A novel means of magnesium protection via alloying with very small percentages of Calcium

[PyroGnome]

TL;DR - Micro-alloying magnesium metal with 0.15% calcium produces an extremely corrosion resistant material which retains all of the other properties of magnesium aside from its ability to reduce everything at normal temperatures and of course rapid corrosion.  The article is worth looking at for the nearly untouched cube of the stuff (aside from a black protective layer) after hanging it in a 3.5% NaCl solution for 180 days vs. two corrosion resistant commercial alloys available.  They also found a flaw in testing of standard resistant alloys that involves the corrosion accelerating between 7 and 21 days (while the standard tests estimate the yearly loss from corrosion by measuring loss of mass over a 1 week period in a salt solution) which is why they ran their tests for so long.  Well probably that, and to show off...  Anyway I'm posting this for anyone who likes to experiment with weird things since the material looks like it has good potential to replace regular Mg in fireworks, which I feel is under-used outside of certain strobe mixes because of the water sensitivity.  MgAl has plenty of uses, some of which are pretty amazing (like Winokur #18 & 20) and other areas where it just can't be substituted in currently known (to me) ways like dragon eggs but doesn't produce the same color depth in regular metal fueled colored star formulas IMO.  There are also all the oft-forgotten colored flash powders which can be very attractive in a timed salute shell but tend to require Mg to get any kind of color depth.  Magnesium stearate coating can probably be used in many cases (one paper I found suggested that it worked nearly as well as dichromate for protection against ammonium perchlorate, but it also slows down the burning rate by about 10%).

Article for those who don't want to read a long research paper:  https://www.chemistryworld.com/news/new-magnesium-alloy-shows-exceptional-corrosion-resistance/4013323.article

DOI:  https://doi.org/10.1039/D0MH01380C

The paper:  https://www.hereon.de/imperia/md/content/hzg/zentrale_einrichtungen/bibliothek/journals/deng_39415.pdf

(supplementary PDF is at the DOI link above)

They used 99.96% ingots of Mg, and 99.9% calcium shot for their test batches. United Nuclear sells calcium metal cheap (really any price is cheap enough when you only need 0.15% though), although I don't know the purity; the information they found on stabilization of Fe impurities indicates that a lower grade Mg can be used, though 1kg blocks of 99.95% aren't hard to find for cheap if you hunt around enough.  

Magnesium has the lowest melting point (650°C/1202°F) of the alkaline earth metals, and in the paper they melted the magnesium then heated to 760°C / 1400°F in a boron nitride coated stainless steel crucible...   I don't know if this is entirely necessary, although the crucible probably can't be silica based or the MgSi producing reduction could theoretically start;  somebody who has made MgAl at home could correct me on that;  and it would probably strip the passivation layer off of stainless and become contaminated with elements from it without boron nitride otherwise) to form the alloy.  It may have been a matter of keeping the contamination non-existent more than a necessity since they needed to test the specific percentages of calcium and they're working with a very small amount, or out of concern for the iron contamination since they hadn't yet done tests and found that calcium was able to make this a sorta non-issue. 

They used an odd argon / sulfur hexafluoride mixture in the furnace during melting to prevent oxidation.  I'm not quite sure why the mix, and usually SF6 isn't used for this type of application AFAIK.  Surprised high temp Mg doesn't start reducing SF6 honestly.  🙂  Nitrogen can't be used because solid calcium is still less dense than molten magnesium so it'll likely float, and will start reacting with nitrogen to form the nitride at 450°C if it's fairly pure.  This is is probably the worst side reaction you could end up with since it releases NH3 when wet which probably won't play well with the alloy in general.  Some other method of doing this is probably possible (can't aluminum be protected from air when melting with a low-melting low-density salt floating on top?).  Anyway they manually stirred for 5 minutes, then left it in for a further 30.  The manual stirring would be another difficult part to work out without proper equipment.  It's possible to use strong enough rotating rare earth magnets or sequentially energizing electromagnets to induce eddy currents in aluminum and stir the molten stuff without ever needing to have anything touch it, which is done on an industrial scale quite a bit these days from what I've read, but I don't know if this would incorporate the floating calcium well enough.  

After that they did regular castings of their 0.05% & 0.1% batches into boron nitride coated SS molds.  Interestingly they did direct-chill casting into a similar cylindrical mold for the 0.15% alloy, which involves chilling the outside of the mold with water at first then just directly immersing the whole thing to expose the still-molten core after a short period.  They didn't give any reason for this, I think it was just faster.  Pouring water into molten Mg doesn't sound like fun but hey. 

Aside from the NaCl test mentioned above they tested reaction with a stream of pure oxygen injected directly next to the material in water to keep the water oxygenated and found the oxygen reduction rate to be extremely low after a brief period where a passivation layer rapidly forms.  They also found that Fe and Si impurities which normally cause rapid corrosion in even very pure Mg are stabilized by the alloying, via formation of a CaMgSi alloy precipitate that stops the more detrimental FeSi from existing.  

The total melting point of the mixture is depressed by roughly a degree, with some partial melts occuring several degrees lower and the MgCa fully melting last, so there shouldn't be a huge change in pyrotechnic properties.  I wouldn't expect the small calcium content to have much effect on flame color especially since Mg isn't used for blue electric stars.  I kinda doubt it loses much of its ability as a reducing agent when it's actually on fire (important for the carbonate containing mixes), either, but this wasn't tested in the paper.  Whether it would stand up to something as nasty as ammonium perchlorate is the big unanswered question, but since the alloy doesn't seem to be commercially available yet someone will have to make it and find out.  I plan on attempting it after I read up on methods of getting it melted relatively cleanly without the need for any insane gas mixtures and a commercial furnace.
 

