Ware's argyrotype

Every iron-silver process before this one had the same problem, and it was not a problem of speed or of formulation. It was the anion. Silver nitrate is the soluble silver salt everybody has, so the siderotypes that put silver on paper all used it, and nitrate in acid conditions oxidises metallic silver — including the finely divided silver that is the picture. The traditional escape was to develop in an alkaline bath, which stops the nitrate attacking the image and, in the same movement, hydrolyses the excess iron(III) into an insoluble hydroxide that stays in the paper and eventually destroys the print from within.

Mike Ware’s answer, published in 1991, was to change the silver salt. Silver sulfamate, NH₂SO₃Ag, is soluble, its anion is not an oxidiser, and it cannot be bought — so it is made in the beaker from silver(I) oxide and sulfamic acid, with the acid deliberately in excess. The whole process then works at about pH 3.5, where the iron washes out cleanly, the silver image is not attacked, and one bottle carries everything the paper needs.

Argyrotype sensitizer — the whole of the light-sensitive chemistry, in one bottle

Ingredient	Quantity	Form the source specifies
Sulfamic acid	7 g	general purpose reagent grade, about 98 per cent, as Ware specifies for every solid here; about 5.87 g of it reacts with the silver oxide and the remaining 1.13 g stays in the bottle as free acid
Silver(I) oxide	7 g	powdered silver(I) oxide, Ag2O; Ware gives an alternative preparation from 10.3 g of silver nitrate and 2.5 g of sodium hydroxide where the oxide cannot be bought, and says 8.4 g of silver carbonate may be used instead
Ammonium iron(III) citrate	22 g	the green variety, about 16 per cent iron; Ware's own comparison table puts the green salt at 14 to 18 per cent iron against the brown at 19 to 28
Glycerol	1 mL	Ware marks it optional in the 2009 handout and does not print it at all in his web account of the process
Water	to make 100 mL	at 70 °C; Purified water — distilled, de-ionised or pharmaceutical. The 70 °C is the temperature the 2009 handout heats the first 70 cc to in order to dissolve the sulfamic acid and then the silver oxide; the web account says 50 to 60 °C. The make-up to 100 cc is made cold, after the solution has been allowed to cool, and it is a make-up volume rather than an added one, so every strength on this page is computed against 100 mL of finished sensitiser.

Clearing bath — de-chlorinated water with citric acid — removes the unused iron and silver salts before the thiosulfate ever touches the image

Ingredient	Quantity	Form the source specifies
Citric acid (anhydrous)	25 g
Water	10000 mL, added	De-chlorinated water and nothing else — rainwater, or water passed through an activated charcoal filter, or water boiled and allowed to stand. Ware states the dose as a tablespoonful, about 25 g, of citric acid to each 10 litres, which his chemicals table gives as about 2.5 g per litre, that is 0.25 per cent w/v; the one-page summary at the end of the same handout writes it as 0.2 per cent, without stating a basis, and the course works to the 2.5 g per litre because that is the figure the chemicals table and the procedure agree on. This is an added volume, not a make-up volume — the source gives the water and then the dose that goes into it.

Toning and fixing bath — 2.5 per cent w/v sodium thiosulphate — dissolves what silver the light did not reduce, and tones what it did

Ingredient	Quantity	Form the source specifies
Sodium thiosulfate (anhydrous)	25 g	Ware writes the anhydrous formula Na2S2O3 in his chemicals table and adds that the hydrated form may be used, giving no second weight for it; in the procedure he calls the material crystals, which in photographic usage means the pentahydrate. The consequence of that ambiguity is worked out under Function of every ingredient.
Water	to make 1000 mL	Ware specifies the bath as 2.5 per cent w/v and defines that in the same handout as X grams of solute in 100 cc of solution, which makes it a make-up volume; his working instruction in the procedure is to dissolve about 25 g in 1 litre of water. At this dilution the two readings differ by well under one per cent.

Used in this order — wet processing of the exposed print, in the order the baths are used

1. Clearing bath — de-chlorinated water with citric acid — 2 to 5 minutes, until the yellow stain is cleared — Preceded by an optional post-exposure humidification, which Ware calls "steaming" — 30 minutes at 100 per cent relative humidity at about 20 °C, or half a minute over water at 40 °C. The first bath is changed for each print.
2. Toning and fixing bath — 2.5 per cent w/v sodium thiosulphate — about 2 minutes — The image intensifies and the colour shifts from red to brown. Overlong treatment costs density, first in the highlights.
3. Water — 15 to 30 minutes, running — Heavier papers take the longer wash. Then drain and air-dry at room temperature.

Clear in a bath of de-chlorinated water with citric acid for 2 to 5 minutes; tone and fix in 2.5 per cent sodium thiosulphate for about 2 minutes; wash in running water for 15 to 30 minutes; drain and air-dry at room temperature. The sensitizer itself is not a bath — it is coated on the paper, dried and exposed before step 1.

The order is the whole of the argument. Ware's own monograph states that excess silver must be washed out of a print-out paper before thiosulfate reaches it, or the image is stained with brown silver sulphide; the acidified clearing bath does that and takes the iron out at the same time, which is the step the kallitype's alkaline developer cannot do.

Purpose

To put a purplish-brown metallic silver image on plain paper from a single sensitiser, and to leave behind neither the residual iron that fades a kallitype nor the oxidising anion that eats the image while it is being made. The sensitiser is the entire light-sensitive chemistry. There is no developer: the picture forms during the exposure, as a printing-out image you can watch appear. The two baths that follow it remove what the light did not use — first the iron and the unreduced silver, then the last of the silver as a soluble complex — and the second of them also changes the colour, which is the part of the process Ware found rather than designed.

Ware’s own framing is worth having because it is unusually modest. He revisited the chemistry, he says, “in the interests of achieving a more ‘user-friendly’ plain paper silver process”, and came up with “an easier (I don’t say ‘better’) process”. The claim is not that an argyrotype is finer than a platinotype. It is that the known faults of the iron-silver family have identifiable chemical causes and that most of them can be designed out.

Recommended uses

Contact printing from a negative with a long ultraviolet density range. Ware asks for at least 2 and as much as 2.4 — the same requirement as his other siderotypes. The sensitiser is far too slow for an enlarger, so the print is the size of the negative.

As the introduction to hand-coated printing, and as a proofing process for the expensive ones. Ware puts the iron-silver processes exactly there: “useful as an introduction to alternative printing for teaching workshops, for proofing large format negatives, and as a starting point for further toning of the colloidal silver image.” A sheet of argyrotype tells you whether a negative has the scale for a palladiotype before you spend palladium finding out.

Where you want a brown print and do not want to keep three bottles. The single-bottle sensitiser with a shelf life measured in years is the practical difference between this and the Van Dyke Brown sensitiser, which is mixed from three solutions, and the kallitype sensitiser, which needs a developer chosen for the colour you want.

Where the water is hard, the paper is buffered, or the darkroom is warm — that is, in the conditions that make a kallitype unreliable. Every one of those problems is a hydrolysis problem, and this formula’s answer to hydrolysis is to work acid throughout.

When another formula is preferable

When permanence is the point, print in a noble metal. Ware says so himself, twice, and the second time without hedging: “If permanence is a paramount issue, it is best to print entirely in platinum, palladium, or gold.” An argyrotype is silver on plain paper with no binder over it, and no amount of care changes what that is.

