Org Prep Daily

August 11, 2009

Dear ACS journal editors – please return from your vacation soon

Filed under: lab destruction, lit highlights — milkshake @ 1:14 am

dovolena credit: National Geographic ‘Photo of the day’

The spectacular “NaH catalytic” oxidation recently published in JACS has been thoroughly covered elsewhere. I would like to bring your attention to another jaw-dropping paper that just came out in Org Letters:

Acetophenones and 1-aryl-ethanols are oxidized to benzamides by heating the material with 3 – 4 equivs of iodine and ammonia in a pressure vessel. There is only a passing reference and footnote that “nitrogen triodide might form in the mix.”

In fact, NI3.NH3 readily precipitates upon mixing iodine with aqueous ammonia. (The products vary; a gradual iodine addition to a large excess of ammonia yields ammonium iodide and nitrogen.) Nitrogen triodide is a notoriously super-sensitive primary explosive. I spent some time hospitalized in eye clinic when I was ten years old – my corneas got burned with iodine and my eardrums ruptured because of playing with a spoonful of nitrogen triodide. (The window pane flew out and I was thrown to the ground by the blast;  it took me half a year to fully recover and this all was from few grams of dry material going poof, unconstrained). I cannot warn strongly enough against mixing iodine with ammonia in a pressure flask and then heating the stuff up!

The authors run these experiments on a 1 mmol scale and they give no details about the order of addition. Since the transformation is pretty useful – with a good substrate scope and it looks simple enough (its done in water) –  sooner or later some innocent person is bound to mix up a big batch in the wrong way – and as he screws on the the pressure vessel cap he is gonna blow himself up into a mauve cloud


  1. Q: was it not your *corneas* that got burned?

    I love the pic.

    I love the concern for people’s safety. What were they thinking??

    Comment by Brian — August 11, 2009 @ 10:24 am

  2. You should email the editors at Org. Lett.

    Comment by Anon — August 11, 2009 @ 10:33 am

  3. You blew yourself out a window and knocked yourself onto the ground with NI3 when you were 10? No wonder you became a chemist.

    I know a woman (with a degree in chemistry) who used to put NI3 in keyholes, under chair legs, etc.

    Comment by Matt — August 11, 2009 @ 10:42 am

  4. Sorry you are right it was corneas – I corrected the typo. (I wrote this post late and I was tired).

    I still remember rolling on the floor with eyes shut and screaming and hearing myself like from a far away – it hurt rather bad; also one of the unpleasanties of getting iodine in your face is that apparently your eyes get extremely light-sensitive long after, and you lose the sense of smell too, for months, so the food tastes very bland

    Comment by milkshake — August 11, 2009 @ 12:42 pm

  5. “so the food tastes very bland”

    How could you tell the difference with Czech food?

    Comment by CowTipper — August 11, 2009 @ 11:28 pm

  6. Milkshake,

    I’m very glad to hear you recovered from your incident. NI3 is scary stuff.

    It is absolutely unbelieveable that this article got in without any safety notes.

    Comment by cookingwithsolvents — August 12, 2009 @ 8:43 pm

  7. I was previously of the impression that NI3.NH3 can only go poof when dry (although I must say I have never played with it myself).

    Comment by syregnask — August 13, 2009 @ 4:48 am

    • There’s a iodination procedure, especially for activated aromatics (eg resorcinols) wherein you add aq KI3 solution (slowly) to a solution of the aromatic compound in 20% aq NH3. I have done this fairly large scale (~50 g), but I get nervous every time when I do this. Also, I am curious about the role of aq NH3 in this reaction. Is it NI3.NH3 that is formed the actual iodinating agent? Thanks

      Comment by Moody Blue — August 13, 2009 @ 11:29 am

      • I suppose they need NH3 as a nucleophile, to form benzamide. I have seen a very similar oxidation like this, with aldehydes and amines to provide amides.

        NI3.NH3 can be handled wet but I would still not recommend to make it on large scale, inside a heavy glass pressure vial and then heat it up. They need to have a detailed investigation in the paper, to see that the stuff does no accumulate, and have a step-by-step writeup where everything is defined. Imagine someone does it on large scale, gets some dark precipitate and filters it and dries it to find out what he got.

        Comment by milkshake — August 13, 2009 @ 1:20 pm

      • Sorry for the confusion. I meant the role of aq NH3 in the iodination rxn I was talking about. I wasn’t very clear as I was digressing…….

        Comment by Moody Blue — August 13, 2009 @ 1:57 pm

      • I don’t know – I suppose it can be there as a mild base. Iodinating resorciols sounds tricky, maybe they worried about selectivity.

        Comment by milkshake — August 13, 2009 @ 2:03 pm

  8. Dear Milkshake, nice catch. I was a fellow garage pyrochemist, and I never made nitrogen triiodide, being more an acetone peroxide kind of guy. Seriously, I hope someone calls everyone’s attention to this risk, preferably in a note to the editors; hey, free publication?


