AP blue formula

[korkanin]

Hi, i decided to make the AP blue firework stars formula from this video: 

 

Its: 50% NH4ClO4

       20% CuO

        15% Shellac

         15% Ba(NO3)2

And since i never made AP stars, im asking you guys, if there is some difference from for example NaNO3, KNO3... Like the safety, or different primes? Or some compositions, which i cant roll the Ap into (chrysantemium to blue AP for exapmle)

 

Thanks

[AustralianPyromaniac]

Must be kept away from chlorates. NH4ClO4 + KClO3 reacts to form NH4ClO3, which is highly explosive. I am no expert on AP safety, just chiming in with one thing I know. 

But I have been very interested in this formula and would love to see your results. I have seen this video and am a bit sceptical about his claims. So please take some quality videos and show us what you find

[Zumber]

1 hour ago, korkanin said:

Hi, i decided to make the AP blue firework stars formula from this video: 

 

Its: 50% NH4ClO4

       20% CuO

        15% Shellac

         15% Ba(NO3)2

And since i never made AP stars, im asking you guys, if there is some difference from for example NaNO3, KNO3... Like the safety, or different primes? Or some compositions, which i cant roll the Ap into (chrysantemium to blue AP for exapmle)

 

Thanks

one thing to say is you cannot directly use a traditional black powder prime (potassium nitrate-based) on an ammonium perchlorate (AP) based star because a chemical reaction forms highly hygroscopic (water-absorbing) ammonium nitrate, rendering the star unignitable.

You need different prime for this AP stars which dont have KNO3 in it.

 

[korkanin]

@Zumber

 

, like  perchlorate prime?

 

Edited September 18 by korkanin

[Carbon796]

No direct contact with kno3, as stated above. Unless you use a boundary layer prime. Preferably also with a non-aqueous binder system.

Also NO aluminum fuel/spark metals, with AP / copper formulas. They will heat up with water contact/moisture.

[Carbon796]

Yes

[korkanin]

I have found this formula, its Shimizu igniter B. 

NaNO3: 75%

C:15%

S:5%

Dextrin: 5%

Do you guys think it is ok for this purpose? I mean, it still contains the NO3...

[Zumber]

It is designed to prime AP stars.

Yes you can use this prime but consider humidity,it should be below 60%.

[PyroGnome]

On 9/18/2025 at 4:31 PM, Carbon796 said:

No direct contact with kno3, as stated above. Unless you use a boundary layer prime. Preferably also with a non-aqueous binder system.

Also NO aluminum fuel/spark metals, with AP / copper formulas. They will heat up with water contact/moisture.

You'll need multiple, fairly thick boundary layers in practice to avoid the risk of things leaching between layers.  I think Shimizu mentioned that this doesn't protect things indefinitely, either, but he was using NC binding and nitrocellulose isn't a fantastic moisture barrier. 

Firstly if you're not trying to mess with strobes I'd just avoid AP entirely...  it's nearly 3x the cost of potassium perchlorate and the safety info is lesser known.  I could write an essay on the subject at this point (after digging through everything I could find for about a year), and plan to eventually, but the main point is that it's far more dangerous than commonly thought.  One of the worst areas is the effect of certain metal salts (copper in particular in the realm of pyro) on sensitizing it to impact detonation, especially with pretty much any fuel added, unless the entire composition is pressed to fairly high pressure.  I've also read up on blues / blue strobes more than I like to admit as well and aside from strobes where there's little choice in the matter (though even in that realm I think there are alternatives) the downsides of dealing with AP don't really outweigh the questionable benefits... low brightness is the worst of the practical issues unless you're using an aluminum fueled blue flare formula.

