Kodak T-21

Four and a fifth litres of bath for half a gram of gold, made in three stages, ripened with silver chloride, left to stand overnight and then worked hot. It is the most elaborate formula in the toner library and every part of that elaboration is in service of one thing: getting gold to the print as gold(I) rather than gold(III).

⚠Read this before the formula

Level C, and the reagent that puts it there is not the gold. Potassium persulfate is a respiratory sensitiser with a 0.1 mg/m³ limit, and its page is explicit about the operation that fixes the classification: weighing the powder, for which the recognised control is engineered extraction. That page states the course’s position in terms — it offers no home procedure using potassium persulfate, and where a gold toner is wanted it points at GP-1 made with sodium thiocyanate, which does the job without a persulfate in it.

This page does not overturn that. It publishes T-21 as chemistry, because the formula is instructive and because the gold chemistry it teaches is the same chemistry the safer formulas use.

Stock Solution A — the thiosulfate bath the gold is reduced in

Ingredient	Quantity	Form the source specifies
Sodium thiosulfate pentahydrate	960 g	crystals
Potassium persulfate	120 g
Water	4000 mL, added	at 50 °C; Water at 50 °C, given as an amount added rather than as a make-up volume. Kodak states no final volume for this stock, so the strengths on this page are computed against the four litres named plus the 64 mL of ripener.
If the mixture does not turn milky, Kodak's instruction is to increase the temperature. The stock is then cooled to 27 °C before the ripener goes in, and Kodak repeats that both solutions must be cool when they are combined.

The silver ripener — the silver-chloride charge, added to Stock Solution A

Ingredient	Quantity	Form the source specifies
Silver nitrate	5 g	crystals
Sodium chloride	5 g
Water	64 mL, added	at 20 °C
Kodak heads this table with no name; the label here is descriptive. The silver nitrate must dissolve completely before the sodium chloride is added - Kodak footnotes it, exactly as it does in T-1a, whose ripener is the same pair of salts at the same 64 mL.

Stock Solution B — the gold

Ingredient	Quantity	Form the source specifies
Chloroauric acid	1 g	printed as "gold chloride"; the course's reagent for that name in the photographic literature is chloroauric acid, for the reasons on both gold pages
Water	250 mL, added	at 20 °C; A 0.4 per cent solution. The mirrored text prints this volume as "250 L"; see the provenance note for why the course reads it as millilitres.
Kodak warns that gold chloride is deliquescent and liquefies rapidly in a normal room atmosphere, and directs that it be kept in a tightly stoppered bottle in a dry atmosphere. Only half of this stock goes into the toner; the rest is the replenisher.

Mixed in this order — completing Stock Solution A

1. Start with the whole of it, cooled to 27 °C of Stock Solution A — The vessel that receives.
2. Then add the whole of it, including the precipitate of The silver ripener — Added slowly, while stirring. Both solutions must be cool when they are combined.

Dissolve the sodium thiosulfate in the water. While stirring vigorously, add the potassium persulfate. If the mixture does not turn milky, increase the temperature. Cool to 27 °C (80 °F). Then slowly add the following solution, including the precipitate, while stirring. Solutions must be cool when you combine them.

Kodak prints the ripener as part of Stock Solution A rather than as a solution in its own right; it is separated here because it has its own water, its own temperature and its own internal order of addition, and because the sheet's own sentence treats it as a solution that is added.

Mixed in this order — the working toner

1. Start with entire amount of Stock Solution A — Stock Solution A as completed above, ripener included. Kodak's own words for the quantity are "Entire amount".
2. Then add 125 mL of Stock Solution B — Added slowly, while stirring rapidly.

Stock Solution A, entire amount. While stirring rapidly, slowly add Stock Solution B, 125 mL. Let the mixed solution stand for about 8 hours before using it. During this time a yellow precipitate will form. Pour the clear solution into another container and discard the precipitate.

The eight hours are part of the formula and not a convenience. The bath is not a bath until the gold has been reduced and the yellow precipitate has come out of it, and the precipitate is discarded rather than stirred back in.

Purpose

To plate a warm-tone print with gold, evenly across the whole scale. Kodak’s description is precise on both counts: T-21 produces a pleasing range of brown tones with most warm-tone papers by “plating” the silver image with gold; it has little effect on cold-tone papers; and it tones the highlights and the shadows of the print at a uniform rate, which is what allows the print to be pulled when it reaches the hue you want.

