Entry tags:

chemistry

Experiments in patination of brass
I was recently given a pair of brass unicorn bookends that look like this:



As it happened I have another pair that look almost identical which I bought some years ago. I got to thinking it'd be cool to color one set of them dark and leave the other pair bright, kind of like white and black unicorns. Painting them is always an option but the real way to give a color to brass or bronze is to use a chemical treatment to apply a patina.

Hundreds of methods have been devised to create different colors on the surface of various metals and alloys, mostly copper-containing ones like brass and bronze. I've got access to a huge book of these recipes, The Colouring, Bronzing and Patination of Metals by Richard Hughes and Michael Rowe, and I've tried a few formulas from it, mostly for cast bronze and with varying success. All the methods given for imparting grey and black colors to cast brass call for immersing the object completely in a solution of various chemicals.

I was reluctant to try anything untested on the bookends directly, although even a horribly failed patination job could probably be scrubbed off with a mild abrasive. The real problem is that submerging both bookends in the smallest suitable container on hand (a small, rectangular garbage can) requires two gallons of liquid and therefore a ton of chemicals, depending on the recipe. A failure could mean wasting a lot of expensive stuff. To try out various recipes, therefore, I bought a small brass, or brass-looking, tchotchke from a thrift shop. I'm not sure what it was supposed to be for, but it had three identical pieces that I could detach, giving three little brass bits like these:



Did I have any luck? Not quite.



All the formulas that I tried from the Hughes and Rowe text contain sodium thiosulfate and a metal salt: a salt of copper(II) in formula A, a salt of iron(III) in formula B, and a salt of antimony(III) in formula C. In the first two recipes I guess that the thiosulfate is there to supply a tiny amount of sulfide by hydrolysis:

S₂O₃^2- + H₂O = S^2- + H⁺ + HSO₄^-

The extent of hydrolysis is very slight, but the precipitation of an insoluble metal sulfide drives it forward. For example, treating a solution of Pb(NO₃)₂ with excess Na₂S₂O₃ gives a clear solution, but boiling it hastens the decomposition of thiosulfate and precipitates black lead sulfide.

I further conjecture that the metal salt in the patination solution is reduced at the surface of the object being coated to a species that precipitates a highly insoluble, dark-colored sulfide. For formula A, that would be cuprous sulfide:

Zn + 2 Cu²⁺ + S₂O₃^2- + H₂O = Zn²⁺ + Cu₂S↓ + H⁺ + HSO₄^-

I'm making the assumption that the zinc in the brass is the active reductant. In formula B, it might be ferrous sulfide:

Zn + 2 Fe³⁺ + 2 S₂O₃^2- + 2 H₂O = Zn²⁺ + FeS↓ + 2 H⁺ + 2 HSO₄^-

Formula C couldn't work this way, however, because the antimony(III) will be reduced directly to the metal. While antimony is bright and metallic in compact form it could look black if it were formed in a layer of tiny particles, as in "platinum black", and this might be all the chemistry that's behind formula C. But then the presence of the thiosulfate is hard to explain.

Now to actual trials. I first determined that one of the brass bits could be entirely submerged in about 100 mL of liquid in a small glass jar that I had on hand. I could make 100 mL batches of various patination solutions quickly and without wasting too much of my limited stock of chemicals. I prepared the brass pieces by scrubbing them with acetone to remove grease and then with fine steel wool.

Formula A:
Sodium thiosulfate pentahydrate (lab grade): 6.0 g
Copper sulfate pentahydrate (commercial "blue vitriol"): 4.2 g
Potassium hydrogen tartrate (commercial "cream of tartar"): 2.2 g
100 mL deionized water

This was all mixed at once in a clean jar (I didn't feel like wasting a proper beaker or flask at the time) and heated in the microwave in 10-second pulses with stirring until the temperature reached 60° C. Not everything dissolved, yielding a pale green, muddy liquid. This was decanted while hot into the small jar containing the first of the brass bits.

What I should have seen, according to the text, is the brass turning first yellow-orange, then bright green. This bright green, when rubbed off, should reveal a black matte surface. Instead the brass bit turned mustard yellow. After drying without rinsing the yellow layer rubbed off (mostly) to give a dull but still brassy surface, nothing like what I wanted. An unexpected change happened, however, after I gave the piece a few minutes' bath in dilute (about 0.25 M) hydrochloric acid. The remaining yellow residue dissolved and the surface dullness disappeared, leaving the metal bright and copper-colored, not at all like the original brass as you can see in the third picture, which does however make the metal look a bit darker than it really is. The effect was intriguing but not what I wanted. I had thought of reusing the patination solution but this was clearly impossible. The spent solution was a yellow-brown sludge probably containing a lot of free sulfur and copper sulfides.

