Classic cyanotype sensitiser

“Classic cyanotype” names a pair of chemicals, not a pair of numbers. Ware compared about sixty published recipes spanning a hundred and fifty years and found them scattered across a wide field of concentrations, with the twentieth-century ones clustering around 13 per cent ammonium iron(III) citrate and 6 per cent potassium ferricyanide in the mixed sensitiser — and with the process working “satisfactorily almost regardless of the proportions chosen”. The weights below are his own, from the section he heads The Classic cyanotype process 1842/1897, and the date in that heading is the whole story of the page: Herschel’s two chemicals of 1842, made up with the green form of the iron salt that Eduard Valenta first prepared in 1897.

Stock solution A — the light-sensitive half

Ingredient	Quantity	Form the source specifies
Ammonium iron(III) citrate	20 g	the green variety, whose iron content Ware gives as 14 to 18 per cent; dissolved in about 70 mL of purified water at room temperature and made up to 100 mL
Water	to make 100 mL	Purified water — distilled, de-ionised or pharmaceutical. Ware's own text allows the same 20 g to be made up to as much as 200 mL, which halves the strength of this bottle; see Variants.

Stock solution B — the precipitant

Ingredient	Quantity	Form the source specifies
Potassium ferricyanide	10 g	General Purpose Reagent grade, 98 to 99 per cent, is adequate; dissolved in about 80 mL of purified water at room temperature and made up to 100 mL of a 10 per cent w/v solution
Water	to make 100 mL	Purified water. Ware states the resulting strength in the same sentence as the make-up volume, which is the useful habit: "make up to 100 cc of a 10% w/v solution".

Mixed in the ratio — the sensitiser, mixed immediately before coating

1 part Stock solution A + 1 part Stock solution B

Before use, mix equal volumes of A and B to give a total volume of sensitizer appropriate for the amount of coating to be done. The surfactant Tween 20 is added at this point. The mixed sensitizer only has a relatively short life. It is best to dedicate a separate, labelled syringe or pipette to each solution, and to use a third one for applying the mixture, in order to avoid cross-contamination during the work.

Halving both stocks, the mixed sensitiser is 10 per cent w/v ammonium iron(III) citrate and 5 per cent w/v potassium ferricyanide. That is more dilute than Ware's own survey average for green-salt recipes, 13 and 6 per cent, and it is deliberate: §7.1.2 records that a modern reappraisal using contemporary papers now favours more dilute sensitisers for image quality.

Purpose

To make a negative-working print in Prussian blue on plain paper, from a contact exposure to ultraviolet light, and to develop and fix it in water. No silver, no fixer, no developing agent: everything that light has not converted is soluble, so the wash that brings up the blue is also the wash that removes what would otherwise go on printing.

It is two bottles rather than one for a single reason, and the reason is worth stating before the recipe. The mixed sensitiser has a short life; the separate stocks keep for months. Every practical inconvenience of this formula — the two syringes, the mixing immediately before coating, the small volumes — follows from that one fact.

Recommended uses

Contact printing from a negative of short density range. The exposure scale of the classic process is about 0.9, which is three stops, and Ware compares it directly with a grade 3 silver-gelatin paper. A negative made for ordinary enlarging paper of grade 2 or 3, or for gum dichromate, prints well; a negative made for platinum or palladium, with a density range of 2 or more, cannot be rendered at all. The contrast of the negative is adapted to the process, not the other way round.

Photograms, sun printing and work with three-dimensional objects, where the length of the exposure is not a problem and being able to watch the image appear is an advantage.

Fabric and other cellulose materials. The sensitiser is applied to cloth by soaking or coating; Jacquard’s own instructions cover cotton, and the same chemistry works because it is the cellulose fibre that has to hold the pigment.

Teaching, and first work in the iron processes. This is the whole iron-based family in its simplest form: light makes iron(II), and something in the paper reacts with that iron(II) to make an image. Change the something and you get the platinotype, the kallitype or the argentotype. Ware’s own assessment is that the chemicals of this formula are not dangerous and can safely be handled by children under supervision, and he says elsewhere that he has no misgivings placing the traditional process in the hands of five-year-olds. The course’s own classification is more cautious and the reason is given under Safety.

Not where reproducibility between workers matters. Ammonium iron(III) citrate is what chemists call an ill-characterised substance: its iron content ranges from 14 to 28 per cent by weight between samples, no single molecular formula can be written for it, and Ware’s blunt conclusion is that no two people working apart can be sure they are using the same chemical.

When another formula is preferable

• For the historical article rather than the modern one, Herschel’s 1842 cyanotype, whose proportions come from his manuscript Memoranda and are the slowest and most ferricyanide-heavy of any published cyanotype.
• For whiter highlights, at the cost of a carcinogen, Ware’s improved Classic cyanotype, which adds oxalic acid and ammonium dichromate. Ware prints it and immediately argues against it: if you are prepared to go to that trouble it would probably be more worthwhile to use the New cyanotype instead, and the improved sensitiser “should not be placed in the hands of children”.
• For a short exposure, a long exposure scale and a maximum density verging on black, the New Cyanotype of 1995, which replaces the citrate with an oxalate. Its exposure scale is around 2.2 against this formula’s 0.9, so one long-scale negative can serve it, salted paper, platinum-palladium and argyrotype alike.
• For contrast control by formulation, the Simple cyanotype or Mike’s cyanotype, which are made to a chosen exposure scale of about 2.7, 2.3 or 1.8 rather than to a fixed one.
• If you would rather not weigh anything, a supplier’s kit: the Photographers’ Formulary and Bostick and Sullivan entries carry those formulations and their own instructions.
• Never by mixing this sensitiser with New cyanotype sensitiser. Ware calls that “a chemical nonsense” and explains why under Incompatibilities.

