Cyanotype (classic)

The cyanotype is the only process in this atlas that needs no fixer, and the reason is worth stating before anything else: its image substance is already insoluble and its unused sensitiser is already soluble, so water alone separates them. Everything below follows from that one asymmetry. The history is established on the Herschel page, the practice on the Part I assignment, and the chemistry of the two reagents in ammonium iron(III) citrate and potassium ferricyanide.

The chemistry

Light acts on the iron, not on the cyanide salt, and Herschel proved it by taking the process apart. In article 210 he left the ferricyanide out, exposed plain ammonio-citrate paper for four or five seconds — an effect “quite imperceptible to the eye” — and only then washed it in the shade with ferricyanide, at which point a strong blue appeared exactly where the sun had fallen. His conclusion is that the cyanide salt is “a mere precipitant on the nascent compounds resulting from that influence”.

So the process runs in three steps and a wait.

1. Ultraviolet and blue light reduce iron(III) to iron(II), using the citrate as the electron donor. Ware puts the quantum yield of the citratoferrate(III) system at about 0.45 at 365 nm and 0.28 at 436 nm, roughly a third of the trisoxalatoferrate figure — which is the whole reason the classic process is slow.
2. The iron(II) meets hexacyanoferrate(III) and precipitates Prussian blue, an intensely coloured, highly insoluble mixed-valence pigment. Ware records that the intermediate rearranges instantly by internal electron transfer, so the product is iron(III) hexacyanoferrate(II) however you made it.
3. Water carries the rest away. Unreacted ferric ammonium citrate and potassium ferricyanide are both freely soluble; the pigment is not.
4. Air finishes it. Prussian white formed by over-exposure is colourless and re-oxidises over some hours, which is why the print deepens as it dries.

The sensitiser is the process’s weak point rather than its chemistry. Ammonium iron(III) citrate is not a compound at all but a family of complexes whose iron content ranges from 14 to 28 per cent, and Ware’s judgement, quoted on that substance’s own page, is that its quality matters more to the outcome than the printer’s choice of concentration ratios.

Historical workflow

Article 223 of Herschel’s 1842 paper is, in outline, the recipe still sold in kits: paper washed with a mixture of ammonio-citrate of iron and ferrosesquicyanate of potash in about equal proportions, exposed under an object or a negative, then “thrown into water and dried”. Ware’s modern statement of the classic formula is a 20 per cent w/v citrate stock and a 10 per cent w/v ferricyanide stock mixed in equal volumes just before coating, with a wetting agent to carry it into the fibres, exposed until the image looks blocked up, and washed face down for twenty minutes.

The commercial history is on the Getty Conservation Institute’s record: the first commercial paper, papier ferro-prussiate, from Marion et Cie in Paris in 1872; the first blueprint machine reaching the United States at the Philadelphia Centennial of 1876; and roughly eighty years from about 1870 in which the cyanotype and its variants were how engineers and architects copied plans, until diazo processes replaced them. The photographic revival dates from the 1960s.

Herschel patented none of it, and Ware notes the consequence twice over: plagiarists had a free run, and nothing slowed the process down.

The image

Colour. One blue, and it is the pigment’s own. Ware gives the absorption as a strong broad band near 700 nm — intervalence charge transfer between an iron(II) and an iron(III) centre bridged by cyanide — so what is left of white light is blue. The maximum shifts between the varieties, 690 nm for the “soluble” form and 730 nm for the “insoluble”, and the print goes on deepening for hours after the wash.

Surface. There is no binder. The pigment sits among the fibres, so the sheet stays matte and the paper’s own texture is part of the picture. That also means the paper is a variable: an alkaline-buffered sheet destroys the sensitiser and the pigment both, and Ware’s remedy is an acid pre-treatment he himself calls tedious and undesirable, because it degrades the surface.

Scale. Short, and this is the classic process’s real limitation. Ware gives it an exposure scale of about 0.9 against about 2.2 for New Cyanotype — the difference between a process that needs a deliberately flat negative and one that can print a normal one. The image also reverses during exposure, darkening to blue and then going back to a pale blue-grey, which is what “blocked up” means and why the process is judged by inspection rather than by a clock.

Permanence

The permanence argument runs opposite to every silver process in this atlas, and the comparison is the most useful thing on this page.

