Choosing the Flux

The Bismuth Trials: A Catalogue of Suffering

When we fire a lustre glaze, we usually don’t reach the temperatures needed to melt, or even soften the base glaze on the piece. That means whatever we apply won’t really bond.

Enter the flux. By adding a low-melting flux, we can locally boost the surface melt just enough for the lustre to fuse to the glaze, creating a far more permanent finish. Without it, the coating would simply wipe away like it was never there.

For low-temperature lustres, the main flux candidates are boron, bismuth, or lead. Lead was out from the start, I like my work hazardous in spirit, not in handling. Boron got a short trial, but as a glass former in its own right, it came with too much baggage. That left me with bismuth and that’s where this particular chapter begins. 

Finding a Soluble Bismuth Compound for Lustre Glazes

Most potters know metal oxides, those dry, dusty colourants we stir into glazes that somehow conjure magic in the kiln. They’re reliable: stable, solid, and hopelessly insoluble. They stay exactly where you put them, and unless a melt drags them along, they’re not going anywhere.

But when you’re building a gold lustre glaze from scratch, oxides simply don’t cut it. In this game, solubility is king. After all, a fired lustre layer is only a few atoms thick. At that scale, your ingredients need to mingle at the atomic level. Grind an oxide as fine as you like, compared to that, the particles are still mountains on a salt flat.

What you really need is a metal compound that:

• Dissolves well in solvents

• Forms a brushable, even film

• Mixes with the other components at an atomic level

• Fires cleanly without speckling or clumps

So began the long, painful journey of finding a bismuth compound that would behave or at least misbehave in the right direction.

Why Not Just Use Bismuth Trioxide?

Because it’s useless here.

• Doesn’t dissolve in any solvent

• Doesn’t react in cold organic systems

• Akin to stirring gravel into treacle

It’s great in high-fire glazes, but useless for low-temp lustres.

The Early Dead Ends

1. Bismuth Subnitrate

• Historically used in old resinate lustre recipes

• Slightly soluble in acids

• Completely insoluble in modern organic solvents

• Refused to dissolve in eucalyptus, toluene, isoamyl acetate, or anything else I threw at it

• Result: Useless in this system

2. Bismuth Nitrate

• Made by dissolving subnitrate or trioxide in nitric acid

• Initially dissolved beautifully, forming a clear solution

• But hydrolysed on contact with air or water turning into a dense fluffy mess

• Converted to bismuth oxynitrate, an insoluble powder

• Result: Unstable, unpredictable, ultimately unusable

Attempts with Metal Carboxylates

Bismuth Acetate and Bismuth Citrate

• Easy to synthesise

• Made by reacting oxide or subnitrate with glacial acetic or citric acid

• Smelled like essence of chip shop and lemon drops

• Still wouldn’t dissolve in the vehicle

• Result: Straight into the “nope” pile

Then came Bimuth Carbonate & Bismuth Hydroxide but after these two I ran out of idea's from the internet so it was time to reflect

Historical Resinate Recipes Revisited

Old 19th-century recipes used pine resin, bismuth subnitrate, and gold chloride. Somehow, it worked, suggesting the bismuth was reacting into a soluble form. So armed with this my focus went onto metal soaps, mentioned in many books as "the wet" method of producing lustres

Bismuth Resinate via Pine Resin Soap

• Saponified pine resin with Sodium Hydroxide

• Filtered and added bismuth subnitrate

• Formed a creamy white precipitate bismuth resinate

• Smelled amazing

• I had high hopes but it sat at the bottom of the beaker like a dead goldfish

It never dissolved in any of my solvents. It’s now proudly stored in a jar labelled "Hope dwindling"." I’m convinced it’s one smart tweak away from brilliance, I'm just waiting on the "smart" bit if any chemists would like to chime in here?

Breakthrough: Bismuth 2-Ethylhexanoate

(Or Bismuth Octoate for an easier name)

This was the turning point.

• Documented in modern lustre patents and SDSs and after all of the above, I can see why this is the go to form of bismuth

• Oil-soluble, mixes well with resins, forms stable films

The only problem: making it.

The Synthesis Process

If you Google bismuth octoate, you'll find a flood of suppliers offering all kinds of industrial mixes, used in plastics, foams, polymers, and more. It’s a common chemical in industry, but here’s the catch: getting a small amount that doesn’t require lab accreditation? Nearly impossible. Sure, if I wanted to buy it by the 55kg drum, easy peasy. But by the vial? Nope. So, I thought, “How hard can it be to make this myself?”

Famous last words.

The process dragged on for months, frustration, late nights, and enough failed batches to fill a small gallery of shame. I began with 500g of raw 2-ethylhexanoic acid and high hopes. By the time I finally cracked it, I had maybe 30g left, a haunted look in my eyes, and a notebook that looked as old as me.

When I first set out on this gold lustre adventure, I was all footwork: books, SDS sheets, patents, scribbled notes in the margins. Calculations done on clunky online tools that felt like they were coded in the ‘90s. AI? Back then, it was just a novelty I used to make rediculous stories about my kids farting.

Then one day, I asked it the most boring chemistry question imaginable:

“How much ethylhexanoic acid do I need to react with 2 g of bismuth trioxide?”

Two seconds later, it not only gave me the exact answer, it offered to “dive into some clean processes for making bismuth octoate.”

My jaw hit the desk. Could this be my dream lab assistant? A sidekick who wouldn’t complain about the smell? I said yes. AI rolled up its digital sleeves and handed me a step-by-step process, complete with warnings and chemical insights.

It was glorious. It was intoxicating. And it was my undoing.

Blinded by the shiny promise of “instant expertise,” I abandoned the careful path I’d been carving with my own research. Big mistake. Looking back, I was a whisker away from finding a clean synthesis on my own. If I’d trusted my gut, I could have saved months, lot of materials and late night sobs.

