modern molecular modeling techniques (quantum calculations).

[Edited on 11-3-2010 by simply RED]

Sauron - 11-3-2010 at 04:18

Sorry but that is rather confused.

The hypothetical structure you posted is not "substituted VX" but rather analog of the 1930s British agent DFP.

Like DFP and unlike any V agent it contains no C-P bond

Its sole common stuctural element to V agents is the O=P-S-CH2-CH2-NR2 unit

VX contains no F

Persistance or nonpersistance in the environment is a deliberate design element of MOPs acccording to tactical requirements.

simply RED - 11-3-2010 at 21:33

I mean exactly this: (Attachment)

Where -O-N=CCl2 could be -O-N=CClF ; -O-CH=CCl2 ; Pinacolyl alcohol, etc - something that "ages" quickly... The first 3 mentioned give higher skin resorption, by the way.

Whether it is P-F or P-CH3 does not matter much in toxicity in this very case. Only P-CH3 variety is much more stable to environmental hydrolysis.

OP.bmp - 653kB

Compounds containing P-O-CH2-CH2-X ; X=Cl, F are generally not very toxic, because they do not match the enzyme. X=N(Et)2 is needed to match the enzyme. The oxygen should be substituted with sulphur that the P-S bond could be easily broken by the enzyme.

[Edited on 12-3-2010 by simply RED]

Sauron - 11-3-2010 at 23:49

Pinacolyl alcohol is slow to hydrolyse for reason os steric hindrance, as in GD soman a German gift to Stalin.

V agents are persistant without steric effects. VX is MeP(=O)(OEt)SCH2CH2N(iPr)2. It has a low vapor pressure, a high bp, the appearance and viscocity of motor oil.

Its toxicity is primarily percutaneous and the high lipophilicity is due to the thioester side chain.

The novichok oxime moiety is there to frustrate some older prophilaxis and treatment protocols.

simply RED - 12-3-2010 at 02:37

"The novichok oxime moiety is there to frustrate some older prophilaxis and treatment protocols. "

Indeed! Post VX OPs are designed to be untreatable by "VX-protocols". As it was mentioned earlier, enzymes phosphorylated by these - age very quickly and are unregeneratable.

I doubt any treatament exists for post VX OPs other than sympthomatic. Is there a new treatment for post VX OPs?

According to the book : "Modern Warfare Toxic Agents" by G.Kotov (1971) - Fast aging OPs like soman are aboslutely untreatable with a combination of attropine and obidoxime! Experiments are cited where attropine and obidoxime were injected in animals 10 minutes BEFORE intoxications with different OPs. For VX like chemicals the effect is 100-700 times higher LD-50s. For fast aging chemicals like soman the effcts is ...... 1.2-1.4 times LD50s.

So, fast aging OPs like Novichoks are absolutely untreatable with a combination of attropine and cholinesterase oxime reactivator.

Anyway, with the structure I posted as an attachment, I just wanted to show a real, complete, fast aging post VX OP formula.

[Edited on 12-3-2010 by simply RED]

Sauron - 12-3-2010 at 06:05

A lot has happened in this corner of military medicine since 1971 (hell since 1991!) and it is well known that atropine and older oximes like 2PAM are useless against VX (I do not recall re soman.)

However as soman has been aroundd for>70 years I am confident that there exist effective therapies and prophylaxis.

About the newcomers I simply do not know.

simply RED - 18-3-2010 at 02:44

Here is a "straight" idea of such "fast aging" compound synthesis:

Phos1.PNG - 15kB

Or directly:

untitled.gif - 3kB

[Edited on 18-3-2010 by simply RED]

simply RED - 18-3-2010 at 22:37

FC(CH3)=N-P(O)(-S-CH2-CH2-N(Et)2) geometry optimized with PM3 (Other methods will give the same conformational geometry).
Hydrogen atoms are deleted that the structure is clear.

Untitled-2.jpg - 26kB

F-P(O)(CH3)(-N=CF-CH2-N(Et)2 geometry optimized with PM3 (Other methods will give the same conformational geometry).
Hydrogen atoms are deleted that the structure is clear.

Untitled-1.jpg - 24kB

[Edited on 19-3-2010 by simply RED]

AndersHoveland - 2-4-2012 at 14:18

also see the thread in this forum: "highly toxic acetylcholinesterase-inhibitor"

While doing research for alternate routes for preparing an energetic plasticizer known as FEFO, I happened to find
some information relevent to the most deadly acetylcholine blockers. I am only posting this because the precursors are virtually impossible to obtain, and there are many non-obvious details being withheld which would be required for the reactions to work. This is for information purpsoes only, and does not contain sufficient information to actually conduct any of the reactions.

The nitration ... of 1,2-dichloro-1,2-difluoro ethylene, ClFC=CFCl, gives chlorofluoronitroacetic acid (40% yield), with a formula HOC(=O)CFCl(NO2). Chlorofluoronitroacetic acid reacts with red fuming nitric acid (HNO3 / NO2) to give chlorofluoronitronitrosomethane, with a formula O=N--CFCl(NO2), and which possesses a deep blue color.
This latter compound reacts with PCl3 to form one of the more potent Novichok agents.
Synthesis and pesticidal activity of chloronitroacetic acid esters. Martynov; Yurtanov; Ivanov; Martynov; Yurtanov; Ivanov; Kulish; Uvarova; Andreeva; Rozhkova; Zhirmunskaya; Veshchestv, Chernogolovka, USSR. Veshchestv, Chernogolovka, USSR. Doklady Akademii Nauk SSSR (1986), 289(1), 109-13 [Chem.]. Doklady Akademii Nauk SSSR (1986), 289 (1), 109-13 [Chem.].

... track down all possible scientific articles about Novichok agents... Zhurnal Obshchei Khimii may be a good source...

in Tobiasons Scientific Principles, the chemical weapons volume, that the Soviets intentionally published large amounts of chemical weapons information in the open literature in the 1950s and 1960s with the hope some rogue nation would use the information to attack the US. The goal here was for the rogue state to finish the job for the Soviet Union, or at least inflict massive American casualities.

I. V. Martynov has published about 500 journal articles in his lifetime to date. Indeed there are many about phosphorus compounds, but those type of articles cease after 1972. He publishes many articles about molecular refraction after that. In 1984 he resumes publication of phosphorus related articles.

Synthesis and anticholinesterase activity of fluorochloronitroacetic acid esters. Ivanov, Yu. Ya.; Brel, V. K.; Postnova; Martynov, I. V.. Inst. Fiziol. Akt. Veshchestv, Chernogolovka, USSR. Khimiko-Farmatsevticheskii Zhurnal (1985), 19(8), 968-71.