That last property is the formula’s real selling point, and it is worth setting against T-26, Kodak’s other gold toner from the same sheet, of which Kodak says the opposite: toning occurs in the highlights first and the shadows last, so partial toning gives blue highlights and untoned shadows. Two gold baths, two different ligands, two entirely different behaviours across the tonal scale.

Recommended uses

Warm-tone papers. Kodak names them and names the exception: cold-tone papers barely respond.

Where the toning is to be stopped by eye at an intermediate point. The uniform rate is what makes that possible — a print pulled half way through is a print that is half toned everywhere, not one that is toned in the highlights only.

Where a bath is to be used for a run of prints. T-21 is replenished rather than discarded, and Kodak publishes the replenishment rate rather than a capacity.

Not for a single print. Bringing four litres to 43 °C, after eight hours’ standing, for one sheet, is the same objection that applies to T-1a and for the same reason.

When another formula is preferable

• In almost every case, GP-1, which the persulfate page names as the alternative by name: the same gold, a thiocyanate ligand instead of a thiosulfate one, mixed cold in a litre, used at 20 °C, and no persulfate anywhere in it.
• Where a blue tone is wanted rather than brown, T-26, which pairs the same gold with thiourea — and which is Level C for the thiourea instead.
• Where the print is a printing-out paper, a salted paper, an albumen or a Van Dyke, the gold-thiocyanate and gold-borax toners, which are made for those processes and are both mixed cold from two liquids.
• Where the aim is a brown print rather than a gold one, any of the sulfide routes, at a small fraction of the cost. Gold is the expensive way to change a colour and the cheap way to make a print last.

Mixing

Three solutions, two combinations, one overnight wait, and a decanting step.

1. Stock Solution A. Dissolve 960 g of sodium thiosulfate in 4 L of water at 50 °C. While stirring vigorously, add 120 g of potassium persulfate. If the mixture does not turn milky, increase the temperature — Kodak treats the milkiness as the confirmation that the reaction has happened. Cool to 27 °C.
2. The ripener. 5 g of silver nitrate in 64 mL of water at 20 °C, dissolved completely, then 5 g of sodium chloride. Add the whole of it, precipitate included, slowly, with stirring, to the cooled stock. Both solutions must be cool when they are combined, and Kodak says so twice.
3. Stock Solution B. 1 g of gold chloride in 250 mL of water at 20 °C — a 0.4 per cent solution. Keep it stoppered and dry; the salt is deliquescent.
4. The working toner. The entire amount of Stock A; then, stirring rapidly, 125 mL of Stock B added slowly.
5. Wait about eight hours. A yellow precipitate forms. Pour the clear solution off into another container and discard the precipitate.

🧮What four litres of bath actually carries

Stock A is 4000 mL of water plus 64 mL of ripener; the working toner adds 125 mL of Stock B, so the bath is about 4189 mL before allowing for the volume of the dissolved solids.

1 g × 125/250 = 0.5 g, in 4.19 L = 0.119 g per litre

Gold chloride

960 g ÷ 4.19 L = 229 g per litre

Sodium thiosulfate

120 g ÷ 4.19 L = 28.6 g per litre

Potassium persulfate

5 g ÷ 4.19 L = 1.19 g per litre of each

Silver nitrate and sodium chloride

Two of those numbers are worth carrying away. The gold at 0.119 g/L sits inside Reilly’s published range for a gold toning bath — 0.1 to 0.2 g of gold chloride per litre for matte salted paper — and within a fifth of GP-1’s 0.10 g/L on the same Kodak sheet. That agreement is what settles the reading of Stock Solution B’s volume; see the provenance note.

The thiosulfate at 229 g/L is the other. This is a fixing bath with gold in it — within five per cent of the hypo strength of F-5 and of plain hypo — which is why the silver ripener is in the formula, for exactly the reason it is in T-1a: a hot concentrated hypo bath with nothing to eat will eat the print.

Behaviour

It has to stand for eight hours and the reason is chemical. What the standing does is take the gold from gold(III) to gold(I) in the thiosulfate, and drop the yellow precipitate that comes out of the process. A bath used before it has stood is a bath in the wrong oxidation state.

It is worked hot, at 43 °C, and the temperature is maintained. A tray in a tempered water bath is Kodak’s published method, as it is for T-1a, and the temperature is six degrees below that bath’s.

Five to twenty minutes is a wide window and that is the point. Because the whole scale tones at a uniform rate, the print can be pulled anywhere in it.

Sediment forms in the tray, especially with several prints at once, and can leave a scum on the print. Kodak’s remedy is a wet sponge or wet cotton immediately after toning.

It is replenished, not discarded. Kodak publishes no capacity and no exhaustion test, only a rate: about 4 mL of Stock B after every fifty 8 × 10 prints when toning to a warm brown, with the amount depending on how many prints and how long each is toned.

Image characteristics

Brown, on warm-tone papers, over a range. Kodak does not name the ends of the range and this page does not invent them.

Nothing much, on cold-tone papers. That is the same limitation the selenium toners have and it comes from the same place: what the toner can do depends on the silver it finds.

Uniform across the scale, which is the formula’s distinguishing characteristic and its practical advantage.

Protection, whether or not the colour moves. The chloroauric acid page carries the argument with its sources: a printed-out or developed silver image is attacked by sulfur-bearing gases and by the sulfur-bearing amino acids in the gelatin itself, and gold coats the silver particles against it. Kodak’s own general claim on the same sheet is that all its toners protect the image whether or not they produce a colour shift.

The mechanism

The whole formula is an apparatus for reducing gold(III) to gold(I). The chloroauric acid page carries the argument and the equations, and this page does not repeat them; what follows is why this formula is shaped the way it is.

⚗Three silvers or one

Gold in the salt you buy is gold(III), and silver oxidises only to silver(I). So a plain acid gold bath spends three silver atoms for every gold atom deposited, which bleaches the print faster than it tones it — Reilly describes the result as flat, lifeless and reddish, and the equation is on the chloroauric acid page.

Every useful gold toner therefore works with gold(I), where the exchange is one for one. Gold(I) is not a stable state you can buy, so making a gold toner is mostly the business of reducing the gold(III) you can. Ware’s two routes are a sulfur ligand — thiosulfate, thiocyanate or thiourea — which both assists the reduction and stabilises the product, or a plain mild alkaline buffer, in which gold(III) slowly oxidises water instead over as much as twenty-four hours.

T-21 is the first route, with thiosulfate as the ligand, and the eight hours’ standing is that reduction happening. Kodak’s own three gold formulas on one sheet are one of each of the sulfur ligands: thiosulfate here, thiocyanate in GP-1, thiourea in T-26. The gold-borax toner is the other route entirely.

What the persulfate is for is the one thing this page cannot source. Kodak’s instruction is to add it while stirring vigorously and that the mixture should turn milky, and to raise the temperature if it does not. A milky thiosulfate solution is precipitated sulfur, so the persulfate is plainly oxidising some of the thiosulfate — but no source read for this course gives an equation for that reaction, or says what the resulting species contributes to the toning, and none is written here. Two readings are available and neither is adopted: that the persulfate is conditioning the thiosulfate into a species that reduces gold(III) more readily, or that the colloidal sulfur is itself part of the toning. Rule 1 forbids choosing between them without a source.

Why there is silver in a gold toner. For the same reason there is silver in T-1a: the bath is 229 g/L of hypo at 43 °C, which is a working fixing bath, and a fixing bath with no silver in it takes silver out of the print. The ripener gives it a charge of its own. Kodak uses the same two salts, at the same 64 mL, in both formulas — five grams here against four there.

Why the yellow precipitate is discarded rather than stirred in. Kodak does not say what it is. What the sheet does say is unambiguous about the handling: pour the clear solution into another container and discard the precipitate. A toner that has been left cloudy is a toner carrying something its maker took out.

Function of every ingredient

Sodium thiosulfate pentahydrate, 960 g, so about 229 g per litre of finished bath. Not the solvent and not a fixer here but the ligand: thiosulfate is one of the three sulfur ligands Ware names as both assisting the reduction of gold(III) to gold(I) and stabilising the gold(I) once made. Without it, or with too little of it, the bath is an acid gold chloride bath that costs three silvers per gold. Its secondary and unwanted property is that at this concentration and 43 °C it is also a fixing bath, which is what the ripener answers. More would strengthen both effects; less risks losing the reduction. The crystals are the pentahydrate at 248.19 g/mol, and 960 g of the anhydrous salt would be a very different bath.

Potassium persulfate, 120 g, so about 28.6 g per litre. The oxidiser that turns the stock milky, and the ingredient this page can describe but not explain — see The mechanism. It is also the ingredient that sets the safety level of the whole formula, and its own page carries the reason: a respiratory sensitiser whose recognised control when weighed is engineered extraction. More or less is not a control any source publishes; Kodak’s only instruction about it is the milkiness test.