[Crazy Swede]

A magnesium alloy containing 0.15 % of calcium could never be used for green or blue flames (I know ordinary magnesium is not used either for blue flames!) because it would disturb the colour too much.

But it could definitely be of interest for the marine distress signal industry that only produces red and white flares. The problem is to have someone making the alloy, turning it to powder and selling it to us at a decent price.

[Mumbles]

I would venture a guess that AP corrosion would still be an issue as you alluded to.  Ammonium salts are very good at dissolving magnesium oxides and hydroxides surface passivation, which is the root cause.  I've been some investigations into dichromate treatment, and it seemed to indicate that it essentially intercalates into the passivation layer or oxidizes itself a new layer on the metal surface, and sort of cross links it together or incorporates itself making it stronger and less susceptible to attack.  As a fair warning, that's all based on memory, and might be figment of my imagination.  The surface is noticeably green however which is consistent with chromium (III) oxide.

I'm not so sure if green stars would be out.  Most barium compounds have a component proportion of calcium as an impurity.  The nicest barium nitrate I've ever used is from Barium and Chemicals. It has an upper spec of 0.05% calcium. If it were at that limit, it'd be contributing essentially just as much calcium into the overall star as this alloy of Mg. I don't have any data on typical calcium proportions lot over lot unfortunately.

[Crazy Swede]

That is probably correct Mumbles, I  overestimated the bad influence from calcium!

[PyroGnome]

Their reasoning for the microalloying seemed to be that for whatever reason CaO forms a secondary barrier / mixture with the MgO which is more resistant than either and the hydroxides don't really end up forming at all because the reductive potential is simultaneously lowered by the alloying.  It loves hovering in water just being cloudy like that.  The center alloy looks like something else is happening with the weird crystal formations though.

On the (di)chromate thing I finally dug out and posted shimizu's 3-component long term version of that coating (which produces a black-coated Mg with the black layer being uncharacterized) in another thread (the stearic acid thing).  He speculates on why any of it works in the first place to some degree in that.  I'll forgive him for not going out of his way to do another 8 years of leaving stars laying around or tracking down somebody to do SEM / mass spec / x-ray diffraction on processed material considering he was 82 years old when he published that article.  I'll be lucky if I can remember my own name when I'm 60.  😉

The shimizu one is the most solid and low additional effort, although the Mg Stearate was specifically for protection with AP and if the slightly lowered burning rate either isn't an issue or can be worked around it's a lot nicer than dichromate for stuff to be fired at a shoot the next day. 

From what I've read in the past (wish I could remember well) it's very difficult to remove much more calcium than that .05% max they listed from barium or strontium by standard means so you might be able to get it as Sigma analytical grade if you want to spend $500 on a 2" shell but the process to get rid of it became too complicated past that to make it worth doing since the results aren't affected much. 

It's only tangentially related but there's an ACS Omega paper out there from 2022 which was posted publically for once (ACS is notoriously bad about locking down single papers behind $50 purchase fees, and they don't participate in the open journal access program I have access to from making 500 sarcastic comments on error- I mean constructive edits to wikipedia) examining spark production from various rare earth metal alloys / eutectic mixtures.  The best one subjectively was a Yttrbium-Copper eutectic which produces a large blue-purple mixed flame which ejects a mixture of deep red, orange, yellow, green, and blue sparks with many of the green ones acting like small glitter spritzels.  One of the more unique things is that sparks will transition from orange to green branching then repeat that cycle sometimes.  Unfortunately this metal isn't a commercially produced item and costs a small fortune thanks to the yttrbium.  Another good one was powdered standard neodymium magnet alloy, which produces tons of chained multiple-branching orange sparks not quite as good as senko-hanabai but with  more repeating sparks.   That stuff is really brittle and can probably be powdered up enough with a hammer...  these days you can buy 100s of the lower grade magnets for next to nothing or just pull them out of random electronics junk like old phone wallet style cases.  The rest of the alloys are mostly either not usable due to high reactivity or beyond either price constraints or equipment most likely.  The SmCo5 they tested is another supermagnet alloy too, I think, but the only reliable place I know of to pull it from are C shaped plates that are mounted above and below the platter read/write assembly on hard drives.  near the corner of the case. 

Customization of the Appearance of Sparks with Binary Metal Alloys.pdf

[Crazy Swede]

Have you read the article "Corrosion Protection of Magnesium Without the Use of Chromates", published by Alenfelt in Pyrotechnica XVI in 1995?

There at least an attempt was made to describe the superficial layer formed from Shimizu's method.

The article also explains why nothing really compares to the chromate coating, since it produces a tight layer that also has a self-healing effect from trapped hexavalent chromium species. No other method can provide that feature!

[Mumbles]

On 10/13/2024 at 5:24 PM, PyroGnome said:

...

From what I've read in the past (wish I could remember well) it's very difficult to remove much more calcium than that .05% max they listed from barium or strontium by standard means so you might be able to get it as Sigma analytical grade if you want to spend $500 on a 2" shell but the process to get rid of it became too complicated past that to make it worth doing since the results aren't affected much. 

...

You might be thinking of my comment in this thread: 

It's probably not so much that it's especially difficult to remove past 0.05% calcium, just that it's not worth it for them. If the customers don't care or specifically need it, they're probably not going to bother. From working with a lot of chemical suppliers, many will work with you to meet whatever specifications you want within their abilities. If you had a specific need for <0.001wt% calcium, you could find it.