There are a few earlier articles about fluorochloronitroacetic acid esters. These are important in the systhesis of Novichoks I would imagine. Samosa did mention in his paper that dihaloformaldoxime are critical parts of Novichok agents, and fluorochloronitroacetic acid should form those.

Here is another article of potential use in the preparation of Novichok agents. This compound is similar to fluorochloronitroacetic acid from which this substance is made:
Synthesis of chlorofluoronitronitrosomethane. Martynov, I. V.; Brel, V. K.; Uvarova, L. V. Inst. Fiziol. Akt. Veshchestv, Chernogolovka, USSR. Izvestiya Akademii Nauk SSSR, Seriya Khimicheskaya (1986), (4), 952-3.
Decarboxylation-nitrosation of ClFC(NO2)CO2H with HNO3 gave 52% ClFC(NO)NO2

Here is another possible tidbit as it relates to insecticides and plant growth regulation. We know they disguised their research under the guise of agrichemicals:
Synthesis and pesticidal activity of chloronitroacetic acid esters. Martynov, I. V.; Yurtanov, A. I.; Ivanov, Yu. J.; Kulish, E. V.; Uvarova, L. V.; Andreeva, E. I.; Rozhkova, N. G.; Zhirmunskaya, N. M. Inst. Fiziol. Akt. Veshchestv, Chernogolovka, USSR. Doklady Akademii Nauk SSSR (1986), 289(1), 109-13 [Chem.].
A series of 31 O2NCRR1CO2R [e.g., R, R1, R2 = H, Cl, n-C7H15 (I); F, Cl, ClCH2CH2; Br, Cl, Et] was tested for insecticidal and, in some cases, plant growth regulatory activity. Eight of the compds., e.g., I, were active insecticides. Twelve of the compds. were new but no preparation details were given.

Here is another possible Novichok variant:
Reaction of phosphorus trichloride with 1,1,2-trichloro-1-nitrosoethane in sulfur dioxide. Martynov, I. V.; Ivanov, A. N.; Epishina, T. A.; Sokolov, V. B. Inst. Fiziol. Akt. Veshchestv, Chernogolovka, USSR. Izvestiya Akademii Nauk SSSR, Seriya Khimicheskaya (1986), (9), 2158.
Reaction of ClCH2CCl2NO with PCl3 in SO2 gave 58% ClCH2CCl=NOP(=O)Cl2.

Here is yet another possible Novichok variant:
Reaction of dialkyl phosphites with 1,1-dichloronitrosoalkanes. Ivanov, A. N.; Epishina, T. A.; Goreva, T. V.; Sokolov, V. B.; Martynov, I. V.. Inst. Fiziol. Akt. Veshchestv, Chernogolovka, USSR. Izvestiya Akademii Nauk SSSR, Seriya Khimicheskaya (1987), (1), 226-8.
(RO)2P(O)ON:CClR1 (R = Bu, Me2CHCH2, pentyl, Me, Et; R1 = Me, Et, Pr, Me2CH, Bu, Me2CHCH2) were prepd. in 44-67% yields by treating (RO)2POH with ONCCl2R1 in EtOH at 20.

Here is a toxicity study done on animals and humans for some "pesticides".
Delayed neurotoxicity from organophosphorus pesticides. Makhaeva, G. F.; Malygin, V. V.; Martynov, I. V.. USSR. Agrokhimiya (1987), (12), 103-24.
A review with 123 refs. on 8 clin. intoxication symptoms, pathmorphol., mechanisms of initiation of delayed neurotoxicity by organophosphorus pesticides (OPP) structure-activity relations of OPP, monitoring of the delayed neurotoxicity of OPP in animals and humans, etc.

Here is another possible Novichok variant:
Reaction of O-alkyl methylphosphonites with 1,1-dichloro-1-nitrosopropane. Sokolov, V. B.; Ivanov, A. N.; Epishina, T. A.; Martynov, I. V.. Inst. Fiziol. Akt. Veshchestv, Chernogolovka, USSR. Zhurnal Obshchei Khimii (1987), 57(4), 952-3.
Reaction of ROP(O)HMe (R = Me2CH, Bu, pentyl) with EtCCl2NO in Et2O gave 50-52% ROP(O)MeON:CClEt (I; same R). Treating MeP(OR)2 with EtCCl2NO also gave I.

Here is an interesting reference, although I doubt this would have very high human toxicity due to the two large aryl groups attached to phosphorus. Still, it gives enlightenment as to where they are headed:
Reaction of diphenylphosphinous acid with 1,1-dichloro-1-nitrosoalkanes. Sokolov, V. B.; Epishina, T. A.; Ivanov, A. N.; Kharitonov, A. V.; Brel, V. K.; Martynov, I. V.. Inst. Fiziol. Akt. Veshchestv, Chernogolovka, USSR. Zhurnal Obshchei Khimii (1987), 57(7), 1658-9.
Treating Ph2P(O)H with RCCl2NO (R = Et, Pr, Me2CH) in Et2O gave 62-75% Ph2P(O)ON:CClR (same R).

Another Novichok possibility:
Synthesis and the structure of dialkylfluoroformiminophosphates. Martynov, I. V.; Brel, V. K.; Uvarov, V. I.; Yarkov, A. V.; Novikov, V. P.; Chepakova, L. A.; Raevskii, O. A. Inst. Fiziol. Akt. Veshchestv, Chernogolovka, USSR. Izvestiya Akademii Nauk SSSR, Seriya Khimicheskaya (1987), (4), 857-60.
Syn- And anti-(RO)2P(O)N:CHF (R = Me, Et, Pr, Bu) were prepd. in 11-25% yields by treating (RO)3P with ClCHFNO2.

Here is some nasty looking stuff that may be of interest:
Reaction of (-aminoalkyl)phosphonates with perfluoro-2-azapropene. Aksinenko, A. Yu.; Pushin, A. N.; Sokolov, V. B.; Gontar, A. F.; Martynov, I. V.. Inst. Fiziol. Aktivn. Veshchestv, Chernogolovka, USSR. Izvestiya Akademii Nauk SSSR, Seriya Khimicheskaya (1987), (5), 1177-9.
(RO)2P(O)CMeR1N:C:NCF3 (R = Me, R1 = Et; R = Et, R1 = Et, Pr, Bu; R = Me2CHCH2, R1 = Et) were prepd. in 40-60% yields by condensing CF2:NCF3 with (RO)2P(O)CMeR1(NH2) in the presence of KF.

