Sciencemadness Discussion Board - NaDIC explodes on heating ???

NaDIC explodes on heating ???

wg48 - 11-2-2018 at 15:54

I was checking how dehydrated about 1g sodium dichloroisocyanunric acid was by heating it in a small test tube. Initially it appeared dry then I noticed what I thought was steam condensing on the upper part of the tube. Then I noticed a change in colour on some of the granules on the hot part of tube from powder white to grey tinged with orange.

It was no long being heated but it appeared to be in thermal runner way with steam rapidly issuing from the loose cap. I was about to dump the tube when its contents rapidly decomposed propelling at least part of tube into the side of a small oven.

I had previously dehydrated about 10g of NaDIC at 120C in a small oven with no problems.

I did not know that NaDIC (if that’s what it is) has a tendency to explode (very rapid decomposition) when heated . Perhaps nitrogen trichloride formed.

Damaged oven, the remains of the test tube is inside

Damaged finger I hope the nail bed is ok

Chemi Pharma - 11-2-2018 at 16:14

I had a similar experience with TCCA, that start to melt and expeled dense fumes inside a blender, while triturating pool tablets to a thin powder.

I was lucky cause I was working inside a fumehood and got to switch the blender power off and the exaustion on, but the plastic of the blender melted and almost catched fire. The hood were filled full of toxic fumes and I spent more than 10 minutes to clean the air inside it.

Until now, I don't know what happened. May be NCl3 as @wg48 said.

I wish somebody here could explain what can it be too.

wg48 - 11-2-2018 at 16:41

Quote: Originally posted by Chemi Pharma

I had a similar experience with TCCA, that start to melt and expeled dense fumes inside a blender, while triturating pool tablets to a thin powder.

I was lucky cause I was working inside a fumehood and got to switch the blender power off and the exaustion on, but the plastic of the blender melted and almost catched fire. The hood were filled full of toxic fumes and I spent more than 10 minutes to clean the air inside it.

Until now, I don't know what happened. May be NCl3 as @wg48 said.

I wish somebody here could explain what can it be too.

The msds say:

"Special Remarks on Explosion Hazards:
Reacts explosively with calcium hypochlorite in the presence of water. May explode from heat or contamination".

Chemi Pharma - 11-2-2018 at 17:09

Thanks @wg48, I know about TCCA MSDS, but I wanna know what's the chemistry behind this behavior.

What happens when you heat TCCA? What chemicals is produced? What chemicals the dense fumes I observed contains? What about the equations?

What have made me astonished is the fact I don't believe triturating pool tablets in a blender less than five minutes produces enough heat to start so big runaway I saw.

I still need to know more about that.

wg48 - 11-2-2018 at 17:28

Quote: Originally posted by Chemi Pharma

Thanks @wg48, I know about TCCA MSDS, but I wanna know what's the chemistry behind this behavior.

What happens when you heat TCCA? What chemicals is produced? What chemicals the dense fumes I observed contains? What about the equations?

What have made me astonished is the fact I don't believe triturating pool tablets in a blender less than five minutes produces enough heat to start so big runaway I saw.

I still need to know more about that.

Yes I would like the details too.

Not much left of the nice Pyrex screw cap test tube and I was going to use one that had maganese oxide burnt on it but I thought that was not a good idea.

Vomaturge - 11-2-2018 at 17:46

Where did you get the NaDIC?
https://patents.google.com/patent/US4118570A/en
Apparently, NCl3 does form during the production process. Just a wild guess, but maybe your sodium dichloroisocyanurate had some leftover precursors in it, some actual trichloramine, or some other reactive impurity. In the "description of the prior art" section of that patent, it said
"In the past, it was necessary to further process the SDCC reaction mixture of the above reactions to remove these undesirable quantities of NCl3. This procedure was time-consuming and costly."
Perhaps the source of your product didn't use the patent, and didn't want to add a "time-consuming and costly" step to their production process. Instead, maybe they let the NCl3 impurity settle to the bottom of the barrel, and sold it to you?
The reason I would suspect that some precursors were left over in the product is that heating the mix could have completed the reaction making some more NaDIC, but also more yellow NCl3, which responded badly to further heating.
I couldn't find the heat of formation for NaDIC, making it hard to estimate whether it itself decomposes exothermically.
Does hypochlorite effect dichlorocyanourates, or just tri?
Because if it effects both, an impurity of it could also be the issue.
In any case, sorry to hear about your finger, test tube, and oven, but glad the 10g batch didn't do this... yet.

[Edited on 12-2-2018 by Vomaturge]

Ozone - 11-2-2018 at 23:06

"...white to grey tinged with orange." I'm wondering about Chlorine dioxide (ClO2), maybe (I've noticed it tends to go dark yellow/orange prior to going kaboom). There are better references, but here is a photo: https://en.wikipedia.org/wiki/Chlorine_dioxide.

Could be NCl3, of course, but it's more yellow, and three times as spooky (blew up a small sample of butylamine, once, this way--best guess at what happened).

O3

woelen - 12-2-2018 at 02:17

TCCA (trichloro cyanuric acid) and Na-DCC (sodium dichloro cyanurate) both are very reactive chemicals. For OTC products I find these remarkably dangerous. I also have experience with extremely violent reactions of these chemicals:
- heating leads to violent decomposition
- mixing with certain organics (e.g. alcohol)
- mixing with ammonia or organic primary amines
- most notably, mixing with hypochlorites

In NL there has been a very nasty accident with Na-DCC some years ago. A certain brand of swimming pool chlorinator sold so-called shock-treatment. It was granular Ca(ClO)2, appr. 70% active chlorine. This same brand one year later had switched to granular pure Na-DCC. The containers were nearly identical, same blue label, pool chlorine shock treatment, water-soluble (something like that on a label). To be mixed with water and then to be poured in the swimming pool. A person had a bottle of this stuff from the previous year, maybe 20% left over, being Ca(ClO)2. He put that in a bucket. He added a similar amount from a new container, being Na-DCC and then added water. He did this inside. The stuff reacted EXTREMELY violently, giving off dense choking fumes and the material charred. He could go outside, otherwise he probably would have died in the dense choking fumes.

I have the impression that Na-DCC and TCCA easily form NCl3, leading to explosions. I do not believe that ClO2 is involved in the danger/risk of these chemicals. These chemicals also react insanely violently with ammonia. Just for fun, add 15% or so ammonia to a pea-sized chunk of TCCA or Na-DCC. You will be surprised by the extreme violence of this reaction. E.g. sodium with water is tame compared to that.

wg48 - 12-2-2018 at 03:09

Quote: Originally posted by Ozone

"...white to grey tinged with orange." I'm wondering about Chlorine dioxide (ClO2), maybe (I've noticed it tends to go dark yellow/orange prior to going kaboom). There are better references, but here is a photo: https://en.wikipedia.org/wiki/Chlorine_dioxide.

Could be NCl3, of course, but it's more yellow, and three times as spooky (blew up a small sample of butylamine, once, this way--best guess at what happened).

O3

I need to check, as well as I can, that it is NaDIC and not significantly contaminated or adulterated in particular with calcium beaching powder. I will not be attempting a melting point test in a test tube on it without blast protection and remote monitoring LOL

It was purchased from a pool chemical suppler on Ebay. I have purchased what was supposed to silica gel but was actual clay and also naphthalene mothballs that where actual a dash of naphthalene, insecticide and mostly inert filler. So I would not be surprised if it is adulterated.

I will also check that the reaction is reproducible but outside this time in an open metal crucible.

Apparently from Woelen's post it may well be unadulterated.
But I would like to know the safe dehydration temperature for the dihydrate, my small oven works just fine with a door.

[Edited on 12-2-2018 by wg48]

Chemi Pharma - 12-2-2018 at 03:45

Quote: Originally posted by woelen

These chemicals also react insanely violently with ammonia. Just for fun, add 15% or so ammonia to a pea-sized chunk of TCCA or Na-DCC. You will be surprised by the extreme violence of this reaction. E.g. sodium with water is tame compared to that.

@Woelen, what do you think about this paper I attached below where researchers mix TCCA with ammonea and alcohol, amines, benzyl halides and aldehydes, as a substrate, to produce nitriles?

Do you think exist a safer way to scale this reaction between milimoles of substrates onto moles?

