The Simple Cyanotype (2019)

The chemical that has made cyanotype unreliable for a hundred and eighty years is ferric ammonium citrate, and this formula does not contain any. It makes its own, in the beaker, in about forty minutes, out of three substances a chemist would call well-behaved: an iron salt of a strong acid, a carboxylic acid, and ammonia. That is the first thing to know about it. The second is stranger and more useful. Once you are building the light-sensitive compound yourself, the amount of ammonia you add is no longer a detail of the preparation — it is the contrast control, and it is the only one any cyanotype sensitiser has ever had.

Ingredient	Quantity	Form the source specifies
Citric acid (anhydrous)	13 g or 12 g (anhydrous citric acid, C6H8O7, which Ware offers as the equal alternative; 12 g of the anhydrous acid and 13 g of the monohydrate are the same 0.062 mol of citrate)	the monohydrate, which Ware writes as C6H8O7.H2O; dissolved first, in 35 cc of pure water, before anything else goes into the beaker
Iron(III) nitrate nonahydrate	12 g	Fe(NO3)3.9H2O, formula weight 404.00, CAS 7782-61-8; added to the citric acid solution a little at a time with stirring, never dissolved in water on its own. Ware warns that the salt is highly deliquescent and that water it has absorbed from the air destroys it irreversibly. This course has no chemical encyclopaedia page for it, which is stated wherever it matters rather than passed over
Ammonia solution (ammonium hydroxide)	33 mL of a 10 per cent w/w solution	specific gravity 0.959, which is how the strength is checked; added slowly with stirring, after the iron has dissolved, whereupon the yellow solution turns green. Ware gives six concentrated laboratory alternatives, each made up to 33 cc with water before use — 14.5 cc of 24 per cent w/w, 13.9 of 25 per cent, 12.6 of 28, 11.8 of 30, 11.1 of 32 and 10.3 of 35 — and warns that household ammonia may carry additives that spoil the chemistry. This volume is the contrast control and is the only quantity that changes between the three published grades: 33.0 cc gives an exposure scale of about 2.7, 38.5 cc about 2.3 and 44.0 cc about 1.8, with the corresponding concentrated volumes printed for each
Potassium ferricyanide	10 g	99 per cent purity preferred; added last, to the green solution, and stirred until every crystal has gone. Ware's later Mike's cyanotype variant grinds it first
Water	to make 100 mL	Purified water — distilled, de-ionised or pharmaceutical. Only 35 cc of it is measured into the beaker at the start; the ammonia solution brings about 30 cc more, and the balance is added at the end in a measuring cylinder or volumetric flask to bring the whole to the 100 cc mark. The make-up volume is what fixes every concentration in the bottle, and it is the only step in the preparation that has to be done in graduated glass.

!The one ingredient the encyclopaedia had to be extended for

Iron(III) nitrate nonahydrate had no page in this course’s chemical encyclopaedia when this formula was written, and it is the substance the whole formula turns on. The register was built from the course’s design files and had never listed it; this page needing it is how that was found. Its entry now exists and the name in the table above is a link like any other.

Read it before you weigh anything, because it carries the one prohibition this formula cannot be worked safely without: the salt hydrolyses and polymerises the moment it meets plain water, so the citric acid goes into the beaker first and Ware’s own instruction is to weigh the nitrate and add it to the acid immediately.

The identity and hazard data are also set out on this page where they are used: Fe(NO₃)₃·9H₂O, formula weight 404.00 — the figure Ware prints — CAS 7782-61-8, EC 616-509-1, a pale violet to greyish-white deliquescent solid melting at 47.2 °C, classified Danger with H272, H315, H319 and H335 in the majority of notifications to the ECHA inventory and with the more severe H314 and H318 in about three in ten. There is no harmonised classification: every one of those statements is a notifier’s opinion, and the encyclopaedia entry says so at greater length than this page has room for.

Purpose

To make a negative-working Prussian blue print on plain paper from a sensitiser you can build yourself out of stock chemicals in under an hour, at a contrast you choose when you mix it.

Two problems are being solved at once and it is worth keeping them apart.

The first is a materials problem. Ammonium iron(III) citrate is not a compound; it is a family of ill-characterised polymeric solids whose iron content runs from 14 to 28 per cent by weight, whose colour runs from light green to dark brown, and for which, as Ware puts it, no single chemical formula can be written. Two printers on opposite sides of the world following the same classic recipe are not necessarily using the same chemical. Ware’s New Cyanotype of 1995 solved that by abandoning the citrate for a crystalline oxalate; this formula solves it a different way, by buying the iron and the citrate separately, in known purity, and combining them in a known molar ratio.

The second is a contrast problem, and it is the one that makes this entry more than a variation. Every other cyanotype sensitiser has a fixed exposure scale, so the rule is that the negative must be made to suit the process. Here the sensitiser can be made to suit the negative, anywhere between about 2.7 and about 1.8 density units, and lower still by changing the development bath.

It is a printing-out process. The image is made by light in the frame, not by a developer afterwards. The bath the print goes into is called development by convention; what it actually does is release iron(II) from its ligands so the blue can finish forming, and wash out everything the light did not use.

Recommended uses

• Printing from a long-scale negative, silver-gelatin or digital. At the low-contrast grade the exposure scale of about 2.7 takes a negative with an ultraviolet density range up to 2.7 or 2.8, which is longer than any silver-gelatin paper can print and about three times the Classic cyanotype’s 0.9.
• Matching a sensitiser to negatives you already have. This is the use that has no equivalent elsewhere in the formulary. If your negatives were made for platinotype or palladiotype, mix the grade whose exposure scale matches their density range instead of remaking the negatives.
• Digital-negative workflows that need repeatability. Ware records that the process was devised at the prompting of John Isner, whose PiezoDN calibrations were being disturbed by batch-to-batch variation of as much as a third of a stop in New Cyanotype sensitiser, and who found this one good to about a tenth of a stop. That is a reported experience from a named practitioner, not a measurement the course has verified.
• Where a Classic sensitiser has been washing away. Ware states that there is little runoff of image substance in the wet processing here, and the processing uses only very dilute citric acid.
• Where mould in the bottle has been the problem. A ferric ammonium citrate solution is an excellent nutrient medium and grows a mat on its surface within a fortnight. Ware states that this sensitiser shows no susceptibility to it and attributes that, as a likelihood rather than a demonstrated fact, to the nitrate.
• Teaching, where the preparation is part of the lesson. Nothing in it needs heating, filtration or crystallisation; it needs a balance reading to 0.1 g, a beaker, a measuring cylinder and a rod.

When another formula is preferable

• When you cannot use concentrated ammonia, or will not. That is the whole reason Mike’s cyanotype exists — Ware wrote it for correspondents who cannot obtain strong ammonia or cannot tolerate handling it, and states that it produces, and produces odourlessly, a sensitiser that should be identical with the first one here. If the ammonia bottle is the obstacle, that is the page to read.
• When you want the fastest and the deepest black. New Cyanotype is more than twice as fast, and Ware’s own one-line summary of it — “faster; finer; higher Dmax” — against this one’s “easy to make; cheap; contrast control” is a fair statement of the trade.
• When your negatives are ordinary. A negative made for a grade 2 or 3 silver-gelatin paper has a density range around 0.9 to 1.2. The classic sensitiser is built for exactly that, and printing such a negative on the low-contrast grade here will waste most of the scale. Use the high-contrast grade, or use the classic formula.
• When you have no balance and no graduated glass. Every quantity here is weighed or measured, and the make-up to 100 cc sets all four concentrations at once. There is no version of this that works by eye.
• When you want the oldest and simplest thing there is. Herschel’s 1842 cyanotype is two salts in about equal proportions and a bath of water, and it is the ancestor of everything on this page.