Here is another variant:
Reaction of polychloronitrosoethanes with phosphorous acid derivatives. Martynov, I. V.; Ivanov, A. N.; Epishina, T. A.; Sokolov, V. B. Inst. Fiziol. Akt. Veshchestv., Chernogolovka, USSR. Izvestiya Akademii Nauk SSSR, Seriya Khimicheskaya (1987), (5), 1086-9.
The title reaction gave 20-93% of 22 o-phosphorylated alkyl chloroformimines. Thus, treating ONCCl2R (R = Me, CH2Cl, CHCl2) with (R1O)3P (R1 = Me, Pr, Bu, Me2CHCH2, pentyl, ClCH2CH2) gave (R1O)2P(O)ON:CClR.

Of all the other compounds I have previously referenced this particular compound looks like it may be the deadliest. It has some similarities to most other nerve gasses in that it uses the simplest alkyl groups, and has a direct alkyl and a direct halogen attachment to phosphorus. I would replace those chlorines with fluorine to increase the toxicity:
Reaction of dichloromethylphosphine with 1,1-dichloro-1-nitrosoalkanes. Sokolov, V. B.; Ivanov, A. N.; Epishina, T. A.; Martynov, I. V.. Inst. Fiziol. Akt. Veshchestv, Chernogolovka, USSR. Zhurnal Obshchei Khimii (1987), 57(7), 1659-60.
MePCl2 reacted with RCCl2NO (R = Et, Pr, Me2CH) in SO2 to give 27-37% RCCl:NOP(O)ClMe.

This compound looks like a good precursor for organophosphorus agents like the previous compound. The chlorines can be replaced by F, and one of the fluorines can form an ester or something else. The second compound is an example of what could be made, and I just bet that stuff is pretty toxic.
Interaction of 2,2,3,3-tetrafluoropropyl dichlorophosphite with 1,1,2-trichloro-1-nitrosoethane. Sokolov, V. B.; Ivanov, A. N.; Epishina, T. A.; Martynov, I. V.. Inst. Fiziol. Akt. Veshchestv, Chernogolovka, USSR. Izvestiya Akademii Nauk SSSR, Seriya Khimicheskaya (1987), (6), 1422-3.
Refluxing CHF2CF2CH2OPCl2 (I) with CH2ClCCl2NO (II) in Et2O gave 67.8% Cl2P(O)ON:CClCH2Cl. Treating I with II in SO2 at 20 gave 48.2% (CHF2CF2CH2O)ClP(O)ON:CClCH2Cl.

We might have a real winner with this one as it has similarities with VX nerve gas. The second compound in particular has a =S group. If that could be isomerized, like it is done in making VX, then we have a thioester. The two isobutyl groups are probably too large to make this particular compound all that toxic. I am sure they could be replaced with methyls instead.
Reaction of diisobutylchlorophosphine with 1,1-dichloro-1-nitrosoalkanes in presence of sulfur dioxide and ethyl mercaptan. Sokolov, V. B.; Ivanov, A. N.; Epishina, T. A.; Martynov, I. V.. Inst. Fiziol. Aktivn. Veshchestv, Chernogolovka, USSR. Izvestiya Akademii Nauk SSSR, Seriya Khimicheskaya (1987), (11), 2586-8.
Treating (Me2CHCH2)2PCl with RCCl2NO (R = Me, Et, Pr, Me2CH) in Et2O contg. SO2 gave 61-74% (Me2CHCH2)2P(O)ON:CRCl (same R). When Et2SH was used instead of SO2, 44% (Me2CHCH2)2P(S)ON:CRCl (R = Me) was obtained.

Another phenyl attached compound:
Synthesis and molecular structure of (O-isopropylchloroformimino) diphenylphosphinate. Martynov, I. V.; Chekhlov, A. N.; Ivanov, A. N.; Epishina, T. A.; Makhaev, V. D.; Sokolov, V. B. Inst. Fiziol. Aktivn. Veshchestv, Chernogolovka, USSR. Izvestiya Akademii Nauk SSSR, Seriya Khimicheskaya (1987), (11), 2595-7.
Treating Ph2PH with Me2CHCCl2NO in C6H6 gave 58% Ph2P(O)ON:CClCHMe2, the structure of which was detd. by x-ray crystallog.

This compound has some VX similarities too:
O,O-Dialkyl O-(dialkylformimino) thiophosphates. Chepakova, L. A.; Brel, V. K.; Pushin, A. N.; Martynov, I. V.. Inst. Fiziol. Akt. Veshchestv, Chernogolovka, USSR. Zhurnal Obshchei Khimii (1987), 57(12), 2716-19.
Twelve (R1O)2P(S)ON:CMeR (R = Me, Et, Pr; R1 = Me, Et, Pr, Bu) were prepd. in 41-62% yields by treating (R1O)2PHS with ONCClMeR or HON:CMeR.

These compounds are similar to the last journal reference except the R and R’ groups are switched. Isomerize that S and we may have something far more toxic.
O-(Alkylchloroformimino) O,O-dialkyl thiophosphates. Martynov, I. V.; Ivanov, A. N.; Epishina, T. A.; Sokolov, V. B. Inst. Fiziol. Akt. Veshchestv, Chernogolovka, USSR. Izvestiya Akademii Nauk SSSR, Seriya Khimicheskaya (1987), (12), 2854-5.
Seven (RO)2P(S)ON:CClR1 (R = Et, Me2CH; R1 = Me, Et, Pr, Me2CH, ClCH2) were prepd. in 33-54% yields by condensing (RO)2PSH with R1Cl2CNO in THF.

Martynov has 64 publications in 1988 alone, his best year. In no particular order here are some highlights:

Molecular and crystal structure of O,O-diethyl 1-[N2-(trifluoromethyl)fluoroformamidino]-1-methylethylphosphonate. Chekhlov, A. N.; Aksinenko, A. Yu.; Sokolov, V. B.; Korenchenko, O. V.; Martynov, I. V.. Inst. Fiziol. Akt. Veshchestv, Chernogolovka, USSR. Doklady Akademii Nauk SSSR (1988), 302(4), 855-8 [Chem.].
The crystal and mol. structure of (EtO)2P(O)CMe2NHCFNCF3 was detd.

Reaction of (N-acetyl-N-ethylamido)alkylphosphonic acid chlorides with cesium fluoride. Krolevets, A. A.; Adamov, A. V.; Popov, A. G.; Martynov, I. V.. USSR. Zhurnal Obshchei Khimii (1988), 58(11), 2628-9.
RP(O)F(NEtCH:CH2) (R = Me, Me2CH) were prepd. in 45, 50% yields, resp., by treating RPCl(NEtAc) (I) with CsF. I were prepd. in 60, 65% yields, resp., by treating RPCl2 with Me3SiNEtAc.