Mixing

Two bottles, made up separately, combined only when you are about to coat. Neither solution needs heat, and neither has a mixing order, because each has exactly one solute — which is unusual in this formulary and is one of the reasons the process spread so far so fast.

1. Stock solution A. Dissolve 20 g of solid ammonium iron(III) citrate, green variety, in about 70 mL of purified water at room temperature, and make up to 100 mL. Float a few crystals of thymol on the surface if the bottle is to stand for more than a week or two; they do not dissolve, and you avoid them when drawing a sample.
2. Stock solution B. Dissolve 10 g of solid potassium ferricyanide in about 80 mL of purified water at room temperature and make up to 100 mL of a 10 per cent w/v solution.
3. Both into brown bottles, in a cupboard or a box. Label and date them.
4. Mix equal volumes of A and B immediately before coating, and mix only as much as the session needs. Ware’s guide to the volume is about 1.5 mL for a 10 by 8 inch print coated by rod, and possibly twice that by brush.
5. Dedicate a separate labelled syringe or pipette to each solution and a third to the mixture. This is not fastidiousness: a pipette that has been in the ferricyanide and goes back into the citrate bottle starts the dark reaction inside your stock.

Coat in dim light — the sensitiser is sensitive to ultraviolet only, so ordinary incandescent or LED room lighting is safe and no safelight is needed — and dry in the dark.

🧮What the mixture actually contains, and why every printed version is a different number

Per cent w/v means grams per 100 mL and nothing else, so the stocks are 20 and 10 per cent. Mixing equal volumes halves both:

20 % ÷ 2 = 10 % citrate · 10 % ÷ 2 = 5 % ferricyanide

The mixed sensitiser, from Ware's §7.1

Now compare that with the field Ware surveyed. His green-salt recipes average 13 per cent citrate with 6 per cent ferricyanide in the final mixture, and the stock strengths he reports as commonly recommended for the green salt are 26 and 12 per cent, which halve to exactly those averages. So §7.1 is a deliberately more dilute sensitiser than the historical norm — about three-quarters of the average strength — and §7.1.2 gives the reason: a modern reappraisal of the process on contemporary papers now favours more dilute sensitisers for the best image quality.

The same section quotes Christina Z. Anderson going further still. She reports having always used a “20/8” mixture, meaning 20 per cent citrate stock with 8 per cent ferricyanide stock, and then finding she preferred 10/10 — 10 per cent of each, mixed one to one, giving 5 per cent of each in the sensitiser. That is half Ware’s citrate. Her stated reasons are a long exposure scale, little grain and a more turquoise blue.

Ware’s own instruction for solution A quietly spans the whole of that argument: the 20 g may be made up “to 100 cc — or more, up to 200 cc”. At 200 mL the bottle is a 10 per cent solution and the mixture is Anderson’s.

The Tween 20 that is in Ware’s chemicals list is not in the formula above, and he says so himself. His §7.1.1 lists 1 mL of a 20 per cent v/v solution of the non-ionic surfactant alongside the two salts, but §7.1.3 introduces it with the words “although it is not part of the traditional formula, I would strongly recommend adding a few drops (2 to 5) of the 20% solution of Tween 20 wetting agent to each 10 cc of sensitizer, to improve its absorption by the paper fibres”. It is his innovation, the appropriate amount depends on the porosity of the paper and is to be found by trial, and this course has no chemical encyclopaedia entry for it, so it is not entered as an ingredient. Two things about it are worth knowing before you use it. First, the two figures Ware gives do not agree with each other: 2 to 5 drops per 10 mL, at the usual 20 drops to the millilitre, is 2 to 5 mL of the 20 per cent solution across the 200 mL his chemicals list makes, not the 1 mL that list states. Second, the drops figure is the one that matches his practice elsewhere — 2 to 5 drops of a 20 per cent solution per 10 mL works out at 0.2 to 0.5 per cent Tween 20 in the sensitiser, which is exactly the 0.25 to 0.5 per cent he specifies for his own Simple cyanotype. The course reports the discrepancy and does not resolve it.

Nothing here is critical. Ware’s preface to the preparation is worth quoting because it is so rarely said in a formulary: the concentrations “are not critical so they do not need to be made up with great precision”. Given the variability of the iron salt itself, weighing to a tenth of a gram would be false precision.

Behaviour

It prints out. The image forms during the exposure, which is why cyanotype was the drawing office’s copying process for eighty years: the print can be inspected in a split-back frame and judged, rather than timed blind.

It is supposed to solarise, and the reversal is your exposure meter. The uncovered sensitiser darkens to blue and then reverses back to a pale blue-grey. Ware’s end point for a correct exposure is that the shadows look somewhat reversed, giving a solarised appearance, the highlights are distinctly green, and the image as a whole looks rather blocked up. A print that looks right in the frame will be too light when dry.

Then the water does three things at once. It completes the formation of Prussian blue, it dissolves out every unexposed trace of both salts, and — this is the formula’s defining weakness — it carries away a significant part of the image. Ware is explicit: there will be considerable leaching of blue pigment, and the extended washing needed to clear the paper will significantly reduce the image and truncate the tonal scale in the high values.

Twenty minutes face down in gently running water is his figure, or several changes of static water, until the yellow has gone from the highlights. Manufacturers’ sheets give shorter times for their own products — five minutes in both the Photographers’ Formulary and Jacquard instructions — and the Photographers’ Formulary sheet states the trade-off in one sentence: a short wash leaves ferric salts in the paper and the print will fade, while prolonged washing lightens the image, particularly if the water is slightly alkaline.