Mixing

Work under dim tungsten or LED lighting, not under fluorescent tubes and not near a window. Ware’s instruction is emphatic and it applies from the moment the ferricyanide goes in. The finished solution is sensitive to the near ultraviolet and the deep blue, which an incandescent or warm LED lamp emits very little of and a fluorescent tube and daylight emit a great deal of.

The order is the formula. These four substances mixed in a different order do not give the same sensitiser, and one of the orders is simply wrong: iron(III) nitrate dissolved in plain water begins to hydrolyse and polymerise at once, which is precisely the defect the preparation exists to avoid. The citric acid goes in first so that the iron never meets water without a ligand and an excess of hydrogen ions waiting for it.

The one-bottle preparation, in Ware's order

1. 113 g citric acid monohydrate in 35 cc of pure wateror 12 g of the anhydrous acid; stir until every crystal has gone. The solution is strongly acid, and that is the point of doing it first.
2. 212 g iron(III) nitrate nonahydrate, added in portionsstirring each small addition to dissolve it before the next. Solution yellow. Weigh it quickly: the salt is deliquescent.
3. 333.0 cc of 10 per cent w/w ammonia, added slowly with stirringthe yellow turns green. This is the volume that sets the contrast: 38.5 cc for medium, 44.0 cc for high.
4. 410 g potassium ferricyanide, stirred in until dissolvedfrom here the solution is light-sensitive and the work is under dim tungsten or LED light only.
5. 5Made up to 100 cc in a cylinder or volumetric flaskmixed, bottled in stoppered brown glass, labelled, dated, stored dark. Left to stand overnight before first use.

Cyanomicon 7.4.2, the low-contrast grade. The two accented steps are the two that cannot be reordered or estimated.

Filter if it needs it. Trace impurities in the chemicals may throw a very small sediment — brown iron(III) hydroxide, or Prussian blue — during the overnight stand. Ware treats this as normal and says it can be removed by filtration if necessary. A sediment that keeps growing is a different matter and is under Troubleshooting.

🧮Why 13, 12, 33 and 10, and what the three grades really are

Ware states the design of this sensitiser in words and leaves the arithmetic to the reader. It is worth doing, because it turns three apparently arbitrary volumes of ammonia into one idea.

The citrate is a two-fold excess over the iron, exactly as stated. 13 g of citric acid monohydrate, formula weight 210.14, is 0.0619 mol; 12 g of iron(III) nitrate nonahydrate, formula weight 404.00, is 0.0297 mol.

0.0619 mol ÷ 0.0297 mol = 2.08

Citrate to iron

Two citrates per iron, which is the ratio in the dicitratoferrate anion Ware is aiming at. The alternative weighing is the same dose: 12 g of the anhydrous acid, formula weight 192.12, is 0.0625 mol, within one per cent of the monohydrate’s 0.0619.

The ferricyanide is equimolar with the iron. 10 g of potassium ferricyanide, formula weight 329.26, is 0.0304 mol against the iron’s 0.0297 — a ratio of 1.02 : 1. Every iron(III) in the bottle has one hexacyanoferrate waiting for it and no more. That is the same “chemically equivalent amounts” principle Ware uses to cure highlight bleeding in the classic sensitiser, where the iron is in more than twofold excess.

The three grades are three, three and a half, and four ammonias per citric acid. Take the low-contrast dose: 33.0 cc of a 10 per cent w/w solution of specific gravity 0.959 weighs 31.65 g and carries 3.165 g of ammonia, which is 0.186 mol.

0.186 mol ÷ 0.0619 mol = 3.00

Ammonia per citric acid, low contrast

Citric acid is triprotic. Three ammonias per acid is exactly its three carboxyl groups, neither more nor less. Repeat the sum for the other two grades and the pattern is unmistakable:

Grade	Ammonia, 10 % w/w	Ammonia	NH₃ per citric acid	Published pH	Published exposure scale
Low contrast	33.0 cc	0.186 mol	3.00	about 4	about 2.7
Medium	38.5 cc	0.217 mol	3.50	about 6	about 2.3
High contrast	44.0 cc	0.248 mol	4.01	about 8	about 1.8

The three volumes are in the ratio 6 : 7 : 8. And the six concentrated alternatives Ware prints for each grade all deliver the same dose to better than one per cent — 14.5 cc of 24 per cent w/w carries 0.186 mol of ammonia, 10.3 cc of 35 per cent carries 0.186 mol, and so do the four in between. The table of specific gravities is not a convenience, it is the check that they are the same formula.

The arithmetic in this box is the course’s. The weights, the volumes, the specific gravities, the pH values and the exposure scales are Ware’s, and so is the statement that the citrate is in two-fold molar excess. Ware does not comment on the 3.00, 3.50, 4.00 pattern anywhere the course has read.

One caution about reading the pH. A solution of triammonium citrate alone is near neutral, and Ware’s measured pH of about 4 at the low-contrast grade is a long way below it. The course’s explanation — offered as an inference and not as anything Ware states — is that the iron(III) has itself displaced further protons from the citrate ligands in forming the complex, since the anion characterised in 1998 carries citrate in its fully deprotonated tetrabasic form. On that reading, the fourth ammonia at the high-contrast grade is what completes the deprotonation and then begins to leave free base in solution, which is why the pH climbs to about 8 for a proportionally small addition.

🧮What is in the bottle, compared with the other two sensitisers

Working the same figures into concentrations for the finished 100 cc:

0.30 M iron(III) · 0.30 M hexacyanoferrate(III) · 0.62 M citrate · 0.89 M nitrate

The one-bottle sensitiser, low-contrast grade

Two comparisons follow, and both are the course’s arithmetic on published figures.

Against New Cyanotype, which Ware states as 0.4 molar ammonium iron(III) oxalate with 0.3 molar ammonium ferricyanide: the ferricyanide is the same and the iron is three-quarters. Three-quarters of the iron does not account for less than half the speed, and it is not meant to: the larger part of that difference is the ligand’s own photochemistry, and the quantum yields are under The mechanism.

Against the classic sensitiser as mixed, 10 per cent w/v ammonium iron(III) citrate and 5 per cent w/v potassium ferricyanide: on Ware’s stated iron content of 14 to 18 per cent for the green salt, that is roughly 0.25 to 0.32 M iron and 0.15 M ferricyanide. So the iron is much the same and the ferricyanide is doubled, exactly as in New Cyanotype, and for the same reason — the classic mixture carries more iron than it has hexacyanoferrate to pair with, and the excess iron(II) made in the shadows migrates into the highlights and prints there.

And the by-product nobody weighs. Every nitrate that came in with the iron is still there, paired with ammonium: 0.089 mol of it, which as ammonium nitrate would weigh 7.1 g in the 100 cc. By weight it is the third-largest thing in the bottle after the citrate and the ferricyanide, it does nothing photochemical, and Ware’s case is that it is doing four other jobs. Those are under Function of every ingredient.

The two-bottle arrangements, of which there are two and they are not the same. If the one-bottle sensitiser does not keep long enough for the way you work, both of Ware’s documents offer a separated version, and they separate it in different places.

• Cyanomicon 7.4.13 stops one step earlier rather than splitting into halves. Dissolve the citric acid in 48 cc of water instead of 35, add the iron, add the same 33.0 cc of ammonia, and bottle the resulting 95 cc without the ferricyanide; that solution should keep for many months at room temperature. Weigh 10 g of ferricyanide into the whole of it when it is wanted, which brings the volume to about 100 cc. For a smaller quantity, Ware gives the rule directly: measure the volume of iron solution you want and divide it by 9.5 for the weight of ferricyanide in grams. The medium-contrast version starts from 43 cc of water and the high-contrast from 38, so that the total lands in the same place after the larger ammonia dose.
• The 2022 sheet gives a true A-plus-B pair. Solution I is the same citric acid, iron and ammonia — but dissolved in only 20 cc of water, warmed to 30 °C, cooled to room temperature before the ammonia goes in dropwise with efficient stirring, and made up to 50 cc. Solution II is 10 g of ferricyanide in 40 cc of water made up to 50 cc. Equal volumes are mixed when needed. Only the concentrated ammonia reagents are listed for this route; the 10 per cent solution would not fit in 50 cc.