Stable alkoxyfluorophosphoranes. Krolevets, A. A.; Popov, A. G.; Adamov, A. V.; Martynov, I. V.. USSR. Zhurnal Obshchei Khimii (1988), 58(11), 2626-7.
RPF2(OR1)2 (R = BuCHClCH2, R1 = Me3C; R = Me2CClCH2, R1 = Et) were prepd. in 45, 40% yields, resp., by treating RPF4 with Me3SiOR1.

O-(Alkylchloroformimidoyl) o-alkyl methylphosphonates. Sokolov, V. B.; Ivanov, A. N.; Goreva, T. V.; Epishina, T. A.; Martynov, I. V.. Inst. Fiziol. Akt. Veshchestv., Chernogolovka, USSR. Izvestiya Akademii Nauk SSSR, Seriya Khimicheskaya (1988), (5), 1128-30.
Nine (RO)MeP(O)ON:CClR1 (R = Et, Pr, Bu, Me2CH, pentyl; R1 = Me, Et, Pr, Bu, Me2CH) were prepd. in 41-67% yields by treating R1CCl2NO with MeP(OR)2 or MeP(O)H(OR).

Reaction of 1,1-dichloro-1-nitrosoalkanes with phosphorus(III) chlorides. Martynov, I. V.; Ivanov, A. N.; Epishina, T. A.; Sokolov, V. B. Inst. Fiziol. Aktivn. Veshchestv, Chernogolovka, USSR. Izvestiya Akademii Nauk SSSR, Seriya Khimicheskaya (1988), (9), 2128-32.
The title reaction was studied. Thus, R1R2P(O)ON:CRCl (R = Me, Et, Pr, Me2CH; R1 = R2 = Cl, Me2CHCH2; R1 = Cl, R2 = Me) were prepd. in 34-74% yields by reaction of RCCl2NO with R1R2PCl in the presence of SO2.

Synthesis and x-ray diffraction study of N-(diisopropoxythiophosphoryl)thioacetamide. Solov'ev, V. N.; Chekhlov, A. N.; Zabirov, N. G.; Cherkasov, R. A.; Martynov, I. V.. Inst. Fiziol. Aktivn. Veshchestv, Chernogolovka, USSR. Doklady Akademii Nauk SSSR (1988), 300(6), 1386-9 [Chem.].
Treating MeCSNH2 with Me3COK in MeCN and then with ClP(S)(OCHMe2)2 gave 15% MeCSNHP(S)(OCHMe2)2, the structure of which was detd. by x-ray crystallog.

Reaction of 1,1-dichloro-1-nitrosoethane with phosphorus oxychloride in the presence of zinc. Sokolov, V. B.; Epishina, T. A.; Martynov, I. V.. Inst. Fiziol. Akt. Veshchestv, Chernogolovka, USSR. Izvestiya Akademii Nauk SSSR, Seriya Khimicheskaya (1988), (7), 1691.
Cl2P(O)ON:CClMe was prepd. in 26.6% yield by treating MeCCl2NO with POCl3 in the presence of Zn.

Comparative studies on the interaction of acetylcholinesterases from human erythrocytes and housefly heads with phosphorylated alkylchloroformoxims. Shataeva, G. A.; Makhaeva, G. F.; Yankovskaya, V. L.; Sokolov, V. B.; Ivanov, A. N.; Martynov, I. V.. Inst. Physiol. Act. Subst., Chernogolovka, USSR. Zhurnal Evolyutsionnoi Biokhimii i Fiziologii (1988), 24(6), 791-6.
Among Valexon analogs, 6 (RO)2P(O)ON:CClMe (I), 6 (RO)2P(O)ON:C(Cl)CH2Cl (II), and 5 (RO)2P(O)ON:C(Cl)CHCl2 (III, R = Me, Et, Pr, iso-Bu, Bu, amyl), and 4 (EtO)2P(O)ON:C(Cl)R1 (IV, R1 = Me, Et, Pr, Bu), I-III (R = Et) were highly selective insecticides, having rate consts. of bimol. reaction with acetylcholinesterase (KII) of human erythrocytes (HE) lower by 1 magnitude order than with that from housefly heads (FL). Inhibition of both HE and FL followed the order I < II < III. Phosphorylation capacity of II 1.6-fold exceeded that of I. Replacing Me by Et, increased the effect of I-III on FL 3-8-fold and decreased that on HE 1.7-4-fold. Further increases in hydrophobicity abolished the specificity of I-III. The selectivity of IV decreased in order of R1: Me > Et > Bu; IV (R1 = Pr) showed no selectivity.

Fluorination of some phosphoric acid derivatives. Zavorin, S. I.; Lermontov, S. A.; Martynov, I. V.. Inst. Fiziol. Akt. Veshchestv, Chernogolovka., USSR. Izvestiya Akademii Nauk SSSR, Seriya Khimicheskaya (1988), (5), 1174-6.
Dialkyl fluorophosphates were prepd. by the title fluorination with Et3N.3HF (I). Thus, fluorination of (EtO)2P(O)ON:CCl2 with I in MeCN gave 83.5% (EtO)2P(O)F.

Reaction of fluorine-containing acetylenic alcohols with phosphorus trichloride. Brel, V. K.; Chekhlov, A. N.; Ionin, B. I.; Martynov, I. V.. Inst. Fiziol. Akt. Veshchestv, Chernogolovka, USSR. Zhurnal Obshchei Khimii (1988), 58(4), 750-7.
Treating RC.tplbond.CCMe(OH)CH2F (I; R = Ph) with PCl3 in Et2O gave 45% Cl2P(O)CR:C:CMeCH2F (II; R = Ph) and 24% E- and Z-Cl2P(O)CHPhCCl:CMeCH2F (III). Under the same conditions, I (R = MeOCH2) gave a mixt. of II (R = MeOCH2) and Cl2P(O)C(:CH2)CCl:CMeCH2F. Treating I (R = Ph) with MeOH and then with Br2 gave oxaphospholene IV. The structure of III was detd. by x-ray crystallog.

Synthesis and anticholinesterase activity of fluorochloronitroacetic acid thioesters. Ivanov, Yu. Ya.; Uvarov, V. I.; Brel, V. K.; Martynov, I. V.. Inst. Fiziol. Akt. Veshchestv, Chernogolovka, USSR. Khimiko-Farmatsevticheskii Zhurnal (1988), 22(5), 538-40.
Treating O2NCFClCOX (I; X = OH) with PCl5 gave I (X = Cl), which reacted with RSH (R = Et, Bu) to give 35-55% I (X = SR; same R) (II). II were less effective acetylcholinesterase inhibitors than I (X = OR; same R) but had comparable activity vs. butyrylcholinesterase with lower toxicity.