A weak acid bath before the water strengthens the scale, and Ware notes it for the classic process without publishing a time: development first in a very dilute weak acid such as 1 per cent citric acid rather than plain water. His warning comes with it — too strong an acid bath blues the highlights and degrades the paper white — and elsewhere he sets the ceiling for the citrate sensitisers at under 1 per cent citric, above which stronger acids cause blue chemical fog. The half-to-one-minute development time that appears in Cyanomicon is published for his Simple cyanotype, not for this formula, and is not transferred here.

It goes on darkening for hours after it comes out. The reversed shadow tones regain density by aerial oxidation of Prussian white back to Prussian blue during drying. Judge nothing until it is dry. If you cannot wait — for densitometry, say — half a minute in 0.3 per cent hydrogen peroxide before the final wash brings the density up at once, and Ware states that this makes no difference to the final result.

Image characteristics

Colour: Prussian blue, brighter than the brown salt gives. Ware’s comparison table puts the image colour of the brown salt at “dull blue” and the green at “brighter blue”. Anderson, at her more dilute 10/10, reports a more turquoise blue again — so within this one formula the hue is a function of concentration as well as of which salt you bought.

Exposure scale: about 0.9, which is three stops. This is the number that decides what negatives the process can print, and it is small. It is not an intrinsic property of Prussian blue; it is largely a consequence of the wash, as the next section explains.

Maximum density: limited, and limited twice over. Solarisation caps what the exposure can build, and peptization removes part of what was built. Ware’s whole list of remedies in §7.2.4 is a list of things the classic sensitiser cannot do, against a sensitiser that can.

Highlights: the first casualty. Loss of gradation in the high values, and in heavily exposed areas the opposite problem — excess iron(II) diffusing sideways and staining adjacent highlights blue, the defect Ware calls bleeding.

Permanence, measured rather than asserted. Ware exposed step tablets made with five different cyanotype sensitisers to artificial daylight and measured the loss of reflectance density, expressed as the fade, 100 times the density change. The row that corresponds to this formula is the one he labels Valenta — typical twentieth-century recipe, 14 per cent green citrate with 12 per cent ferricyanide in the mixture:

Sensitiser tested	Maximum fade, 100 ΔD	Over initial density range
Smee — ferricyanide alone	13	0.90–0.95
Herschel — his own modified recipe, aquapel-sized paper	11	0.55–1.05
Lietze — typical nineteenth-century recipe, brown salt	10	0.65–0.85
Valenta — typical twentieth-century recipe, green salt	10	0.35–0.85
Ware — New Cyanotype	22	0.85–1.15

Two readings of that table matter. The classic green-salt sensitiser is at the good end, tied with the brown-salt recipe and better than either Herschel’s own or the New Cyanotype. And the fade peaks in the mid-tones, across a broad plateau, not in the deepest blue — so a permanence test that measures only maximum density will report a stability the picture does not have. Note also that Ware’s summary comparison of the two salts calls the brown “fairly resistant” and the green “more susceptible” to light fading, while his own measured maximum fades for the two are the same number. The course reports both and does not reconcile them.

Washing loss, also measured. Sets of nearly identical strips were washed in gently running tap water at pH 7.2 and in static purified water at pH 6.5, and remeasured after 5 minutes, 20 minutes, 1 hour and 4 hours. The losses are again 100 ΔD:

Sensitiser and water	5 min	20 min	1 hour	4 hours
Valenta, tap water	1.2	3.9	9.6	30.7
Valenta, purified water	0.7	1.0	2.2	—
Herschel, tap water	0.4	2.3	6.4	13.9
Herschel, purified water	0.1	0.4	2.2	3.2
Lietze, tap water	0.1	2.9	7.0	17.5

The green-salt formula is the most vulnerable of the three in tap water and among the least vulnerable in purified water. The variable that matters is the water, not the formula, and the factor between the two columns at four hours is more than tenfold. Earlier work Ware cites points the same way: Ian and Angela Moor reported an average image density loss of 18 per cent after 15 minutes in tap water at pH 7.5–8.5 against 4 per cent in distilled water, and Sarah Wagner at the Library of Congress reported dramatic lightening in tap water and slight lightening in deionised water.

Surface: none. There is no binder. The image sits in and on the paper fibres, which is why the paper’s sizing changes the result more here than in any silver process.

The mechanism

⚗One photochemical step, one chemical step, and a third if you overexpose

Step one: light reduces the iron, and the citrate pays for it. All the photochemistry belongs to the iron salt. The photoactive species is a dimeric citratoferrate(III) anion, which forms above about pH 2; the monomer that predominates below pH 1.5 is photo-inactive, which is why the pH of an iron sensitiser is a real variable and not a detail. Ware gives the quantum yield at 365 nm as 0.45 at pH 4, falling to 0.28 at 436 nm — that is, nearly half of every absorbed near-ultraviolet photon does chemical work.

He writes the change as a scheme, not a balanced equation, and says why in the same sentence: the organic photoproduct has been identified as acetone dicarboxylic acid, which may decarboxylate further to acetone, but the nature of the iron(II) photoproduct remains unknown.

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

Ware's scheme for the photolysis

This course does not print that as an equation with a balanced left and right, because completing it would mean naming an iron species the source declines to name. What is established is the direction and the cost: light absorbed, iron(III) reduced to iron(II), citrate oxidised, carbon dioxide lost.

Step two: the iron(II) precipitates the blue, and no light is needed. It is a plain electron transfer between two iron centres, which is what makes the whole iron-based family work.

Fe²⁺ + [Fe(CN)₆]³⁻ → Fe³⁺ + [Fe(CN)₆]⁴⁻

The oxidation states switch

The iron(III) and the hexacyanoferrate(II) then precipitate together as Prussian blue. Ware is careful about the naming: the intermediate ferrous ferricyanide is highly unstable and transforms very rapidly to ferric ferrocyanide, so what looks like a Turnbull’s blue reaction gives Prussian blue, and the two names are the same substance.

Step three, on over-exposure: the blue is reduced in its turn.

light + Prussian blue + ferrous ions → Prussian white + ferric ions

Solarisation

Ware makes the point that this is not simply a loss. A printed-out blue masks itself: the pigment already formed absorbs the light that would form more. When the shadows reverse to white, that masking is reduced, more light gets in, more product forms, and the final density after reoxidation can be higher than it would have been. This is why the correct exposure looks over-exposed.

And back again, in the tray or on the line.

Prussian white + hydrogen peroxide → Prussian blue + hydroxide ions

Reoxidation

Air does the same thing more slowly, over several hours of drying.

Deeper: why the exposure scale is only 0.9

The short scale is not a property of Prussian blue and it is not caused by the ferricyanide. It is caused by the wash, and the argument is worth following because it is the single most useful thing on this page.

The image substance the light makes is a nanoparticle pigment. If the sensitiser has soaked into the interfibrillar space inside the surface cellulose fibres, those particles are trapped where they form. If it has only sat in the coarse pores between the fibres — which is what happens on a heavily sized modern paper, because ammonium iron(III) citrate is a large, polymeric, poorly penetrating molecule — then the wash takes the particles away. It takes proportionally more from the thin deposits than the thick ones. The light tones therefore lose a larger fraction of their density than the shadows do, the top of the curve collapses towards the paper white, and what is left is a short, contrasty scale. Ware’s own words for the result are “a serious loss of gradation in the high values, truncating the tonal scale, and conferring an artificially high contrast”.

That also explains the two remedies practitioners reach for, and why neither is satisfactory. Gross over-exposure puts down more pigment everywhere so that more survives, at the cost of image quality and of any ability to judge the exposure. Double coating puts down more sensitiser, with the same objection. The real remedy is to get the sensitiser into the fibre, which is what the surfactant is for, and the structural remedy is a smaller molecule, which is what the New Cyanotype is.

Deeper: how much iron is in excess, and why the highlights bleed

Ware lists “bleeding” — excess iron(II) diffusing out of heavily exposed areas and staining the adjacent highlights blue — as the sixth shortcoming of the classic sensitiser, and says of the oxalate replacement that it “can be mixed in chemically equivalent amounts with ferricyanide to avoid excess iron(II)”. The implication is that the classic sensitiser is not in chemically equivalent amounts. The arithmetic below is this course’s, from Ware’s own figures, and it says how far out it is.

🧮The molar ratio of iron to ferricyanide in the mixed sensitiser

The mixture holds 10 g of the citrate and 5 g of ferricyanide per 100 mL. Ferricyanide is a defined compound of molar mass 329.24 g/mol, so:

50 g/L ÷ 329.24 g/mol = 0.152 mol/L

Ferricyanide

The iron salt is not a defined compound, so it has to be reckoned through its iron content, which Ware gives for the green form as 14 to 18 per cent by weight in present-day commercial material:

100 g/L × 0.14 ÷ 55.85 g/mol = 0.251 mol/L

Iron, at 14 % Fe

100 g/L × 0.18 ÷ 55.85 g/mol = 0.322 mol/L

Iron, at 18 % Fe

So there are between about 1.7 and 2.1 moles of iron(III) for every mole of ferricyanide, and which of those two numbers applies depends on a batch specification most suppliers do not print. The electron transfer that makes the image is one iron(II) to one ferricyanide. Long before every ferricyanide ion has been used, there is still iron(II) being made, and it is free to diffuse.

Two consequences follow. Where the exposure is heavy the surplus iron(II) migrates and blues the highlight beside it. And the ratio the reader actually has is uncertain by a factor of 1.3 even when both weights are correct, which is a large part of why two printers following the same recipe report different prints. Ware’s suggestion for the brown salt — add about one third of its weight of citric acid, to bring the citrate-to-iron ratio closer to that of the green form — is a partial correction of the same problem from the other direction.

Deeper: what Valenta’s green salt actually changed

The only significant chemical advance in the entire history of the process, on Ware’s account, was Valenta’s preparation in 1897 of a green, somewhat acidic form of ammonium ferric citrate in place of the brown, basic form that had been used since about 1840. The difference is the iron-to-citrate ratio: Valenta’s analyses give the brown form about 19 per cent iron and the green about 14 per cent, so the green salt carries more citrate per iron atom.

Property	Brown salt	Green salt
First preparation	about 1840	Valenta, 1897
Iron content	19–28 %	14–18 %
Nature of the salt	basic	acidic
Speed of printing	slow	about one stop faster in the mid-tones
Solarisation	relatively slight	stronger, and a higher maximum density
Edge etch effect	distinctive	absent
Image colour	dull blue	brighter blue
Light fading	fairly resistant	more susceptible

⁇Eight times faster, one stop faster, or no different at all?

The disagreement. The claim made for the green salt on its introduction was that it gave a sensitiser eight times faster than the brown. Ware’s own table says about one stop — a factor of two — in the mid-tones. The Getty Conservation Institute’s atlas says only that the green form “was found to increase the light sensitivity” and adds that the final chemical composition of the image is identical whichever form was used.

Who holds what. The eightfold figure is the nineteenth-century claim, reported by Ware from the period literature. Ware himself qualifies it in the same paragraph: the apparent speed depends on the criterion by which it is assessed, and there is less speed difference in the mid-tones than in the shadows, where solarisation is at work. The Getty’s statement is about the image, not the speed, and is not in conflict with either.

What this course uses and why. Ware’s one stop in the mid-tones, because it is stated with the criterion attached and the criterion is the whole of the difference between the two claims. An eight times faster shadow is a shadow that has solarised sooner, which is a different fact from an eight times faster print. Where a reader needs a number for planning an exposure, one stop is the honest one.

Function of every ingredient

Ammonium iron(III) citrate, green variety, 20 g per 100 mL of solution A, giving 10 per cent w/v in the mixed sensitiser. What it is: the ammonium salt of an iron(III)–citrate complex, and not a compound in any strict sense. It is made by precipitating ferric hydroxide, which is a highly polymeric and often colloidal substance, and dissolving it in citric acid; the product cannot help being variable, and its iron content is found anywhere between 14 and 28 per cent by weight. Chemists call it ill-characterised, and that word is a verdict. It reached photography from pharmacy, where it was Ferri et Ammoniae Citras, an iron tonic. Why it is here: it is the only light-sensitive substance in the formula. What it does: absorbs near- ultraviolet and blue light, and its iron(III) is reduced to iron(II) at the expense of the citrate ligand, which is oxidised to acetone dicarboxylic acid with loss of carbon dioxide. Photographic consequence: it sets the speed, and — because the surviving iron(II) is what the ferricyanide converts — it also sets how much blue there can be. More of it: faster printing, more blue available in the shadows, and more of the excess iron(II) that bleeds into the highlights; on Whitaker’s 1883 testing, reported by Ware, an excess of the citrate also shortens the storage life of coated paper. Less of it: slower, cleaner highlights, and — this is Anderson’s report — a longer exposure scale, less grain and a more turquoise colour. What it interacts with: the ferricyanide, which it must not meet until you are ready to coat; the paper, which it penetrates badly; humidity, because it is deliquescent and the coating goes tacky; and any alkali, which will attack the image it forms.

Potassium ferricyanide, 10 g per 100 mL of solution B, giving 5 per cent w/v in the mixed sensitiser. What it is: the red prussiate of potash, potassium hexacyanoferrate(III), a defined crystalline compound of molar mass 329.24 — which makes it the only substance in this formula you can weigh with confidence. General Purpose Reagent grade, 98 to 99 per cent, is adequate. Why it is here: to be the precipitant. It is unchanged by the exposure and waits for the iron(II) that the light makes. What it does: accepts an electron from iron(II), becoming hexacyanoferrate(II), while the donor becomes iron(III); the pair precipitates as Prussian blue. It has a photochemistry of its own — Herschel’s first cyanotype used it alone — but with quantum yields around 0.01 against the citrate’s 0.45 it contributes almost nothing to the speed here. Photographic consequence: it sets the ceiling on how much blue can form, and it simultaneously slows the print, because it absorbs blue and ultraviolet light strongly and so acts as an internal filter between the light and the molecules that can use it. More of it: a deeper potential blue and a longer exposure; Whitaker found that an excess gives a finer quality and greatly lengthens the exposure. Less of it: faster printing, a lower maximum density, and a larger surplus of iron that bleeds. What it interacts with: acid, which must never reach it; light, which slowly decomposes it in the bottle; and every silver material in the building, because ferricyanide is the bleach in Farmer’s reducer.

Water, twice, and it is not a neutral ingredient. Why it matters: both stocks are salt and water, so the water defines the strength; and the water of the wash decides how much of the image survives, which the table under Image characteristics quantifies. Ware and every supplier sheet read for this page specify purified water — distilled, de-ionised or pharmaceutical — for making the stocks. What happens with more or less: the proportions are the formula. A stock made up to the wrong volume changes both the speed and the maximum density, and there is no development stage later in which to compensate. What it interacts with: calcium. Hard water damages the Prussian blue image, and alkaline water above pH 7 must not be used at any stage. Tap water at pH 7.2 cost more than ten times as much image density over four hours as purified water at pH 6.5 in Ware’s own measurements.

What the traditional formula does not contain, and why each omission has consequences. There is no wetting agent, which is the single most-recommended modern addition: Ware’s 2 to 5 drops of a 20 per cent Tween 20 solution per 10 mL of sensitiser, giving 0.2 to 0.5 per cent, exists to get the sensitiser inside the fibre rather than into the pores between fibres, and so to stop the image washing away. There is no acid, so the development is the long, soft, low-density one that plain water gives; 1 per cent citric acid is the standard modern refinement and its own risks are set out under Behaviour. There is no preservative, so the citrate bottle grows mould, and thymol is the usual answer. There is no dichromate, and there must not be: see Safety. And there is no fixer, which is the point of the whole process — everything unexposed is soluble in water.

Interactions

Solution A with solution B, before you want them to meet. They mix without precipitating, because the iron(III) is locked up in the citrate complex and is not available as free Fe³⁺ to react with ferricyanide. Nothing stops the slow dark reaction between them; the complex only slows it. That is why the mixed sensitiser has “a relatively short life”, why the formula is two bottles, and why a supplier who premixes nothing tells you the sensitiser is stable for 2 to 4 hours.

Iron(II) with ferricyanide, which is the image. Instantaneous, and independent of light.

Light with the Prussian blue already formed, which is solarisation during the exposure and fading afterwards. They are the same photoreduction on two timescales. Ware’s explanation of why pure Prussian blue does not bleach while the Prussian blue of a cyanotype does is that an oxidisable impurity in intimate contact with the pigment — trapped in its lattice — gives the excited state time to accept an electron instead of relaxing thermally.

Air with the print, which finishes it, over several hours of drying; hydrogen peroxide at 0.3 per cent for half a minute, which finishes it in half a minute and changes nothing else.

Alkali with the finished image, which destroys it. This is the most important interaction on the page. Ware measured the effect of a pH 9.4 buffer for ten minutes on prints from five sensitisers; the green-salt formula lost 94 hundredths of a density unit, 71 per cent of its image, the worst of the five. Prussian blue is hydrolysed irreversibly by alkali, and pH 9.4 is roughly the pH of a saturated calcium carbonate solution — the buffer routinely added to archival board and to many conservation-grade papers. A cyanotype needs an unbuffered sheet and an unbuffered mount.

Calcium with the image, more slowly and by the same route. Hard wash water carries it; so does hard tap water used to make up the stocks, which is why every sheet specifies distilled water.

Phosphates and detergents with a finished print on fabric, which is the same alkali problem arriving through the laundry. Jacquard’s own note is to wash cyanotype fabrics in cold water with non-phosphate detergent, and warns that sweat and hand oils can discolour prints too.

Citric acid with the sensitiser during development, which strengthens the tonal scale, and — above about 1 per cent — fogs the highlights blue and degrades the paper white.

New Cyanotype sensitiser with this one, which is the interaction to avoid entirely. See Incompatibilities.

Variants

There is no canonical set of numbers here, so the honest way to present the alternatives is as a table of what has actually been printed. All concentrations are per cent w/v; the “in the mixture” columns assume the equal-volume mixing that every one of these sources specifies.

Printing	Stock A, citrate	Stock B, ferricyanide	In the mixture
Ware, Cyanomicon §7.1 (this page)	20 % green	10 %	10 % / 5 %
Anderson’s preference, quoted by Ware	10 %	10 %	5 % / 5 %
“20/8, as recommended by the books”, quoted by Ware	20 %	8 %	10 % / 4 %
Photographers’ Formulary kit 07-0090	20 % (100 g to 500 mL)	8 % (40 g to 500 mL)	10 % / 4 %
Jacquard detailed instructions	about 25 % (100 g in 400 mL)	about 10 % (40 g in 400 mL)	about 12.5 % / 5 %
Ware’s reported common recommendation, green salt	26 %	12 %	13 % / 6 %
Ware’s reported common recommendation, brown salt	20 % brown	16 %	10 % / 8 %
Ware’s survey average, green-salt recipes	—	—	about 13 % / 6 %
Ware’s survey average, brown-salt recipes	—	—	about 10 % / 8 %
Whitaker’s 1883 optimum, brown salt	—	—	15 % / 10 %
Herschel’s own, 13 August 1842	9 % brown	33 %, saturated	4.7 % / 16.6 %

A note on the Jacquard row. Its instructions give the water as an amount added — “40 g in 400 mL water” — not as a make-up volume, so the strengths are nominal rather than stated. That distinction runs through this whole formulary and it is not pedantry: a make-up volume fixes a concentration, and an added volume does not.

Three things that table shows. The ferricyanide varies by a factor of four across the modern printings and by a factor of two among sources that all call themselves classic. The historical drift is unmistakable — more citrate, less ferricyanide, decade by decade — and Herschel’s own mixture is at the far end of it. And Ware’s §7.1, the formula on this page, is at the dilute end of the modern range rather than in the middle of it.

Ware’s improved Classic sensitiser is a separate entry, here: 25 g of green citrate with 0.5 g of oxalic acid in solution A, and 10 g of ferricyanide with a further 0.5 g of oxalic acid and 0.2 g of ammonium dichromate in solution B. It is claimed to keep the highlights whiter. Ware prints it, says it is widely circulated and that he does not know where it originated, and states that it should not be placed in the hands of children.

The brown salt with added citric acid. Where only the brown form is available, Ware suggests adding about one third of its weight of citric acid to the solution, to bring the citrate-to-iron ratio closer to that of the green form. He gives it as an advantage rather than a formula, and no quantities for a finished sensitiser accompany it.

A first bath of dilute ferricyanide, for more contrast. The Photographers’ Formulary sheet gives an initial wash bath of 0.2 per cent potassium ferricyanide — 2 g per litre — before the running-water wash, to raise contrast, and a 5 per cent oxalic acid solution for spot-clearing blue from the whites. Both are treatments rather than changes to the sensitiser.

No course variant of these proportions is offered. There is nothing to make safer by changing them, and the better-documented alternatives — New Cyanotype, Simple cyanotype and Mike’s cyanotype — already exist as separate formulas with their own evidence.

Safety

Level B, and the classification comes from the potassium ferricyanide alone. Ammonium iron(III) citrate is Level A and Prussian blue is Level A. Ware’s own assessment is more relaxed than the course’s: he rates the citrate 1 for health on the four-point HMIS scale, notes that it has been prescribed as an iron tonic and is a permitted food additive, states that the chemicals of this formula are not dangerous and can be safely handled by children under supervision, and writes that “from all points of view, there is no safer photographic process”. The course keeps each chemical’s own page as the governing document and takes the formula’s level from the highest of them, which puts this one at B. Where a source and this course differ, the difference is stated rather than smoothed.

The specific points that follow, none of which replaces the chemical pages:

• Ferricyanide is a cyanide complex, not a cyanide salt. The cyanide is bound to iron and is not free. It is not the hazard that potassium cyanide is, and the two must never be spoken of as the same thing.
• It becomes a real hazard in contact with acid, which is why acid must never enter this workflow. See Incompatibilities. Ware’s qualification of the frequently quoted warning is worth having exactly: the heating has to be very strong and the acid very concentrated to get much hydrogen cyanide, and neither is done in cyanotype; the traditional process using only ammonium ferric citrate and potassium ferricyanide “will not emit any significant, perceptible or harmful amount of hydrogen cyanide during a normal exposure and processing”.
• The exception is the ferricyanide-only process, and it is not this one. Where paper coated with ferricyanide alone is given the hours of ultraviolet it needs, a perceptible odour of hydrogen cyanide, and possibly cyanogen, can be detected when the frame is opened. Ware states that the quantity is not life-threatening and that there is no particular reason to make cyanotypes that way. That warning belongs to Herschel’s proto-cyanotype, and it is not generalised to this page in either direction.
• Never use a dichromate bath to re-oxidise a print. It was once common practice, especially for fabrics. Ware states that he strongly advises against it, that the consequences to one artist who used a dichromate bath for quilting fabrics are described in the medical literature, and that victims of dichromate exposure can become particularly sensitised to it. Potassium dichromate and ammonium dichromate are chromium(VI) compounds and carry the course’s chromium ruling. Use air, or hydrogen peroxide.
• The exposure is the hazard people forget. This is an ultraviolet process. A twenty- to thirty-minute exposure in direct sun is a skin and eye exposure; a UV lamp is worse, because it is closer and does not feel warm. Cover the skin, protect the eyes, and enclose a lamp.
• Thymol is harmful, and it is the one addition to this formula that is not benign. Its own page carries the handling.
• Gloves, and no dust. Both solids are handled dry when the stocks are made up. Weigh without raising dust, and keep everything that has held either salt out of the kitchen. Suppliers’ instructions say the same in plainer terms: wear gloves, and do not store the solutions in a refrigerator that holds food.

Storage

Two brown bottles, in a cupboard or a box, never one bottle. The mixed sensitiser has a short life; the separate stocks keep for months. A supplier selling the same two solutions premixed as a kit puts the shelf life of the separate bottles at about a year from purchase, kept dry and away from daylight.

Mix only what the session needs. Jacquard’s instruction for its own chemistry is that the sensitiser is stable “just 2 to 4 hours”, which is the only number any source read for this page attaches to Ware’s “relatively short life”.

The citrate bottle grows mould. Within a week or two in many environments its surface will carry a thick furry growth, because a citrate solution is an excellent nutrient medium. A few crystals of thymol floated on the surface at the outset inhibit it; they do not dissolve, and you avoid them when drawing a sample. A supplier’s note adds that a little mould does not necessarily spoil the solution, but the course’s position is that a bottle you are unsure of should be replaced rather than filtered and hoped for.

The solid citrate is deliquescent, goes sticky in humid air, and on long storage can compact into an intractable mass. Keep it stoppered, dry and cool. The same deliquescence follows the sensitiser onto the paper: in humid conditions a coated sheet can be tacky enough to damage a negative pressed against it.

The ferricyanide solid keeps well; its solution keeps less well, decomposing slowly in light, which is the second reason for a brown bottle.

Coated paper keeps, within limits. Jacquard states that sensitised paper and fabric may be stockpiled and stored, that six months is the figure for best results, that a sealed bag in a cool dry place is the way, and — usefully — that coated material which has darkened is not necessarily expired and should be tested, since it may simply need a longer rinse in hotter water. Ware records Whitaker’s finding that an excess of citrate shortens the storage life of coated paper, and gives no keeping time of his own.

Label both bottles with the strength and the date. Neither solution announces its age, and given the variability of the iron salt, the batch is worth recording too.

Incompatibilities

Acid, with the ferricyanide, absolutely. No acid stop bath, no acid fixer, no vinegar on the bench, nothing acidic in the sink the trays drain into. The single exception is the deliberate, dilute, weak development bath — 1 per cent citric acid or less — which is a controlled use and not a licence.

Alkali, with the finished print: buffered board, alkaline mountboard, alkaline sizing, alkaline tap water, ammonia, household cleaners. The pH 9.4 figure under Interactions is the measurement behind this rule.

Buffered paper, before you even coat. The same chemistry arriving earlier: an alkaline reserve in the sheet attacks the Prussian blue as fast as light makes it. Suppliers’ sheets converge on 100 per cent cotton rag, unbuffered.

Hard water, at every stage, because calcium salts damage the image.

Metal, in trays, tongs, clips and sinks. An iron–cyanide chemistry finds a trace of iron and prints it as a stain.

Silver, and every silver bath in the darkroom. Ferricyanide is the oxidising half of Farmer’s reducer; a shared cylinder or a contaminated tray will attack a silver print.

New Cyanotype sensitiser. Ware calls mixing the two “a chemical nonsense”, and the reason is specific: the whole point of the New Cyanotype preparation is to eliminate potassium ions, because sparingly soluble potassium ferric oxalate crystallises out on the paper and wrecks the print — and this sensitiser is full of potassium. He adds a second incompatibility: the oxalate sensitiser wants a mineral-acid developer for its best maximum density, while the citrate sensitisers tolerate only a very weak acid.

Dichromates, for the reason given under Safety.

Household bleach and anything chlorine-releasing, in the same room and the same drain, for the same reason as the acid rule.

Waste

Two streams, and neither is a silver stream. The unused stocks are dilute ammonium iron(III) citrate and dilute potassium ferricyanide; the wash water carries unreacted sensitiser plus the peptized Prussian blue you can see in the tray, which is image substance rather than dye.

Do not acidify anything, at any stage, including in the waste bottle. This is the disposal consequence of the incompatibility above and it is the reason cyanotype waste is not simply “iron salts”.

⁇How hazardous is cyanotype effluent?