Both give the same working sensitiser. The 2022 arrangement keeps better — Ware says the separate stocks should keep indefinitely — and both share the same weakness, which is that once mixed the sensitiser has an unpredictable life of a few weeks or even days.

Tween 20 is added to the sensitiser and never to the stock bottle. It is a non-ionic surfactant, polyoxyethylene sorbitan monolaurate, and it is not in the ingredient table for two reasons: Ware states that it does not last well when dilute and that the right amount depends on the paper and must be found by trial, and this course has no chemical encyclopaedia entry for it. His doses all come to the same final concentration of about 0.25 to 0.5 per cent, by three different routes:

• one drop, about 0.05 cc, of a 5 per cent v/v stock per cc of sensitiser (Cyanomicon 7.4.6);
• one drop of a 15 per cent v/v stock per 3 cc (the 2022 sheet);
• one drop of a 25 per cent v/v stock per 5 cc, for larger coating volumes (both).

Do not substitute Tween 80, which Ware says foams.

Coating, drying and the paper. About 1.5 cc for a 10 by 8 inch area, five to eight passes of a glass rod, on an unbuffered, internally AKD-sized pure cellulose paper; blot off any excess, which would otherwise crystallise and damage a negative. Dry one to two hours at room temperature in the dark, or let the coating soak in until the surface is no longer reflective and then dry it with a uniform air stream at about 40 °C for five to ten minutes — but note that heat-drying raises the contrast by about a stop, so it is a variable and not a convenience. Ware’s named papers are Arches Platine, Hahnemühle Platinum Rag, Magnani or Legion Revere Platinum, Bergger COT 320, Ruscombe Mill’s Buxton and Herschel, Crane’s Platinotype and Weston Diploma Parchment; 160 gsm up to A4, 240 gsm or more at A3 so the wetted sheet does not belly away from the negative.

Behaviour

It prints out, and what you see in the frame is close to what you get. Because there is little loss of image substance in the wash, the appearance under the negative is a usable guide: high values light green, mid-tones firm blue, deepest shadows reversed to a pale blue-grey. Ware adds a detail that costs nothing and buys a stop: leaving the exposed sheet in the dark for some hours before processing may bring out another stop of highlight detail.

The contrast is set twice, in two different places, and they are not equivalent. In the bottle, by the ammonia dose, the exposure scale runs from about 2.7 down to about 1.8 — with intermediate values available by mixing the low-contrast and high-contrast sensitisers proportionally, which is exact and repeatable. In the tray, by using water instead of 1 per cent citric acid, it drops further to about 1.3, with higher contrast and no fogging but a weakened maximum density. Use the bottle for the contrast you want and the tray for the last resort, because the tray costs you black.

The shadows come back on their own. The reversed shadow tones are Prussian white, and air reoxidises them during processing and drying. Ware’s statement for this sensitiser is that they regain density fairly rapidly and without needing peroxide, and he credits the nitrate for part of that: on development in mild 1 per cent citric acid at about pH 4, he writes, the nitrate also helps to reoxidise the solarised Prussian white. Peroxide is for when you cannot wait — for a densitometer reading the same afternoon.

It gasses during exposure. Carbon dioxide is evolved as the citrate ligand decarboxylates, and if it has nowhere to go it forms bubbles between the negative and the paper and costs acutance. Ware’s answer is to back the sheet in the frame with papermakers’ felt or another porous material. This is not a fault of this formula in particular — it is what every iron-carboxylate print-out process does — but it is the one place a cyanotype’s sharpness is decided by something other than the negative.

Heat-drying is a lever, not a shortcut. About one stop of contrast, gained or lost depending on whether you use it; and over-rapid drying diminishes the penetration of the sensitiser into the fibres and therefore worsens the loss of image substance in the wash. Choose one drying regime and keep to it, or your exposure scale is not the number you calibrated.

The bath goes off before it wears out. One litre of the 1 per cent citric acid bath does two or three prints. What ends it is the Prussian blue it has taken out of the paper, which will otherwise stain the next sheet.

Image characteristics

Tonal scale. The property this formula exists for. About 2.7 density units at the low-contrast grade — nine stops — with what Ware describes as delicate gradation in the high values. Set against the Classic process’s exposure scale of about 0.9, that is three times the range, and it is why a negative made for platinotype prints here and will not print on a classic cyanotype at all.

Maximum density. About 1.4 in Cyanomicon and in Ware’s web article; about 1.5 in the 2022 sheet. Either figure is below New Cyanotype’s 1.7, which Ware describes as verging on black, and well above what a classic cyanotype retains after its image has partly washed away.

Speed. Less than half New Cyanotype, about four times a typical Classic. In practice, 5 to 10 minutes under an average 365 nm source against the Classic process’s 20 to 30.

Colour and surface. Prussian blue, in paper, with no binder and therefore no surface of its own: the print has the surface of the sheet it is on. The course records one thing it cannot tell you. Ware’s account of New Cyanotype turns on the difference between a potassium Prussian blue and an ammonium one, the ammonium form being much less readily dispersed by the wash; this sensitiser contains both cations in quantity, about 0.09 mol of potassium and about 0.19 mol of ammonium per 100 cc, and neither Ware nor any other source the course has read says which ends up in the lattice. The observable fact he does state is that there is little runoff of image substance.

Reversal in the shadows. Pronounced, because this is a print-out process taken to a solarised end point on purpose, and self-correcting in air.

Staining. None in the silver-gelatin sense, but the sensitiser stains skin, clothes, wood and absorbent surfaces, and a yellow residue left in the paper after an inadequate wash is a stain and not a fog. The difference matters when diagnosing a fault and is set out under Troubleshooting.

The mechanism

The chemistry falls into two halves that never happen at the same time. The first half is done in the beaker, in the light, by you; the second is done on the paper, in the dark and then under the lamp, by photons. Almost everything unusual about this formula belongs to the first half.

In the beaker: making the light-sensitive compound

Iron(III) in plain water is a trap. The hydrated Fe³⁺ ion is small and triply charged, so it pulls electron density off the water molecules bound to it and lets a proton go, and the products condense into hydroxy- and oxo-bridged polymers:

Fe³⁺ + 3 H₂O → Fe(OH)₃ + 3 H⁺

Hydrolysis, which begins the moment the salt dissolves in water alone

That polymerisation is exactly what has been wrong with commercial ferric ammonium citrate since the 1840s. The pharmacopoeial preparation of that substance starts by precipitating iron(III) hydroxide and then dissolving it in citric acid, so what you buy is a citrate-capped hydroxide polymer whose iron content the supplier reports as anything from 14 to 28 per cent. Ware’s answer is to keep the iron out of that condition from the first second: the citric acid is already dissolved, and already holding the solution strongly acid, before the nitrate crystals go in.

Then the ammonia is titrated in, and the complex forms. Ammonia takes the protons off the carboxyl groups:

C₆H₈O₇ + 3 NH₃ → 3 NH₄⁺ + C₆H₅O₇³⁻

Neutralising the citric acid; the ammonium ions stay in solution as spectators

and as the citrate deprotonates it grips the iron. The target is the anion that Matzapetakis and co-workers isolated and characterised by X-ray structure determination in 1998 — the first mononuclear water-soluble iron-citrate complex — as (NH₄)₅Fe(C₆H₄O₇)₂·2H₂O:

Fe³⁺ + 2 C₆H₄O₇⁴⁻ → [Fe(C₆H₄O₇)₂]⁵⁻

The monomeric dicitratoferrate(III) anion: one iron, two tetrabasic citrate ligands

The colour change Ware tells you to watch for — yellow to green as the ammonia goes in — is that reaction happening. The 1998 authors precipitated the salt out with ethanol in order to analyse it. Ware’s insight, which is the whole of the invention, is that a photographer does not need the solid: the solution is the sensitiser.

⚗Why the nitrate is not a spectator

Ware’s own account of the by-product is the most interesting paragraph in the source, and it is carefully hedged. Every nitrate ion that came in with the iron is still in the bottle, about 0.089 mol of it against 0.030 mol of iron, and he offers four things it does. The first is stated as a prediction from chemistry, the second as an observation, and the last two as likelihoods:

1. It is the preservative. Nitrate is a mild oxidant at low pH, so it should prevent iron(II) forming during dark storage — and iron(II) in a bottle that also contains ferricyanide means Prussian blue, which is how a cyanotype sensitiser dies. This is the job ammonium dichromate does in New Cyanotype, done by an ion that is not a chromium(VI) compound.
2. It helps the shadows come back. On development in 1 per cent citric acid at about pH 4, the nitrate assists the reoxidation of solarised Prussian white to Prussian blue.
3. It may assist print-out, because ammonium nitrate in the dried coating is hygroscopic and holds humidity in the paper, as it is observed to do in the Malde-Ware print-out platinum and palladium process.
4. It probably inhibits mould, which is the standing nuisance of a ferric ammonium citrate solution.

Ware also notes that adding ammonium nitrate to improve the stability and quality of blueprint paper sensitiser was the subject of a United States patent, number 2,113,423, as long ago as 1938. The course records that as his citation; it has not read the patent.

He publishes no equation for any of this, and neither does this page. The nitrate/nitrite and nitrate/nitric oxide couples are strongly pH-dependent and the kinetics in a citrate-buffered iron solution are not something the sources establish. What is established is the effect and the reasoning; the balanced chemistry is not, and inventing it would be the one thing this course does not do.

On the paper: making the picture

One: light reduces the iron. The citratoferrate(III) absorbs in the near ultraviolet, an electron moves from a citrate ligand to the iron, and the iron(III) becomes iron(II) while the ligand decarboxylates. Ware identifies the initial organic photoproduct as acetone dicarboxylic acid, which may lose two more carbon dioxides to end as acetone, and states plainly that the nature of the iron(II) photoproduct remains unknown — it is likely to be a citrato complex of some kind. Because he leaves that half of the equation open, this page leaves it open too and gives no balanced reaction for the step. What it can give is the two numbers that matter: the quantum yield at 365 nm is about 0.45, falling to 0.28 at 436 nm.

⚗The two numbers that separate this sensitiser from the New Cyanotype

The citrate system’s quantum yield of about 0.45 at 365 nm is what a photon buys here. Ware’s own appendix on the oxalate system gives it as approximately 1.2 between 250 and 420 nm, because the oxalate mechanism delivers two electrons for one photon — one from the excited complex and one from the radical anion it releases, which reduces a second, unexcited complex. That ratio of roughly two and a half to one is already enough on its own to account for “less than half the speed of my New Cyanotype”, and the iron concentrations pull the same way, four moles per litre of sensitiser there against three here. The course does not multiply the two into a predicted ratio, because Ware publishes no measured one to check it against.

The second number is the one that matters for this page. The oxalate’s quantum yield is largely independent of pH; the citrate’s is not. It peaks at 0.45 at pH 4 and falls away above it. That is why the New Cyanotype has no contrast control in the bottle and this sensitiser does — and it is also the reason the deeper section below treats the mechanism of that control as an open question rather than a settled one.

Two: the iron(II) reduces the ferricyanide, or reaches it. Ware’s Appendix III.8 sets this half out for the blueprint process generally, and works its numbers on the oxalate system because that is where the constants have been measured — he says of the citrate that the identity of the iron(II) photoproduct is not known but that it is likely to be a citrato complex. What carries across is the shape of the argument, not the figures. The iron(II) is still wearing carboxylate ligands and cannot enter the Prussian blue lattice until they are stripped off, and hydrogen ions are what strips them, by competing for the carboxylate to make the undissociated acid. That is the whole reason there is an acid bath at all, and the reason developing in water instead costs maximum density. Meanwhile a simple electron transfer from the iron(II) complex to the hexacyanoferrate(III) can happen without any stripping, and in the oxalate system the redox potentials favour it — +0.356 V for the ferricyanide/ferrocyanide couple against +0.02 V for the iron oxalato pair. The equivalent potentials for the citrato complexes are not published in the sources this page rests on, so the course states the direction and not a number.

Three: the pigment forms. Free iron(II) is competed for by both oxidation states of hexacyanoferrate present:

Fe²⁺ + [Fe(CN)₆]³⁻ → Fe[Fe(CN)₆]⁻

Prussian blue, where exposure has been light and ferricyanide predominates

Fe²⁺ + [Fe(CN)₆]⁴⁻ → Fe[Fe(CN)₆]²⁻

Prussian white, where exposure has been heavy and the ferricyanide is largely reduced

Four: the white goes back to blue. Prussian white reduces atmospheric oxygen, which is why the reversed shadows recover on their own over hours, and why a dilute peroxide bath does the same thing in half a minute:

4 Fe[Fe(CN)₆]²⁻ + O₂ + 2 H₂O → 4 Fe[Fe(CN)₆]⁻ + 4 OH⁻

Aerial reoxidation during processing and drying

2 Fe[Fe(CN)₆]²⁻ + H₂O₂ → 2 Fe[Fe(CN)₆]⁻ + 2 OH⁻

The same conversion, in half a minute, with dilute hydrogen peroxide

Both consume protons and generate hydroxide, which is why Ware says the oxidations are best carried out under acid conditions: enough hydroxide will start hydrolysing the pigment it has just made.

Deeper: why the ferricyanide is not the light-sensitive ingredient

It is light-sensitive; it is just very bad at it. Its own photolysis begins with a photoaquation:

UV + [Fe(CN)₆]³⁻ + 2 H₂O → [Fe(CN)₅(OH₂)]²⁻ + HCN + OH⁻

Photoaquation of hexacyanoferrate(III), the first step of its own slow photolysis

and the reported quantum yields for the aquation and reduction reactions are of the order of 0.01 — roughly one fortieth of the citrate’s 0.45. That single ratio is why Herschel’s ferricyanide-only “proto-cyanotype” needs hours where this needs minutes, and it is also why the ferricyanide in this formula can be treated as photochemically inert on the timescale of a print.

It is not optically inert, though, and that is a different problem. Ferricyanide absorbs strongly in exactly the band the iron salt has to be exposed in, so a large part of the light entering the coating is captured by a molecule that does nothing useful with it. Ware calculates the fraction of incident light actually absorbed by the photoactive component of a typical cyanotype coating as 0.65 at 365 nm and 0.03 at 420 nm. Two consequences follow for this page. The process wants a genuine 365 nm source, not a violet one; and the equimolar ferricyanide here — twice the classic sensitiser’s concentration — buys pigment at some cost in speed.

Deeper: the speciation the contrast control may be built on

This is the most interesting open question on the page, and the course does not close it.

Ware’s own appendix on citratoferrate photochemistry reports the solution chemistry: a photo-inactive monomer predominates between pH 0.5 and 1.5, a photo-active dimer forms above pH 2, and the quantum yield at 365 nm reaches its maximum of 0.45 at pH 4 before falling away as the pH rises further, possibly because the iron(III) begins to hydrolyse and oligomerise again. He also notes that equilibria in this system become slow above pH 3 — which is very likely why the finished sensitiser has to stand overnight before first use.

Now put the three grades beside those numbers. Their published pH values are about 4, about 6 and about 8. The low-contrast grade sits precisely at the quantum-yield maximum; the other two sit progressively beyond it, in the region where hydrolysis and oligomerisation are setting in again.

Ware does not draw that connection anywhere the course has read, and the course does not draw it either. It is recorded here because the two sets of facts are in the same book, they are obviously related, and a reader is entitled to see the question. What would settle it is one measurement Ware does not publish: the relative speed of the three grades. His own step-tablet comparison exposed all three for the same 8 and a half minutes, so it shows the change in scale and says nothing about the change in sensitivity. Until somebody prints the three grades to matched maximum densities and records the exposures, the mechanism of this contrast control is a plausible hypothesis and not an explanation. There is an experiment for it at the foot of this page.

Function of every ingredient

Citric acid, 13 g of the monohydrate or 12 g anhydrous — the ligand, and the reason the iron behaves. It is a triprotic carboxylic acid, and it is doing three jobs at once. Before the iron arrives it holds the solution strongly acid, which is what suppresses hydrolysis of the iron(III) that is about to be dissolved into it. As the ammonia neutralises it, its citrate anion binds the iron in a defined, monomeric, two-ligand complex — the same anion Matzapetakis and co-workers characterised in 1998. And it is the ligand that light destroys: the electron that reduces the iron comes off the citrate, which decarboxylates.

More or less of it. The 0.062 mol is a two-fold molar excess over the iron and that ratio is the design. Less than two citrates per iron and there is not enough ligand to keep every iron monomeric, so the hydrolysed polymers this formula exists to avoid start to appear. More citrate is not neutral either, because the ammonia dose is calibrated against the acid: three ammonias per citric acid at the low-contrast grade, so adding acid without adding ammonia moves the pH down and adding both changes nothing except the total salt concentration. The same substance also does duty as the development bath at 1 per cent w/v, ten times more dilute than in the sensitiser, and Ware’s rule for a highlight that has come up blue is to make that bath weaker still.

Iron(III) nitrate nonahydrate, 12 g — the iron, and the preservative, in one crystal. This is the substance that has no encyclopaedia page in this course, so its identity is given here: Fe(NO₃)₃·9H₂O, CAS 7782-61-8, EC 616-509-1, formula weight 404.00, a pale violet to greyish-white solid that melts at 47.2 °C and dissolves freely in water. It is chosen because it is an iron(III) salt of a strong acid: it brings a clean, monomeric, hydrated Fe³⁺ and no competing ligand of its own, so the citrate already in the beaker gets the iron uncontested. The trick is a century older than this formula: the course’s iron(III) chloride page sets out, with its source, an alternative blueprint sensitiser from a 1924 manual that built an ammonium iron(III) tartrate in the bottle from iron(III) chloride, tartaric acid and ammonia, and warned in the same breath that excess ammonia must be avoided. Same reasoning, different carboxylate, different halide.

Its second job is the one nobody weighs. The three nitrate ions per iron stay in solution, paired with ammonium — about 7.1 g of ammonium nitrate equivalent in the 100 cc — and Ware’s case for them, set out in the callout above, is that they are the oxidising preservative that keeps iron(II) from forming in the dark. That is why this one-bottle sensitiser keeps at all without a dichromate in it.

More or less of it. Less iron and the ratio to the equimolar ferricyanide breaks: there is hexacyanoferrate with no iron to pair with, which costs density. More iron and you have the classic sensitiser’s defect back — excess iron(II) in the shadows diffusing sideways into the highlights and printing there, which is bleeding. It also changes the citrate ratio away from two, and the nitrate concentration with it. This is the least adjustable number in the formula.

Weigh it fast and keep the jar shut. Ware’s 2022 sheet is blunt about it: the salt is highly deliquescent, and any water it absorbs “will react and destroy it irreversibly”. Its melting point of 47.2 °C is the reason a jar that has taken up moisture goes to a syrup and never comes back. Weigh straight from a dry, tightly closed container into the beaker.

Ammonia solution, 33.0 cc of 10 per cent w/w — the titrant, and the contrast control. It has no photographic function of its own. What it does is remove protons from the citric acid so that the citrate can bind the iron, and the extent of that removal is what sets the pH of the finished sensitiser and therefore its exposure scale. Three ammonias per citric acid — exactly its three carboxyl groups — gives pH about 4 and a scale of about 2.7; three and a half gives about 6 and about 2.3; four gives about 8 and about 1.8.

More or less of it. This is the one quantity in the formula that a reader is invited to change, and it is also the one that must be measured accurately, which is why Ware asks for a graduated pipette in preference to a cylinder. Below the low-contrast dose the citrate is not fully deprotonated and the complex does not form completely; above the high-contrast dose you are past the point where the sensitiser was characterised at all and heading for the region where iron(III) hydrolyses again. Intermediate contrasts are obtained by mixing the finished low- and high-contrast sensitisers proportionally, which is exact, rather than by guessing an intermediate volume.

Strength matters more than volume. Ammonia leaves solution steadily, so an old bottle is weaker than its label and will under-neutralise. Ware’s remedy is a hydrometer and the table of specific gravities — 0.959 for 10 per cent w/w, 0.912 for 24, 0.909 for 25, 0.900 for 28, 0.894 for 30, 0.890 for 32 and 0.880 for 35 — and his warning about household ammonia is that commercial brands may contain additives that spoil the chemistry.

Potassium ferricyanide, 10 g — the precipitant, and the internal filter. It supplies the hexacyanoferrate that becomes half of the Prussian blue lattice. It contributes almost nothing to image formation, its own quantum yield being about 0.01 against the citrate’s 0.45, and almost everything to image substance. Its second, unwanted role is optical: it absorbs strongly in the printing band and shades the photoactive iron beneath it.

More or less of it. At 10 g it is equimolar with the iron, which is the “chemically equivalent amounts” condition Ware invokes against highlight bleeding. Less of it and there is iron with no partner — which is the classic sensitiser’s condition, and brings back both of its defects at once: a maximum density limited by how much pigment there is to make, and excess iron(II) in the shadows diffusing into the highlights and printing there. More of it thickens the internal filter and slows the print without adding pigment, because there is no more iron for the extra to pair with.

Purified water, to make 100 cc — the thing that makes the numbers mean anything. Distilled, de-ionised or pharmaceutical. Only 35 cc is measured at the start; the make-up at the end, in graduated glass, is what fixes all four concentrations and therefore the exposure scale you calibrated. Tap water is excluded for the same reason it is excluded from the wash: calcium damages Prussian blue.

What is deliberately not in the ingredient list. Tween 20, which Ware’s chemicals table gives at about 1 cc per 100 cc, is added to the working portion just before coating and never to the stock, on his own instruction, and the course has no encyclopaedia entry for it; it is a wetting agent whose job is to get the sensitiser into the interfibrillar space of the surface cellulose fibres, where the pigment is trapped beyond the reach of the wash, instead of leaving it in the coarse pores where the wash takes it away. And there is no dichromate and no oxalate anywhere in this formula, which is the difference Ware most wants noticed: the preservative role is played by the nitrate that arrived free with the iron.

Interactions

Citric acid with iron(III): the interaction the formula is built on. Complexation, in a defined two-to-one ratio, in place of hydrolysis. Order and acidity are what decide which of the two happens, which is why the mixing order is not a convenience.

Ammonia with citric acid: the interaction that sets the contrast. Acid-base, three protons available per acid molecule, and the sensitiser’s pH is the direct consequence. Everything the reader controls in this formula, they control through this one titration.

Citrate with light: the interaction that makes the picture, and the one that also makes carbon dioxide, which needs somewhere to go during exposure or it will lift the negative off the paper in patches.

Nitrate with iron(II): the interaction that keeps the bottle alive. Ware’s proposition, stated as a prediction from the oxidising nature of nitrate at low pH rather than as a measurement. Its photographic consequence is a sensitiser that keeps for weeks without a chromium salt in it, and shadows that recover their density without a peroxide bath.

Iron(II) with hydrogen ions: the development. The acid bath is not developing anything in the photographic sense; it is dissociating carboxylate ligands off the iron(II) photoproduct so the free ion can build pigment. That is why the bath deepens the blacks and why replacing it with water costs maximum density.

Prussian blue with alkali: the destruction. Alkali hydrolyses the pigment to weakly coloured iron compounds. This is why the paper must be unbuffered, why the wash water must be below pH 7, and why buffered mountboard above about pH 9 will slowly consume the print.