Synthesis and antiesterase activity of sulfur-containing phosphorylated oximes. Chepakova, L. A.; Bret, V. K.; Makheva, G. F.; Yankovskaya, V. L.; Beznosko, B. K.; Malygin, V. V.; Martynov, I. V.. Inst. Fiziol. Akt. Veshchestv., Chernogolovka, USSR. Khimiko-Farmatsevticheskii Zhurnal (1988), 22(2), 143-6.
Reaction of (RS)2POEt (R = Et, Pr, iso-Bu, Bu or amyl) with O:NCFCl2 gave the corresponding (RS)2P(:O)ON:CClF (I). An increase in the hydrophobicity of I did not alter the anticholinesterase activity of I, while the butyrylcholinesterase and carboxylesterase activity were increased.

O-substituted alkylchloroformoximes as substrates and inhibitors of cholinesterases. Ivanov Iu Ia; Sokolov V B; Epishina T A; Martynov I V Doklady Akademii nauk SSSR (1990), 310(5), 1253-5.

Inhibition of cholinesterase activity with fluorine-containing derivatives of alpha-aminophosphonic acid. Kuusk V V; Morozova I V; Agabekian R S; Aksinenko A Iu; Epishina T A; Sokolov V B; Kovaleva N V; Razdol'skiy A N; Fetisov V N; Martynov I V Bioorganicheskaia khimiia (1990 Nov), 16(11), 1500-8.
A series of O,O-diethyl-1-(N-alpha-hydrohexafluoroisobutyryl)aminoalkylphos phonates (APh) has been synthesized and their interaction with human erythrocyte acetylcholinesterase (AChE) and with horse serum butyrylcholinesterase (BuChE) studied. Most of the APhs inactivated the cholinesterases irreversible through formation of the enzyme-inhibitor intermediate. The inactivation rate constants and the enzyme-inhibitor intermediate dissociation constants are calculated. The quantitative structure-activity relationships including both hydrophobic and calculated steric parameters of substituents are developed for APh--ChE interactions. Molecular mechanics (programme MM2) was used for determining steric parameters (Es). On the basis of QSAR models analysis it was concluded that hydrophobic interactions play an essential role in APh--AChE binding, whereas for APh--BuChE binding steric interactions are essential. Presence of at least two APh binding centres on the surface of AChE and BuChE is suggested.

Reaction of 1,1-dichloro-1-nitrosobutane with (N,N-dimethylamido)dichlorophosphite. Sokolov, V. B.; Ivanov, A. N.; Epishina, T. A.; Martynov, I. V.. Inst. Fiziol. Akt. Veshchestv, Chernogolovka, USSR. Izvestiya Akademii Nauk SSSR, Seriya Khimicheskaya (1989), (6), 1416-18.
Reaction of PrCCl2NO with Me2NPCl2 in Et2O or in SO2 gave 36% Me2NPCl4 or 30% Me2NP(O)ClON:CClPr, resp. Treating Me2NPCl4 with SO2 gave 91% Me2NP(O)Cl2. Reaction of PrCCl2NO with Me2NPCl2 in Et2O, and then with Ph3P and distn. gave Ph3PO and PrCN.

Alkyl chlorofluoroformimino perfluoroalkylphosphonates. Chepakova, L. A.; Brel, V. K.; Martynov, I. V.; Maslennikov, I. G. Inst. Fiziol. Akt. Veshchestv., Chernogolovka, USSR. Zhurnal Obshchei Khimii (1989), 59(6), 1455-6.
Treating RP(OR1)2 (R = CF3, R1 = Pr, Bu; R = CF3CF2, R1 = Me, Bu) with CFCl2NO in Et2O gave 76-88% title compds. R1OP(O)RON:CFCl.

Synthesis of dialkyl (3-alkyl-1,3-alkadien-2-yl)phosphonates. Brel, V. K.; Abramkin, E. V.; Martynov, I. V.; Ionin, B. I. Inst. Fiziol. Akt. Veshchestv, Chernogolovka, USSR. Zhurnal Obshchei Khimii (1989), 59(9), 2142-3.
(RO)2P(O)C(:CH2)CR1:CMe2 (R = Et, Pr; R1 = Me, Et) were prepd. in 41-73% yields by the Grignard reaction of (RO)2P(O)C(CH2OMe):C:CMe2 with R1X (X = halo).

Synthesis and antiesterase activity of O,O-dialkyl S-(ethoxycarboxyl)chloromethyl thiophosphates. Khaskin, B. A.; Makhaeva, G. F.; Torgasheva, N. A.; Ishmuratov, A. S.; Yankovskaya, V. L.; Fetisov, V. I.; Malygin, V. V.; Martynov, I. V.. Inst. Fiziol. Akt. Veshchestv, Chernogolovko, USSR. Izvestiya Akademii Nauk SSSR, Seriya Khimicheskaya (1989), (12), 2741-6.
The title compds. (RO)2P(O)SCHClCO2Et (I; R = alkyl homologs) were prepd. in 82-95% yields in the reaction of (RO)2P(O)SCl with N2CHCO2Et at -25 (in Et2O) or 6-7 (in benzene), presumably via a noncarbene mechanism. I irreversibly inhibited acetylcholinesterase, butyrylcholinesterase, and carboxylesterase; antibutyrylcholinesterase activity increased in the homologous series of R, with max. at R = Bu. An antiesterase MSBAR of I was fulfilled with parameters representing hydrophobicity and steric properties of R.

Synthesis and cholinesterase hydrolysis of O-acylated alkylchloroformoximes. Sokolov, V. B.; Ivanov, Yu. Ya.; Epishina, T. A.; Agabekyan, R. S.; Martynov, I. V.. Inst. Fiziol. Akt. Veshchestv, Chernogolovka, USSR. Khimiko-Farmatsevticheskii Zhurnal (1989), 23(11), 1317-20.
The title compds., RCO2N:CClR1 (R = Me, Et, Pr or CH2Cl and R1 = Me, Et, Pr, or iso-Pr) were prepd. e.g., by the reaction of 1,1-dichloro-1-nitrosobutane with AcCl in the presence of Zn. These compds. were good substrates for acetyl- and butyrylcholinesterases. The kinetic parameters (Km, Vmax and ac) of these compds. in the hydrolysis reactions were comparable to those with acetylcholine. The acute toxicity was 79-381 mg/kg in mice given drugs orally.

Synthesis and structure of O,O-dialkyl 2-[(ethoxycarbonyl)amino]hexafluoroisopropylphosphonates. Aksinenko, A. Yu.; Chekhlov, A. N.; Korenchenko, O. V.; Sokolov, V. B.; Martynov, I. V.. Inst. Fiziol. Akt. Veshchestv, Chernogolovka, USSR. Zhurnal Obshchei Khimii (1990), 60(1), 61-5.
The title compds. (RO)2P(O)C(CF3)2NHCO2Et (I; R = Me, Et, CHMe2) were prepd. in 54-76% yields in the reaction of (RO)2P(O)H with (CF3)2C:NCO2Et. The crystal and mol. structure of I (R = Et) was detd.