Silver nitrate, 5 g, so about 1.19 g per litre. The ripener’s silver. It is never used as silver nitrate: the salt converts it in the beaker to silver chloride before it goes near the hypo. Without it, a hot 229 g/L thiosulfate bath takes silver out of the print. More has not been published as a control by any source read here.

Sodium chloride, 5 g, so about 1.19 g per litre. Table salt, and it makes the silver chloride. At 85.6 mmol against the silver’s 29.4 mmol it is in nearly threefold molar excess, and the free chloride left over is what keeps the fresh silver chloride from redissolving in all that hypo. The order is footnoted by Kodak and matters more than the quantity: the silver nitrate dissolves completely, and only then does the salt go in.

Chloroauric acid, 1 g in 250 mL of Stock Solution B, of which 125 mL reaches the bath — so about 0.119 g per litre of toner. The gold, and the only ingredient that ends up in the print. Kodak prints it as “gold chloride”, and the course’s reading of that name in the photographic literature is this substance rather than gold(III) chloride proper; both pages carry the reasoning, and the practical consequence is that the two are not interchangeable by weight — the true chloride is 64.93 per cent gold against about 50 per cent for the acid, so a gram of the wrong one is about thirty per cent more gold than the formula intended. More gold is not better: Reilly warns that too strong a bath deposits gold superficially and the layer is then largely removed in the fixer. Less tones more slowly and eventually not at all, which is what the replenishment rate exists to prevent. The salt is deliquescent and Kodak says to keep it tightly stoppered in a dry atmosphere.

Stock Solution A, the entire amount of it. Kodak’s own quantity for the working toner is the words “Entire amount”, which is why the schema carries a field for a source’s own wording rather than a number: there is no ratio here, only an instruction to use all of it.

Stock Solution B, 125 mL of it — half. The other half is the replenisher, and that is a deliberate feature of the formula rather than a leftover. Writing a gold weight in place of this line would hide the fact that the formula makes exactly twice as much gold stock as it uses.

Water, at two temperatures and in three places. Four litres at 50 °C for the thiosulfate and persulfate, because the reaction needs heat; 64 mL at 20 °C for the ripener; 250 mL at 20 °C for the gold. Kodak states no final volume for the stock or the bath, so every strength on this page is computed against the water named and says so.

Interactions

With acid, through the persulfate and the thiosulfate both. No stop bath, no acid fixer, nothing acidic into the waste.

With a fixing bath carried in on the print. The chloroauric acid page is blunt on this for gold toners generally: Reilly says the toner is ruined by even a trace of fixer, which is why a print is washed before toning and why toning tongs and fixing tongs are never the same tongs. T-21 is the awkward case, because the bath is mostly hypo — but the rule still holds for the print, which must be thoroughly washed.

With metal, of any kind. Ware records that these gold compounds are corrosive and oxidising, attack most metals and organic matter, and will stain skin a spectacular purple. Plastic or glass spatulas, stirrers and trays; Kodak’s general toning guidance forbids metal trays for every toner anyway.

With organic matter and with light, both of which reduce dissolved gold(III) to the metal. Reilly’s storage instruction is to keep gold stock solutions in the dark.

With heat, which the formula requires twice — 50 °C to make Stock A and 43 °C to work the bath — and once forbids: the ripener and the stock must both be cool when they meet.

Variants

Kodak GP-1, on the next page of the same sheet: the same gold at 0.10 g/L, a thiocyanate ligand instead of a thiosulfate one, mixed cold, made to a litre, used at 20 °C for 10 minutes, and no persulfate and no silver ripener anywhere in it. See its page. It is the formula the potassium persulfate page names as the route to take instead of this one.

Kodak T-26, the blue toner, on the page after that: the same gold at 0.4 g/L with thiourea and tartaric acid, worked cold or hot, and toning the highlights first. See its page.

The alt-process gold toners. Gold-thiocyanate and gold-borax are the two toners the printing-out processes use, at 0.1 g/L and 0.16 g/L of gold respectively — the same order of gold as this bath, from suppliers who publish their kits’ compositions.

No course variant is offered. The objection to T-21 is one ingredient and it cannot be reformulated away without changing what the formula is; the alternative is GP-1, which is a different formula and already exists.

Safety

Level C, from potassium persulfate, on the first of the rubric’s Level C criteria: a sensitiser at the concentrations used, where engineered extraction is the recognised control. Read that page. It also names the one operation that fixes the classification — weighing the powder — and states in terms that the course offers no home procedure using this substance and that where a gold toner is wanted, GP-1 does the job without a persulfate in it.