Aside from that I don't know about this formula... if you want a blue that's shifted more towards the upper end of <500nm instead of closer to 450nm you don't need barium nitrate.  I have some questions about the logic in general (I didn't watch the whole video because it's not worth the effort, but skipped to the part where he's babbling about paris green) since it seems to be related to some assumption that arsenic's primary emission spectrum is green with calomel and that was the reason for using barium in the other formula, but NIST is telling me that As(II) is almost exclusively blue between 419-454nm in the visible spectrum (https://pml.nist.gov/PhysRefData/Handbook/Tables/arsenictable2.htm) with a single weaker line up in red-orange at 617nm...  As(I) is irrelevant since it seems to only emit in UV-B and IR.  Heck, there's a flame test picture of it at https://en.wikipedia.org/wiki/Flame_test and it matches with that range (sort of UV-purplish)...   I'll ignore the fact that lighting loose piles of blue compositions on a spoon is possibly the worst possible way to test them.

They apparently missed the fact that NH4Cl and copper ammonium chloride are contributing more chlorine to the first formula than the one with calomel.  Mercury (I) emits primarily from 398-436nm along with one spike at 546nm, along with tons of UV.  Its function as a chlorine donor was probably less significant than the blue emission.  The original formulas used barium chlorate, not barium nitrate, which had the effect of lowering combustion temperature further...  and having some green emission was probably better than broadband emission from potassium so balancing it away somewhat couldn't have hurt things.  I might have missed some important part since I f'ing hate most youtubers but the part of the premise I caught was incorrect.   Another reason to push things towards green is to make them brighter to the human eye since we're terrible at picking up blues as they approach UV, but that sorta runs contradictory to the reasons for using ammonium perchlorate in the first place (which is to get a deeper blue color than possible otherwise.)  If you're doing that you can just use KClO4 and somewhat higher / more apparently bright dominant wavelength in that sky blue range they're going for.

Some visual composition comparisons (along with intensity, primary emission frequency, and luminance comparisons) are at  https://onlinelibrary.wiley.com/doi/full/10.1002/prep.202000114

As with the video itself, the BT.709 / sRGB colorspace is incapable of properly reproducing deep blues so the colors in the paper will not look right on your screen no matter what screen you're viewing it on.  The closest you can get to reproducing them correctly involves a fully calibrated color pipeline and use of BT.2020 (HDR10) throughout, then playing the whole mess back on one of the 3-laser RGB projectors than can cover that entire space... and even those can't display blues below 454nm since that's the frequency of the blue laser.  Expensive pigment printers can likely reproduce the colors more accurately from a high bit depth digital photo.

Note how much lower the brightness of the AP comps is than the potassium perchlorate comps, and the microstar perchlorate comp (#8) still beats all of the AP formulas in color purity while having a higher luminance than the best of them, which was #1, an unstable formula using copper metal, and they made #8 into the same 10g pellets as everything else when it was intended for star cores / microstars which they admitted probably negatively affected performance.

In practice the AP + Cu Benzoate formula (#4) is probably the best to use if you're dead set on AP as it won't go unstable on you (though it can detonate like pretty much all AP blues) and as a core for charcoal streamer stars it's probably bright enough, you just need to use multiple layers of acetone / alcohol bound layers between it and KNO3.  Having no metals means at worst you'll get stars that turn to sludge eventually, it won't auto-ignite like a metal fueled star would.  You could make it somewhat better by binding the core itself with PVC or Chlorowax so that binding system would also exclude moisture.  Copper benzoate can be made from copper sulfate / sodium benzoate IIRC or bought from FWCB...  but if you're just making cores for streamer stars, why not use the microstar comp instead and get better color?

As an alternative to traditional charcoal streamers, for an AP color core, you might want to look into potassium perchlorate / lampblack / charcoal streamer comps.  I know I've seen a few floating around and they should produce a suitably subtle effect to replace the standard KNO3 versions and you won't need to screw with a bunch of intermediate layers between the two and the layer of the streamer should be thick enough that you won't need to worry about just using BP as an outer prime. 
 

 

 

[AustralianPyromaniac]

I want to stay by saying this is an amazing reply and a really great opener into this discussion. I have thought about these ideas a lot, and also done my own tests using my potassium chlorate/ CuCO3 formula with added BaCO3. I do not have easy or cheap access to AP.

COMMENTS ON THE VIDEO

The video is mostly decent science, but it is definitely flawed. Scientifically, he makes a few claims that are dubious. He says all halogens turn copper blue, which is untrue. Although most of the other info is at least decent. But I know his other videos on red and green contain a lot more mistakes and incorrect info. 