O-Substituted alkylchloroformoximes as substrates and inhibitors of cholinesterases. Ivanov, Yu. Ya.; Sokolov, V. B.; Epishina, T. A.; Martynov, I. V.. Inst. Fiziol. Aktivn. Veshchestv, Chernogolovka, USSR. Doklady Akademii Nauk SSSR (1990), 310(5), 1253-5 [Biochem.].
The ability of O-substituted alkylchloroformoximes to serve as substrates for acetylcholinesterase (ACE, EC and butyrylcholinesterase (BCE, EC and to inhibit acetylcholine hydrolysis by these enzymes was detd., along with the LD50 of these compds. in mice. The compds. tested were O-acylated alkylchloroformoximes of the general formula R1C(O)ON:C(Cl)R2 [R1 = R2 = Me; R1 = Me, R2 = Et; R1 = Me, R2 = Pr; R1 = Et, R2 = Me; R1 = Et, R2 = iso-Pr; R1 = Pr, R2 = iso-Pr; R1 = CH2Cl, R2 = Pr (I); R1 = CH2Cl, R2 = iso-Pr (II)], O-carbonylated alkylchloroformoximes of the general formula EtOC(O)ON:C(Cl)R [R = Me (III), iso-Pr (IV)], and O-carbamoylated alkylchloroformoximes of the general formula (Me)2NC(O)ON:C(Cl)R [R = Me (V), iso-Pr (VI)]. All of the compds. except for I and II were good substrates for the enzymes, with Km values for ACE ranging (0.3-11.0)  10-4M and for BCE ranging (0.5-13.0)  10-4M (the Km values of ACE and BCE with acetylcholine were 1.3  10-4 and 5.4  10-4M, resp.). III and IV were competitive (Ki 1.6  10-4M) and mixed-type (Ki 4.2  10-4M) inhibitors, resp., of ACE. V and VI were effective inhibitors of both ACE and BCE, with bimol. rate consts. for inhibition (kII) of 5.7  103 and 1.4  105 M-1 min-1, resp., for ACE, and 9.8  103 and 5.4  106 M-1 min-1, resp., for BCE. The LD50 values for the tested compds. ranged 60-381 mg/kg body wt.

O-(alkylchloroformimino)(methyl)thiophosphonic acid chlorides. Lyashenko, Yu. E.; Sokolov, V. B.; Martynov, I. V.. Inst. Fiziol. Akt. Veshchestv, USSR. Izvestiya Akademii Nauk SSSR, Seriya Khimicheskaya (1989), (12), 2865-6.
Treating the adduct from RCCl2NO and MePCl2 with H2S gave 21-35% MeP(S)ClON:CRCl.

Interaction of 1,1-dichloro-1-nitrosoalkanes with S-ethylmethylphosphonous chloride in the presence of sulfur dioxide. Sokolov, V. B.; Ivanov, A. N.; Epishina, T. A.; Martynov, I. V.. Inst. Fiziol. Akt. Veshchestv, Chernogolovka, USSR. Izvestiya Akademii Nauk SSSR, Seriya Khimicheskaya (1990), (2), 464-5.
EtSP(O)MeON:CClR (R = Me, Et, Pr) were prepd. in 42-47% yields by treating RCCl2NO with EtSPMeCl in the presence of SO2.

O-(alkylchloroformimino)-O-alkylphosphoric acid chlorides. Sokolov, V. B.; Ivanov, A. N.; Goreva, T. V.; Epishina, T. A.; Martynov, I. V.. Inst. Fiziol. Akt. Veshchestv, Chernogolovka, USSR. Izvestiya Akademii Nauk SSSR, Seriya Khimicheskaya (1990), (5), 1122-5.
Reaction of (ON)CCl2R with (R'O)2PCl (R, R' = alkyl) afforded the title compds. (R'O)ClP(O)ON:CRCl (I) in up to 69% yield. Hydrolysis of I led to substitution of P-, and not C-bound Cl, resulting in (R'O)(NH4O)P(O)ON:CRCl.

Reaction of the adduct of methyldichlorophosphine and 1,1-dichloro-1-nitrosoethane with thioacetic acid. Lyashenko, Yu. E.; Sokolov, V. B.; Martynov, I. V.. Inst. Fiziol. Akt. Veshchestv, Chernogolovka, USSR. Zhurnal Obshchei Khimii (1990), 60(8), 1923-4.
Treating MePCl2 with MeCCl2NO in PhMe, followed by addn of 1 or 2 equiv AcSH gave 56% MeP(S)ClON:CMeCl or 32% MeP(S)ClON:CMeSAc, resp.

Inhibition of cholinesterase activity by fluorine-containing derivatives of -aminoalkylphosphonic acids. Kuusk, V. V.; Morozova, I. V.; Agabekyan, R. S.; Aksinenko, A. Yu.; Epishina, T. A.; Sokolov, V. B.; Kovaleva, N. V.; Razdol'skii, A. N.; Fetisov, V. I.; Martynov, I. V.. Inst. Physiol. Act. Subst., Chernogolovka, USSR. Bioorganicheskaya Khimiya (1990), 16(11), 1500-8.
A series of O,O-diethyl-1-(N--hydrohexafluoroisobutyryl)aminoalkylphosphonates (APh) has been synthesized and their interaction with human erythrocyte acetylcholinesterase (AChE) and with horse serum butyrylcholinesterase (BuChE) studied. Most of the APhs inactivated the cholinesterases irreversible through formation of the enzyme-inhibitor intermediate. The inactivation rate consts. and the enzyme-inhibitor intermediate dissocn. consts. are calcd. The quant. structure-activity relationships including both hydrophobic and calcd. steric parameters of substituents are developed for APh-ChE interactions. Mol. mechanics (program MM2) was used for detg. steric parameters (Es). On the basis of QSAR models anal. it was concluded that hydrophobic interactions play an essential role in APh-AChE binding, whereas for APh-BuChE binding steric interactions are essential. Presence of at least two APh binding centers on the surface of AChE and BuChE is suggested.