    (I posted this over at Finchsigmate’s blog too, but it was on the relevant thread about the old Fe3+ Grignard couplings by Kochi and so I think it’s buried. One of you gurus who have done this kind of chemistry mind taking a look?)

    I tried Kochi coupling with 5 mol % Fe(acac)3 to prepare p-dihexylbenzene from p-dichlorobenzene and a freshly prepared, titrated solution of n-HexMgBr, and got bugger-all. NMR shows nothing but p-DCB starting material, with only the tiniest hint (seriously <1%) of the benzylic triplet around 2.5 ppm.

    The only thing I can see that I did differently (other than the p-dichlorobenzene starting material, obvs, but the 2-chloro-6-tosylpyridine in the Org Prep synthesis should stand in for it) was that I ran the reaction at RT, and it actually got rather warm as the Grignard was added. Any thoughts?

    Comment by opsomath — August 17, 2009 @ 9:34 pm

    • The only thing you did differently is using a very unreactive starting material. What makes you think that you can freely swap 2-chloro-6-tosylpyridine for 1,4-dichlorobenzene, and still expect your reaction to work?? 2-halopyridines are exceptionally reactive. Get para-dibromobenzene and try again. When looking for a lit precedent you need to find a close one.
      (By the way I played with acetonperoxide as well, as a kid, and got myself almost killed in a mishap. That finally convinced me to stop.)

      Comment by milkshake — August 18, 2009 @ 10:28 am

      • Yeah the p-diCl is likely very poor substrate and I have had mixed results with Fe catalysts in similar chemistry, particularly if purchased or not made in glove box. Heck Pd conditions (of which many variants to try) might be better choice if intent is to just make the product although still likely to messy.

        Comment by CMCguy — August 18, 2009 @ 11:11 am

      • Actually I tried p-dibromobenzene before and it worked equally badly. Good point about the 2-position of pyridines, though. The Org Synth study did a dibenzofuran (2-triflate) substrate, on the other hand, which looks like it should be terminally unreactive, but that goes.

        Anyway, I can make the dihexyl from the dichloro or dibromo easily using Ni(dppp)Cl2 or the like. I just wanted to try this cool retro chemistry…

        I think I will still try doing it cold. This would be a handy one to have in the box.

        Comment by opsomath — August 18, 2009 @ 4:33 pm

      • And yeah, acetone peroxide is a terrible idea, agreed. It is said that God watches out for children, drunks, and fools… “neophyte chemists” should be on the list as well.

        Comment by opsomath — August 18, 2009 @ 4:34 pm

      • You can also try CuCN or Co(acac)2 as a catalyst for Kumada coupling with p-diBr-benzene

        Comment by milkshake — August 18, 2009 @ 4:37 pm

  9. It you have a big exotherm —> decomposition is very possible

    Try controlling internal T with cold external bath and slow addition rate. Then, if conversion is not complete let it reach RT or even heat a bit.
    Good luck

    Comment by vasili — August 18, 2009 @ 3:14 am

  10. Yes, I have done the NI3 prep. It’s…uhh…fun.

    Comment by Curious Wavefunction — August 18, 2009 @ 1:51 pm

  11. Just thought of checking yor view on the gaint molecular model kit that Aldrich sells. Do you have any suggestions for the molecular model kit for the demos in the class, especially organic?

    Comment by xyz — August 22, 2009 @ 6:35 pm

  12. I don’t – since I never had to teach. For building models of organic molecules for research projects I really like HGS Stereochemistry Molecular Model Set (Aldrich Z277835-1EA/$291) but the price is atrocious, especially since one needs two kits for making large molecules (and you cannot do cyclopropanes). I suppose the C-C bond distance (about 3 cm) with this model is good enough for showing the model across a table but too small for a classroom demonstration.

    Comment by milkshake — August 23, 2009 @ 2:48 am

  13. I’ve always understood NI3 to be hazardous when dry; presumably it needs to reach a critical concentration to go poof?

    Anyway, I’ve been doing oxidations of aldehydes to nitriles for some time using aqueous NH3/iodine/THF on multigram scales without any ill-effects. I always make sure to add the iodine gradually, and only use a slight excess, quench the reaction afterward with thiosulfate, etc.

    I always wear pretty thick goggles just in case though.

    Comment by The Main Gauche of Mild Reason — August 23, 2009 @ 10:00 pm

  14. True; they have one which is pricy but would like to go with them than trusting the other comapnies. Thanks.

    Comment by xyz — August 23, 2009 @ 10:57 pm

    • My Grad School Prof had an old (British or German?) model set that was very nice (and expensive) although he only used that to make “target molecules” for display on shelves and had another less expensive set for class/group work. Never taught other than labs however I always liked the cheap plastic kind, probably because they were less forgiving and did not break when you built “impossibly strained” structures. I could also paint the balls for additional elements not in standard kit, notably sulfur.