Attachment: alcohols, aldehydes, amines and benzyl halides to nitriles with TCCA + NH4OH.pdf (115kB)
This file has been downloaded 401 times

woelen - 12-2-2018 at 04:29

I think that this is useful for synthesis of specialty nitriles in small quantities, but when you want to have bulk0production at a mol-scale, then I see no practical use for this reaction. The authors themselves already state that adding 0.75 mmol of TCCA to the reactants must be done in very small portions to avoid explosion. If you want to add e.g. 2 moles, then it would takes ages before all of it can be added to the reactants.
The reaction between TCCA and many other chemicals scales badly. Heat-production vs. heat transfer out of the reaction mix scales cubically vs. quadratically. So, if e.g. 8 times as much heat is produced by taking 8 times as much reactant, then only 4 times as much heat transfer can take place. This leads to a strong increase of temperature in the 8-fold reaction mix. Things become much worse when larger amounts (e.g. 1000 times as much) are needed.

Chemi Pharma - 12-2-2018 at 08:58

Thanks for the explanation @woelen.

I believe have found something digging old threads about the products of TCCA decomposition and why it can be so smelly (the fumes) and even explosive:

Quote: Originally posted by Formatik

Quote: Originally posted by halogenstruck

careful:Hazzard:[heating TCCA]maybe explosive decomposition due to NCl3 prescence although dilution by NaCl reduces the possible danger

Yes, be cautious dry heating trichloroisocyanuric acid especially with oxidizables. Thermal degradation of it in inert gas is a known synthesis to carbonyl diisocyanate (colorless liquid which smells choking like SO2Cl2, probably extremely poisonous) alongside some NCl3, to a lesser extent Cl2 and CO2.

Ozone - 12-2-2018 at 09:42

Hmm. Now that I think more on it, NCl3 (and other nastiness) seems more likely than ClO2.

I've run into a few very stable things that become unstable when rigorously dried--that is, some amount of water (be it crystallization, solvation, or what have you) stabilizes the compound. For example, sucralose (the sweetener), is so stable some are regarding it as a potential environmental pollutant. BUT, in it's pure form, it will spontaneously eliminate HCl(g) when bone-dry (in a drybox at rt). This was shockingly apparent when the bottle was opened and I was gassed with HCl and noted that the product had turned black. Upon opening, some moisture got in there, and the reaction began to auto accelerate (pressurizing and rounding the square container). Now, this was quenched in a sink full of water and that was that--But, of course I bought more to try and figure out why I was gassed by a non-nutritive sweetener.

The result is completely reproducible, and can be measured via TGA/DSC and the tar examined via GC-MS. It turns out, the stuff is thermodynamically stable, but when dry, is entropically unstable--this is why it's heavily diluted onto maltodextrin and kept in aluminized mylar (to maintain moisture rather than keeping it out). Funny thing is that Tate and Lyle give a melting point for the stuff, the patent indicates that the product was isolated as a syrup, and said MP (which is impossible because it decomposes) has been parroted ad-infinitum ever since. I suppose a product that decomposes to yield HCl would be hard to market as a food ingredient...

I suppose the point of this anecdote is to point out that chemicals, even those that we eat, can rear up and bite you when treated in new and creative ways--and that the literature sometimes omits (for commercial purposes, I think, in the case above) details that might mitigate these risks.

It also makes me wonder if rigorously drying this stuff might have an effect? My guess is probably not, but you never know--and that is the point.

I'd consider writing up a hazard note for publication.

Heal up, and carry on,

O3

happyfooddance - 12-2-2018 at 10:12

This is very interesting.

I have about half a kilo of sucralose that I bought from a japanese company a couple years ago. It is a fine crystalline solid, supposed to be molecular and have no fillers. I haven't checked on the bulk of it but I can guarantee it isn't gassing HCl because nothing is damaged or corroded in the container. I have a small jar in my spice cabinet which has 5-10 grams in it. Even just opening the container to remove the smallest amount I can taste it strongly.

Once, years ago I extracted sucralose from Splenda (separated it from maltodextrin) using dry IPA. I didn't take notes, but I remember drying it with liberal heat and didn't notice any HCl.

I am curious as well to these mechanisms and underlying processes.

In wg48's case, I wonder if his compound was dried first over a strong dessicant, if it would react the same? Either way, thanks for the urging of caution with these inherently dangerous reagents.

Edit: spelling

[Edited on 2-12-2018 by happyfooddance]

Ozone - 12-2-2018 at 11:54

Try drying some of the sucralose in a desiccator over drierite for a month or so. In the normal atmosphere, the stuff isn't deliquescent, or anything, but it does maintain an equilibrium with water adsorbed from the air which keeps it stable.