When the budget rules and the colour does not, the Van Dyke is a quarter of the silver. Its 3.8 g of silver nitrate carries 2.41 g of silver per 100 mL against this formula’s 6.52 g. If you are learning to coat, learn on the cheaper one.

When you want to choose the image colour by changing the developer, use a kallitype. Its three published developers — sodium citrate, Rochelle salt and borax — are three different image colours from one coating, and the argyrotype has no developer to change. Its colour controls are humidity and heat.

When the negative is short-scale, a process with a bath-based contrast control is easier to steer. The argyrotype’s control is real but it lives in the sensitiser, so it has to be decided before the paper is coated rather than after the exposure is judged.

When you want blue, or gold, or an image that cannot be attacked by sulfur at all, this is the wrong family entirely: see the classic cyanotype and the New Cyanotype.

Mixing

⚠Read Safety before you weigh anything

Silver(I) oxide is classified by the majority of its ECHA notifiers as a strong oxidiser that may cause fire or explosion and as causing serious eye damage. This preparation heats a solution to about 70 °C and stirs a solid into it for up to an hour. Both facts belong to the procedure and not to a footnote.

Work under tungsten light, not fluorescent and not daylight. Ware repeats the instruction at the head of the preparation and again in the numbered steps. The finished solution is sensitive to ultraviolet, and ordinary domestic filament lighting is the one common source that emits almost none.

The order is not a convention, it is a synthesis, and each step exists because the one before it has happened:

1. Weigh 7 g of sulfamic acid into a 250 cc Pyrex beaker and dissolve it in about 70 cc of purified water at about 70 °C. The acid has to be in solution and hot before the silver arrives, because the next step is a reaction and not a dissolution.
2. Add 7 g of powdered silver(I) oxide to the hot solution in small portions, stirring vigorously, until it has all dissolved. Ware’s warning is that this may take up to an hour. Silver oxide is not appreciably soluble in water; what is happening is that the acid is converting it to silver sulfamate, which is, and the rate is set by the surface of the solid.
3. Add 22 g of green ammonium iron(III) citrate in portions to the warm solution, stirring until it has dissolved. Allow it to cool. The iron goes in last of the solids because the solution it goes into must already be acid — see Interactions for what happens if it is not.
4. Add 1 cc of glycerol and mix well. Optional, and it is the image colour that depends on it.
5. Make up to 100 cc with purified water in a measuring cylinder and filter through a Whatman grade 1 paper in a conical funnel. The finished solution should be a clear olive-green.
6. Store in a labelled brown bottle, in the dark, at room temperature.

🧮Ware's stoichiometry, checked

The handout states that 7 g of sulfamic acid is about a 20 per cent excess, and that “the stoicheiometric amount needed to react with 7 g of silver oxide is 5.87 g”. That is a claim a reader can test, and it is worth testing because it is the only independent check on the two weights.

Silver(I) oxide is Ag₂O, relative molecular mass 231.74; sulfamic acid is H₃NO₃S, 97.10. The reaction is one oxide to two acid molecules.

Ag₂O + 2 NH₂SO₃H → 2 NH₂SO₃Ag + H₂O

Silver sulfamate, made in the beaker

2 × (7 ÷ 231.74) × 97.10 = 5.87 g

Acid required for 7 g of the oxide

Ware’s figure to three significant figures, exactly. The excess is therefore 7 − 5.87 = 1.13 g, which is 19.3 per cent — his “about 20 per cent”. The same 7 g of oxide carries 2 × (7 ÷ 231.74) × 107.87 = 6.52 g of silver, so the sensitiser is 6.52 per cent w/v silver, or 0.604 mol/L.

Two quantities that check each other to three figures is a good sign that the weights on the sheet are the ones the author meant, and not a typographic descendant of something else.

⚗Where the pH of 3.5 comes from — and where it does not

The 1.13 g of free sulfamic acid left over is 0.0117 mol in 100 mL, that is 0.117 mol/L. Sulfamic acid is a strong acid — the encyclopaedia’s own page puts it with nitric and hydrochloric rather than with citric. A 0.117 mol/L solution of a strong acid on its own would sit near pH 0.9.

Ware measures about 3.5. The difference is the citrate. Twenty-two grams of ammonium iron(III) citrate per 100 mL is a very large buffering load, and citric acid’s three dissociation constants — the encyclopaedia’s citric acid page records pK_a1 2.87, pK_a2 4.35 and pK_a3 5.68 from the IUPAC compilation — bracket the working pH on both sides. So the “excess acid” is not free strong acid sitting in the bottle; it is a reserve that has titrated part of the citrate system and left the mixture parked in the middle of citrate’s buffer region.

That is the course’s reading of two published facts rather than a statement Ware makes, and it matters practically for one reason. It is why the contrast control works the way it does. Adding another gram of sulfamic acid to 100 cc does not crash the pH the way adding a gram of a strong acid to unbuffered water would; it moves it a little, along a buffered curve, which is exactly what makes “mix the two sensitisers to fine-tune the contrast” a usable instruction rather than a knife edge.

If silver(I) oxide cannot be bought, Ware gives two routes to it. Neither is a different formula; both end at the same 7 g of oxide.

• From silver nitrate. Dissolve 2.5 g of sodium hydroxide in about 40 cc of purified water; dissolve 10.3 g of silver nitrate in about 40 cc in a 200 to 250 cc beaker; add the alkali slowly to the silver with stirring, to precipitate a dark brown sludge of silver(I) oxide, 7 g in total. Filter it off, wash it on the paper with purified water three times, draining between, then transfer the wet paper and sludge into the hot sulfamic acid of step 1 and stir gently until the brown solid has gone; remove the filter paper with tweezers and continue at step 3.

  2 AgNO₃ + 2 NaOH → Ag₂O + 2 NaNO₃ + H₂O

  Precipitating the oxide, so that the nitrate can be washed away

  The whole point of the washing is that the nitrate leaves with the wash water. A reader who skips it has reinvented the problem the process exists to solve.

• From silver carbonate. 8.4 g may be used instead of the oxide, but dissolved at room temperature in a tall vessel — a 250 or 500 cc measuring cylinder — added in small portions with the effervescence allowed to die down each time. The gas is carbon dioxide from the carbonate.

The wetting agent is not mixed into the bottle. Ware is explicit in the 2009 handout: “Do not add Tween to the stock sensitizer solution: it doesn’t last very well, and the appropriate amount will depend upon the paper.” Keep a separate 10 per cent or 5 per cent v/v stock of Tween 20 and add one drop (0.05 cc) of the 10 per cent stock per cc of sensitiser immediately before coating, for a final concentration of about 0.25 to 0.5 per cent. His earlier web account does put 0.2 cc of Tween in the 100 cc; the later instruction supersedes it and gives its reasons.

Coat about 1.5 cc per 10 by 8 inch area, five or six passes of a glass coating rod on a levelled plate, and blot off any excess, which would otherwise crystallise on the surface and damage a negative. Then dry: about an hour at room temperature and an ambient relative humidity between 50 and 90 per cent, or about five to ten minutes in a stream of warm air at 40 °C. Ware warns that over-rapid drying worsens the loss of image substance in the wet processing.

