Choosing the Flux

The Bismuth Trials: A Catalogue of Suffering

Most potters are familiar with metal oxides — those colourful powders we mix into glazes with the vague hope that something magical will happen in the kiln. Oxides are the ceramic workhorses: stable, solid, and stubborn. They don’t dissolve; they suspend. And unless fired into a glaze melt, they mostly stay put.

But when you're building a gold lustre from scratch — something that needs to behave like paint before firing — solubility becomes king. You need metal compounds that will dissolve into organic solvents or at least form a fine, stable film. That’s why I turned away from bismuth trioxide early on. As useful as it might be in a high-fire glaze, it’s practically inert in the solvent systems used for lustre work.

Instead, I focused on bismuth salts — compounds where the bismuth is paired with reactive anions like nitrates, acetates, or carboxylates. These salts have a much better chance of reacting, dissolving, or forming soluble complexes in organic mediums — which is exactly what I needed.

Like some tragic chemist version of Goldilocks, I went through every bismuth compound I could buy or make, trying to discover one that was just right.

The two common bismuth starting points are bismuth trioxide and bismuth subnitrate, as these are readily available off the shelf. I started with bismuth subnitrate, a classic in historical resinate lustres. It shows up in 19th-century recipes alongside pine resin, gold chloride, and a general lack of health and safety. It seemed promising — slightly soluble in acids, white and fluffy, and historically used as the main flux.

Getting it to dissolve in any modern organic vehicle proved impossible, however. The fine powder simply refused to budge in the solvents I was using — eucalyptus, toluene, isoamyl acetate, you name it. At this stage, I was still chasing clean, measurable solubility and didn’t yet understand why the Victorians had succeeded where I was failing.

Next: Bismuth Nitrate – The Flaky Barsteward

After failing to tame the subnitrate, I moved on to bismuth nitrate — a tantalisingly “soluble” form of bismuth, at least on paper.

The process seemed simple enough: dissolve bismuth subnitrate or bismuth trioxide in concentrated nitric acid. The reaction proceeded smoothly — the solid vanished into solution just as expected. In theory, I’d simply evaporate off the excess acid and be left with a fine, white, water-soluble powder.

But reality had other plans.

What I actually got was a white solid that seemed to dissolve in water… until it didn’t. It played a frustrating game of “now you see it, now you don’t.” Even freshly prepared bismuth nitrate began to lose its solubility after just a few days.

That’s when I re-learned an old word: hydrolysis.

Bismuth nitrate, it turns out, is highly prone to it. It reacts with even trace amounts of moisture in the air, converting into bismuth oxynitrate, an insoluble compound. This transformation happens quickly and unpredictably, making it nearly impossible to work with in any system that involves water or even atmospheric humidity. Since avoiding all sources of water in my lustre vehicle was essentially impossible, bismuth nitrate joined the growing list of promising but ultimately unusable dead ends.

And Still We Persist

Anyway, onwards and downwards — we’re not beaten yet. There were still plenty of bismuth options to explore.

Next on the list: bismuth acetate and bismuth citrate. Both are metal carboxylates, and given that the pine resin experiment had shown some promise with metal soaps, I figured it was worth revisiting this family of compounds.

As Nicola’s saint-like patience began to fray ever so slightly, I set about placing yet another chemical order — this time for glacial acetic acid and pure citric acid to synthesise these two hopefuls.

With ever-growing skills and the same eternal optimism, I prepared both salts with relative ease. The synthesis went smoothly: bismuth oxide or bismuth subnitrate reacting with the respective acids under gentle heat.

Then came the solubility tests with my suite of organic solvents. But as experience had already taught me, optimism alone doesn’t get results. And sure enough, both bismuth acetate and bismuth citrate showed poor solubility in my glaze vehicle — ruling them out for any practical use in this system.

Revisiting the Ancients

Somewhere between scrubbing glassware and swearing at yet another insoluble compound, I found myself pondering a very simple question: why did the old resinate lustres work at all?

Those 19th-century recipes were gloriously vague — “a spoon of this, a dram of that, boil until it smells odd” — but they worked. They used bismuth subnitrate, pine resin, gold chloride, and a healthy disregard for respiratory safety. Yet somehow, the mixture stayed paintable, stable, and fired to a lustrous finish.

That got me thinking.

Pine resin isn’t just sticky tree gloop — it’s a complex cocktail of abietic acid and related carboxylic acids. And when heated with a metal salt like bismuth subnitrate, it can form metal carboxylates — specifically, bismuth resinate. This was the missing link. Perhaps the secret wasn’t in the bismuth salt itself, but in what it became after reacting with the resin.

That insight sent me off in a new direction, chasing down the elusive family of bismuth carboxylates — compounds that might finally offer solubility, stability, and compatibility with the delicate alchemy of my gold lustre glaze.

Bismuth Carboxylate: Hope on the Horizon?

It made sense to explore bismuth carboxylate as a stand-alone product, so I set about making a variant of it.

The first process began with the saponification of pine resin — a straightforward reaction where pine resin dust is combined with sodium hydroxide to produce a resin soap. This soap was then diluted in water, filtered, and bismuth subnitrate dropped in. This immediately yielded a lovely looking light-coloured bismuth resinate.

Hopes were high — but soon dashed when, instead of dissolving in my solvents, it just sat at the bottom like a dead fish. Oh well. Move on.

This soapy mixture does have a well-established history in the production of metal soaps, which are often more soluble in organic solvents, so I’ve kept a dried batch on hand for future experimentation.

While the synthesis of bismuth carboxylate itself went smoothly, its solubility in my preferred solvent system was disappointing. Ultimately, it proved to be another dead end — at least via this particular route.

The Breaking Point and a Flicker of Hope

This was the turning point where things started to get a bit sticky. I began looking into more exotic bismuth complexes, and progress noticeably slowed. The growing cost and time investment of this obsession finally hit home, forcing me to take a step back and ask myself if this was all really worth it.

But whether it was the sting of failure fading or the lure of a breakthrough, eventually I found myself back at the laptop, trawling through obscure chemical literature.

That’s when I stumbled across another set of promising carboxylates: bismuth 2-ethylhexanoate (also known as bismuth octoate) and bismuth octanoate.

Now this was exciting. Bismuth octoate actually had documented use in modern gold lustre formulations, making it a highly relevant candidate.

The catch? It requires 2-ethylhexanoic acid, a branched-chain carboxylic acid widely used in the plastics industry. On paper, it seemed easy to source — but unless you’re buying by the shipping container, it turns out to be surprisingly hard to get in small quantities.

That’s when my attention turned to bismuth octanoate, made using caprylic acid (octanoic acid), a straight-chain fatty acid easily derived from purified coconut oil and much more accessible in small-scale quantities.

And I mean, what’s the difference between a branched chain and a straight chain? They’re practically twins, right?

Once again, the order book came out, and one more chemical was added to my ever-growing collection.