      Comment by CMCguy — August 25, 2009 @ 3:51 pm

  15. I’ve been following this blog ever since I discovered this around July and I love how informative this is (especially the comments section!). I just wish there were more updates though.

    Speaking of comments, I have a question for all you synthetic vets out there: what’s a good experimental organic book you’d recommend? I know the Vogel one’s a classic but I’m looking at something slightly more modern, was told Advanced Practical Organic Chemistry by Leonard, Lygo & Procter as well as Modern Organic Synthesis in the Laboratory by Jack Li are both great picks. Any suggestions welcomed!

    Comment by Rich — August 26, 2009 @ 3:13 am

    • Experimental org chem book: Unfortunately I cannot help – I hope someone else can come up with a good suggestion.

      Update frequency: The main problem is that oftentimes there is not enough interesting stuff to write about, chemistry-wise or in general, and juicier synthetic bits sometimes have to be saved for a later publication, especially if the work involved more people or related to an active medchem project in a way that’s sensitive to my employer’s intel property.

      I started this web page as a private version of OrgSyn, to post my own procedures, and I kind of hoped that by the time I run out of decent synthetic procedures that accumulated in my old notebooks over the years I could attract few occasional contributors; when that did not happen I progressed into putting up a more personal stuff as well – like lab disaster lore and rants about industry life – but in the end there is only as much that can be said about these things without getting repetitive or cranky.

      Comment by milkshake — August 26, 2009 @ 3:54 am

      • Nah… you wouldn’t get cranky…

        Comment by Taitauwai — August 31, 2009 @ 1:00 am

  16. Totally Synthetic’s liveblogged attempt to repeat the “NaH oxidation” made C&EN…did anyone see that article? They admit in there that “the proposed mechanism may need rethinking.”


    Comment by opsomath — August 28, 2009 @ 6:44 pm

  17. one thing that can keep things rolling until your new procedures or other contributors come in, is to give a personal touch for the well known and important reactions.

    For example: For Mitsonubu reaction, Wittig reaction etc as to what to do, what not to do and any latest modifications of the old and beaten procedures.

    Comment by xyz — August 29, 2009 @ 2:56 pm

  18. Some time ago I wrote up stuff like you are suggesting, in “Notes on terrible reactions” and two things happened: I got discrete responses pointing out that I was out of my depth on some of the subjects, and second, it got re-posted all in Chemical Know-how forum in like 12 separate posts – all posted there as if by me – which really made me look like a self-promoting blowhard. So I better write only about things that I know well – and for example Wittig and Misunobu I did only few times; any total synthesis jock has probably more experience with these than I do

    Comment by milkshake — August 30, 2009 @ 2:18 pm

  19. Anybody knows a good software that can predict metabolites?

    Comment by Quinine — September 5, 2009 @ 10:54 pm

  20. I am not aware of any commercial package, and I would not trust any software (that pretends to know the metabolites) because from what I have seen things are extremely dependent on small structural changes. I would rather look up some monograph/textbook (and maybe also consult with someone who worked in DMPK field)

    Comment by milkshake — September 5, 2009 @ 11:46 pm

  21. There have many attempts at in silico metabolism prediction in the past 10-15 yrs, but ASFAIK they have all been a dismal failure. Human metabolism is just *way* too damn complex for them them to provide anything close to reality.

    Comment by Jose — September 8, 2009 @ 12:21 am

  22. In a similar vein, 10.1002/anie.200704772
    in Angewante last year used NH4I/NaOCl/NH3(aq) to make aziridines.

    These people don’t mention the potential hazards of NI3 or chloramines but they also don’t report concentrating their reaction mixtures! GC yields only.

    Comment by Nemo — September 9, 2009 @ 1:18 pm

  23. NI3! In my university days a friend of mine was using the lab spare time to produce it in batches of 25-50 g (the production was on dayly basis, the material was used for jokes). No problem in manipulating the wet product. When dry, is a low potency explosive (orders of magnitude less potent than mercury fulminate) at NORMAL TEMPERATURE AND PRESSURE.
    Milkshake, are you sure that was NI3 the substance of your accident? I saw myself an explosion of about ten grams of dry product in the hands of my mate (the product was on filter paper): the bang was not so loud, nobody got injured, the most notable risk was associated to the cloud of iodine vapours.

    Comment by processchemist — September 18, 2009 @ 4:36 am

  24. it was that bad because I detonated it on the window still few inches from my face, without glasses, and all the iodine got into my eyes. At that time my height was the same as the window still so I was eyeballing it as I used a pocket knife trying to scrape off a part of the dried material

    Comment by milkshake — September 18, 2009 @ 1:51 pm

  25. It’s also unstable even wet or in water. Once I has some NI3 covered in dilute ammonia and exploded in my face without touching it.

    Comment by Zir0 — December 17, 2012 @ 6:24 pm

RSS feed for comments on this post. TrackBack URI

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Blog at

%d bloggers like this: