Lead, thallium and nickel cyanotype toners

Every other toner in this formulary either converts the image substance into something else or lays a second substance beside it. These three do neither. They leave the Prussian blue where it is and walk a foreign cation in through the holes in it — and what comes out is the same pigment with a different guest inside, a different colour, and in one case ten times the resistance to the thing most likely to destroy it.

The chemistry is the most interesting in the whole of cyanotype toning. The three metals that do it are lead, thallium and nickel.

☠Level D: this page is not a procedure

Lead(II) acetate, thallium(I) sulfate and nickel(II) sulfate are each classified historical study only, and the Level D policy governs what this page may contain. It may give the chemistry, the practice in outline, the image characteristics, the measured results and the hazards.

It gives no procedure. There is no sequence of steps here, no immersion time, no bath temperature, no agitation, no wash, and nothing below is addressed to a reader as something to do. Where a source publishes one of those numbers, this entry says that it does and does not repeat it.

The strengths are the exception, and the reason is worth stating rather than assuming. They appear because the schema of a formulary entry cannot hold a formula without them; because a formula that cannot be read cannot be identified, and identification — of a print, of a paragraph in a book a reader already owns — is the use this entry has; and because the safety argument on this page is quantitative. A nickel bath at 5 per cent is a hundred times the concentration at which an already-sensitised person reacts to a patch test, and about fifty times the limit at which a British waste assessment calls a waste carcinogenic. Neither of those sentences can be written, or checked, against a number the reader cannot see. The strengths are here as evidence. No other part of the method is.

No protective-equipment list is offered, because the reason for the classification is not one that equipment answers. See Safety.

The toned cyanotype this course does teach is the tannic acid toner and the bleach-and-tone chemistry of the toned cyanotype. The nearest thing to these three that a reader may sensibly mix is none of them, and that is said plainly under When another formula is preferable.

Ingredient	Quantity	Form the source specifies
Lead(II) acetate	50 g	the trihydrate, Pb(CH3COO)2·3H2O, which is the form Ware's own table of solutions names; 50 g per litre is his "5% w/v"
Ammonia solution (ammonium hydroxide)	20 mL of a concentrated (S.G. about 0.9) solution	"Approximately 20 cc ... may be needed per litre", added to the made-up 5 per cent solution until it reads pH 7.5 to 8 on paper; a dose to an end point rather than a weighed quantity
Water	to make 1000 mL	5 per cent w/v of the lead salt, which Ware's notes define as grams of solute per 100 cc of solution and then write as 50 g dissolved in a litre of water. No temperature is recorded here. The finished bath is about 1020 mL, because the ammonia goes into the made-up litre.

Purpose

To change what a finished cyanotype is, by putting a different cation inside the pigment that is already in the paper. Ware states the general case in one sentence at the head of the section: treatment of a Prussian blue image with a solution of a heavy metal salt may cause the incorporation of the metal cation into the Prussian blue lattice, and thus modify its colour.

Three metals, three purposes, and they are not equally interesting:

Lead, for violet. The only one of the three whose shift its own author does not call slight, and the only one a supplier still prints as an option in a kit sheet sold today. Ware describes it as a shift towards a more violet blue, “sometimes rather inaccurately described as ‘lilac’ or ‘lavender’” — a hedge worth keeping, because it is a warning against the word a reader will otherwise carry away.

Thallium, for a bluer blue. A “slight but agreeable colour shift … to a ‘cornflower blue’”. Ware’s next clause is the refusal: this cannot be recommended as a toning practice for the general public, owing to the severe toxic hazard presented by the substance. That is the author of the standard monograph declining to recommend a method in the same sentence in which he describes it, and this entry does not improve on him.

Nickel, for permanence rather than for colour at all. The colour shift is “rather slight — towards a more greenish-blue”; what the bath actually buys is resistance to alkaline hydrolysis, by at least an order of magnitude. That is the single most serious weakness of the cyanotype answered by the only treatment in the literature this course has read that answers it. It is also a Category 1A carcinogen by inhalation, which is why this page exists rather than a lesson.

Recommended uses

What the sources record, as history and as reported experiment. The course recommends none of them, and one of them its own author declines to recommend.

A finished, fully processed cyanotype, and nothing else. All three are post-treatments. Nothing here is part of making a print.

A print made by one of the newer sensitisers, for the lead bath specifically. Ware records that the result may depend on the cyanotype formula and that it works well with the new cyanotype recipes. That is a real qualification and it cuts both ways: a reader identifying a violet print has a hint about how it was made, and a reader reading the method across from a classic-sensitiser print has been told, quietly, that it may not behave the same.

An object whose alkali resistance matters more than half a stop of mid-tone density, for the nickel bath. Ware tested it “in case it could offer a possible conservation procedure”, which is the frame it belongs in: a conservation judgement about a specific object, made by somebody qualified to make it, and not a darkroom preference.

Not on a print you cannot make again, for any of the three. The lead treatment is reversible by dilute nitric acid on Ware’s account, and the nickel one is not described as reversible at all; the density the nickel bath takes out does not come back.

Not for anything that has to be mounted, stored or handled by somebody who was not told. A print carrying a heavy metal in its image layer is an object with a hazard in it, and the person who frames it in forty years will not know.