Technically, his use of powders burning on a spoon is very inaccurate, and the fact that so many of the formulas look so good leads me to believe there is some heavy filtering on the colours. Loose star powder would not look that good, especially filmed up close in that way.

We do, however, have to admit one thing, which has been said before this guy by many authors. Blue stars with asenic look better than those without it. So the arsenic must be doing something, and if we understand it, can we recreate it.  And this point he makes is very valid. 

COMMENTS ON MY OWN TESTS

Firstly, we have to ask, assuming this colour mixing theory is correct as to what the asenic is doing, what wavelength of light is perceived as the most true blue by the human eye? We use this to mix to that colour. Sources from the laser community suggest the answer is 460nm. 

I spoke to Drake Anthony who is a member here and runs the YouTube channel Styropyro. His laser experience means he has actually seen pure 450nm, 460nm, 490nm etc light, and which looked the most/ richest blue - so I sent a message through to him. He confirmed what reports seem to suggest, 440nm blue is a blue/ violet, 460 blue is "true blue" and 480 is a sky blue. Looking at the table for CuCl emission in F.A.S.T. it is clear that CuCl alone produces blue with peaks clustered around 445/450 and only a small amount of light at 460nm. Also note here that the peak sensitivity of the blue cone is 430nm, which does not align with this true blue wavelength. 

I did some calculations, which showed the correct molar ratio of Ba to Cu is 9:1, to shift the perceived colour of a star from the dimly perceived 445nm of CuCl to the true blue 460nm. This assumes scotopic (night-adjusted) viewing conditions and takes into account the intensity of light perception by the eye. In my formula, this corresponded to a ratio 11/2 CuCO3/BaCO3 (w/w). I also made several other batches either side of this ratio. The formula in his video is 81% copper 19% barium, which corresponds to a lighter sky blue/ aqua colour in my tests. 

RESULTS

The formula with barium looks different. It is brighter, and maybe you can argue looks more washed out. But really it is not washed out, it is just brighter, like when you turn up a blue light, the eye perceives the colour more as white and less as blue. But the actual base colour is not less blue. As more barium is added, the colour begins to turn aqua and so deteriorates. Below 10% barium, the stars become progressively dimmer and back towards the pure CuCO3 formula. There is a change, but it is very slight to my eye. Some people may perceive the colour to be better, some worse. Most people I showed it to either could not tell the difference or slightly preferred the one with Ba, but were hard-pressed to tell the difference without careful examination. These tests were of primed cut stars thrown through the air at night by hand at a distance from the observer. As they would be perceived in a shell. For what it is worth, I, who has spent hours throwing these things through the air, can see almost no difference, and dont have much of a prefernce between them. 

Considering all the evidence I have, I think the mixture with Ba is slightly "bluer" than the formula with Cu alone, which appears "violet" by comparison. Mixtures above the 90/10 look brighter and less blue, and mixtures below 90/10 look more violet. I think 90/10 is a good goal, but the ratio is not super critical. I think 15/85 is also ok. The pure Cu formula also appears more "saturated" than the mixture, which is something Drake mentioned. A mixture of two colours will never appear as saturated as the pure spectral wavelength, even if the perceived colour is the same. My main conclusion is that there is very little difference at least in the potassium chlorate/ carbonate formula I tested with, but there is enough to call a slight improvement. 

[AustralianPyromaniac]

18 hours ago, PyroGnome said:

if you want a blue that's shifted more towards the upper end of <500nm instead of closer to 450nm you don't need barium nitrate

18 hours ago, PyroGnome said:

If you're doing that you can just use KClO4 and somewhat higher / more apparently bright dominant wavelength in that sky blue range they're going for.

 

These two comments seem to imply there is another species that you would suggest using to do the colour tuning. But what species do you suggest? We don't have many options.

18 hours ago, PyroGnome said:

As(II) is almost exclusively blue between 419-454nm in the visible spectrum (https://pml.nist.gov/PhysRefData/Handbook/Tables/arsenictable2.htm) with a single weaker line up in red-orange at 617nm...