Via TGA, you can watch the water come off, and then away it goes. It also poofs up enough to blow the lid off the crucible...sub-optimal...

O3

happyfooddance - 12-2-2018 at 13:13

I have no idea what TGA is, my friend... I tried searching, but found nothing...

wg48 - 12-2-2018 at 14:17

Dehydrating the dihydrate:

From: http://www.merckmillipore.com/GB/en/product/Dichloroisocyanu...
"Loss on Drying (105°C) 9.0 - 16.0 %" Probably to constant mass. The dihydrate is 14% H2O

Probagation testing:

From DOI: 10.1002/prs.680220202
Estimated decomposition enthalpy -1000cal/g
self-sustaining thermal decomposition reaction approximately .06 to 0.1 cm/s. Max temperature 600C
Onset of decomposition approximately 90C, with maximum rate at 150 to 200C

Decomposition products

From https://pubs.acs.org/doi/pdf/10.1021/ie502154b
"Starting from these results, we have characterized two decomposition reactions for SDIC, both giving two molecules of chlorine isocyanate (Cl-N=C=O) and one molecule of sodium isocyanate (Na-N=C=O). However the first path corresponds to a two steps mechanism while the second is composed by a single step. The first reaction, shown in Scheme 1, involves an initial breaking of two nitrogen - carbon bonds (G≠ = 44.1 kcal mol-1) to form a molecule of chlorine isocyanate. The rest of the initial cycle decomposes with a low energetic barrier (G# = 12.6 kcal mol-1) to form another chlorine isocyanate and a sodium isocyanate molecule. The second path involves a concerted breaking of N1-C2, N3-C4 and N5-C6 bonds of SDIC (see Figure 1) but its energy is too high to be competitive (G = 48.0 kcal mol-1) with the first proposed reaction."

[Edited on 12-2-2018 by wg48]

Ozone - 12-2-2018 at 14:39

TGA = thermogravimetric analysis. https://en.wikipedia.org/wiki/Thermogravimetric_analysis

When coupled with differential scanning calorimetry (DSC) https://en.wikipedia.org/wiki/Differential_scanning_calorime... , you can get a lot of information about the thermodynamic behavior of the material.

Aside the SDS for DCCA (for whatever they're worth) indicates that it decomposes at 240°C (no other info given), and that "Strong heating (decomposition)...shock and friction are Conditions to avoid." They seem more concerned about environmental hazards.

Aside, the Wiley Guide to Chemical Incompatibilities indicates NCl3 (see attached).

Hope this helps,

O3

[Edited on 12-2-2018 by Ozone]

wg48 - 12-2-2018 at 15:20

Quote: Originally posted by Ozone

TGA = thermogravimetric analysis. https://en.wikipedia.org/wiki/Thermogravimetric_analysis

When coupled with differential scanning calorimetry (DSC) https://en.wikipedia.org/wiki/Differential_scanning_calorime... , you can get a lot of information about the thermodynamic behavior of the material.

Hope this helps,

O3

I considered diy thermo analysis on a sample diluted with clean sand. I have multi temperature reader and pid controller that can interface to a pc, but I need an appliction to display the results and manage the interface. Thats why I was so interested in Nemo's open source GC software but it has no hardware interface only file input. Still hoping to get something like labview free.

wg48 - 12-2-2018 at 15:38

Quote: Originally posted by Ozone

Aside, the Wiley Guide to Chemical Incompatibilities indicates NCl3 (see attached).

But it said that was the reaction with water which makes no sense to me.

Ozone - 12-2-2018 at 16:38

We know it's tame in water. I'm thinking this was a typo (or some kind of brainfart) and they meant heating. The SDS cautions against heating the stuff, Wiley mentions NCl3. My mind puts those two together (only because I know it doesn't kaboom when mixed with water up to 250 g/L @ STP).

The carbonyl diisocyanate (CDIC)-thing, however, is intriguing. It's been made from TCCA (Crazy Klapotke, again: https://pubs.acs.org/doi/abs/10.1021/acs.jpca.6b04245), so that's maybe not too much of a stretch. See also from DCCA and phosgene: https://www.google.com/patents/US3941873, where, "The reaction is accompanied by side-reactions and is difficult to control due to the simultaneous formation of the highly explosive compound, NCl3."