The gold is Level B and its page carries its own controls: nitrile gloves whenever the solid, the stock or the bath is handled; splash goggles for the solid and the stock, because the classification gives severe skin burns and serious eye damage unanimously; plastic rather than metal for every spatula and stirrer, because these compounds attack most metals; and hands washed before eating.

The silver nitrate is weighed once, in five grams, and converted immediately. Its own page carries the handling.

The bath is worked at 43 °C, which is a scald and spill hazard in four litres before it is anything else, and 43 °C over a hot thiosulfate bath is also the condition under which thiosulfate decomposes towards sulfur dioxide — the same warning that governs T-1a, from the sodium thiosulfate page. Ventilation over the tray, not a cracked window.

Storage

Stock Solution B in the dark, tightly stoppered, in a dry place. Kodak’s own footnote is that gold chloride is deliquescent and will liquefy rapidly in a normal room atmosphere; Reilly’s is that light reduces dissolved gold. Half of this bottle is the replenisher and it has to survive as long as the bath does.

Stock Solution A publishes no keeping figure and none is invented here.

The working toner is kept and replenished, which is unusual in this section. Kodak gives no discard point, only a replenishment rate, and no shelf life.

Label the gold, in words, with its strength. A 0.4 per cent gold solution and a 1 per cent gold solution look identical and cost the same to make a mistake with.

Nothing metal touches any of it, including the container’s cap liner.

Incompatibilities

Acids, with the thiosulfate and the persulfate both.

Metals, with the gold: spatulas, stirrers, trays, tongs. Ware’s warning is that these compounds attack most metals.

Organic matter and light, with the gold stock, both of which reduce it.

Combustible material, with the persulfate, which its page classifies as an oxidiser.

Fixer carried in on a print, per Reilly’s warning for gold toners generally.

Developer, in either direction.

Waste

A gold-bearing and silver-bearing thiosulfate stream, and it is worth money. At 229 g/L of hypo the spent bath is a fixer by composition, and it carries the gold that did not reach the prints along with silver taken out of them.

It is not acidified. Adding acid to a hot spent thiosulfate bath is the sulfur dioxide reaction, and this one carries a persulfate as well.

It does not go in with a sulfide toner’s waste.

Collect it, label it as gold- and silver-bearing, and follow the silver-bearing waste SOP and the disposal ruling. Local regulation decides, and this course cannot tell you what it says where you are.

Troubleshooting

Recorded from the source for the sake of understanding the formula, not as instructions for running it.

The stock did not turn milky. Kodak’s own instruction is to increase the temperature. It is the sheet’s only test that the persulfate step has worked.

The toner will not tone. The most likely cause is that it was used before it had stood eight hours, so the gold is still gold(III). The second is exhaustion, for which Kodak gives a replenishment rate rather than a test.

The print is bleaching rather than toning. Gold(III) rather than gold(I), spending three silvers per gold — which is the same fault as above, seen from the print’s side. Reilly’s description of that result is flat, lifeless and reddish.

The highlights have gone. The ripener was omitted or under-measured, and a hot 229 g/L hypo bath found something to fix.

Scum on the dried print. Sediment from the tray. Kodak’s remedy is a wet sponge or wet cotton immediately after toning, and it is worse with several prints in the bath at once.

Nothing happens on this paper. Cold-tone paper. Kodak says so in the product description rather than the troubleshooting.

The bath went cloudy again after decanting. More of the same precipitate. Kodak’s instruction is to use the clear solution.

Experiments

For a laboratory that has the controls the persulfate needs.

Prove the eight hours. Split a freshly mixed working toner and tone identical prints at one hour, four hours, eight hours and twenty-four. Kodak gives a single number with no tolerance and no test; the colour of the bath and the behaviour of the print should both change across that series.

Test the uniform-rate claim. A step wedge toned in T-21 and an identical one in T-26, both pulled at the same fraction of full toning, both read step by step. Kodak claims one tones the whole scale evenly and the other tones the highlights first; that is a claim about the shape of the change and it can be plotted.

Measure what the ripener prevents. The same bath with and without the silver charge, on identical prints, read for maximum density. The claim is that a hot hypo bath with no silver in it takes silver out of the print.

Compare the three ligands at matched gold. T-21, GP-1 and T-26 all carry gold in the same order of magnitude with three different sulfur ligands. Matched to the same gold concentration and the same paper, what remains is the ligand’s own contribution, and no source in this corpus isolates it.