Prussian blue with calcium: the same destruction, arriving through the tap. Hard water damages the image, and chalk-buffered paper attacks it from underneath.

Ferricyanide with strong acid: the interaction that is not happening here, and the reason it is not. A cyanide complex plus acid sounds alarming and the claim circulates widely. Ware addresses it directly and his qualification is quantitative: the heating has to be very strong and the acid very concentrated to liberate much hydrogen cyanide, and cyanotype does neither. The strongest acid on this page is 1 per cent w/v citric acid — one part in a hundred of a weak organic acid. See Safety.

Ferricyanide with the New Cyanotype sensitiser: the interaction that ruins a print. Ware’s caveat on incompatible cyanotype processes applies to this formula from the other side: New Cyanotype has had its potassium removed at some trouble, and this sensitiser is full of it, so mixing them invites potassium iron(III) oxalate to crystallise in the coating. He gives a second reason as well — New Cyanotype wants a mineral-acid developer for its best maximum density, while the Simple and Classic sensitisers tolerate only a very weak acid, above which they fog blue in the highlights.

Variants

The three contrast grades are the formula, not variants of it. One number changes, the volume of ammonia, and it is printed for each grade in three concentrations by Ware and in seven by his 2022 sheet. They are set out in full under Mixing.

Intermediate contrasts, by mixing. Ware’s own method is to mix the finished low-contrast and high-contrast sensitisers in proportion, which gives any exposure scale between about 2.7 and about 1.8 without a second titration. It is the accurate way to match a sensitiser to a measured negative.

The water-developed version. Omit the citric acid bath and process in water alone. The exposure scale drops to about 1.3, the contrast rises, there is no chemical fogging, and the maximum density is somewhat weakened. Ware offers this as a fourth contrast option rather than a fault, and his own step-tablet plate includes it as the fourth strip.

The separated versions, for keeping. Two of them, and they are not the same arrangement; both are described under Mixing. Cyanomicon 7.4.13 bottles the ammonia-neutralised iron solution without the ferricyanide and adds the solid when it is wanted, with the rule “divide the volume by 9.5 for the weight in grams”. The 2022 sheet gives a conventional two-bottle A plus B of 50 cc each, mixed in equal volumes.

Mike’s cyanotype, the ammonia-free route to the same solution. Ware wrote it for people who cannot obtain or cannot tolerate strong ammonia. It replaces the citric acid-plus-ammonia pair with 15 g of triammonium citrate, keeps the same 12 g of the iron salt and 10 g of ferricyanide, and Ware states that it should produce a sensitiser identical with the first one here. What it necessarily gives up is the contrast control, because the neutralisation is no longer something you titrate.

Ware’s own five-way summary, which is the shortest guide to where this entry sits among its neighbours. His adjectives are his:

Formulation	Ware’s own one-line summary
Classic cyanotype	easy to prepare; uncertain; limited tonal range
Classic improved	more toxic; better range and density
New cyanotype	faster; finer; higher Dmax; toxic; chemically tricky
Simple cyanotype (this page)	easy to make; cheap; contrast control; non-toxic — the last adjective is his summary and is not a hazard classification, and this course does not adopt it; see Safety
Mike’s cyanotype	Simple cyanotype without using ammonia

Toning. A finished Simple cyanotype is Prussian blue on paper like any other, so the tannic acid toner and the rest of the toned cyanotype repertoire apply. Nothing in the sources this page rests on tests toning against this sensitiser specifically.

⁇What the medium-contrast grade's exposure scale actually is

The disagreement is inside one author’s own work, and it is small but real. Cyanomicon 7.4.2 heads the middle grade “Medium contrast: Exposure Scale ~2.4” and its step-tablet plate at 7.4.12 labels the second strip “ES ~ 2.4”. Twelve pages later, 7.4.13 heads the same middle grade in the separated preparation “Medium contrast: ES ~2.3”. The 2022 instruction sheet gives 2.3 in all three places it appears.

What the course states. About 2.3, because that is the figure the later and more specialised of the two documents gives consistently, and it is one of the two figures the monograph gives. The difference is a tenth of a density unit, which is a third of a stop and well inside the variation any reader will meet between papers and drying regimes.

Two neighbouring numbers move in the same way and are recorded rather than reconciled. The water-developed scale is 1.2 in Cyanomicon 6.7.6, 1.3 in Cyanomicon 7.4.10 and in both practical statements in the 2022 sheet, and 1.4 in that sheet’s introduction. The maximum density is about 1.4 in the monograph and Ware’s web article and about 1.5 in the 2022 sheet. Nobody outside Ware’s own laboratory has published measurements of this sensitiser, so there is no independent value to prefer.

Safety

Level B, and the level is set by the ammonia bottle and by the ultraviolet lamp — not by the ferricyanide and not by the finished sensitiser. Under the course rubric, Level B covers the handling of concentrated acids or alkalis and the use of ultraviolet exposure units, and both are here. This is the same level as the classic sensitiser and New Cyanotype, reached by a different route.

Ware’s own view, recorded as his claim and not adopted as a classification. His claim is that the sensitiser is “fairly non-toxic” because it contains neither dichromate nor oxalate, and that it may therefore be used by children. That is the originator’s judgement about the finished solution. This course does not adopt it, because it says nothing about the concentrated ammonia and the deliquescent oxidising nitrate that a person has to handle in order to make that solution. His own instruction sheet is the evidence for that distinction: it opens the preparation with “PROTECT YOUR EYES with safety goggles” and “Avoid inhaling ammonia vapour!”, and then distinguishes preparation from use when it comes to gloves. The course’s position is that the preparation is Level B work, that the coating, printing and processing of the made-up sensitiser is the ordinary Level A alternative-process routine, and that the ultraviolet source is a control of its own wherever it appears.

The substances, with the classification and the source for each.

• Ammonia solution is the reason for the level. The notified classification aggregated across 116 notifications is signal word Danger with H314 (severe skin burns and eye damage) in 99.9 per cent of the reports carrying hazard codes, H400 (very toxic to aquatic life) in 99.8 per cent, H335 and H318 in the minority. A 25 per cent solution boils at 38 °C and has a vapour pressure of 48 kPa at 20 °C, so the headspace of the bottle is loaded before you tip it. Ware’s condensed HMIS rating for the 28 to 30 per cent reagent is health 3, contact 3. Splash goggles, not spectacles; gloves and an apron; extraction or an open window with a fan running before the bottle is opened; an eyewash within reach. Add the ammonia slowly, to the iron solution and not the reverse, at room temperature.
• Iron(III) nitrate nonahydrate carries signal word Danger in the aggregated ECHA notifications, with H272 (may intensify fire; oxidiser) in 60.8 per cent of reports carrying codes, H315 and H319 (skin and serious eye irritation) in 70.1 per cent each, H335 in 61.9 per cent, and the more severe H314 and H318 in about 29 per cent. The hazard summary PubChem carries from Haz-Map adds that it may cause liver damage and may induce methaemoglobinaemia, and the threshold limit value is 1.0 mg/m³ as iron for soluble iron salts. Ware’s own condensed rating is health 1, reactivity 2, contact 2, with “may intensify fire; oxidizer”. Treat it as an oxidising solid: keep it off paper, cloth and wood, weigh it on a clean glass or plastic surface, and keep the jar closed against the deliquescence.
• Citric acid is the Level A member of the group: Warning, with H319, H335 and H315. Its eye irritancy is the hazard nearly every notifier agrees on, and Ware adds that prolonged or repeated exposure may provoke an allergic reaction in some people.
• Potassium ferricyanide is classified as a low toxic hazard that may act as a skin, eye and lung irritant. The one thing to know about it is under the next heading.