Interestingly, CDIC will decompose to 3CO + N2, which could be frisky: http://www.chemistryviews.org/details/ezine/9824551/Stepwise... but this was photochemical. But, where would that phosgene be coming from, or what else might do the trick?

Was the sample exposed to light just prior to the explosion, or was it illuminated the whole time?

In any case, here: http://datasheets.scbt.com/sc-236901.pdf it's indicated that, "Combustion products include: carbon monoxide (CO), carbon dioxide (CO2), hydrogen chloride, phosgene, nitrogen oxides (NOx), other pyrolysis products typical of burning organic material." So maybe you get a little CDIC and NCl3. Maybe you somehow got a lot?

An assay of the starting material would be helpful.

Aside, if you have a PID your experiment seems plausible--although it might not behave the same way on sand. DAQ or not, with the PID (assuming calibration) you should know what temperature it's at when it goes. The tricky bit is that heating rate can have a significant effect on this "characteristic" temperature. I've seen things go at much lower T when heated rapidly.

As I've said, once something blows up in my face I want to figure out why, so...

I'll keep thinking on it,

O3

Now that I re-read your original post, and knowing that decomposition can spread through the stuff from an initiating hot-spot, I'm wondering if it was just a hell of a lot faster once a hot spot formed at the tube wall and spread into an already hot mass?

[Edited on 13-2-2018 by Ozone]

wg48 - 13-2-2018 at 01:31

O3: The idea of mixing sand with the NaDIC to reduce the possibility of thermal runaway by reducing the temperature rise for a given amount heat of decomposition. It would also make it more difficult to detect the temperature rise.
.
The link I gave for the decomposition products said it was a theoretic decomposition because they could not find any definitive measured decomposition products. So although they gave justifications for the reaction from bond energies, chemistry frequently is not that simple or it would be impossible to electroplate chromium from an aqueous bath.

The propagation speed of the decomposition front given in the link was 1mm/s max but gave no temperature or conditions, was it in a tube or trough, what atmosphere? I guess it was a trough in which case the propagation in a confining tube could be much higher and as you suggest if the all NaDIC had been heated to near its decomposition temperature before decomposition started.

Ozone - 13-2-2018 at 11:02

Yep, I get what the sand was for--but, because you'd maybe be trying to recreate the incident (in a controlled way) in order to study it, the safer option may not give the information you want (and, as you say, would confound measurement).

I hear you on the decomposition products. The way I'm seeing it, now (which is subject to change with new/better information), is that the stuff decomposes around 240 °C (it's likely many changes happen before that). As this occurs, some phosgene is made that reacts with remaining DCCA to make CDIC and NCl3 (both of which are quite bad). Perhaps if the stuff will propagate at 1mm/s (assuming rt), then it might do so much more quickly (making more nasties, in less time, markedly increasing the concentrations, perhaps critically so).

It would be very interesting to test propagation rates through a trains of DCCA equilibrated to various temperatures, e.g. 25, 75, 125, 175, and 225 °C. I suppose they could be initiated with a glowing splint at one end (on a looong stick...or electronically), and the rates measured with a high speed camera. No way I can set up to try that, though :/

O3

AJKOER - 23-2-2018 at 08:38

My understanding of the reaction trichloroisocyanuric acid (Cl-CNO)3 with water (and correspondingly, salts of this acid on hydrolysis, or with acids):

(Cl-CNO)3 + 3 H2O --> 3 NH2Cl + 3 CO2 (g) (implied by one of the reaction scheme paths of R-CNO on page 9 at http://cdn.intechopen.com/pdfs/38589/InTech-Polyurethane_an_... )

NH2Cl + H2O = HOCl + NH3

HOCl + NH2Cl = H2O + NHCl2

HOCl + NHCl2 = H2O + NCl3

which clearly suggests the possible creation of some unstable and explosive NCl3.

[Edited on 23-2-2018 by AJKOER]

Ozone - 23-2-2018 at 16:52

Hmm. I'd have thought it was drier than that?

O3

Vomaturge - 23-2-2018 at 18:19

I think AJKOER is talking about the warning of SDIC or TCCA forming traces of NCl3 with water. As far as wg48's original accident, I think it was caused by the self-sustaining exothermic thermal decomposition of SDIC.

