'Had a really FUN Saturday

[lloyd]

Every once in a while, I need a 'diversion' from the everyday design and making of stuff.

 

I was given a handful of green smoke stars by a fellow pyro who likes them, and wants to duplicate them. He asked me if I could analyze them.

 

Now... I LOVE lab chemistry, and even have a few appurtenances to the endeavor, but I am NO CHEMIST. So this sort of thing is always a challenge; fun, but hard.

 

First, I burnt a couple. They gave a black ash about twice the size of the original star, and a good gout of green smoke, with a burn time of about two seconds. They'd look good in the air. Next was trying to figure out the contents of them.

 

After 16 hours of fooling around (both Friday night and yesterday), I've found that the formula is about 30% water-insoluble dye. I asked a retired chemist friend how to tell what was in there.

 

He had to remind me of paper chromatography, which I've done before.

 

I happened to already have some test strips which, along with the beaker I needed, I had to 'hunt down' in my stash. After finding the stuff, I did a quick run, and the colors separated nicely. I don't know WHAT dyes they are, but I now know it's a mix of yellow and blue, and that it's not at all water-soluble, slightly ethanol and isopropanol soluble, and totally soluble in acetone.

 

Then I went to figure out what the flame package was. Almost the whole content of the star - except the dye - was completely soluble in an excess of water. That gave some clues. One thing that was left-over after cleaning out the dye from the residue is what appears to be about 2% elemental sulfur. At the same time I was doing that, I was trying another strategy.

 

I washed one star clean of dye in acetone. DAMN! There was enough dye in that one little 5/16" star to dye a white tee shirt a good green! It took nearly a quart, and about 25 changes of acetone to get that thing broken down to a fairly clean powder. Then, for some analysis of it.

 

Most "toy" smokes contain a phlegmatizer like sodium bicarbonate (required by APA 87-1), but this wasn't from a consumer item. I tested a small sample of the cleaned powder with acid -- no evolution of gas/CO2, so no bicarb in it.

I burnt a small sample. It gave a pure whitish/violet flame, indicating a potassium oxidizer. It left a clean white residue, evident of some sort of chloride or oxide from the oxidizer.

 

Could it be the oxidizer was KNO3? Well, KO is yellow, so probably not that. K2CO3 can result from burning carbon fuels with KNO3, and is white. Hmmm... but so is KCl. And I don't have any qualitative tests for KCL or K2CO3!

 

Ah... yes I do!

 

I know KCl is 'salty' in taste. This was. Oops! 'Turns out, so is K2CO3. THAT didn't work. Ah... but K2CO3 and acid evolves CO2 gas.

 

A drop of acid on a small bit of it revealed no gas being evolved. So, it must be KCl. Even that doesn't tell me everything.

KCl could result from either perchlorate or chlorate as the oxidizer. And at this point, I only have documents to go by. Almost all colored smokes are oxidized by potassium chlorate (because it's cool-burning with carbon fuels). Because these contained no phlegmatizer, and didn't consume the dye during their burning, I have to assume it was the chlorate.

My conclusions are not quantitative -- I don't have the lab setup to do that.

 

But I have books just loaded with colored smoke comps. So, in order to duplicate these stars, all I have to do is identify what dyes were used. I've done a LOT of internet searches, and cannot identify any yellow and blue dyes that are water-insoluble, barely soluble in alcohols, and completely soluble in acetone. (Maybe one of you might know!)

 

But, what a FUN diversion from my usual work! It was a good Saturday!

 

Lloyd

 

PS... subsequently, a retired chemist in Great Britain informed me of a test that will differentiate between chlorate and perchlorate, and I have the required reagents.

 

Also, a chemist friend here in the states told me about some on-line color charts used with paper chromatography to help identify specific dyes by their separation colors.

 

And a friend here who's a commercial printer informed me that phthalocyanine blue and oil yellow (which appear in one of the Shimizu smokes) are soluble in 'press wash' fluid, which contains polar solvents like (if not actually) acetone, and that the particular dye package of those two is NOT water-soluble.

 

So... we might be narrowing it down as to what the dyes are. I only had about ten of the VERY wide chromatography strips that are designed for multiple side-by-side samples on a single strip... so last night I ordered another sleeve of single-sample narrow strips, so as not to waste the few big ones I have left. The narrow ones will be here in tomorrow morning's mail, so I'll delay doing any more chromatography until tomorrow.

L

[dave321]

Lloyd,

 

the two yellow dyes I have used in the past are Solvent Yellow 33 and Solvent Yellow 93, there are probably others.

 

the Blue I use is Solvent Blue 78, but I think there are some Solvent blue 8? s also which can be used.

 

sometimes Solvent green 3 is also used with a solvent yellow.

 

alternatives to sod. bicarbonate would be either magnesium oxide or magnesium carbonate.

 

for burn rate control china clay can also be used which is largely inert

 

I am sure you know most of this

 

I did a similar exercise myself on the screening smoke composition used in the british screening smoke L132a,

but unfortunately wasn't able to nail it down perfectly, the biggest problem being reducing the "chuffing"

 

dave

[lloyd]

Yes, Dave, and thanks.

 

I ruled out inert (clay) burn-rate suppressants and magnesium oxide because everything but the dye and a tiny trace of sulfur are completely water soluble at 20C with only enough water to ensure dissolution of the amount of flame package I had, IF it contained the chlorate. The oxide is only slightly soluble at that temperature. The Mg carbonate would have reacted with an acid to evolve CO2.

 

In fact, my test partly-revealed that the oxidizer is probably the chlorate, because the perchlorate wouldn't have fully dissolved at that temperature and quantity of water.

 

Besides, there are quite a number of smokes (especially military) that don't use one, they simply increase the dye amount to make the mixture especially oxidizer-poor.

 

My brief foray into looking up Solvent Green 3 seemed to say it is also water-soluble, so I tend to rule that one out. It, too, is actually a mixture of blue and yellow, so it would have left a chromatography trace similar to the quick test I did.

 

Tomorrow's more-precise tests with various blends of solvent/non-solvent should give me better separations on the strips, and Tom's color charts will at least help narrow it down.

 

But Mike's clue about phthalo blue BS and oil yellow seems promising enough for me to buy some of those two, and see if they behave similar to what I have in the stars.

 

Thanks for the additional suggestions.

 

LLoyd

[dave321]

i always thought that phthalocyanine blue gave a rather washed out (relative to other blues) "weak" blue,

but that was on its own for blue smoke.............maybe different with a yellow to make green.

 

I know some of the other blues can be heat labile when used in a tube, smoke turning from blue to greenish to black,

but that probably is less of a concern for a small diameter star.

 

I guess a smoke star would have to have a high burn rate to generate a sufficiently coloured smoke plume moving through the air.

 

please post your findings, i'm always interested in this stuff.

[lloyd]

I shall. Likely, I'll not positively identify the dye package, without submitting it to professional chemical analysis. I won't do that! This is a "fun" project, not a big commercial deal.

 

But there are so many successful green smoke formulae (I have at least 21 of them), that I'm sure an approximation - tuned for burn rate, which I know how to do - will be good-enough to satisfy my friend.

 

The 'big deal' for me, is I get to do something purely for the art, instead of designing tools or making them, or consulting on regulatory matters. I make money doing those things, but I'd rather PLAY at pyro! <grin>

 

Lloyd

[WSM]

The cool burning mixture typically used is potassium chlorate and lactose (or other cool burning fuel). Other sugars may work and I've seen folks use sorbitol for better KNO₃ and sugar rocket propellants which are less hygroscopic. The main thing , of course, is to create a smolder reaction which doesn't flare up and burn the dye. I'm sure you know that but I state it for the benefit of those who haven't worked with smoke compositions yet.

 

I don't know if that helps, but good luck with your fun project.

 

WSM

[ThisisMarks]

Wow, you guys know your stuff. Perhaps you can help me with my colored smoke or lack of colored smoke mixture. I came across several pounds of pigment dyes while I was helping to remove wooden interior structure from a warehouse that was recently vacated. The former tenant did printing and silk screening, he left behind pallets of inks and I saw that these were pigments, powdered. The brand on the can is Day-Glo Thermoset Pigments, I retrieved technical information from their website dayglo.com about their T/GT fluorescent pigments.

*Insoluble in water, hydrocarbons, and many common solvents

*Decomposition Point - Approximately 195oC (383oF)

*Average Particle Size - 4.0-5.0 microns (by volume)

*Can be used with coatings or inks made with nitrocellulose, polyvinyl chloride-acetate, polyamide, acrylic, chlorinated rubber

 

This is the formula I have been trying with no luck in producing a colored smoke

% Ingredient

27 Potassium Chlorate

18 Lactose

16 Magnesium Carbonate

3 Sodium Bicarbonate

36 Pigmented Dye Day-Glo T-15 Blaze Orange

 

I feel it's burning too hot and destroying my dye as you stated in your post. Any ideas on changing recipe values or ingredients?

I am new at this, and I have a mind that needs to understand the "why's" not just the "what's", if I understand why I can figure out what.

Thanks for your info, enjoyed reading.

[Arthur]

Firstly, not all dyes will make coloured smoke, some just burn up or char. Second, every part of a smoke canister is important, you need it to burn inside the canister so that smoke is expelled -once the flame gets into the air, atmospheric oxygen helps to destroy everything by over burning everything.

[ThisisMarks]

Thanks Arthur.

So I mixed up another batch last night and took your advice on enclosing vessel better. Used a 1' cardboard tube with plug hot glued on one end and plug with a 1/4' hole on the other with fuse going into hole. Burn was long and steady but no color and smoke seemed thin, not real thick.

Is there anything I can try to reduce the temperature of the burn and does anyone have tech data on dyes that do produce the bright smoke? 383 deg decomposition point seems very low for something on fire.

Thanks

[dave321]

sod bicarbonate is a more aggresive coolant than magnesium carbonate,

sod bicarbonate will give a slower burning composition compared with magnesium carbonate

 

the reformulated military yellow smoke is as follows

yellow 33 37

pot chlorate 34.5

sucrose 21.5

mag carbonate.4H₂0 5.5

stearic acid 1

fumed silica 0.5

Edited March 29, 2019 by dave321