Synthesis and anticholinesterase activity of O-carbamoylated alkylchloroform oximes. Sokolov, V. B.; Ivanov, Yu. Ya.; Epishina, T. A.; Martynov, I. V.. Inst. Fiziol. Akt. Veshestva, Chernogolovka, USSR. Khimiko-Farmatsevticheskii Zhurnal (1991), 25(4), 33-4.
Treating ClCO2N:CClR (R = Me, Et, Pr, CHMe2) with NHR1R2 (R1 = R2 = H, Me, Et; R1 = H, R2 = Me) in Et2O gave 50-69% R1R2NCO2N:CClR (same R-R3), which are acetyl- and butyrylcholinesterase inhibitors (k11 = 1.1  10-2 to 5.4  10-6 M-1 min-1). Acute oral toxicity in mice ranged from 32 to 565 mg/kg.

O-Alkyl O-methylchloroformimino phenylphosphonates - effective inhibitors of the hen brain neurotoxic esterase. Makhaeva, G. F.; Kononova, I. V.; Malygin, V. V.; Lyashenko, Yu. E.; Sokolov, V. B.; Martynov, I. V.. Inst. Fiziol. Akt. Veshchestv, Chernogolovka, USSR. Doklady Akademii Nauk SSSR (1991), 317(4), 1009-12 [Biochem.].
The title phosphonates were effective inhibitors of neurotoxic esterase; with increasing hydrophobicity the compds. showed pronounced and selective biol. activity towards brain neurotoxic esterase compared to acetylcholinesterase. Thus, the structure of phenylphosphonate played a major role in the inhibitory effects of these potential pesticides towards neurotoxic esterase or acetylcholinesterase.

Synthesis and anticholinesterase activity of fluorine-containing -aminophosphoryl compounds. Korenchenko, O. V.; Ivanov, Yu. Ya.; Aksinenko, A. Yu.; Sokolov, V. B.; Martynov, I. V.. Inst. Fiziol. Akt. Veshchestv, Chernogolovka, Russia. Khimiko-Farmatsevticheskii Zhurnal (1992), 26(6), 21-3.
Reaction of R2P(O)H (R = MeO, EtO, PrO, Me2CHO, Ph) with (CF3)2C:NCOR1 (R1 = OEt, OCH2Ph, OPr, OBu, OCH2CH2CHMe2, CF3) in Et2O gave 44-93% R2P(O)C(CF3)2NHCOR1. Treating a 1,4,2-oxazaphospholine deriv. with alcs. gave Me(R)P(O)C(CF3)2NHCO2Et (R = BuO, Me2CHO). Bimol. rate consts. for inhibition of cholinesterases by these compds. were detd.

Synthesis and insecticidal and acaricidal activity of O-alkylchloroformimine O,O-dialkyl phosphates and O,O-dialkylthiophosphates. Ivanov, A. M.; Ivanova, G. B.; Sokolova, V. B.; Epishina, T. N.; Goreva, T. V.; Beznosko, B. K.; Martynov, I. V.. Inst. Fiziol. Okl. Veshchestv., Chernogolovka, Russia. Fiziologicheski Aktivnye Veshchestva (1991), 23 58-62.
Of 26 title compds., those having ethoxy group at P were both insecticides and acaricides, whereas those having their methoxy group showed insecticidal activity only. Increasing hydrophobicity of the alkoxy substituents decreased i.m. toxicity to mice, but also the effectiveness. O replacement by S also decreased toxicity. Synthesis is indicated.

Paradoxical toxic effect and calcium antagonism of the cholinesterase inhibitors O-(N-arylcarbamoyl)acylhydroximoyl chlorides. Ivanov, Yu. Ya.; Sokolov, V. B.; Epishina, T. A.; Martynov, I. V.. Inst. Fiziol. Akt. Veshchestv, Chernogolovka, Russia. Doklady Akademii Nauk (1993), 328(6), 744-6 [Biochem.].
N-phenylcarbamates and aliphatic analogs of the formula R R1N(O)ON::C(Cl)R2 [where R = Et, Me, and Ph; R1 = H, Me; R2 = Et, Pr, iso-Pr] were examd. for their acetylcholinesterase and butyrylcholinesterase inhibition, for their acute toxicity and their action on selective organs. The enzyme inhibition depended on their molecular structure. Paradoxical effects (higher dose and low toxicity and vice versa) were noted.

Crystal and molecular structures and synthesis of O,O-diisopentyl 1-(phenylsulfonamido)-1-(trifluoromethyl)-2,2,2-trifluoroethylphosphonate. Chekhlov, A. N.; Aksinenko, A. Yu.; Sokolov, V. B.; Martynov, I. V.. Inst. Fiziol. Akt. Veshchestv, Ross. Akad. Nauk, Chernogolovka, Russia. Doklady Akademii Nauk (1995), 345(3), 360-363.
Reaction of (CF3)2C:NSO2Ph and (Me2CHCH2CH2O)2P(O)H in Et2O gave 85% title compd. (Me2CHCH2CH2O)2P(O)C(CF3)2NHSO2Ph, the structure of which was detd. by x-ray crystallog.

[Edited on 2-4-2012 by AndersHoveland]

AndersHoveland - 2-4-2012 at 21:30

One more comment, the chlorofluoronitrosomethane, that was previously mentioned as a precursor to some of the Novichok agents, is very similar to another chemical weapon, phosgene oxime, which is a nettle agent (it is fast-acting and causes chemical burns upon skin contact)

chlorofluoronitrosomethane ClCHF(-N=O)
phosgene oxime Cl2C=NOH
(phosgene oxime is formed by the reduction of chloropicrin, Cl3CNO2, with stannous chloride SnCl2.

The nitroso compound may likey be just a tautomer of the oxime, meaning the two could be the same compound.
ClCHF(-N=O) <==> ClCF=NOH

Formaldoxime is known to have similar equilibrium (usually existing in the trimer actually, although this is likely not the case of phosgene oxime), sometimes forming a nitrosomethane monomer in solution (although the nitroso groups themselves can potentially dimerise).

With some of these experimental agents, it seems they were trying to somehow combine two different chemical weapons onto the same molecule. This does not really seem like a logical approach, as there are two separate mechanisms of action, and the acetylcholinesterase inhibitor effects are usually much more potent than the other.

Also, to comment on the (CH3)2NCH2CH2S- group bonded to the phosphorous atom, I think the tertiary amine helps facilitate the hydrolysis of this group with the alcohol group on the target enzyme. The amine grabs the hydrogen ion, while the sulfur attaches to the oxygen of the former alcohol.* I am not really sure what advantage there is to this over a conventional P-F bond.

*Although the amine is fairly separated on this molecular group from the sulfur atom, it can still have a big effect. An excellent example of this is found in the nitrogen mustard agents. Usually chloroalkanes are relatively unreactive as alkylating agents, but when an tertiary amine is also present on the molecule like this, it becomes more potent. A typical nitrogen mustard has the structure ClCH2CH2N(CH3)CH2CH2Cl. (a sulfur atom can also be in place of the nitrogen, as in normal mustard gas, but in this case the sulfur takes a positive charge, which would be just the opposite of the sulfur in the above mentioned groups in the Novochok agents, which could likely come off with a negetive charge) The precise chemistry is more complex, but will not be mentioned here.