What it does not have. No silver, therefore no sulfiding, no mirroring, no residual thiosulfate and no fixing step to get wrong. The image is a pigment.

What it has instead. Two specific weaknesses, both from Ware. Alkali destroys Prussian blue, and the destruction is not reversible: a buffer at pH 9.4 — no more alkaline than saturated calcium carbonate, which is the alkaline reserve built into archival board — decolourises it in one to ten minutes, and 0.25 mol/L sodium carbonate at about pH 10.7 in under half a minute. A cyanotype wants an unbuffered enclosure, which is the one case in this course where the standard conservation precaution is the wrong precaution. Light fading is real and reversible: where an oxidisable impurity such as citrate is trapped in the lattice the excited pigment picks up an electron and becomes colourless Prussian white, which air oxidises back. A faded cyanotype recovers in the dark.

A third behaviour is not fading at all. Ware is emphatic that “soluble” Prussian blue is a misnomer and that what looks like a solution is a colloidal dispersion; that peptization is why a cyanotype bleeds blue into its wash water long after you expect it to have stopped.

Hazards

The controls belong to the ferricyanide, not to the iron salt. Ammonium iron(III) citrate is classified irritant and is a Level A material; potassium ferricyanide carries the signal word Warning with serious eye irritation on 62.8 per cent of notifications and aquatic toxicity with long-lasting effects on 45.3 per cent, and is a Level B material. That is what sets the level for mixing and coating.

⚠The one incompatibility to carry out of this page

Potassium ferricyanide is a hexacyanoferrate: its cyanide is bound to iron, not free. Ware’s sentence is the one this course repeats — it “is incompatible with concentrated acids, which may release deadly hydrogen cyanide gas.” So cyanotype paper, and the water it was rinsed in, never meets an acid, a bleach or a heat source. Ware’s own assessment of the traditional process is that a normal exposure and wash release no significant or perceptible amount of hydrogen cyanide; the course states the incompatibility anyway, because a reader improvising with lemon juice or household bleach has left the process that assessment was made for.

One further prohibition, from the same author: a dichromate bath must not be used to force the re-oxidation of Prussian white, although it has been recommended in the past. Dilute hydrogen peroxide at 0.3 per cent for about half a minute does the same job.

Where the course teaches it

Three places, at two different levels.

Part I, at Level A, on paper somebody else coated. The botanical photogram assignment is the first permanent picture the course makes, and it is Level A precisely because the reader never mixes a sensitiser: a pack of pre-coated paper, a sheet of glass, a tray of water. Cost band £.

Part XXI, at Level B, from the two bottles, and it is a whole part rather than a lesson. Iron, Light and Prussian Blue is the chemistry: the sensitiser as a complex rather than an iron salt, the ligand-to-metal electron transfer, why the process must have ultraviolet and why an enlarger will not do, Prussian white and the print that deepens overnight, and the argument that the wash is chemistry rather than rinsing. What These Processes Need supplies the three things every process in the cluster depends on — the negative’s density range, the paper, and a dose you cannot see — with the numbers and their measurement conditions. Lab: Mixing, Coating and Printing a Classic Cyanotype is the practice, coating by rod and by brush, with a dose established on your own lamp because nobody else’s minutes are yours. Experiment: The Exposure Scale, Maximum Density and Contrast of a Cyanotype is where the 0.9 in this entry’s tonal-scale field stops being a published figure and becomes a measurement of your own paper. The New Cyanotype and the Other Ways to Make a Blue Print sets five formulations against each other and is honest about the answer for most readers still being this one. Lab: Toning the Prussian Blue Image uses the alkali sensitivity deliberately, and Break/Fix: The Cyanotype That Went Wrong works six failures as diagnoses, with the rule that the fault is usually in the paper or the water rather than in the bottle. The part closes with an assignment at scale — a print bigger than the tray, a botanical sequence made the way Atkins made hers, and cloth whose wash-fastness you measure instead of believing.

Everywhere the chemistry recurs. The iron(III) photochemistry here is the same photochemistry as the siderotype family — argentotype, chrysotype, Van Dyke Brown, kallitype, platinotype — with a different ending each time. What changes is what the iron(II) reduces, and Part XXIV is where that sentence is turned into a redox argument.