The disagreement. Ware argues at length that the processing solutions are very dilute and that treating the effluent as chemical waste is “quite unnecessary from an environmental viewpoint, and an expensive and needlessly troublesome consumption of vital resources”. His grounds are specific, and each is a claim about one named substance at the dilution a cyanotype tray actually reaches rather than a general licence: ammonium ferric citrate is taken medicinally as an iron tonic and is permitted as food additive E381; ferricyanide reduces in the environment to the more stable ferrocyanide, which is permitted as E536 and used as an anti-caking agent in table salt; and Prussian blue is so insoluble that, on his account, a human can eat and excrete grams of it a day without ill effect, which is why it is a clinical antidote to thallium and caesium poisoning. He names and rebuts a published article that reached the opposite conclusion. None of that touches the two controls this page does insist on — the acid rule, and no dichromate — because neither is about dilution.

Who holds what. That is the position of a chartered chemist writing about his own process. It is not a statement about what any particular sewer authority permits, and Ware does not present it as one; he raises it in the context of practitioners working with septic tanks and land drainage.

What this course does and why. The course does not publish a jurisdiction-specific disposal instruction anywhere, and does not make an exception here. Collect the waste, label the containers with what is in them, follow the general chemical waste SOP and the course’s disposal ruling, and comply with local regulation, which is the only document that can actually answer the question. Ware’s argument is recorded because it is well-founded and because a reader is entitled to know that the chemistry is benign relative to much of what else is in a darkroom — not because it settles what may go down a particular drain.

Troubleshooting

A pale, flat print with empty highlights. The classic failure, and it is peptization: too much of the image washed out. Suspect, in order, a heavily sized paper the sensitiser never penetrated, no wetting agent in the mixture, a wash longer than it needed to be, and hard or alkaline wash water. The fixes in that order are a different paper, Tween 20 at 0.2 to 0.5 per cent, a shorter wash judged by the clearing of the highlights rather than the clock, and distilled water.

No image at all, or one that goes and does not come back. Check the paper first: an alkaline buffer will take the blue apart as fast as light makes it. Then check the water. Then check the age of the citrate bottle.

Blue fog across the highlights and the paper base. Either the coated paper was fogged before exposure, or the acid development bath was stronger than about 1 per cent, or excess iron(II) migrated during the wash. On these proportions the iron is in large excess over the ferricyanide, so bleeding is the likely one; the arithmetic is under The mechanism.

Blue stains unrelated to the negative. Iron from a tray, a clip or a sink; or sensitiser left standing where a brush rested twice. Herschel’s own warning still applies: the wash “cannot be in too thin a film”, and a brush passed twice over one part gives the effect of a coarse and ill-printed woodcut.

Small white spots, and a gritty feel to the coating. Crystals on the surface. A stock made up at a warmer temperature and used cold will deposit; so will a sensitiser mixed too strong for the paper to absorb before it dries.

The coated paper is tacky and marks the negative. Deliquescence plus humidity. Dry the coated sheet properly, store it sealed, and use an interleaving sheet in a humid climate.

The print is right in the frame and too light when dry. Expected: it needed to look blocked up, with reversed shadows and green highlights, and it lost more in the wash. Print for the dried result.

The print keeps changing colour for hours after drying. Also expected. Prussian white needs air to become Prussian blue again. Peroxide at 0.3 per cent will do it in half a minute if you cannot wait, without changing the final density.

The scale is too short for the negative. It is 0.9, about three stops, and no adjustment to this formula will make it 2. Either make a lower-contrast negative or use a formula with a longer scale.

Two batches of the iron salt behave differently. They may genuinely be different substances. Iron content between 14 and 28 per cent is the published range, “green” material is often not green and sometimes contains 18 per cent iron rather than 14, and no supplier is obliged to tell you. Record the supplier and the batch alongside your exposure times, and re-test when either changes.

Mould in the citrate bottle. Thymol at the outset prevents it. Once established, the safest course is to make the solution again.

Experiments

Print the three modern proportions side by side. Coat one sheet each at 20/10 (this page), 10/10 (Anderson’s) and 20/8 (the books’), from stocks made on the same day with the same salt, and expose all three together under one step tablet. Read the maximum density and the exposure scale from each. Ware’s survey says the process works “almost regardless of the proportions chosen”; this is the test of that claim on your paper, and it takes one afternoon.

Measure what the surfactant is worth. Two identical coatings, one with 0.3 per cent Tween 20 and one without, on a heavily sized paper where the effect should be largest. Densitometer the same three steps before and after the wash. The prediction from Ware’s account is that the difference is almost entirely in the high values.

Reproduce the wash-loss table. Four identical strips through a step tablet, washed for 5 minutes, 20 minutes, 1 hour and 4 hours, dried and measured. Do it once in tap water and once in distilled. The figures under Image characteristics are what Ware got; the interesting result is your own water’s number, because that is the one that governs your prints.

Water against 1 per cent citric acid. Two prints from one negative, one developed in plain water and one in the dilute acid before the same wash. Judge maximum density, highlight cleanliness and where the scale gained or lost. Then repeat the acid one at 2 per cent to see the fogging Ware warns about, deliberately, once, so that you recognise it.

Test the alkali rule instead of believing it. Two prints, one mounted on buffered board or given a brief bath in a mildly alkaline solution, one not. This is the fastest demonstration in the course that an archival material can be the wrong material, and it is the same chemistry as the first bath of the tannic acid toner.

Check the peroxide claim. Two prints from one exposure, one given half a minute in 0.3 per cent peroxide before the final wash and one left to oxidise in air overnight. Measure both the next morning. Ware states the final densities are the same; a densitometer will tell you whether that holds on your paper.

Compare two batches of the iron salt. If you can obtain material from two suppliers, or a “green” and a “brown” grade, print them under identical conditions. This is the experiment that makes the phrase “ill-characterised substance” mean something, and it is the reason the New Cyanotype exists.