DDTea - 10-4-2012 at 22:50

To quantitatively discuss the potency of Military Organophosphates (MOPs), it's important to understand the basics of their molecular toxicology.

Inside the binding pocket of acetylcholinesterase (AChE), the hydroxyl group on the active site serine residue displaces the leaving group on the organophosphate by nucleophilic acyl substitution [1]. Like any nucleophilic substitution reaction, the kinetics are determined both by the accessibility of the electrophilic site (P, in this case) and the leaving group's ability to stabilize the negative charge transferred to it. The basic idea is an extension of nucleophilic acyl substitution on esters, which is discussed in any introductory organic chemistry textbook.

Up until this point, the phosphorylated enzyme is inhibited reversibly. Reactivation is slow under ordinary circumstances but can be achieved through oxime reactivators--basically, another nucleophilic acyl substitution in which the active site serine is the leaving group. Secondary reactions, collectively called aging, result in irreversible inhibition. One of the more important aging reactions is the loss of the alkyl group to form a formally negatively charged O=P-O(-) moiety [1]. In the case of Sarin, Soman, and DFP, which all possess branched R groups, it has been proposed that this involves the heterolytic cleavage of the O-R bond to form an alkyl carbenium ion [1]. In the case of soman, the pinacolyl carbenium ion can undergo a methyl shift to place the positive charge on a tertiary center and gain added stability. The formed negative charge on the phosphoryl group inhibits further nucleophilic attack (i.e., reactivation) [4].

According to this mechanism, any interactions in the alkyl group that could stabilize the nascent positive charge ought to increase the rate of the dealkylation reaction. This consideration would be mitigated by steric factors such as how well the organophosphate molecule fits inside the binding pocket of AChE and how accessible the central phosphorus atom is to the serine hydroxyl group.

So with my simplified explanation, we have a measurable, quantifiable idea of "potency." It is a combination of the following factors:
#__The ability of the molecule to reach the target site (toxicokinetics), which is affected in part by its lipophilicity.
#__Kinetics of the phosphorylation of the catalytic serine residue in AChE, which is affected by the leaving group ability, electrophilicity of the P center (more on this later), and sterics around the central P atom.
#__Kinetics of the aging reaction that prevents reactivation of the enzyme-substrate complex.

The specific enzyme-substrate interactions of different AChE's has some unexpected variability: serine is not the only residue involved ([5] and [6]). In the case of chiral nerve agents, R and S isomers have slighly different molecular toxikodynamics but the end result is the same. For that matter, AChE is not the only enzyme targeted by organophosphate nerve agents. This discussion is basically a summary of a thread I'd started in the Biochemistry section a few years ago [7].

With some of these experimental agents, it seems they were trying to somehow combine two different chemical weapons onto the same molecule. This does not really seem like a logical approach, as there are two separate mechanisms of action, and the acetylcholinesterase inhibitor effects are usually much more potent than the other.

Dihaloformaldoxime looks like one of those funky groups on molecules with highly tailored end-uses that have resulted from lots of structure-activity relationship studies. It actually reminds me a lot of the structure of the organophosphate insecticide DDVP. For all we know, this may not even be a moiety in the fabled Novichok agents.

Wikipedia has a page on the Novichok agents now [3]. It mentions a book by V.S. Mirzayanov, published after the start of this thread, State Secrets: An Insider's Chronicle of the Russian Chemical Weapons Program (ISBN 9781432725662), which gives different structures for the Novichok agents. They're duplicated at the bottom of the wiki page and I'm including them in this post (see attachment). There are a few subgroups here: phosphoriminofluoridates, methyl phosphoniminofluoridates, phosphonothionofluoridates, and methyl phosphonoselenofluoridates (!).

So we're back to the P-F bond. I guess F (HF, really) is just an awesome leaving group for volatile organophosphorus compounds intended to shutdown AChE. Maybe the O=P-N=C-N(R2) moiety rearranges to form a group that is particularly inert toward oxime reactivating agents or is itself inert to reactivation?

That P=Se group seems especially interesting. Phosphorothionates (OP's with P=S) groups are unable to inhibit AChE in and of themselves because the P is not electrophilic enough. Instead, they undergo a rearrangement (thiolo-thiono rearrangement) via an intramolecular Sn2 mechanism in which the phosphoryl S attacks the carbon in the P-O-C moiety, displacing the oxygen and forming the oxon analog. This, in turn, makes the central P atom electrophilic enough to be attacked by the serine hydroxyl group. A bit of background: Pearson's hard bases (e.g., -OH in the serine residue) preferentially attack the phosphorus atom while Pearson's soft bases (e.g., -SH, -SeH) preferentially attack tetrahedral carbon atoms in OP's [4]. Selenium, being larger and squishier than sulfur, may allow phosphoroselenates to readily undergo an analogous rearrangement while maintaining the lipophilicity/hydrophobicity of the original molecule.

(And on a further tangent, there was a great article on the use of HSAB Theory for describing toxicant-target interactions in Chemical Research in Toxicology : )

And this is another tangent but the actual chemistry is too interesting to ignore:

A typical nitrogen mustard has the structure ClCH2CH2N(CH3)CH2CH2Cl. (a sulfur atom can also be in place of the nitrogen, as in normal mustard gas, but in this case the sulfur takes a positive charge, which would be just the opposite of the sulfur in the above mentioned groups in the Novochok agents, which could likely come off with a negetive charge) The precise chemistry is more complex, but will not be mentioned here.

Sulfur mustards undergo an internal Sn2 reaction in which the S atom displaces a beta Cl and forms an intermediate with a three-membered ring. This strained ring, in turn, is opened by --NH2 group of a protein residue in a second Sn2 reaction. [2] Another cool use of three-membered heterocycles in biological chemistry!


1. Biochemistry 1999, 38, 7032-7039.
2. McMurry, John. Organic Chemistry, 6th edition, pp 380-381. Thomson Brooks/Cole: 2004.
4. "Organophosphorus Insecticides. " Plimmer, Jack R.; Gammon, Derek W.; Ragsdale, Nancy N. (2003). Encyclopedia of Agrochemicals, Volumes 1-3.. John Wiley & Sons.
Online version available at:
5. Biochemistry 2006, 45, 74 81
6. J. Am. Chem. Soc. 1999, 121, 9883 9884
7. "Toxicity of AChE Inhibitors on a Quantitative, Molecular level."

Novichoks-Mirzayanov.png - 40kB

[Edited on 4-12-12 by DDTea]