From what I understand, in the heat of the flame, the species are always taken to be in the +1 oxidation state, i.e As(I) and coupled with a single chlorine atom to produce the emitter species AsCl/CuCl/SrCl/BaCl. We ignore the normal valence of the cations. The emission lines of these species with chlorine can be vastly different to the species without it. Barium, for example, is a broad-spectrum white without chlorine and a strong green with chlorine. I do not think the info given on that website is valid for our uses.

The flame test may be more valid, because the flame test is often done using a sample dissolved in or including HCl, but the test in that photo may be the pure metal. As you have said, arsenic without chlorine, for example, the gas AsH3, burns blue. That may be another theory as to how Paris Green works. At a time when chlorine donors were limited and poorly performing, the asenic burned blue even with little chlorine in the flame, supplementing the blue emission from copper in the form AsOH AND tuning the colour up 460nm in the form AsCl.

18 hours ago, PyroGnome said:

and having some green emission was probably better than broadband emission from potassium so balancing it away somewhat couldn't have hurt things.

Not a big fan of this idea. The potassium emission is somewhat problematic, I agree, but adding the ultra-high intensity BaCl emission in anything but the smallest amount seems like it would only make things worse. Broad-spectrum potassium emission, mostly falling in the UV, would be better. I am also unsure about the barium chlorate lowering the flame temp, but sceptical about that also. 

19 hours ago, PyroGnome said:

Another reason to push things towards green is to make them brighter to the human eye since we're terrible at picking up blues as they approach UV

For me, this is partially the idea, even though he does not mention it. Hard to say why we perceive 460nm as "true blue" but it must be a combination of conditioning and perceived brightness, seeing the 460nm blue does not line up with the peak sensitivity of the blue cone. 

19 hours ago, PyroGnome said:

Its function as a chlorine donor was probably less significant than the blue emission

Maybe a valid point, but it is used as a chlorine donor in other formulas, where its blue emission is not needed. Most chloride salts are stable well above flame temps; only a few - notably mercury chloride and ammonium chloride - can be decomposed by flame yielding Cl2/HCl, and these have been used as chlorine donors. So it would also be working as a chlorine donor. 

[korkanin]

@PyroGnome

I actually wanted to avoid using AP, since it is not so easy to find, and pretty dangerous. but i couldnt find a KCLO3/KCLO4 formula, which i liked enough, and also had chemicals which i could get. (I dont have any source of hexachlorbezneze, or dechlorane, which some formulas calls for). Than i saw this video, and really liked the AP formula, and didnt seep until i finded a way to make AP relatively cheap at home. But its definitly easier for me to use KCLO4, or KCLO3. So if u have some good formula without AP, i would be very grateful.

[AustralianPyromaniac]

7 hours ago, korkanin said:

@PyroGnome

I actually wanted to avoid using AP, since it is not so easy to find, and pretty dangerous. but i couldnt find a KCLO3/KCLO4 formula, which i liked enough, and also had chemicals which i could get. (I dont have any source of hexachlorbezneze, or dechlorane, which some formulas calls for). Than i saw this video, and really liked the AP formula, and didnt seep until i finded a way to make AP relatively cheap at home. But its definitly easier for me to use KCLO4, or KCLO3. So if u have some good formula without AP, i would be very grateful.

Have you tried out some formulas and been unhappy? Which ones did you try, and if so, what did you not like? 

What chlorine donors do you have access to? 

[korkanin]

I havent tryed any, bcs pyro chemicals (needed for blue) are pretty expensive in my coutry, and hard to get. (czechia). So im trying to pick some formula before i buy them, and design the shopping list depending on which formula i choose. I can buy parlon, pvc, (red gum), NH4CL, and thats it from chlorine donors i guess. Idk, if i will be able to buy them tho, cuz blasttec is closed currently, I hope that it will open soon...

I originnaly wanted to try making this formula: 65% KClO4, 10%S, 16% CuCl, 11%Parlon, 7% Red gum

Also considered this one: KClO4: 63%, CuO: 13%, Parlon: 14%, red gum: 6%.