When another formula is preferable

Each of the three effects has a modern answer in this course, and they are three different answers because the three effects are three different things.

• For a warm or a dark print, the tannic acid toner. It reaches brown through to black — a far larger change than any of these three make — from a plant polyphenol and a bottle of household ammonia. It works by destroying the Prussian blue and putting something else in its place, which is the opposite mechanism, and it costs the print its reversibility rather than its safety.
• For violet specifically, the mild borax bath the same kit sheet prints beside the lead one. Photographers’ Formulary’s cyanotype kit offers Violet Tones by either a mild borax solution or a warm 5 per cent lead acetate solution, with or between them and no way to tell them apart on hazard. They are not equivalent. The borax route is an alkaline bath on an alkali-sensitive image and would need its own reading before this course published it as a formula; it is still the one to read first, and the lead(II) acetate page sets out the comparison and its sources in full. Borax is not innocent — that page records the reproductive-toxicity statement it carries, which is why this course does not treat it as a Level A material either — but it is not a cumulative heavy-metal poison, and it does not leave a waste for which a household has no route.
• For a bluer blue, change the sensitiser rather than the print. This is the answer to the thallium bath, and it is not one toner substituted for another. Ware’s own hue coordinates make the point without being intended to: an untreated Herschel-sensitiser print and an untreated Ware-sensitiser print already sit a hue page apart on the Methuen scale, and a tone step apart as well, before anything is done to either — the same order of change as the one bath in this family whose colour shift anybody has measured. Whether thallium’s shift is larger than that is not established, because it is the shift nobody has put a number on; what is established is that choosing the process costs nothing and is a lever the reader already has. See Image characteristics.
• For alkali resistance, avoidance instead of protection. Unbuffered paper, unbuffered mounts, unbuffered enclosures and attention to the wash water are what this course teaches, and the troubleshooting entry on alkaline bleaching is where it teaches them. That is a different answer rather than the same answer more safely, and the difference is worth being honest about: avoidance protects a print from an environment you control, and the nickel treatment protects it from one you do not. Nothing else in the corpus does the second job.
• For a cyanotype that must not change at all, no toner. Every entry in this family begins by interfering with the one substance the picture is made of.

Mixing

This section gives no mixing procedure. What can be described without giving instructions is the shape of these formulas, and the shape is unusually informative, because two of the three have almost no shape at all.

Two of the three are one salt in water. The thallium bath is thallium(I) sulfate at about 5 per cent w/v; the nickel bath is a nickel(II) salt — Ware names the sulfate, the chloride and the nitrate as examples of one — at 5 to 10 per cent w/v. There is no second ingredient, no order of addition, no target pH and no published temperature for either. A single-salt bath is a strength and not a recipe, which is why this entry can print both in a sentence and has nothing further to withhold.

The lead bath has a second half, and it is the interesting one. A 5 per cent w/v solution of lead(II) acetate made up from the solid is acidic, at about pH 5, and Ware’s finding is that at that pH it does not tone at all. Bringing it to pH 7.5 to 8 with ammonia is what makes it work. So the formula in the table is not “lead acetate solution”; it is lead acetate solution plus the thing that admits the lead, and a page that printed only the first would be printing the bath that fails.

🧮Why 5 per cent w/v and specific gravity 1.24 cannot be the same bath

Ware specifies the historical bath and the modern working strength two sentences apart and in different units: Bolle’s is a solution of specific gravity 1.24, and the one whose pH he discusses is about 5 per cent w/v. Those look like two descriptions of one thing. They are not, and the arithmetic is one line.

A litre of a solution of specific gravity 1.24 weighs about 1240 g. A litre of water weighs about 1000 g. If the dissolved salt took up no room whatever — if the litre of solution were exactly a litre of water with salt hidden inside it — the extra 240 g would all have to be salt:

1240 g − 1000 g = 240 g per litre = 24 % w/v

The lower bound, assuming the solute occupies no volume

Any real solute takes up some room, so the water in that litre weighs less than 1000 g and the salt must weigh more than 240 g. Putting in the trihydrate’s density of 2.55 g/cm³, from its own page, and assuming volumes simply add:

m × (1 − 1 ÷ 2.55) = 240 g, so m ≈ 395 g per litre ≈ 40 % w/v

Salt per litre, m, on additive volumes

So Bolle’s bath is somewhere between five and eight times the strength of the one Ware works at, and it is a perfectly possible solution: the trihydrate dissolves to about 63 g per 100 mL in cold water, so 40 per cent is well inside its solubility.

The arithmetic is the course’s; both figures are Ware’s, and he does not remark on the difference between them. Volumes are not exactly additive in a concentrated salt solution, so the 40 per cent is an estimate; the 24 per cent is a floor and needs no assumption beyond the salt not shrinking the water. Neither reading brings the two specifications together, and what follows from that is an open question rather than an answer: whether a bath eight times stronger needed the pH correction at all is not something either printing addresses.

The hydrate is stated in one of Ware’s two printings and not in the other, and it matters here more than it usually does. The Cyanomicon writes “lead(II) acetate”; the workshop notes name Pb(CH₃COO)₂·3H₂O in the table of solutions. The lead(II) acetate page carries the arithmetic — the same weight of the trihydrate delivers about 86 per cent of the lead the anhydrous salt delivers — and the same trap sits under the nickel bath, where the gap is far wider: Ware’s notes name NiSO₄·6H₂O, and a reader who weighed the anhydrous salt to the same stated percentage would put 1.70 times as much nickel into the bath.

Nothing tells the reader the order to dissolve things in, and for once the order is not a free choice. The ammonia goes into the made-up lead solution and not the other way about: a lead(II) salt poured into ammonia meets a large excess of base and comes down as the hydroxide wholesale, where the same ammonia added to the lead solution to an end point throws only the slight precipitate Ware says can be filtered off. That is chemistry rather than instruction, and it is why the dose is stated per litre of the finished 5 per cent solution.

Behaviour

Recorded from the sources so that the formulas can be understood, not so that they can be run.

Times and temperatures are published and are not repeated here. Ware gives a bath temperature for Bolle’s original lead bath, a different one for the alkalised version, an immersion time for both, and an immersion time for the nickel bath; he gives none of the three for thallium. That inventory is what a reader is entitled to — the record is complete and the omission is this page’s — and the conditions themselves are the part that only helps somebody run the bath.

The lead bath fails silently on the wrong side of neutral. This is the behaviour worth carrying away from the whole page. It does not tone weakly at pH 5; it does not tone. A recipe that omitted the pH would look like a recipe that does not work, and for most of a century that is exactly what it was: Ware’s endnote dates the pH finding to his own unpublished observations of 1999, roughly a century after Bolle.

The nickel bath costs density while it works, and the source disagrees with itself about where. The toning section says the nickel solution “may cause a slight loss of highlight density”. The conservation chapter, reporting Ware’s own measurements, calls it a distinct loss from the mid-tones and puts numbers on it. Highlights and mid-tones are not the same part of a print and a reader deserves to know that one book says both.

Both changes survive washing, and one of them reverses in acid. The colour a nickel bath gives is not the green of nickel solution left in the paper — that would rinse away — and the nickel(II) sulfate page sets out why. The lead treatment, Ware records, “can apparently be reversed by treatment in a bath of dilute nitric acid, which removes the lead”. Note the hedge: apparently.

Nothing is published about capacity, keeping or exhaustion for any of the three. No source read for this course says how many prints a bath will take, how long it keeps, or what a spent one looks like. Those fields are empty in the data above because the sources are empty, not because the page declined to fill them.

One thing that is not behaviour and is easy to mistake for it. Ware’s Figure 8.2 shows four cyanotypes — untoned, lead(II), nickel(II) and thallium(I) — so all three of these are prints somebody made and photographed, not descriptions carried across from an older book. The evidence for the colours is a plate, and the evidence for the alkali resistance is a densitometer.

Image characteristics

Only one of the three colour claims is measured, and it is the smallest of the three. That is worth stating first, because it inverts the order a reader would expect. Ware adopts the Methuen atlas as his colour notation — a hue page numbered 1 to 30, a tone letter A to F and an intensity digit 1 to 8 — and he uses it exactly once in this family:

Specimen	Untreated	After the nickel bath
Herschel sensitiser	21D8	23D8
Ware sensitiser	20E8	21E7

📷What a two-page hue shift looks like, and what it does not

Read against Ware’s own key to the atlas, the Herschel specimen moves from page 21 — the page that carries Prussian blue itself, at 21F7 — to page 23, the page he labels with cyan, cerulean blue and sapphire blue. The Ware specimen moves one page, from 20, the page of lapis lazuli, to 21.

So “rather slight” means one to two hue pages out of thirty, with the tone letter unchanged in both specimens and the intensity dropping by one step in the second. That is a real, repeatable, describable change, and it is nothing like the difference between a blue print and a brown one.

Now set the other two against it. The lead violet and the thallium cornflower blue are described in words and measured in nothing — no Methuen coordinate, no spectrum, no colorimetry, in either of Ware’s printings. The one shift that carries a number is the one its author calls slight; the two that carry adjectives are the two he calls agreeable and violet. A reader deciding what any of these baths would do to their own print is reading, in two cases out of three, a colour word.

Violet, for lead, and the word to distrust is lilac. Ware’s phrasing is careful and this page keeps it: the shift is towards a more violet blue, and calling the result lilac or lavender is “rather inaccurate”. His workshop notes, writing for people about to do it, allow themselves “beautiful violet-blue”. Both are the same claim with different enthusiasm.

A bluer blue, for thallium. Cornflower blue is a blue: this is a shift within the hue family and not out of it. Set against the substance that produces it, it is the least favourable exchange in the toning literature — which is a judgement, and it is Ware’s before it is this course’s, since he declines to recommend it in the sentence after he describes it.

Ten times the resistance to alkali, for nickel, and this is a permanence characteristic rather than a visual one. Untreated cyanotypes lose between half and seven tenths of their shadow density to ten minutes in a pH 9.4 buffer; treated ones lose between 3 and 13 per cent. The full table, the sensitiser types it covers and the three cautions the course attaches to it are on the nickel(II) sulfate page, which owns that measurement.

About four times the light-fastness, for lead, on Ware’s unpublished observations of 1998 — “an exposure factor of about 4x compared with untreated images” — and no improvement whatever against alkali, which he states in the same sentence. A lead-toned cyanotype is still killed by an alkaline mount. The two heavy metals therefore protect against two different enemies and neither protects against both.

The metal is in the image and not in the paper. For lead, the evidence is an X-ray spectrometric measurement in which the lead signal correlates with the iron signal of the Prussian blue — a stronger claim than “the print changed colour”, because it locates the metal in the image substance. Two things about that evidence belong in the same breath: it is convincing, and Ware’s endnote records it as a private communication from David Saunders of the National Gallery, London, rather than as a published measurement. For thallium there is no such measurement at all, and Ware’s own account of the mechanism says “very probably”.

The mechanism

A pigment with holes in it, and three guests of three different sizes and charges.

⚗The cavity, and the cations that fit it

Prussian blue is not a dense solid. Its lattice is a cubic framework of iron atoms bridged by cyanide, and in the “soluble” form the potassium of KFe[Fe(CN)₆] sits in a cubic cavity inside it. Ware gives the number that governs everything on this page: the interior radius calculated for that cavity is 182 pm. Against it, the six-coordinate crystal radii of Shannon and Prewitt:

Cation	Radius (pm)	What the lattice does with it
Li⁺	90	Cannot be incorporated at all — its bound hydration shell makes it effectively far larger
Na⁺	116	Enters, and can be irreversibly exchanged for potassium
NH₄⁺	151	Enters, and diminishes the pigment’s tendency to peptize
K⁺	152	The native occupant
Tl⁺	164	Enters
Rb⁺	166	Enters
Cs⁺	181	Enters, and is irreversibly trapped

The whole of the thallium toner is in one row of that table. A thallium(I) ion is 164 pm and singly charged; a potassium ion is 152 pm and singly charged. Chemically, thallium(I) is a very heavy potassium, and the lattice cannot tell the difference — which is why the same pigment is the recognised antidote for thallium poisoning, and why the affinity is strong enough that somebody has built an analytical electrode for thallium(I) out of it. The thallium(I) sulfate page is built around that loop.

Lead is too highly charged to walk in as itself, and Ware’s explanation of how it gets in is the reason the pH matters. A lead(II) ion carries two charges and the cavity holds one. Ware’s suggestion, from the observation that lead incorporation improves in slightly alkaline conditions, is that it enters as the singly charged hydroxo species PbOH⁺, which is known to be present in the hydrolytic equilibria of lead(II):

Pb²⁺ + H₂O ⇌ PbOH⁺ + H⁺

The species the lattice can accept, and the equilibrium that pH moves

Read the equilibrium and the recipe together and they explain each other. Adding base removes H⁺ and drives the equilibrium to the right, making more of the singly charged species; the bath tones. An acidic bath drives it to the left, and the lead stays as Pb²⁺, which the cavity will not take. The toning is not the lead’s arrival but the lead’s admission, and pH is the door.

Ware marks this as a suggestion and so does this page. The correlation between lead and iron in the X-ray spectrum establishes that lead ends up in the image; that it entered as PbOH⁺ through a cubic void is an inference from the pH dependence and the lattice geometry, not a measurement.

Nickel is a different mechanism in the same lattice, and Ware treats it in a different paragraph. Nickel(II) is doubly charged and cannot be a one-for-one substitute for K⁺: one Ni²⁺ must displace two potassium ions, or arrive with something else to balance the books. Ware’s wording is that “soluble” Prussian blue may be rendered less liable to peptize by treatment with multiply-charged cations, most notably Fe³⁺ and Ni²⁺, “which are also believed to be partially exchanged for the K⁺ ions in the lattice” — and he does not resolve the stoichiometry. Neither does this page.

⚗Why one treatment does two apparently unrelated things

The link a reader needs is that peptisation resistance and alkali resistance are the same property here, and Ware states it in a single clause: rendering the blue “insoluble” is important because it confers a greater degree of resistance to decomposition by alkalies.

Prussian blue disperses in water as nanoparticles, and the tendency is believed to be associated with the loose cations in the lattice — which is why replacing K⁺ with the larger, less mobile NH₄⁺ already diminishes it, and why the new cyanotype is built on an ammonium salt rather than a potassium one. A doubly charged guest that bridges two cation sites at once ties the framework together harder still.

And a pigment that disperses less is a pigment that hydrolyses less, because hydroxide attacks the surface, and a solid that stays a coherent solid inside a paper fibre presents less surface than one that keeps breaking up into particles. One treatment, two effects, one reason.

The unpacking of that clause is the course’s; the clause is Ware’s. The nickel(II) sulfate page carries the same argument at greater length, together with the apparent contradiction in Ware’s own text between the nickel soak causing peptisation and preventing it.

⁇Why these three metals and not cheaper ones

Ware offers a rule for which cations shift the colour and labels it a conjecture: the metal may need two accessible oxidation states to intervene in the electronic charge-transfer transition that gives Prussian blue its colour. Lead (+2, +4), thallium (+1, +3), nickel (+2, +3) and copper (+1, +2) all produce a perceptible effect; alkali and alkaline-earth cations of similar size and charge do not.

There is no dispute in the literature this course has read, because nobody else has published on it; the disagreement is between the conjecture and the evidence for it, and Ware says so himself by calling it one. It is worth keeping for two reasons. It is falsifiable — it predicts which cations should fail, and the table above shows several that fit the cavity perfectly well and are said to do nothing to the colour. And it explains why this family is a family of heavy metals rather than a cheaper and safer set: if the requirement were only size and charge, sodium and rubidium would tone cyanotypes and nobody would ever have reached for thallium.