[rquote] it appeared to be in thermal runner way with steam rapidly issuing from the loose cap. I was about to dump the tube when its contents rapidly decomposed propelling at least part of tube into the side of a small oven.
[rquote]

assuming that the tube had a cap that had not yet been removed when it burst, this is exactly what I'd expect. I don't know what the cap was like (rubber stopper? Threaded?), but the fact that a big portion of the tube flew apparently in one piece like a rocket, busted a hole in the oven, then shattered? That seems like a deflagration to me. The exact details of the incident are important, but now that we know SDIC creates enough heat to sustain its own decomposition, it seems plausible that no other energetic materials were present.

Quote: Originally posted by wg48

Probagation testing:

From DOI: 10.1002/prs.680220202
Estimated decomposition enthalpy -1000cal/g
self-sustaining thermal decomposition reaction approximately .06 to 0.1 cm/s. Max temperature 600C
Onset of decomposition approximately 90C, with maximum rate at 150 to 200C

Decomposition products

From https://pubs.acs.org/doi/pdf/10.1021/ie502154b
"Starting from these results, we have characterized two decomposition reactions for SDIC, both giving two molecules of chlorine isocyanate (Cl-N=C=O) and one molecule of sodium isocyanate (Na-N=C=O). However the first path corresponds to a two steps mechanism while the second is composed by a single step. The first reaction, shown in Scheme 1, involves an initial breaking of two nitrogen - carbon bonds (G≠ = 44.1 kcal mol-1) to form a molecule of chlorine isocyanate. The rest of the initial cycle decomposes with a low energetic barrier (G# = 12.6 kcal mol-1) to form another chlorine isocyanate and a sodium isocyanate molecule. The second path involves a concerted breaking of N1-C2, N3-C4 and N5-C6 bonds of SDIC (see Figure 1) but its energy is too high to be competitive (G = 48.0 kcal mol-1) with the first proposed reaction."

[Edited on 12-2-2018 by wg48]

parameters like burn rate also depend on wether the material is a solid block, wether the burning surface is able to vent gases (or wether the gas gets forced deeper into the powder, heating it further), what pressure the reactant is under, etc. It seems logical that preheating the material would make it burn faster too, because it would require less heat transfer to reach ignition temperature deeper in the powder. It is my uneducated opinion that preheating made less than a gram of ETN DDT in an open test tube, causing this accident here:
https://www.sciencemadness.org/whisper/viewthread.php?tid=16...
By getting a large piece of energetic material within a few degrees of its ignition point, you make it more sensitive to further temperature increase. When one part does ignite the burst of heat/pressure effects the rest of the sample much more than it would at room temperature.

By the way, I would kind of expect the same damage pattern from a NCL3 accident as from ETN. The glass would probably fly in all directions, with a stronger explosion making smaller, much faster fragments. It would be more radial, and less coherent. Of course, I don't know. Maybe the tube was against the oven, and the big hole was made by a dense shower of fragments, or even the blast pressure itself?

[Edited on 24-2-2018 by Vomaturge]

wg48 - 24-2-2018 at 00:37

The oven was 2 to 3ft away. Only three fragments of glass (two with threads) and one piece of the threaded cap were found in the opposite direction of the oven. That suggests the capped end failed first injuring my finger and propelled a mostly intact tube into the oven where the all but the three pieces of the tube were found. The oven is old. I suspect the plastic is brittle from repeated heating cycles.

I think tests will confirm rapid decomposition due to the temperature and partial confinement of the decomposition products. I may perform some outside tests soon undercover of burning garden rubbish. I don't want to alarm my neighbours.

I guess the morale of the story is if this experiment unexpectedly goes bang have I mitigated the risks. I think I was lucky my hand or more could have been seriously injured. I only had eye protection.

An other one would be don't use a loose screw cap as a vent use a cottwool or glass wool plug or none.

From my camping days: If you wrap a strike anywhere match head in Al foil then throw it in to a camp fire it does not go pop. If you place it near the edge of the fire in just the right position such that it heats up slowly, it will frequently make a bang. Occasionally matches thrown in to a camp fire that land near the fire or on the cooler side of wood such that they heat up slowly can ignite so quickly that they are propelled out of the fire.