I really cant pick one, because i have low budget, so if i choose the wrong one, my blue stars wont exist this year

[korkanin]

@AustralianPyromaniac

Wilbur's Winning Blue Pyrotechnic Formula
Potassium perchlorate	65%
Copper oxide	14%
Red gum	7%
Parlon	5%
Dextrin	5%
Dechlorane	4%

 

I have been thinking about this too, but i dont know where the heck to take dechlorane from. Can i just substitute it by parlon based on chlorine %?

[johnnypyro]

There are two broad categories of blue (or any colour for that matter) star: organic and metallic. For organic blues, the colourants I've tried are Copper Oxide and Copper Chloride. They both need good primes and are slow to burn. They also have low light output so need to be combined with other effects that are not too bright - C6/C8/TT are all good, but not Ti based starts. There is not much to choose between them in terms of colour and quality because of the low light output.

Metallic blues, like Veline's blue, are brighter but more washed out in their blue. They can compete when combined with brighter effects, but I found the blueness to be a bit disappointing.

[AustralianPyromaniac]

I use very simple chlorate formulas that I developed myself for OTC pyro. You can give either of these a go. You can either use PVC powder or PVC cement from the hardware store. The pressure-rated cement has more PVC in it. Use clear, not coloured cement if you can find it. Shellac is very easy to buy for woodworking, but if you can buy the ultra-fine hammer-milled powder, that is much better than flakes. Copper carbonate is used in pottery. The colour is very good - deep and rich. The PVC cement formula is a little better because it does not have dextrin to disturb the flame colour. To the PVC powder formula, you can remove the dextrin and add in equivalent parts of shellac + bind with alcohol. Else bind with water. 

OTC BLUE (PVC CEMENT) 

KClO3 - 68

SHELLAC - 9

COPPER CARBONATE (CuCO3) - 15

PRESSURE PVC CEMENT - 35

OTC BLUE (PVC POWDER) 

KClO3 - 68

SHELLAC - 6

COPPER CARBONATE (CuCO3) - 15

PVC POWDER - 8

DEXTRIN - 3

[AustralianPyromaniac]

On 9/23/2025 at 7:54 PM, korkanin said:

originnaly wanted to try making this formula: 65% KClO4, 10%S, 16% CuCl, 11%Parlon, 7% Red gum

Also considered this one: KClO4: 63%, CuO: 13%, Parlon: 14%, red gum: 6%.

Where does this formula come from? I have never seen it. CuCl is a very difficult-to-come-by chemical, and reverts to copper oxychloride over time and particularly when in contact with water. The parts do not add to 100, and it seems to be very fuel rich. Many things wrong. 

The second formula I like much more, would work well. See no issues. I think KClO3 blues are better. And CuO generally does not produce as nice a colour as other compounds; however, there is a lot of chlorine in that formula which should make it turn out fine. 

6 hours ago, johnnypyro said:

For organic blues, the colourants I've tried are Copper Oxide and Copper Chloride.

Do you mean copper carbonate?? And if you do mean copper chloride, do you mean CuCl2 or CuCl? 

[johnnypyro]

These are the blues I have tried myself.

 Pihko KP #2 – this is my go to composition. Not easy to light, but then none of them are. Needs good prime (I used monocapa) to ignite in the air.

KP 63 

Parlon 14 

Copper (II) Oxide 13 

Red Gum 10

(and dextrin + 0.04)

 

Shimizu Blue Star II – very similar in colour quality to Pihko KP #2, but a bit harder to ignite.

KP 66.5

Red Gum 9.9

CuO 13.4

Parlon 5.4

Dextrin 4.8

 

Pihko KP #3 – I found this very hard to ignite, so gave up after one batch

Potassium Perchlorate 65 

Copper(I) Chloride 16 

Parlon 11 

Sulfur 10 

Red Gum 7

  

Veline blue – bright, but not so blue.

KP 53

Red gum 9

Magnalium -200 mesh 6

Copper (II) oxide 14

Parlon 14

Dextrin 4

 

More info here:

https://www.amateurpyro.com/forums/topic/3047-best-kp-blues/

https://www.skylighter.com/blogs/how-to-make-fireworks/blue-stars-part-2