Formula B:
Sodium thiosulfate pentahydrate (lab grade): 6.25 g
Ferric ammonium sulfate dodecahydrate (homemade): 6.0 g
perhaps 2 to 3 mL of 5% acetic acid
DI water to 110 mL

The original recipe actually calls for 5.0 g of ferric nitrate to each 100 mL of water, but I had none of this nor any easy way to make any since exhausting my supply of nitric acid. Some while ago, however, I had prepared a small quantity of ferric ammonium sulfate by oxidation of ferrous sulfate with ammonium persulfate. It had crystallized at the time into beautiful lavender crystals but, when I retrieved the bottle, they had effloresced into a white powder. In any case, on the hypothesis that the nitrate took no part in the reaction, I thought I could replace 5.0 g of ferric nitrate nonahydrate (0.0124 mol) with 6.0 g of ferric alum (0.0124 mol) without trouble.

In efflorescing, though, the ferric ammonium sulfate had also become partly insoluble. I risked adding a little acetic acid to the water to help it along, but even after this and many minutes of shaking there was still a white residue. The sodium thiosulfate was dissolved in a separate portion of water, then the two solutions mixed and brought up to about 110 mL with water. The mixture was at first dark brown, almost black,, but upon heating to 60° C with short bursts in the microwave as above it lightened considerably.

The text says that brass immersed in the warm solution turns black within sixty seconds. I saw nothing of the kind. Even after twenty minutes the brass bit was only somewhat darker and nowhere near black. I decided to let the reaction continue, though. After several hours the brass object had significantly darkened and the patination solution had mostly lost its brownish color and acquired the bluish tinge of copper(II). The next day the solution was obviously completely spent.

The brass bit, when washed and dried, was dark grey and somewhat dull. When finished with wax, as one is supposed to do to, the patina would no doubt be almost black, which is what I wanted. Unfortunately the layer was not even. It's not as obvious in the photo as it is to the eye, but even in the picture you can see patches that are more reddish than grey. The surface was perhaps not clean enough at the start or was of uneven composition. Maybe it wasn't even proper brass. Or perhaps there was not enough room around the brass piece in the small jar I was using for testing. You're supposed to leave as much space around the object as possible, but I was trying to keep the amount of patination solution kept to the absolute minimum needed for testing. Perhaps messing with the recipe, adding the acetic acid for example, prevented it from behaving. All the same, it gave the best result of the three experiments.

Formula C:
Sodium thiosulfate pentahydrate: 30. g
Potassium antimonyl tartrate hemihydrate (homemade "tartar emetic"): 1.5 g
Ammonium chloride (lab grade): 2.0 g
100 mL deionized water

This recipe is stated to produce a grey rather than a black surface but I figured it might be dark enough to pass for black after waxing. In any case I was curious to try a recipe calling for a more exotic starting material. I had some months before prepared my own tartar emetic from cream of tartar and antimony trioxide for experiments in permanganate titration and I still had a fair amount on hand. It had effloresced slightly but not horribly.

The mixture began to turn opaque and yellowish almost immediately. At first I thought this was from deposition of sulfur, a product of the decomposition of thiosulfate, but the color deepened on warming and it was clear that antimony(III) sulfide was separating. I couldn't see how this was good, because it meant that the active ingredient was crashing out of solution in a colorful but useless form. All the same, I poured the 60° C solution over the third brass bit and hoped for the best. The recipe called for a twenty minute immersion. After this time, the surface did not look much changed, so I let the solution stand as I did with formula B. After several hours the Sb₂S₃ had mostly settled out of solution. The brass piece had turned steel-grey but only in patches, so I replaced it to stew in what was left of the patination solution overnight, not optimistic of any change. If there was any, it was slight.

The photo does a bad job of reproducing the actual appearance, making it look much darker than it really is. For the most part the piece has much the same dark, metallic, coppery appearance that I got with formula A, except with greyer patches resembling polished hematite, only not quite so dark. Not a horrible effect but, again, not what I wanted. I'm still really puzzled as to how this method is supposed to work. I've confirmed in multiple sources that sodium thiosulfate precipitates antimony(III) sulfide with extreme ease. How could this be of any possible use in patination?

There are many other recipes that I could try. Some are also thiosulfate + metal salt recipes, but there are some different ones: two using potassium permanganate with metal salts; a couple using copper(II) carbonate and ammonia; several using even more exotic metal compounds that I'm pretty sure I'd have no chance of obtaining, like lead dioxide and arsenic trioxide. There's still room for experimentation. Who knows, maybe I'll come up with something on my own.

Or maybe I should just paint the damn things.