[Edited on 24-2-2018 by wg48]

Vomaturge - 24-2-2018 at 11:05

Thanks. Sounds like the stuff deflagrated as described in the link you gave. Heating the substance to 90-250 C makes it produce gases, and more than enough heat (600C! ~4 kj/gm!) to propagate the reaction. Like a fuel/oxidizer mixture, or an unstable compound like nitrocellulose, this substance has variable burn rate. The experience with match heads supports the theory that burn rates can be increased by preheating. The fact that a propellant/low explosive can cause a closed container to explode is hardly surprising. A loose cap is probably good enough for an experiment involving slow endothermic decomposition or evaporation, but energetic materials (including slow burning thermites and KNO3/sugar mixes) if they must be heated, likely need to be in an open pan, preferably well away from the operator!

What did surprise me, though, is that this material decomposes energetically, and that this property is not well known. The Wikipedia page on sodium dichloroisocyanurate talks about it as a cleaning agent, mentions a luminescent reaction with hydrogen peroxide, but says nothing about the fact that it is a low explosive! Of course, nitrocellulose, ammonium nitrate, and picric acid all have (or had) non explosive uses. I still wish the MSDS would let you know when something is known to decompose violently on heating:
http://www.essef.be/images/documenten/004800_vf1_E.pdf
They describe it as producing toxic fumes on heating, but nothing about how it decomposes, or what to do if a not-quite-sealed barrel or shipping container of it is hissing and smoking. Run!

wg48 - 24-2-2018 at 14:18

Vomaturge: The MSDS was written in 2015 and states that it is “presented in good faith and believed to be correct as of the date of publication” but that accident report was written in 2003!!!

AJKOER - 27-2-2018 at 06:29

Quote: Originally posted by Vomaturge

I think AJKOER is talking about the warning of SDIC or TCCA forming traces of NCl3 with water. As far as wg48's original accident, I think it was caused by the self-sustaining exothermic thermal decomposition of SDIC.......

[Edited on 24-2-2018 by Vomaturge]

I found a prior comment by Woelen relating to some NH2Cl formation, which is described as, not normally expected, and which I have suggested as a possible minor reaction path upon hydrolysis:

Quote: Originally posted by woelen

The TCCA is pure. The bottle says "92% active chlorine" and this is near the theoretical value for 100% pure TCCA.

I also have severe doubts about CO. CO is strongly reducing and how could such a gas be formed in such a strongly oxidizing environment? I myself tend to believe more and more that it is NH2Cl, but unfortunately very little information can be found about this compound. Most information is about swimming pools and about water treatment with very dilute solutions of NH2Cl, but of course that information does not tell anything about flammable or explosive properties of the more or less pure gas.

The smell of the gas mix is really bad. It is not as strong as the smell of chlorine, but its strength comes close to it (the actual smell, however, is quite different from the smell of chlorine). You must very carefully whiff some of the gas to your nose and certainly should not stick your nose in the test tube. If you do that, then you'll certainly get the mother of all caughs.

I can imagine formation of NH2Cl from TCCA as follows:

TCCA contains -N(-Cl)-C(=O)- chains. These could hydrolyze to NH2Cl and H2CO3, but normally this does not occur. The oxidizing nature of the hypochlorite may lead to oxidative breakage of the ring and then the above mentioned hydrolysis may occur. At the same time, part of the N-atoms must end up in oxidized species, most likely N2. The -C(=O)- part would end up in carbonate ion, while the -N(-Cl)- part hydrolyzes and picks up hydrogen from water, the OH-group from the water then goes to the -C(=O)- part and becomes part of a (bi)carbonate ion. So, the gas produced would be a mix of N2 and NH2Cl, with at least one molecule of N2 for 4 molecules of NH2Cl, but probably more of the N-atoms are oxidized and a larger part of the nitrogen ends up as N2. What remains behind in solution is carbonate ion and chloride.

Cyanuric acid does not contain -N(-Cl)- parts. Its triol-form contains -N(-H)- parts and its triketol-form contains aromatic -N-C-N chains. I can only imagine oxidative cleavage of the ring structure to occur with formation of N2 and hydrolysis to NH3 and carbonate ion. The NH3 then immediately is destroyed by excess bleach and nearly all is converted to N2.

[Edited on 18-6-13 by woelen]

Link: http://www.sciencemadness.org/talk/viewthread.php?tid=24659

[Edited on 27-2-2018 by AJKOER]