⚠Cyanide complexes, nitrates and ultraviolet, in one place

Ferricyanide and concentrated acid. The cyanide in potassium ferricyanide is bound to iron in a complex; it is not the hazard that potassium cyanide is and the two must never be spoken of as the same substance. The international chemical safety card records that it decomposes on contact with acids, and Ware’s qualification is that the heating must be very strong and the acid very concentrated to liberate much hydrogen cyanide. Nothing on this page approaches either condition — the strongest acid used is 1 per cent citric — but the rule that follows is absolute: no mineral acid ever goes near the ferricyanide bottle, the sensitiser, or the trays. That is also the reason this sensitiser cannot borrow New Cyanotype’s nitric acid developer. The course’s fuller treatment of the question, including one genuine disagreement between a university safety guidance document and established darkroom practice, is on the potassium ferricyanide page, and it is stricter than Ware.

Citric acid and a metal nitrate. This one is on the page rather than beside it, because the formula deliberately brings the two together. CAMEO’s reactivity profile for citric acid names metal nitrates among the substances it reacts with and says those reactions are potentially explosive. Ware puts the two in one beaker on purpose and wants the nitrate there. The two statements are reconciled by how it is done, and every part of the method matters: the nitrate is added as a solid, in small portions, to a concentrated aqueous acid solution, stirred, at room temperature, never heated, never mixed dry, and never scaled up beyond the published quantity. The course has found no source that draws the boundary explicitly, so the rule it gives is the conservative one — follow a published order and quantity exactly, do not improvise with a nitrate, and do not make ten times the batch.

Ultraviolet is on the same list, and this process uses a great deal of it. The guidance that warns about acid also warns about strong ultraviolet, which a cyanotype exposure is. Ware’s answer is specific: the traditional process, ferricyanide with an iron(III) salt, “will not emit any significant, perceptible or harmful amount of hydrogen cyanide during a normal exposure and processing”, whereas paper coated with ferricyanide alone — Herschel’s proto-cyanotype, which needs twenty to fifty times the exposure — can give a perceptible odour when the frame is opened. The practical rule is the one this formula follows anyway: the iron salt is always there, the exposure is minutes rather than hours, and the frame is opened in a ventilated room.

Ultraviolet. The exposure is 5 to 10 minutes at close range under a source that does not feel warm and gives no warning. Enclose the lamp, protect the eyes, cover the skin, and switch it off before reaching into the box. A facial tanning unit is a domestic appliance being used as laboratory equipment.

Working light. Dim tungsten or LED from the moment the ferricyanide is added. This is a photographic control rather than a hazard control, but it belongs with the rest of the setup: the preparation and the coating happen in the same low light, so lay out the glassware where you can find it.

Staining is a real nuisance rather than a hazard. The sensitiser marks skin, clothing, wood and anything absorbent, and Prussian blue is not easily removed once it has formed.

Storage

The solid iron salt is the one that will not wait. Keep it in a tightly closed container in a dry place, and buy it in a size you will use. Ware’s warning is unambiguous: the salt is highly deliquescent, and water it absorbs from the air destroys it irreversibly. A jar that has gone damp or begun to slump has not simply caked — its contents are no longer what the formula asks for, and its melting point of 47.2 °C means a warm room finishes the job.

The ammonia bottle loses strength every time it is opened. Check it with a hydrometer against Ware’s table of specific gravities if a batch of sensitiser comes out at the wrong contrast; a bottle that has weakened will under-neutralise, and a sensitiser that has been under-neutralised is not a grade Ware ever published.

The made-up sensitiser: brown glass, stoppered, labelled, dated, dark, cool. Several weeks at room temperature, much longer refrigerated at about 5 °C, and the date on the label is the only way to know which of those you are looking at. Look at the bottle before you coat. A clear yellow-green liquid is sound. Darkening towards blue, or a blue deposit, is Prussian blue forming in the bottle and the end of that batch’s useful life.

Let a fresh batch stand overnight before its first use, and expect a very small sediment of brown iron(III) hydroxide or Prussian blue from trace impurities, which is filtered off if it is in the way. A sediment that keeps returning is a different problem and is under Troubleshooting.

Separate the two halves if weeks are not enough. Both of Ware’s separated arrangements are under Mixing. The 2022 pair should keep indefinitely as long as they are apart; once mixed, the working sensitiser has an unpredictable life of a few weeks or even days.

Neither the Tween nor anything else goes into the stock bottle. Ware states that the surfactant does not last well when dilute and that the right dose depends on the paper.

Coated paper is a short-term store and the paper decides how short. Within a few hours if possible; longer in a cool, dark, desiccated enclosure. Light yellow is sound, green or blue is chemical fog and a verdict on the sheet rather than on the coating.

Finished prints keep away from alkali. Ware’s permanence note is short and applies to every cyanotype: buffered wrappings and mounts above about pH 9 must be avoided; prints can fade somewhat under daylight or bright gallery lighting at around 2000 lux, but the loss is regained on dark storage in air over a few days; and exhibition at 50 to 200 lux should cause no measurable fading.

Incompatibilities

Alkali, at every stage and in every form — buffered paper, alkaline sizing, alkaline tap water, household cleaners, buffered mountboard, and any sheet sold as “permanent” to ISO 9706, which by definition carries an alkaline reserve and is therefore exactly the wrong paper. Where a buffered sheet cannot be avoided, Ware’s pre-treatment is a bath of 5 per cent v/v hydrochloric acid or 5 to 10 per cent w/v sulfamic acid to destroy the calcium carbonate, followed by a wash. He rules out oxalic acid for that job on chemical grounds: calcium oxalate is as insoluble as calcium carbonate, so it removes nothing.

Hard water, for the calcium, at development, wash and everywhere else.

Mineral acids, anywhere near the ferricyanide, as set out under Safety. This includes nitric acid, which is the standard New Cyanotype developer and must not be brought to this one — Ware states that the Simple and Classic sensitisers tolerate only a very weak acid, above which they fog blue in the highlights.

New Cyanotype sensitiser, mixed with this one, for the potassium reason and the developer reason under Interactions.

Silver, and everything in a darkroom that has touched it. Ferricyanide is the oxidising half of Farmer’s reducer, and a shared cylinder, tray or sink will attack a silver image. Keep the glassware separate and labelled.

Metal trays, tongs, clips and sinks. An iron-and-hexacyanoferrate chemistry will find a trace of dissolved iron and print it.

Combustible material, in contact with the dry nitrate. It is a classified oxidiser; paper, cloth, sawdust and a wooden bench are not where it is weighed or spilled.

Gelatin-sized papers with Tween 20, which Ware notes elsewhere may interact unfavourably, and Tween 80, which foams.

Waste

Three streams and none of them is silver. The spent development bath, 1 per cent citric acid carrying peptized Prussian blue and dissolved sensitiser, replaced after two or three prints per litre; the wash water, which carries the rest of the unexposed sensitiser, the ammonium nitrate and the potassium salts out of the paper; and the occasional failed batch of sensitiser, which is the only concentrated item in the list.

Test the pH before anything else. Ware’s own general instruction for cyanotype waste is that if it seems very acidic — test it with pH paper — it can be neutralised with a cheap alkali such as lime or chalk. Do that in a labelled container rather than in the sink, and expect any excess alkali to destroy the Prussian blue in the bath, which is chemically expected and visually startling.

Never acidify a ferricyanide-bearing stream, and never send it down a drain that is also taking a chlorine-releasing household cleaner.

The nitrate is a nutrient, not an inert. A failed 100 cc batch carries roughly 7 g of ammonium nitrate equivalent, and nitrate is what eutrophies watercourses. Nothing from this process goes onto a garden, into a soakaway or into a watercourse. Bottle it, label it, and follow the general chemical waste SOP and the course’s disposal ruling. Check your local regulations, which are what actually govern and are the only document that can answer the question.

⁇How hazardous is cyanotype effluent?