🧮How much silver is on the sheet

A 10 by 8 inch area is 80 square inches, which is 516 cm² or 0.0516 m². At 1.5 cc of a sensitiser carrying 65.2 g of silver per litre, the sheet receives 1.5 × 0.0652 = 0.098 g of silver, a coating weight of about 1.9 g/m².

Almost none of that ends up in the picture. Ware’s own analysis of salted paper prints — a different process, so the figure is a scale rather than a prediction — puts the image coating weight at about 0.1 g/m², so that a whole-plate print holds around 3.3 mg of silver in total, against a pinhead of silver metal at about 10 mg. The rest of what you coat leaves in the clearing and fixing baths, which is why the waste from this process is worth catching and why over-coating is expensive rather than generous.

The arithmetic is the course’s; the coating volume, the concentration and the salt-print figures are Ware’s.

Behaviour

It prints out, and that is the single most useful thing about it in practice. At an ordinary ambient humidity the image appears during the exposure — orange-brown on a yellow ground — so the exposure is judged by inspection through a hinged-back contact frame rather than by test strip. Ware’s web account says outright that this “gives a good indication of the correct exposure, making test strips unnecessary.”

Two things then change the print after you stop the exposure, and both are in the same direction. Some further development occurs in the wet processing — Ware puts it at half to one stop in the high values — and the colour deepens markedly in the thiosulfate bath. On top of that the print dries down by “at least one Zone”. So a print judged correct on the frame will be dark when it is dry. Ware’s advice is to err towards overexposure, because a dense image can be reduced in the fixer and a thin one cannot be rescued.

Exposure is of the order of five minutes under an average ultraviolet-A source; Ware’s own figure with an 800 W high-intensity-discharge lamp and digital negatives is about one minute. Like every iron process, it is far too slow for anything but contact printing.

Humidity is a working variable, not a nuisance. Three separate humidity decisions change the result: the relative humidity the coated sheet dries at, whether the dried sheet is humidified before exposure, and whether the exposed sheet is “steamed” before processing. The reason all three matter is in The mechanism.

The clearing bath is judged by the disappearance of the yellow stain, two to five minutes, and the first bath is changed for each print. If a red-brown stain runs off the dense areas into the water, the paper fibres are failing to hold the silver particles: that is the bleeding fault, and the remedies are more Tween or a different paper.

The fixing bath is short and weak on purpose. Two minutes at 2.5 per cent w/v. For comparison, the course’s plain hypo fixing bath is 400 g of the pentahydrate per litre — 1.61 mol/L against this bath’s 0.158 mol/L as the anhydrous salt, about ten times the molar strength. This is not a fixer that has been diluted for economy. It is a bath doing two jobs at once and it is held at the strength where the second job stops before it goes too far.

Image characteristics

Colour. Purplish-brown, deepening to “a rich mahogany brown” in the thiosulfate bath. The starting colour of the print-out image is orange-brown on yellow; the shift from red to brown happens in the fixer and is dramatic enough that Ware treats it as the process’s signature.

Four ways to move the colour, all of them cheap.

Change	What it does	Ware’s account
Leave the glycerol out	More yellowish-brown	The humectant is included “to ensure a pleasing purplish-brown colour”
Humidify the coated sheet to 100 per cent RH for 30 minutes before exposure	An attractive purplish-grey print-out image	“A very economical method of colour control”, with a warning that a humidified sheet can damage a negative unless thin polyester is interposed
Dry the sensitised sheet in warm air rather than at room temperature	Warmer tones	Stated as a shift towards warmer tones on drying
Heat-dry the finished print — ferrotype plate, dry-mount press or an iron	A more neutral blackish brown	Given as an option after the final wash

Gradation. Ware claims “a finer gradation than the traditional iron-silver processes”. That is the originator’s assessment of his own process and the course records it as such; no independent sensitometry of the argyrotype was found in the sources read for this page.

Contrast. Set by the acid in the sensitiser, not by a bath. The standard formula suits a negative of 2 to 2.4 density range in the ultraviolet. An extra 1 g of sulfamic acid per 100 cc gives a more contrasty sensitiser, and because the two are miscible in any proportion the contrast is continuously adjustable between them.

Surface. Plain paper, matte, with the image among the fibres rather than in a coating. There is no binder layer, no gloss and no tackiness, and the sheet does not curl. Ware notes that retouching is easy on the receptive surface with good watercolour paints.

Staining. None inherent to the formula. The yellow of the unexposed sensitiser is removed in the clearing bath, and a highlight that stays yellow is a fault rather than a characteristic.

The mechanism

Four things happen in order, and only the first of them is photochemistry.

1. Light reduces the iron

The only light-sensitive substance in the sensitiser is the iron(III) citrate complex. Ware’s general statement of the iron chemistry is that under ultraviolet, iron(III) is reduced to iron(II) at the expense of the carboxylate ligand, which is oxidatively decarboxylated. For the oxalate he writes it dissected as

2 Fe³⁺ + C₂O₄²⁻ → 2 Fe²⁺ + 2 CO₂

Ware's dissected form of the light reaction in an iron(III) carboxylate

and he is careful to add that the citrate’s chemistry is “similar in principle, but rather more complicated”. It is worth knowing exactly how much more complicated, because the argyrotype’s pH is chosen on the strength of it.

⚗The citrate photolysis, as far as it is known

From Ware’s Cyanomicon, Appendix III.6, drawing on the speciation literature: in aqueous iron(III)/citrate solutions a photo-inactive monomer predominates between pH 0.5 and 1.5, and a photo-active dimer forms above pH 2. Two such dimeric anions have since been crystallised and solved by X-ray diffraction, [Fe₂cit₂(OH₂)₂]²⁻ and [Fe₂(Hcit)₃]³⁻, with the iron centres oxo-bridged through the alkoxide groups of the ligands.

The quantum yield at 365 nm has a maximum of 0.45 at pH 4 and falls above it, probably because the iron(III) hydrolyses and oligomerises further; at 436 nm it is 0.28. The initial photoproduct has been identified as acetone dicarboxylic acid, which may decarboxylate further to acetone, and — in Ware’s words — “the nature of the iron(II) photoproduct remains unknown”.

He therefore prints the reaction as a scheme rather than as a balanced equation:

hν + [Fe₂(C₆H₄O₇)₂(OH₂)₂]²⁻ → 2 [Fe^II] + CO(CH₂COOH)₂ + CO₂

Balancing the products he names, and leaving the iron(II) as the free ion because nobody has identified what it actually is, gives the dissected form the rest of this page uses. This equation is the course’s arithmetic on Ware’s stated products, not an equation he prints, and the reason he does not print it is the one just given.

2 Fe³⁺ + C₆H₄O₇⁴⁻ → 2 Fe²⁺ + C₅H₄O₅²⁻ + CO₂

Citrate photolysis, dissected — two electrons from one citrate, one carbon lost as CO2

So the pH of 3.5 is not an accident and it is not only about keeping things in solution. It sits above the pH 2 threshold where the photo-active dimer exists at all, and close to the pH 4 where the quantum yield is highest. Ware states plainly in the handout that pH 3.5 “is also the optimum for the photosensitivity”; the speciation work is why. This is one of the places where an alternative process turns out to have been designed rather than inherited.

2. Iron(II) reduces silver(I)

The iron(II) that light has made is a reducing agent strong enough to put silver(I) down as metal, one electron at a time.

Fe²⁺ + Ag⁺ → Fe³⁺ + Ag

The image: silver reduced by the iron light made

Nothing in this step requires a developer, and nothing in it requires a bath. It happens in the sheet, during the exposure, which is what makes the process a printing-out one. But it requires ionic mobility: in a bone-dry coating the ions cannot reach each other. That is the whole reason humidity is a control variable here. Ware makes the same point about the print-out platinum route, where the sensitiser is worked at around 70 per cent relative humidity so that the paper carries about 8 per cent of water by weight and the photoproduct can react in place. The glycerol in this formula is there to keep that water in the coating.

3. Everything unused is washed out, acid

At the end of the exposure the sheet holds metallic silver where the light was, and everywhere else unchanged iron(III) citrate, unreduced silver sulfamate, and the iron(III) generated in step 2. All of it has to leave, and the order and the pH decide whether the print survives.

⚗Why the first bath is acid and the second is not the first

Iron(III) must come out before the print dries, and it comes out in acid. Ware’s Platinomicon sets out what residual iron does in a siderotype: iron(III) chemisorbs to the hydroxylic functions of cellulose, hydrolyses above about pH 4 to a polymeric colloidal hydroxide, and — if it is not removed before the print dries — undergoes an irreversible transformation to an insoluble oxyhydroxide that no ordinary bath will lift. Every yellow-stained kallitype in every drawer is that reaction. An acid first bath is the direct answer to it.

Excess silver must come out before the thiosulfate. His Argyronomicon gives this as the first of three pitfalls of fixing a print-out paper: unless the excess silver salt is washed out first, “the image will be seriously stained with brown silver sulphide resulting from the reaction of thiosulphate with silver nitrate”. Ware records that Herschel recognised the importance of the step. It is why the argyrotype’s first bath is acidified water and not a fixer.

And the water must not be chlorinated. Ware’s instruction is rainwater, charcoal-filtered water, or water boiled and allowed to stand, with citric acid added “to scavenge any traces of chlorine”. The reason is in the same monograph’s account of back-reactions: hypochlorous acid oxidises silver, and the nanoparticles are at their most vulnerable at exactly this moment.

4. The thiosulfate bath fixes, and tones

What is left of the silver leaves as the soluble bis(thiosulfato) complex, which is ordinary fixing chemistry.

Ag⁺ + 2 S₂O₃²⁻ → [Ag(S₂O₃)₂]³⁻

The unreduced silver leaving as a soluble complex

But something else happens in the same two minutes, and it is the reason the bath is weak and short.

⚗Colour, plasmons and a monolayer of sulfide

A brown silver image is not brown because silver is brown. The particles are nanoparticle silver, on the order of 10 to 100 nm, far smaller than the wavelength of visible light, and their colour comes from surface plasma resonance — light exciting collective oscillations of the conduction electrons. Ware reproduces Wiegel’s calculated and experimentally confirmed table: silver spheres of 10 to 20 nm transmit yellow; 25 to 35 nm, red; 35 to 45 nm, purplish red; 50 to 60 nm, violet. He adds that linear aggregation of a yellow 10 to 20 nm sol brings in a long-wavelength absorption, so it “becomes red-brown”, and that raising the refractive index of the surrounding medium shifts the band to longer wavelengths.

Ware closes the loop himself, in the handout: “The colour of the silver print is determined by the size of the metal particles, which is affected by the humidity of the coating during the printing-out exposure.” Humidity governs how the particles nucleate and grow; particle size, shape and aggregation govern the plasmon absorption; and the plasmon absorption is the colour.

Now the fixer. Ware’s monograph states the chemistry generally: “Silver sulphide can form directly by the reaction of thiosulphate ions with nanoparticle silver, especially under acidic conditions. The minute spheres of silver metal initially acquire only a surface-coating of the sulphide — like the skin of an apple — but the colour of such particles — again just like the apple — is very sensitive to the condition of the surface.” A thin layer of silver sulfide “deepens the colour of nanoparticle silver to a more satisfying brown, but prolonged contact with thiosulphate causes complete conversion to silver sulphide, which weakens the colour again to a drab buff.”

The size of that second effect is the number to remember. Nanoparticle silver has an extinction coefficient of about 16,000 dm³ mol⁻¹ cm⁻¹ at its absorption maximum; nanoparticle silver sulfide has been estimated at about 560. Complete sulfiding therefore drops the maximum optical density by a factor of nearly thirty. A little sulfide enriches the image; total sulfide destroys it. Two minutes and two and a half per cent is where Ware puts the line.

⁇Is an argyrotype sulfide-toned? Ware says probably, and says so as a belief

The claim. Ware writes in the 2009 handout that the procedure “does not prescribe a toning step for Argyrotypes, because I believe that they already become partially sulphide-toned when they pass through the thiosulphate bath, where the dramatic colour change takes place from yellowish-red to a richer mahogany-brown.” He offers a mechanism — sulfide a few atoms deep, “perhaps even a monolayer”, modifying the plasmon absorption — and one piece of instrumental evidence, energy-dispersive X-ray analysis of argyrotypes by Ellie Young at the Royal Melbourne Institute of Technology, showing the presence of sulfur as well as silver in the image.

What supports it. His own 2019 monograph states the same chemistry as established for print-out silver generally, without hedging, and quotes Henglein’s finding that dilute sulfide turns a silver sol from yellow to brown by adsorbing a monolayer.

What is not established. Nobody has published a quantitative account of how much sulfide an argyrotype acquires, or shown that it is enough to protect the image over decades. Ware’s own conclusion is careful: argyrotypes “should enjoy a good lifetime without additional toning — but it cannot be guaranteed, because they are silver.”

What the course says. The colour shift in the fixer is an observation anyone can repeat. The sulfide explanation is the originator’s interpretation, supported by his own general chemistry and by one instrumental result he reports rather than publishes in full, and it is the best explanation available. It is not a permanence guarantee, and this page does not treat it as one. Where a print matters, tone it in gold or palladium before the thiosulfate bath, which is Ware’s own instruction for anyone who wants to.

Deeper: why nitrate and acid cannot share a print

This is the argument the whole formula is built on, so it is worth following to the end.

Ware’s Argyronomicon lists the oxidation of silver by nitrate in acid conditions as the first of the back-reactions that destroy print-out silver:

3 Ag + 4 H⁺ + NO₃⁻ → 3 Ag⁺ + 2 H₂O + NO

Nitrate eating the image, in acid

Read the equation as a mechanism and the design follows. The reaction consumes four protons for every nitrate, so its driving force falls as the acidity falls — Ware’s own table of standard potentials gives the nitrate couple as +0.96 − 0.079 pH volts against silver’s +0.799 V, which is to say that nitrate is a strong enough oxidant to attack silver in strong acid and a progressively weaker one as the pH rises. That is precisely why the kallitype is developed in an alkaline bath: alkali is the cheapest way to switch this reaction off.

But alkali switches something else on. Above about pH 4, iron(III) hydrolyses to a colloidal polymeric hydroxide that lodges in the paper and eventually oxidises the image itself. So the traditional iron-silver process is caught between two failure modes and can only choose which one to have.

Ware’s move is to refuse the choice by removing one of its terms. Take the nitrate out, and the acid side of the trade costs nothing: sulfamate has no oxidising chemistry, so the print can be worked at pH 3.5 where the iron clears properly and stays cleared. That is the entire content of the invention, and every other difference between this formula and a Van Dyke follows from it.

🔬A second reason to avoid the back-reaction

The nitrate back-reaction is slow and diffusion-controlled, and Ware invokes it to explain something subtler than fading: why very long, weak exposures give worse prints rather than merely slower ones. Silver deposited slowly is continuously being re-oxidised as it forms, which keeps the particles smaller than they would otherwise be — and small particles are both weaker in colour and more vulnerable to everything that follows. A sensitiser with no oxidising anion in it does not do this, which is a plausible reading of Ware’s claim of finer gradation. The course marks that as a reading rather than as a demonstrated result.

Function of every ingredient

Sulfamic acid, 7 g — the reagent that is also the control. It does three jobs and they are worth separating. First, 5.87 g of it converts the silver oxide into soluble silver sulfamate; without it there is no silver in solution at all, because the oxide is not appreciably soluble in water. Second, the 1.13 g left over sets the working pH at about 3.5, which Ware says suppresses hydrolysis of the iron(III), helps keep the silver in solution — “otherwise silver citrate may precipitate out” — and is the optimum for the photosensitivity. Third, its anion is the point of the entire formula: sulfamate is not an oxidiser, so the acid conditions the iron needs cost the silver image nothing. More of it raises contrast, because the added acid “tends to dissolve silver in the highlights”; Ware’s contrasty variant is an extra 1 g per 100 cc, and mixing the two sensitisers gives any intermediate contrast. Less of it lowers contrast, and Ware’s floor is about 6 g, “but it might cause problems of precipitation” — below that the free acid no longer holds the silver and the iron in solution. Ware’s stated reason for choosing this acid is its anion: a few soluble silver salts with non-oxidising anions are known, and most of them, silver fluoride among them, have “unacceptable properties or a level of toxicity that debar them from ‘home chemistry’”, while the sulfamate is “little-known and relatively innocuous” and can be made in the beaker. A second property, which he does not give as a reason here but which matters at the bench, is that this is a strong acid you can weigh: a solid, odourless, non-volatile powder rather than a fuming concentrate.

Silver oxide, 7 g — the image, delivered without its usual anion. Ag₂O, relative molecular mass 231.74, CAS 20667-12-3. It is not the light-sensitive substance and it takes no part in the exposure except as the thing waiting to be reduced; light acts on the iron, and the silver(I) is reduced afterwards by the iron(II). The reason silver arrives as an oxide rather than as the usual nitrate is negative rather than positive: the oxide’s counter-ion is O²⁻, which becomes water on reaction and disappears from the formula entirely, so that the finished sensitiser contains silver, sulfamate, citrate, ammonium and iron, and no oxidising anion whatsoever. More of it would mean more silver than the iron can reduce and more to wash out, at the direct cost of the most expensive ingredient; less would run the iron-to-silver ratio away from unity and cost maximum density. Its practical property is that it must be reacted rather than dissolved: powdered, added in small portions to hot acid, with vigorous stirring, over as much as an hour. Its encyclopaedia entry was written because this formula needed it, and carries the oxidiser handling in full.

Ammonium iron(III) citrate, green, 22 g — the only thing light acts on. It is also the least well defined substance in the formula, and the page linked above explains why: this is not a compound but a variable mixture of iron, citrate and ammonium sold under one name, whose iron content runs from 14 to 28 per cent depending on grade. Ware specifies the green variety at about 16 per cent iron — the acidic form, first prepared by Eduard Valenta in 1897, which his own comparison table puts at 14 to 18 per cent against the brown salt’s 19 to 28. The citrate has two roles at once: it is the ligand whose oxidative decarboxylation supplies the electrons that reduce the iron, and it is the chelating agent that keeps the iron soluble and lifts it out again in the clearing bath. More of it increases speed and the risk of a tacky, deliquescent coating; less starves the reduction of silver and costs density. The variability is the practical problem: a new jar of citrate is a new sensitiser, and a printer who changes supplier should expect to recalibrate.

🧮Iron to silver, and why the grade of the citrate matters

The 7 g of silver oxide carries 6.52 g of silver, which is 0.0604 mol.

At Ware’s stated 16 per cent iron, 22 g of the citrate carries 3.52 g of iron, or 0.0630 mol. The molar ratio of iron to silver is then 1.04 to 1 — very nearly the one-electron-per-silver-atom stoichiometry the reduction requires.

At the ends of the published range for the green salt the same 22 g carries 0.0551 mol of iron at 14 per cent and 0.0709 mol at 18 per cent, so the ratio runs from 0.91 to 1.17. At the bottom of that range there is not enough iron to reduce all the silver even in principle.

The arithmetic is the course’s; the 16 per cent and the 14-to-18 range are Ware’s. It is the clearest illustration this course has of why the citrate’s variability is not a pedantic complaint.

Glycerol, 1 cc — the humectant, and therefore the colour control. Glycerol is hygroscopic: it holds water in the dried coating instead of letting the sheet equilibrate down to the room’s humidity. Ware includes it “to ensure a pleasing purplish-brown colour”, and states that without it “the image may be a more yellowish-brown”. The chain from molecule to picture is short and complete — glycerol keeps the coating damp, dampness lets the ions move during the printing-out exposure, ion mobility governs how the silver particles nucleate and grow, particle size and aggregation govern the plasmon absorption, and the plasmon absorption is the colour. More of it would leave the coating tacky enough to threaten a negative; none gives the yellower image, which is what his earlier web formulation, containing no glycerol at all, produced. Ware marks it optional. Glycerol’s encyclopaedia entry was written alongside it, and records with it the contested history of glycerine development.

Citric acid, about 25 g per 10 litres of the clearing bath. Two jobs, both stated by Ware: it scavenges traces of chlorine from the water supply, and it provides slight acidity. The acidity is the important one — it is what stops iron(III) hydrolysing to the insoluble colloidal hydroxide that ruins a print from the inside, and it is the step that the kallitype’s alkaline developer structurally cannot perform. Citrate is also the same chelating ligand that is already in the sensitiser, so the bath is lifting iron with the ligand it arrived on. More of it is not obviously harmful but is not published; none of it is what Ware’s earlier account specified, a plain water bath, and the 2009 revision is an improvement on it rather than a variant of it. Ware describes the bath as giving “slight acidity (pH 4)”. The course has not measured it, and notes only that 2.5 g/L of citric acid in pure water calculates nearer pH 2.5 on its first dissociation constant; the figure to work to is the published one, and what matters photographically is that the bath is on the acid side rather than what its second decimal is.

Sodium thiosulfate, 25 g/L — the fixer that is also the toner. Its ordinary job is to complex the silver the light did not reduce and carry it out of the paper as [Ag(S₂O₃)₂]³⁻. Its second job, on Ware’s reading, is to lay a molecular skin of silver sulfide on the image particles, which is where the shift from red to mahogany comes from. Longer or stronger completes that conversion and destroys the image — the thirtyfold loss of optical density set out above — which is why the bath is short and weak and why Ware offers overlong immersion as a deliberate, if uneven, way of reducing an overexposed print. Shorter or weaker leaves unreduced silver in the sheet, which is a permanence fault rather than an aesthetic one. The bath is not stored and re-used, and its published capacity is about ten 10 by 8 inch prints per litre.

!Anhydrous or crystals — a difference of nearly 40 per cent

Ware’s chemicals table writes the formula Na₂S₂O₃ and adds that “hydrated form may be used”, giving no second weight; his procedure calls the material “crystals”, which in darkroom usage means the pentahydrate. The two are not interchangeable by weight. Twenty-five grams of the anhydrous salt (158.11 g/mol) is 0.158 mol; 25 g of the pentahydrate (248.19 g/mol) is 0.101 mol — 36 per cent less thiosulfate. His earlier web account asks for 20 g of “the crystals” per litre, which is 0.081 mol/L and lower again.

For an ordinary fixer that spread would not matter. For a bath whose whole design is to stop part-way through a sulfiding reaction it plainly might. The course records the ambiguity rather than resolving it, notes that all three readings fall between 0.08 and 0.16 mol/L — an order of magnitude below a working fixing bath — and suggests that anyone chasing a colour should record which salt they used.

Tween 20, about 0.25 to 0.5 per cent v/v in the coating solution, and not in the bottle. It is not in the quantity table above because this course has no chemical encyclopaedia entry for it, and because in the 2009 instructions it is not part of the sensitiser: it is added at the moment of coating. Its function is genuine and not cosmetic. Ware states that “the argyrotype sensitizer does not penetrate the cellulose fibres of paper effectively unless a surfactant (wetting agent) is present”, and that without one the nanoparticle silver may fail to be retained by the fibres and bleed out during wet processing, taking image density with it. More of it is the first remedy for bleeding; too much, or any at all on a gelatin-sized paper, may interact unfavourably. The right amount depends on the paper and is found by test.

Interactions

Acid with iron(III), which is the reason the whole formula is acid. Iron(III) hydrolyses. Left above about pH 4 in contact with cellulose it forms a polymeric colloidal hydroxide, and if it dries in that state it becomes an insoluble oxyhydroxide that will not come out at all. Holding the sensitiser at 3.5 and the clearing bath on the acid side keeps the iron in a form that can still be lifted. This is the interaction that decides whether the print is still there in twenty years.

Acid with silver citrate. Ware’s stated reason for the excess acid includes keeping the silver in solution, “otherwise silver citrate may precipitate out”. So the acid is holding two different equilibria at once, and the 6 g floor he gives is the point below which the second one starts to fail.

Acid with contrast, through the highlights. The mechanism Ware gives for the contrast control is that added acid “tends to dissolve silver in the highlights” — that is, the extra acid attacks the smallest and least aggregated silver particles, which are the ones in the faintest tones, so the toe of the curve is eaten away and the scale shortens. This is a chemical contrast control acting on the image rather than an optical one acting on the exposure, and it explains why it works on the highlights specifically.

Sulfamate against nitrate, in the presence of everything else. Set out in full under The mechanism. The short form: nitrate plus acid oxidises image silver; alkali stops that and hydrolyses the iron; sulfamate removes the conflict.

Thiosulfate with acid. The clearing bath is acid and the fixing bath follows it. Thiosulfate is decomposed by acid, and Ware’s own monograph notes that silver sulfide forms from thiosulfate and nanoparticle silver “especially under acidic conditions”. A print carried wet from an acid bath into a small volume of weak thiosulfate is carrying acid with it. That is one plausible reason why the bath’s capacity is only ten prints and why Ware forbids storing and re-using it; he does not say so himself, and the course marks the inference as its own.

Chlorine with the image, at the worst possible moment. Tap water in most supplies carries free chlorine. Hypochlorous acid oxidises silver, and the clearing bath is the moment when the silver particles are smallest, most exposed and least protected. Ware’s de-chlorination instruction and the citric acid are both aimed at this, and it is the reason he says “the purer the water, the better” even for the final wash.

Calcium with everything. Calcium carbonate buffered into an “acid free” paper neutralises the sensitiser’s acid where it meets it, which promotes exactly the iron(III) hydrolysis the formula is designed to avoid. Ware’s remedy is a pre-treatment soak in dilute hydrochloric acid at 2 to 5 per cent v/v, or in sulfamic acid at 5 to 10 per cent w/v, for ten minutes, followed by washing before coating.

Variants

The contrasty sensitiser. Dissolve an extra 1 g of solid sulfamic acid in each 100 cc of finished sensitiser. Ware presents this as a second formulation rather than an adjustment, and the useful part is the third option it creates: “By mixing this acidified sensitizer with the ‘standard’ one, the contrast can be ‘fine-tuned’.” Two bottles and a syringe give a continuous contrast scale. A softer version by dropping the acid to about 6 g per 100 cc is offered with a warning about precipitation.

The 1990s formulation, which is a different sensitiser in three respects. Ware’s web account of the process — undated, on a site whose copyright line runs from 1996 — differs from the 2009 handout in ways that a reader comparing the two should know about:

	Web account	2009 workshop handout
Glycerol	not present	1 cc per 100 cc, marked optional
Tween 20	0.2 cc, in the sensitiser	0.25 cc as a separate stock, added at coating
Water temperature for the acid	50 to 60 °C	about 70 °C
Stated working pH	2 to 3	about 3.5
Clearing bath	plain water, 5 minutes	de-chlorinated water with citric acid, 2 to 5 minutes
Thiosulfate bath	2 per cent, 3 minutes	2.5 per cent, about 2 minutes
Final wash	20 minutes	15 to 30 minutes
Sensitiser keeping	at least a year	several years

The course treats the 2009 handout as the formula, because it is later, fuller, and gives reasons for each of its changes. The differences are not recorded as a range: they are two printings by the same author, and the schema has no field for a disagreement between printings — the shortfall noted as open item 5 in docs/FORMULARY-SCHEMA.md. Where the two conflict, this page says so.

Alkaline clearing, for delicate highlights. The web account offers one variation the handout does not: “if very delicate highlight detail is desired, a little ammonia may be added to the clearing bath to make it distinctly alkaline (pH 9 to 10); this inhibits the dissolution of silver, but may raise the level of residual iron in the image.” That is the kallitype’s bargain, offered deliberately and with its cost stated. Anyone taking it is choosing highlight detail over permanence, and should know that is the trade.

Toning, if you want it. Ware’s position is that the process “does not prescribe a toning step”, because he believes the thiosulfate bath already sulfide-tones the image. But he records that “modern makers of Kallitypes and Van Dykes seem to agree that toning with platinum, palladium or gold is essential to their preservation, and Argyrotype is in the same category”, and gives the rule for doing it: like all print-out silver images, gold or noble-metal toning is done before the thiosulfate bath, not after. See the gold thiocyanate toner and the platinum toner for printing-out papers. He notes elsewhere that the image is receptive to the usual toning agents, selenium and gold among them, though possibly at lower concentrations than usual, and — writing in the first person — that he has not tested them all and would be glad of others’ experience. That is an honest gap and this page leaves it as one.

The relatives, for comparison. The Van Dyke Brown sensitiser carries the same iron salt and the ordinary silver nitrate, in three bottles, at a quarter of the silver. Herschel’s argentotype of 1842 is the ancestor of all of them and puts the silver on after the exposure. Nicol’s kallitype uses ferric oxalate and a developer. The iron-silver clearing sequence and the alkaline thiosulfate fixer are the general iron-silver processing entries; this formula’s own two baths are on this page because they are part of what Ware published and are not the general ones.

Safety

Level B, and the classification comes from the silver(I) oxide and the hot preparation together. The course’s rubric puts silver nitrate at Level B for its oxidising and caustic behaviour, and silver(I) oxide’s aggregated classification is at least as serious. Sulfamic acid is a skin and eye irritant; ammonium iron(III) citrate is Level A, an irritant; citric acid is an eye irritant; glycerol carries no classification at all. The formula takes the level of the highest of them, and the preparation adds a hazard none of the substances has on its own.

• Silver(I) oxide is classified as an oxidiser and as damaging to the eyes. PubChem’s aggregation of the ECHA C&L Inventory, from 409 company reports across 25 notifications, gives signal word Danger with the oxidiser, corrosive and environmental pictograms: H271, may cause fire or explosion, strong oxidiser (85.6 per cent of reports); H272, may intensify fire (13 per cent); H318, causes serious eye damage (88 per cent); and H400 and H410, very toxic to aquatic life with long lasting effects (85.3 and 87 per cent). A second, much smaller group of notifications sits on the same PubChem page and is not this substance: it carries EC 628-958-0 and CAS 11113-88-5, which ECHA names silver oxide (predom. silver(II) oxide), and its milder classification — H272, H315, H319, H335 — belongs to the silver(II) compound. The encyclopaedia entry classifies silver(I) oxide against EC 243-957-1 alone, and so does this page. Treat the solid as an oxidiser: keep it away from paper, wood, organic solvents and anything else reducing, and never put it in a wastepaper basket. Eye protection is not discretionary while it is being weighed or added.
• The preparation involves hot liquid and a long stir. Seventy degrees is hot enough to scald, and the silver oxide goes in over as much as an hour, which is a long time to be leaning over a beaker. Use a hotplate with a stable base or a bath of very hot water rather than a naked flame, keep the beaker on a tray, and stand back while adding solid to hot acid.
• The finished sensitiser is toxic and stains. Ware’s own words: “The solution is toxic and will stain skin and textiles”; and in the web account, “The sensitizer solution is irritant and toxic, and will stain skin and fabrics: wash away any spillages with plenty of cold water.” Gloves at every wet stage. The glove guidance covers the choice, and the silver nitrate handling SOP is the nearest procedure the course has for a silver solution of this strength; a silver stain on skin is metallic silver bound to protein and wears off rather than washing off.
• Weigh all three solids without raising dust. The citrate is deliquescent and its sheet asks for eye and face protection, gloves and exhaust ventilation where dust is formed. The weighing SOP is the routine.
• This is an ultraviolet process. Five minutes under a UV bank is a skin and eye exposure, and a UV lamp is more dangerous than sunshine because it is close and does not feel warm. Ware’s own note is that “an ultra-violet lamp must of course be shielded from accidental viewing”. Enclose the unit, use a timer and follow the UV unit SOP.
• If you make the oxide yourself, you are handling two more hazards. Solid sodium hydroxide is corrosive and its dissolution is exothermic, and silver nitrate is an oxidiser and a caustic. The alkali spill and silver nitrate spill procedures both apply.

⚠What is not a hazard here, and why

There is no cyanide chemistry and no chromium. The argyrotype’s contrast control is an acid, not a dichromate, which is a real difference from the Van Dyke kit and from Ware’s own New Cyanotype. The chromium ruling does not bear on this page at all, and nothing on it has been omitted under it.

No step evolves a gas you need to ventilate for. The photoreduction liberates carbon dioxide, but it does so inside the paper and in micromole quantities for a sheet. Ventilation here is a control for dust while weighing, not for vapour while working; sulfamic acid and citric acid are both solid, odourless and non-volatile, which is a large part of why Ware chose the first of them. The one exception is the silver carbonate route, where the effervescence is carbon dioxide and the instruction to use a tall vessel is about spray rather than gas.

The finished print is not a hazard. What is on it is metallic silver, in milligram quantities, among paper fibres. The hazards on this page all belong to the bottle and the tray.

And the dilute solution is not the solid. Seven grams of an oxidising solid in a hundred millilitres of water, on a sheet of paper, is not the object that came out of the jar. The oxidiser precautions above are for the jar and the balance; the sensitiser precautions are for a toxic, staining silver solution. Keeping the two apart in your head is how a hazard assessment is actually made.

Storage

One labelled brown bottle, in the dark, at room temperature, and it will outlive most of what else is on the shelf. Ware’s 2009 statement is that the sensitiser “should keep for several years”; his web account says at least a year. Either way this is the formula’s practical advantage over its relatives, and it exists because there is nothing in the bottle that reacts with anything else in it at room temperature — no dichromate to keep a reductant in check, no unstable complex, and no oxidising anion.

A small black precipitate is expected and is not the end of the bottle. Both printings say so, and both give the same remedy: re-filter it. What has come down is almost certainly silver, and the loss is small; what would be alarming is a large deposit or a solution that has gone from clear olive-green to turbid.

Date the bottle and record the citrate batch on the label. The iron content of ammonium iron(III) citrate varies between grades and suppliers, and the label is the only place that information survives. The labelling SOP and the batch record SOP are the routine.

Sensitised paper is a short-lived thing. Best used within a few hours of coating, though Ware says longer storage in a dark desiccated box is quite possible, and the web account gives at least a week in a dry environment. The reason for the desiccation is the citrate: it is deliquescent, and a coating that has picked up water is tacky enough to damage a negative in a contact frame.

Keep the Tween 20 stock separate and replace it when it shows a deposit. It does not keep well, which is one of Ware’s two reasons for not putting it in the sensitiser bottle.

Store the solid silver oxide as an oxidiser, away from combustible and reducing materials, in the original container, and not on a wooden shelf above the paper.

Incompatibilities

Alkali, at every stage after the sensitiser is mixed. This is the formula’s defining incompatibility and it is not a safety matter but a chemical one. Raising the pH above about 4 hydrolyses the iron(III) into a colloidal hydroxide that lodges in the paper. That rules out an alkaline wash aid, an alkaline fixer such as the course’s Reilly alkaline fixing bath, a carbonate-buffered water supply, and a chalk-buffered paper used untreated.

Chalk-buffered “acid free” papers, for the same reason. Ware’s remedy is decalcification: ten minutes in dilute hydrochloric acid at 2 to 5 per cent v/v or sulfamic acid at 5 to 10 per cent w/v, then washing, before coating.

Chlorinated water, at the clearing stage above all. Hypochlorous acid oxidises silver and the particles are at their most vulnerable while the print is in the first bath.

Hard water and dissolved iron. Calcium promotes the hydrolysis this process is built to avoid, and a supply that already carries iron is adding to the load the clearing bath has to remove.