Now, AI and I have an understanding: it’s my calculator, not my co-pilot. Numbers? Great. General info? Sure. But the thinking, the decisions, the chemical gut-feel that part stays human.

And so, after many low-yield disasters, murky extracts, and more cleaning than any sane person would tolerate, I finally have it: a clean, repeatable process for producing bismuth octoate. No mystery, no mess, no drama. Just the straightforward method I’d been chasing from the start. It doesnt required high end lab equipment or a PHD. Proof, if ever it were needed, that sometimes the old formula still works: patience + experience + good science = gold.

Moral of the story: trust your brain, use the calculator, and never, ever let a chatbot convince you it’s the captain of your ship.

Ingredients:

• Bismuth subnitrate

• Glacial acetic acid

• 2-ethylhexanoic acid

• Toluene (for dissolving final product)

• Watch Glass

• 100ml Borocilicate beaker

• A magnetic stirrer hotplate will really help you out here 

• Fume hood is essential

Method:

• Dissolve 4g of bismuth subnitrate in 25ml pure acetic acid, between 80-110 degrees (the higher the temp, the quicker this step)

• When the solution is completely clear with no cloudyness reduce the temperture to 80-90 degrees stirring continuously

• Slowly add 7.5g of 2-ethylhexanoic acid, drop by drop, dont let the temperature drop during this. This will give 25% excess of 2-ethylhexanioc acid, Ive ran this synthesis so may times weighing starting and ending products and this is about right to account for evaporative losses over this long running synthesis.

• Pop a watch glass over the beaker and continue stirring for a couple of hours.

• The solution should now have took on a pale yellow tinge, slide the watch glass off a little, just enough for a little steam to escape and continue.

• Maintain temperatures between 80-90 degrees and allow the acetic acid to evaoprate off really slowly, probably overnight.

• If the solution clouds up your evaporating the mix too agressively, add an ml or two of fresh acetic, bump the temperature to 100 for a minute or two, when the solution clarifies again back down the temperature to 80 degrees and start evaporating again but less agressively

• Eventually you will be blessed with a thick viscouse pale yellow oil, use your nose, if you still detect acetic acid, continue cooking. The end product should smell a little fatty with no vinigar/acrid smells

• Generally I dont just flick the off button, i will decrease the temperature incrementally over and hour or two

• When your happy the reaction is complete (No acid smell, no steam) slowly add 4-5 times the weight with toluene, warm back up and filter. Then evaporate down to your desired strength, i aim for 60% octoate to 40% toluene. If you attempt to filter at this ratio the octoate with change phase and block the filter

Key Observations:

• Any clouding up during the evaporation process is bismuth acetate falling out of solution, temperature too low or evaporating too fast. Ensure your at least 80 degrees and evaporate slower.

• Water is your enemy in this reaction, if its not removed during the process, progress can stall or reverse, or worse still your bismuth hydrolyses and becomes a dead fish in the bottom of the beaker

• Stirring helps keep things smooth with even heating, a magnetic stirrer is pretty much essential

• The end product wont catch your nose or throat, just a fatty smell, any acrid/acid smell means the reaction isnt complete, or you've overcooked it

• During the evaporation stage i find this can easily run overnight, 8-9 hours, the whole process can take 12-24 hours, with the slower scale being easier to control

• A slight haze in the end product is normal, this can be filtered off using papers or 0.45um PTFE syringe filters and a glass syringe (Plastic and toluene do not mix well)

I’ll be honest, this process took many attempts, months of trial and error. It would’ve been far easier with proper kit. A reflux setup with a Dean-Stark trap would’ve allowed continuous water removal during the reaction, pushing it toward completion without frying the organics. Cleaner, safer, and involving significantly fewer creative swear words. But by that point, the thought of telling Nicola that yet another bit of lab equipment was required pushed me over the finish line using what I had to hand.

Antidote to Impatience

Best prepared slowly, with occasional swearing

Ingredients:

• 1 tsp Perspective (aged 50+ years)

• 2 drops of "Remember Why You Started"

• 1 handful of Humility (preferably ground from past failures)

• 3g Unexpected Progress (keep frozen)

• 0.5 ml Nicely Aged Sarcasm

• 1 slow breath per minute (15 mins minimum)

• Optional: 1 dram of single malt whisky

Method:

• Combine perspective and motivation in a well-ventilated space (ideally far from your bench)

• Add humility - expect fizzing

• Thaw and stir in progress while muttering vague prayers

• Begin slow breathing

• Stir in sarcasm to catalyse patience into dry humour

• Add whisky if the plan fails

Caution: May cause insight, or at least the ability to walk away and try again another day.

The Final Product

After months of suffering, the final product is:

• A 60% bismuth octoate solution in toluene

• Fully compatible with PMIB and dodecanethiol-based golds

• Filtered to 0.45 µm

• Stable in storage (glass amber bottle recommended)

• Forms a smooth, even film when applied

No specks. No sludge. No gels. Just a beautiful flux that works.

Notes on Use

• Can be stored as a concentrated viscose solution

• Easily diluted further with toluene

• 1-5% in final glaze recipes appears ideal

• Best to store in an amber dropper bottle, dilute to a known percentage, count how many drops from the bottle makes a gram of product. Use these numbers to work out exactly how much bismuth octoate is supplies from one drop

Closing Thoughts

If you’ve ever wondered how many ways bismuth can break your heart, this page is the answer. From oxides to nitrates to weird organic soaps, the journey has been long, messy, and surprisingly fragrant.

But the result? A clean, reproducible bismuth flux suitable for modern lustre glazes and a growing suspicion that, deep down, I enjoy this level of chemical masochism.