Note also what the conjecture does not claim. It is about the colour shift alone. Nickel’s alkali resistance is explained by the peptisation argument above and needs no second oxidation state at all, so the two halves of the nickel entry rest on two independent pieces of reasoning.

Function of every ingredient

Lead(II) acetate, 50 g in the litre, so 5 per cent w/v. The guest cation and the whole of the colour. Lead(II) is what enters the lattice and shifts the absorption that makes Prussian blue blue, and there is no other source of it in the bath. More would put more lead into a solution that already carries fifty grams of a harmonised category 1A reproductive toxicant per litre; less is not published as a control by anybody, and neither Ware printing offers a strength series or says what one would show. The form matters twice over. Its own page is the anhydrous salt, because that is what the binding European classification attaches to; Ware’s table names the trihydrate, and the same weight of the two differs by about 14 per cent in lead. And the counter-ion is not inert scenery: acetate is the conjugate base of a weak acid, but the lead cation’s own hydrolysis outweighs it, which is why a solution of this salt comes out acidic and why the next ingredient exists. That page carries the classification, the exposure regime and the reason lead is measured in the person rather than in the air, and none of it is restated here.

Ammonia, about 20 mL of the concentrated solution per litre, and this is a dose to an end point rather than a weight. Its function is not to make the bath alkaline for its own sake but to convert enough of the lead into the singly charged hydroxo species the lattice will accept; the target is pH 7.5 to 8, checked on paper. More and less both fail, in opposite directions and at different speeds. Too little and the bath stays acidic and does nothing at all — which is the failure mode that hid this formula’s central fact for a century. Too much precipitates lead hydroxide out of solution; Ware records that a slight precipitate does not matter and can be filtered off, which is a statement about a slight one and not a licence. Note the form: 20 mL is not a quantity and 20 mL of concentrated ammonia is, which is why the strength travels beside the volume. Ware gives that strength as a specific gravity of about 0.9 and glosses the same figure elsewhere as 30 per cent, while his workshop table calls concentrated ammonia about 27 per cent; the entry records the specific gravity, which is what the toning paragraph actually prints.

Water, to make the litre. It is the reason the salt sits at 5 per cent w/v rather than at some other strength, and 5 per cent w/v is defined in Ware’s own notes as grams of solute per 100 cc of solution. His instruction sentence says 50 g in a litre of water, which is a different statement; for a 5 per cent solution the two differ by under 2 per cent and nothing here turns on it. No source read for this course specifies distilled water, a temperature, or a water quality for any of the three baths — which is an odd silence in an iron process, where tap water brings its own iron and its own alkalinity, and it is recorded as a silence rather than filled in.

The two ingredients this formula does not have, because each is a bath of its own

Thallium(I) sulfate, at about 5 per cent w/v in water, and nothing else whatever. That is the entire published composition of the thallium toner: one salt, one strength, no second ingredient, no pH, no temperature and no time. Its function is the same as the lead salt’s — it is the guest cation — and it does the job more simply, because Tl⁺ is already singly charged and needs no hydroxo species to get through the door. More or less is not published as a control. What the encyclopaedia entry adds is the part Ware compresses into the words “severe toxic hazard”: a harmonised classification of fatal if swallowed, damage to organs through repeated exposure, a skin notation in both the British and American occupational limits because the substance is absorbed through skin, and a salt that is colourless, odourless and tasteless. The symmetry of this one is hard to improve on: the pigment being toned is the licensed antidote for the substance doing the toning.

Nickel(II) sulfate, at 5 to 10 per cent w/v, with the chloride and the nitrate named as alternatives. Its function is the one that is not about colour: it makes the pigment less liable to peptize and therefore harder for alkali to hydrolyse, and the colour shift towards green is a side effect its own author calls slight. More, within Ware’s own range, is the direction his conservation experiment ran — it is a 10 per cent bath there against the 5 per cent of his toning notes — and the direction the harm runs too, because the classification that governs this substance is a carcinogen and a respiratory sensitiser by inhalation, and the bath is already a hundred times the concentration at which a sensitised person reacts. The hydrate trap here is the worst of the three: Ware’s table names NiSO₄·6H₂O, and the anhydrous salt weighed to the same percentage delivers 1.70 times the nickel. The alternatives are not equivalent for that reason either — a nickel(II) chloride or nitrate at “5 per cent” is a third figure again, and no source converts between them.

Interactions

With alkali, in three different roles, which is the thing to hold on to about this whole family. Alkali is the reagent that admits the lead; it is the reagent that destroys the pigment the lead is being admitted into; and it is what the nickel treatment exists to defend against. The same substance is the tool, the threat and the reason for the treatment, at three different concentrations and on three different timescales.

With the print’s own paper and mount. A cyanotype on chalk-buffered paper has already lost part of its image, and a saturated solution of calcium carbonate — the buffer built into archival board — sits at the pH 9.4 at which Ware, reporting Holtzman, records Prussian blue as completely decolourised by irreversible hydrolysis in one to ten minutes. A stronger case again is 0.25 molar sodium carbonate at about pH 10.7, which does it in under half a minute. The lead treatment does nothing about this; the nickel one is the only thing in the corpus that does.

With the cyanotype’s own sensitiser residues. Lead(II) is thrown out of solution by carbonates, phosphates, sulfates, chlorides, citrates and tartrates, which is most of an alternative-process shelf and includes the ammonium iron(III) citrate of the classic sensitiser. A print that went into a lead bath carrying unwashed citrate is a print delivering a precipitant into the tray.

With the other toner on the same kit sheet. Tannic acid is listed as an incompatibility of lead acetate — tannin, phenol, resorcinol, salicylic acid, vegetable infusions and tinctures all appear in the same line of the hazard record. The two toners a beginner is offered side by side are chemically incompatible with each other, which is not a hazard so much as a sign that they belong to different chemistries and cannot be combined or run in sequence without thinking.

With chloride, thiocyanate and sulfide, against thallium, which precipitate thallium(I) out of solution — the first two sparingly, the third almost completely. With alkali, carbonate and sulfide, against nickel, for the same reason. Neither is a dangerous reaction; both are the reason a heavy-metal bath and the rest of a darkroom cannot share water, and both are traps of a particularly unpleasant kind, because they move the metal out of the beaker while a worker still thinks it is in there.

With ammonia, against nickel, in the opposite direction. Ammonia does not precipitate nickel(II) — it dissolves it, replacing the water ligands with ammine ones. A nickel residue that has met ammonia is in solution and passes through a filter, a rinse and a drain, where the same residue met by carbonate is a solid you can see. Since ammonia is also the reagent the lead bath needs and the bleach of the tannic acid toner, this is a realistic collision rather than a theoretical one.

With potassium ferricyanide left in a print, or with an unfixed sensitiser. Ware’s other metal-toning method — Murray’s patents, under Variants — depends on exactly that reaction: a metal salt meeting the unreacted ferricyanide of a freshly printed cyanotype throws an insoluble metal ferricyanide in the highlights. On a properly washed print there is no ferricyanide left to do it. On an imperfectly washed one there is, and the highlights are where it shows.

With the person, over years rather than over a session. This is an interaction and it belongs in this list. None of the three does anything to a worker on the day. Lead accumulates; nickel sensitises permanently and the asthma arrives hours after the exposure; thallium’s own safety card records delayed symptoms. A darkroom notices trouble by noticing it, and none of these three offers anything to notice.

Variants

Copper, the fourth member of the family, which Ware describes and this course does not tabulate. His §8.6.4 gives a toner derived from Harold Whiting’s United States patent 397,480 of 5 February 1889: a deep blue stock of tetraamminecopper(II) nitrate, made by dissolving copper nitrate in water and adding aqueous ammonia until the copper hydroxide precipitate just redissolves, then diluted sixteen times with more ammonia added to taste. The colour is a purplish-brown Ware reports as described in the period as a “Bartolozzi red”, and he says it is believed to contain some copper(II) ferrocyanide. That last clause makes it a different mechanism, and it is why copper is not simply a fourth column of this page: a ferrocyanide of copper is a new compound formed in the paper, not a cation lodged in an existing lattice. It is named here and not tabulated because copper(II) nitrate has no entry in this course’s chemical encyclopaedia, and the formulary does not publish a formula containing an ingredient a reader cannot look up.

Murray’s insoluble metal ferricyanides, 1933 and 1935. Ware records British and United States patents of Humphrey Desmond Murray in which a salt of a metal that forms an insoluble, lightly coloured ferricyanide is put in the first wet bath, immediately after printing out, at 10 to 20 per cent. It colours the highlights — where unreacted ferricyanide remains — leaves the shadows blue, and fixes the print by removing the light-sensitive material at the same time. Manganese(II) gives a salmon pink and nickel(II) a bright yellow. This is worth knowing for two reasons: it is the same metals doing something completely different, and it is a reminder that “a nickel salt on a cyanotype” names two unrelated procedures in one book.

Holtzman’s own preferred route, which is not a toner at all. Ware records that incorporating nickel into Prussian blue is best done at the time the blue is prepared, by forming the pigment in the presence of nickel salts, and that the retrospective soak is the second-best option. The course notes that neither Ware nor this page has read Holtzman’s 1945 paper directly; everything attributed to him here is Ware’s account of it, and the outstanding document is named on the nickel(II) sulfate page.

The lead bath as a supplier prints it today. Photographers’ Formulary’s kit sheet offers violet tones from “a warm 5 per cent lead acetate solution”, used until the desired colour is achieved and then washed for fifteen minutes. Its strength agrees with Ware’s and it says nothing at all about pH — which, on Ware’s finding, is the difference between a toner and a tray of nothing happening. The kit’s own entry carries what the sheet publishes.

Neighbours on the same page of Ware’s workshop notes, for a reader holding that document: a tannic acid toner for purplish-brown, which is the chemistry this course does teach, and a trisodium phosphate bath that bleaches a heavily printed cyanotype to a golden yellow, probably iron(III) phosphate. Neither is heavy-metal incorporation, and the second has no entry here because its reagent has no page.

No course variant is offered and none could be. There is no version of a lead toner without lead. The substitutes named under When another formula is preferable are not variants of these formulas: they are different chemistry reaching a different result by a different mechanism, and each has its own entry.

Safety

Level D, three times over, and for three different reasons. The classifications come from the three chemical pages, which own the GHS records, the occupational limits and the first aid; the Level D policy owns the rule those facts are read against. Neither is repeated here.

⚠Three substances, one level, three arguments

The classification rubric assigns a level to a procedure and not a rank to a substance, which is why three very different hazards land in the same place.

Lead(II) acetate — a cumulative poison with a harmonised classification. Repr. 1A, may damage the unborn child, legally binding across Europe rather than notified by companies. What rules out a home darkroom is not an event but the absence of one: there is no splash to catch and no vapour to extract, the harm is a body burden accumulating over years, and every occupational regime that governs lead measures the person — NIOSH ties its limit to a blood lead concentration, and Great Britain regulates lead under its own regulations rather than through the ordinary workplace exposure limits. A page of glove advice against that would be a lie.

Thallium(I) sulfate — acute, and undetectable by any sense. A harmonised Acute Tox. 2, fatal if swallowed, marked as a minimum classification whose true severity may be higher; STOT RE 1; skin absorption, with the Sk notation in EH40 and the skin notation on the American limits. The salt is colourless, odourless and tasteless, the symptoms are delayed, and its safety card’s control for a spill is a fitted respirator and complete protective clothing. It is also the one of the three a reader is very unlikely to be able to buy: its own page carries the record of the rodenticide’s withdrawal and the restrictions that followed it. The reason this entry exists is therefore not that somebody might purchase the salt. It is that they will meet the name, in a numbered section of the standard monograph, between two toners the same book does recommend, and have nowhere to look it up.

Nickel(II) sulfate — the one whose acute reading is misleading. Swallowing it is classified harmful, which is the exact phrase the rubric’s Level A criterion permits; and the same harmonised entry is Carc. 1A by inhalation, Muta. 2, Repr. 1B, STOT RE 1, a respiratory sensitiser and a skin sensitiser. Its safety card’s inhalation control is “use closed system or ventilation (not if powder)”, which excludes by name the one operation a darkroom would perform. And sensitisation does not undo: the card’s instruction to anyone who has reacted is to avoid all further contact with nickel and with copper, chromium and cobalt compounds, for the rest of their life.

All three meet Level C on hazard. All three are written at Level D because a home darkroom cannot supply Level C’s controls and because the course teaches a different route to each effect — which is Level D’s second criterion, and it is satisfied for the colour without argument and for the permanence only partly. See When another formula is preferable, where that gap is named rather than papered over.

🧮The strengths, weighed against two limits that have nothing to do with photography

This is the arithmetic the Level D note at the head of the page promised, and the reason the strengths are printed at all.

Against a patch test. The minimum level of nickel needed to provoke a reaction in already-sensitised subjects is recorded as of the order of 112 ppm of nickel, which the nickel page shows to be 0.05 per cent of the hexahydrate.

5 % ÷ 0.05 % = 100 times the minimum eliciting concentration

Ware's toning bath, in units of a patch test

At the 10 per cent of his conservation experiment it is two hundred times. Both figures describe a person who is already sensitised, and say nothing about the dose that sensitises somebody in the first place — which is the hazard the classification is actually about.

Against a waste limit. The Environment Agency’s worked assessment applies a concentration limit of 0.1 per cent for a waste to be classified HP7, carcinogenic, and 0.3 per cent for HP10, toxic for reproduction.

5 % ÷ 0.1 % = about 50 times

The same bath, against the carcinogenicity limit

A five per cent lead acetate bath meets the other limit on the same table: it is about seventeen times the 0.3 per cent at which a waste is classified toxic for reproduction, and it carries 50 g of a harmonised Repr. 1A substance in every litre. None of these is a marginal case, and none of them turns on a judgement about how careful a worker is.

Nothing on this page is a control and none is offered. The classification is not a statement about how difficult the controls would be; it is a statement that a page cannot assume them.

One of the three is still put in front of beginners. That is not a rhetorical flourish, it is the reason the lead entry was written: a cyanotype kit on sale now prints a lead acetate toning option in a paragraph as matter-of-fact as the one above it about borax, and a reader who buys that kit has the strength, the immersion and the wash in a sentence each, with nothing to tell the two violet options apart on hazard.

Storage

None of the three is stocked in a home darkroom, and the three chemical pages say so for the substances. What their safety cards have in common is worth reading as a group: original container, well closed, separated from food and feedstuffs, and — for all three — in an area without drain or sewer access, with provision to contain the effluent from fire fighting. A storage instruction that specifies the absence of a drain is a storage instruction written for a substance whose escape route is the thing being guarded against.

No keeping figure is published for any of the three baths, and none is invented here. The lead bath is the only one with an obvious reason to change on standing, since it is held at a pH it was adjusted to and ammonia leaves an open bottle; nothing in either Ware printing addresses it, and a reader should treat the silence as silence.

An old jar of any of the three is not what its label says. The lead trihydrate effloresces and takes up carbon dioxide, so an aged jar holds basic carbonate as well as acetate and no longer dissolves completely; the nickel hexahydrate is somewhat efflorescent, and its own records state plainly that a sample’s water content varies with whatever moisture it has met before, so a mass weighed out of such a jar is not the mass a strength assumed. Thallium(I) sulfate is recorded as very stable, which means the hazard of an old jar of it is simply the contents.