Nitrate, from any source. It is worth stating explicitly on this page of all pages: do not top up an argyrotype sensitiser with silver nitrate, do not use a tray that has held a silver nitrate salt print sensitiser without washing it, and if you make your own silver oxide, wash the precipitate the three times Ware asks for. The whole design is the absence of that anion.

Anything reducing, near the solid oxide. Paper, wood, organic solvents, sulfur, powdered metals. The incompatibilities reference has the general rule.

Other alternative processes’ rods, trays and brushes. Iron and silver contamination shows up as spots and stains in a print with no binder to hide them: cross-contamination between alt processes covers it.

Waste

Every bath from this process is silver-bearing, and the first one is the richest. That is not the usual arrangement and it is worth being clear about. About 98 mg of silver goes on to a 10 by 8 sheet and only a small fraction stays in the image; almost all of the rest leaves in the clearing bath, as dissolved sulfamate and citrate complexes, before the thiosulfate has touched it. Do not treat the clearing bath as rinse water. Collect it with the fixer.

The fixer carries the silver that is easiest to recover, as the thiosulfate complex, and this is the stream conventional silver recovery is designed for. Ten 10 by 8 prints per litre is a concentrated bath by the standards of a domestic darkroom.

The iron is the other half of the load. Iron(III) citrate is not a heavy metal problem but it is an oxygen demand and a discoloration, and there is a great deal of it: 22 g per 100 cc of sensitiser, nearly all of it washed straight out again.

Silver is classified as very toxic to aquatic life with long lasting effects, which is the justification for all of the above and is on the silver oxide’s own hazard classification. The disposal guidance sets out the chemistry and the general practice, and the general chemical waste SOP the routine.

Local regulation governs, and this page does not know where you are. Collect, label, and dispose through the route your jurisdiction provides; check your local regulations before you pour anything away. Nothing here is a permission to drain any of it.

Troubleshooting

Highlights stay yellow after the clearing bath. Iron that has not come out. Extend the first bath, change it more often — Ware changes it for every print — make sure it is genuinely acid, and check whether the paper is chalk-buffered. Once the print has dried with iron in it the change may be irreversible. See iron-silver highlights that will not clear.

A red-brown stain runs off the dense areas in the first bath, and the print loses density. Bleeding: the fibres are not holding the nanoparticle silver. Ware’s diagnosis is insufficient Tween 20 or an unsuitable paper; his immediate palliative is to process face down so that the running silver does not stain adjacent areas, and he says plainly that there will be some density loss anyway. A contributing cause he names separately is drying the coated sheet too fast.

The print is much lighter dry than it looked wet. Expected, and large: at least one zone of dry-down, on top of the development that occurs in processing. This is a calibration problem rather than a fault. Expose for the dried print, not the wet one.

Highlights weaken during the fixer, or the whole print goes pale. Overlong immersion in thiosulfate, or a bath used past its capacity so that it has become a silver-loaded solution rather than a fixer. Two minutes, then out. Ware notes this can be used deliberately to reduce an overexposed print, “but maybe unevenly”, which is the honest description of every chemical reduction.

Highlight fog after steaming. Overlong post-hydration. Thirty seconds over water at 40 °C, or 30 minutes at 20 °C, and no longer.

The image is yellowish-brown rather than purplish. Either there is no glycerol in the sensitiser, or the paper dried too dry, or the ambient humidity is below the 50 to 90 per cent Ware asks for. The fastest test is the humidification route: 30 minutes above water before exposure should swing it towards purplish grey.

The negative comes out of the frame damaged or stuck. Two causes, both documented: excess sensitiser that was not blotted off and has crystallised on the surface, and a sheet humidified before exposure without a thin polyester interleaf. Ware asks for film under 20 microns for the second case.

A print made months ago has faded or gone yellow in a drawer. Residual iron oxidising the image silver, residual thiosulfate, or sulfur from the environment attacking a binderless silver image. See the iron-silver print that faded and residual silver and yellow staining; the diagnostic is whether the loss is in the highlights, the shadows, or everywhere.

Uneven, blotchy or streaked coating. A coating fault rather than a chemistry one: blotchy and streaked hand coating and mottled coating from humidity.

The sensitiser has thrown a black precipitate. Re-filter it, as Ware instructs. If it recurs quickly, or if the solution is no longer a clear olive-green, mix a new batch and record what was different.

Experiments

The acid contrast series. Make 100 cc of the standard sensitiser and 100 cc of the contrasty one (an extra 1 g of sulfamic acid). Coat five sheets with mixtures at 100:0, 75:25, 50:50, 25:75 and 0:100, expose all five under the same step tablet at the same exposure, and process them together. Read the number of distinguishable steps on each. The one variable is the free acid concentration; the control is the 100:0 sheet. This measures the only contrast control the process has, and it tells you which mixture matches your negatives.

The humidity and colour experiment. Coat six sheets from one batch. Dry three at room humidity and three in warm air. Expose one of each pair straight, one after 30 minutes at 100 per cent relative humidity, and one after the same humidification with a polyester interleaf. Process identically and compare the colours against a grey card in daylight. The prediction from the mechanism is that the more humid the coating during printing-out, the smaller and less aggregated the particles and the colder the colour; the point of the third pair is to check that the interleaf does not itself change the result.

The glycerol test. Two 50 cc batches from the same beaker, split before step 4, one with 0.5 cc of glycerol and one without. Everything else identical. This isolates the humectant, which is the one ingredient Ware marks optional, and it is the cheapest way to find out whether it earns its place on your paper in your climate.

The fixing-time series, which is really a sulfiding series. Print six identical sheets. Fix for 30 seconds, 1, 2, 4, 8 and 20 minutes in fresh 2.5 per cent thiosulfate, wash all six identically, dry and read maximum density and colour. The mechanism predicts a rise then a fall: enrichment as the sulfide monolayer forms, then loss as conversion completes, with the highlights going first. If you can borrow a densitometer, this is the experiment on this page most likely to produce a curve worth publishing — and if the fall is much smaller than the thirtyfold that complete sulfiding would give, that tells you the reaction never went anywhere near completion, which is Ware’s claim.

The anhydrous-against-crystals control. Make two fixing baths at 25 g/L, one from the anhydrous salt and one from the pentahydrate, and run the fixing-time series above in both. The molar strengths differ by 36 per cent. If the colour and the density track the molar strength rather than the weight, the ambiguity in the published instruction matters and should be resolved on the label of your own bottle.

The nitrate control, which is the argument of the whole page. Coat two sheets from the standard sensitiser. Before drying, add one drop of a dilute potassium or sodium nitrate solution to a marked corner of the second. Expose, process and compare that corner with the same corner of the control. This is a qualitative test and it is meant to be: what it can show is whether an oxidising anion, at this pH, visibly costs shadow density in a print-out silver image. Record the concentration you used, because the interesting result is the one where nothing happens.

📓What to record

Batch and supplier of the ammonium iron(III) citrate, and its stated iron content if the jar carries one. Which sodium thiosulfate — anhydrous or pentahydrate. Ambient temperature and relative humidity at coating, at drying and at exposure. Coating volume per sheet and number of rod passes. Exposure source, distance and time, and the appearance of the print-out image at the moment you stopped. Clearing bath time and whether the water was rainwater, filtered or boiled. Fixing time. Wet appearance against dry appearance, in stops. Nothing else on this list is as useful as the last one.