The disagreement. Ware argues that cyanotype 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 substance by substance: ferricyanide reduces in the environment to the more stable ferrocyanide, a permitted food additive; Prussian blue is so insoluble that it is used clinically as an antidote to thallium and caesium poisoning; nitrate is of low toxicity and ammonium nitrate is spread on farmland in quantity as a fertiliser. He is writing for practitioners working in the countryside with septic tanks and land drainage.

Who holds what. That is a chartered chemist writing about his own process, and his argument is well founded on the toxicology of the individual substances. It is not, and he does not offer it as, a statement about what any particular sewer authority permits, and it says nothing about the quantity or the concentration at which a discharge stops being trivial.

What this course does and why. It publishes no jurisdiction-specific disposal instruction anywhere and makes no exception here. Collect, label, follow the SOP, comply with local regulation. Ware’s argument is recorded because a reader is entitled to know that this chemistry is benign relative to much of what else is in a darkroom, not because it settles what may go into a drain. The one place the course is firmer than Ware is the nitrate load of a discarded batch of sensitiser, which is a concentrated solution and not a tray of rinse water.

Troubleshooting

The sensitiser has darkened towards blue, or has thrown a blue deposit. Iron(II) has formed in the bottle and made Prussian blue with the ferricyanide. Filtering will clear the liquid but will not restore what has been consumed. Make a fresh batch, keep it colder and darker, use it sooner, and consider one of the separated arrangements under Mixing. If a fresh batch does this within days, suspect the iron salt: a deliquesced jar is the commonest cause.

A brown sediment on the first day. Iron(III) hydroxide from trace impurity, which Ware treats as normal and says may be filtered off. If it keeps returning, the neutralisation has gone too far — check the ammonia volume and the strength of the ammonia bottle.

The exposure scale is not the grade you mixed. Three sourced causes, in the order worth checking. The ammonia bottle has weakened, so the sensitiser is under-neutralised — check it with a hydrometer against Ware’s specific gravities. Or the make-up to 100 cc was estimated rather than measured. Or the coated sheet was heat-dried, which Ware records as raising the contrast by about a stop; heat-drying and air-drying are not interchangeable once you have calibrated.

Blue highlights. Ware’s first remedy is to make the citric acid development bath more dilute than 1 per cent, not to shorten the time. His second is a bath of 1 per cent ammonium oxalate for a few minutes, for which this course has no encyclopaedia entry. If the highlights were already blue before development, this is fog and not a processing fault — see the next entry.

The coating goes green or blue before it is exposed. Chemical fog. Ware’s diagnostic algorithm for fogged highlights, Table 7.1 in Cyanomicon 7.6, separates six causes by asking when the fog appeared, and the procedure needs two reference areas on every sheet: a margin of uncoated paper, and a border that was coated but masked during exposure. Fog visible on the coating itself means the sensitiser is already decomposed, or the paper is very hostile. Fog that appears during drying means a hostile paper chemical if it happens in total darkness and a faulty safelight if it does not. Fog that appears only after exposure means the negative’s maximum density is too low. Distinguish fog from stain: fog is unwanted image substance and reads grey to blue, stain is residual iron salts and reads yellow.

Yellow left in the sheet after washing. Stain, not fog: the sensitiser has not cleared. Wash longer, in water that is neither alkaline nor hard. This is the failure that ends as a print that goes on darkening in the mount.

Weak blacks. The development bath was water rather than 1 per cent citric acid, which lowers the maximum density by design. Or the sheet was underexposed, which on a print-out process you can see — the deep shadows should be reversed to a pale blue-grey before processing. Or the sensitiser has aged and part of its iron is already committed to blue in the bottle.

Blue speckles in the image. A two-bottle working solution that has begun to decompose and is carrying precipitated Prussian blue. Ware says filtering may sometimes rescue it. Mix smaller quantities more often.

The image weakens during washing. Less likely here than with a classic sensitiser, and the causes Ware names are alkaline or hard water, over-rapid drying of the coating before exposure, or a paper whose sizing holds the sensitiser in the surface pores instead of inside the fibres. A wetting agent addresses the third.

Soft, slightly doubled detail, worse in large flat areas. Carbon dioxide evolved during exposure, trapped between negative and paper. Back the sheet in the frame with papermakers’ felt or another porous material.

The shadows stay reversed. They are Prussian white and they need oxygen. Air will do it over hours; 50 cc of 6 per cent hydrogen peroxide per litre of the first wash bath does it at once, and changes only when the density arrives.

A whole session prints flat, on good negatives. The negatives are probably right for a different process. This sensitiser at the low-contrast grade wants an ultraviolet density range of 1.8 to 2.8, and a negative made for grade 2 or 3 silver-gelatin paper carries about a third of that. Mix the high-contrast grade, or develop in water.

Experiments

1. The contrast control, calibrated. Mix the low-contrast and high-contrast sensitisers, coat six sheets — 100 per cent A, 80 : 20, 60 : 40, 40 : 60, 20 : 80 and 100 per cent C — and print one step tablet through all six at the same exposure. Read the scales. Ware’s endpoints are about 2.7 and about 1.8 and his claim is that intermediate values can be matched accurately by proportional mixing; what nobody has published is whether the relation is linear in the mixing ratio. One afternoon settles it, and the answer is the calibration you will use every time you match a sensitiser to a negative.

2. The speed of the three grades, which is the missing measurement. Coat three sheets, one of each grade, and print each to the same maximum density rather than for the same time, recording the exposures. Ware’s published comparison holds the exposure constant, so this number does not exist anywhere the course has read — and it is the number that would show whether the contrast control is riding the quantum-yield maximum at pH 4 described under The mechanism. If the high-contrast grade at pH 8 needs substantially more exposure than the low-contrast grade at pH 4, that is evidence for the speciation explanation. If it does not, the explanation is wrong and something else is setting the scale.

3. Order of addition, demonstrated rather than asserted. Make two 20 cc batches at one-fifth scale. In the first, follow Ware’s order. In the second, dissolve the iron(III) nitrate nonahydrate in the water first and add the citric acid afterwards, then proceed identically. Compare the two solutions by eye against a white card for colour and clarity, leave both overnight, and coat and print both. The prediction from the chemistry is that the second is cloudier, browner, slower and lower in maximum density, because part of its iron is locked into hydrolysed polymers. This is the cheapest possible demonstration of why a mixing order is part of a formula.

4. Simple against Classic, in the wash water. Coat one sheet with this sensitiser and one with classic sensitiser, expose both to full density, and wash them in the same volume of the same water for the same time, catching the wash in a white tray. Ware’s claim is that there is little runoff of image substance here, and the classic process’s chief defect is that there is a great deal. You can see the difference in the tray before you see it in the print.

5. The nitrate as preservative. Split a fresh batch. Keep half in the dark at room temperature and half at about 5 °C, and read a coated step tablet from each at weekly intervals against a strip coated on day one. This measures the shelf life Ware gives as “several weeks” and “about a month” and turns it into a date you can put on a label. A stricter version, for anyone with the chemicals, prints alongside a batch of Mike’s cyanotype made from triammonium citrate — which contains the same nitrate — and a classic sensitiser, which contains none.

6. Acid against water, on the same sheet. Expose two identical step tablets, develop one in 1 per cent citric acid for half a minute and the other in water alone, and read both. Ware’s published figures are an exposure scale of about 2.7 against about 1.3 and a somewhat weakened maximum density for the water-developed print. This is the experiment that shows what the development bath is actually doing — stripping ligands off the iron(II) photoproduct so it can build pigment — and it is worth doing before you accept the common statement that a cyanotype has no development stage.

7. The internal filter, if you have two ultraviolet sources. Print the same step tablet under a 365 nm source and under one weighted towards 400 to 420 nm, and record the time to a fixed maximum density. Ware’s calculated absorbed fractions for a typical cyanotype coating are 0.65 at 365 nm and 0.03 at 420 nm, which predicts a very large difference. It is also the explanation for why two printers with “a UV box” can report exposures for the same formula that differ by a factor of ten.