An inherited bottle is a waste question, not a find. See Waste.

Incompatibilities

Alkalis and carbonates, against a finished print — the pigment’s own vulnerability, and the reason unbuffered mounts and enclosures are the rule for every cyanotype whether it has been treated or not.

Carbonates, phosphates, sulfates, chlorides, citrates and tartrates, against lead(II), all of which precipitate it; and tannin, phenol, resorcinol, salicylic acid and vegetable infusions, which its hazard record lists in the same breath.

Bromates, against lead(II) acetate, which is the one genuinely violent entry in this family: the double salt that can form while preparing lead bromate from the acetate is explosive and very sensitive to friction, and its preparation has killed. Nothing in photography brings the two together, and the entry is here because the substance’s own record puts it first.

Chloride, thiocyanate and sulfide, against thallium(I); alkali, carbonate and sulfide, against nickel(II); and ammonia, against nickel(II) in the opposite direction, dissolving it rather than precipitating it.

Strong oxidants, against all three, and heat against all three, which decomposes each of them to fumes carrying the metal.

The dust of each of them, against itself. All three safety cards record that a harmful concentration of airborne particles can be reached quickly when the substance is dispersed, and lead acetate dust is recorded as able to form an explosive mixture in air.

Food, feedstuffs and anything that goes near them, which the safety cards and the transport rules treat as an incompatibility rather than as advice. See incompatibilities.

Waste

Nothing here is generated in this course, because none of these baths is made up.

There is no neutralisation to describe, and the reason is the same for all three. An acid can be neutralised because the proton is consumed; a dichromate can be reduced because chromium(VI) becomes chromium(III). Lead(II), thallium(I) and nickel(II) are elements in their stable oxidation states, and every treatment that exists moves the metal out of a solution and into a solid that is still lead, thallium or nickel waste.

All three are heavy metals by the definition the classification uses. Great Britain’s List of Waste defines “heavy metal” as any compound of a named set that includes lead, nickel and thallium so far as they are classified as hazardous substances, so a waste bearing any of these salts is assessed as a heavy-metal waste before anything else is considered. The concentration limits in the worked assessment — 0.1 per cent for carcinogenic, 0.3 per cent for toxic for reproduction — are orders of magnitude below the strengths these baths are made at.

All three carry aquatic classifications, and all three safety cards instruct that the substance must not be allowed to enter the environment.

There is no domestic route. Council hazardous-waste collection is the nearest thing GOV.UK points a householder at, and that page says on its own face that the service exists in England and Wales only. Where such a solution or such a jar exists, it is a matter for a licensed hazardous-waste route, and the disposal ruling explains why this course goes no further than that. The general chemical waste SOP is the course’s own procedure for the wastes it does produce. Local regulation decides, and this course cannot tell you what it says where you are.

Troubleshooting

There is no procedure here to troubleshoot. What follows is what the sources record going wrong, kept because each item explains a piece of the chemistry and because a reader working from a period manual, a kit sheet or a modern monograph will meet them.

“The lead bath did nothing at all.” The bath was acidic. A 5 per cent solution made up from the solid sits at about pH 5, and at that pH the lead stays as Pb²⁺, which the lattice will not accept. This is the single most useful sentence on the page and it is a century younger than the recipe it corrects.

“A precipitate appeared when the ammonia went in.” Lead hydroxide. Ware records that a slight one does not matter and can be filtered off. A heavy one means the base went in faster than the solution could take it, and what is on the filter is the lead that was meant to be in the bath.

“The print lost density in the nickel bath.” Expected, and the source disagrees with itself about where — highlights in one chapter, mid-tones in the other. Whether the loss is the price of the nickel or simply the price of an hour in a tray of water is an open question that the nickel(II) sulfate page sets out; the control experiment that would settle it does not appear in the text this course has read.

“The green washed out of the nickel-treated print.” Then it was not toning, it was staining. The nickel aqua ion is green in its own right, and a print merely carrying bath in its fibres rinses back to blue. The hue shift Ware measures survives washing, because the change is to Prussian blue’s own absorption rather than to what is lying on top of it.

“The toned print faded in the mount.” For a lead-toned print this is expected and Ware says so: the light-fastness improves by about four times and the vulnerability to alkali is undiminished. Buffered board is buffered board whatever colour the print is.

“I want the violet taken off again.” Dilute nitric acid, on Ware’s account, “apparently” removes the lead. Note both hedges — his and this one — and note that the reagent then contains the lead.

“There is an unlabelled jar of a white salt in an inherited darkroom.” Not a troubleshooting question, and not a tasting question either: sugar of lead earned its name honestly and is the rarest kind of poison, one that tastes pleasant. See Waste.

Experiments

No experiment on this page involves making or using any of these three baths.

Measure the blue you already have. Ware’s own hue coordinates put two untreated cyanotypes made with different sensitisers a page apart on the Methuen scale and a tone step apart as well — the same order of difference as the one bath in this family whose colour change anybody has measured. Print the same negative on the classic sensitiser, the improved classic, the new cyanotype and the simple cyanotype, dry them together, and describe the four blues against any colour atlas you can get hold of. That is the experiment the thallium bath is competing with, and it costs nothing but paper.

Run the control Ware did not. His nickel result has a hole in it that a single strip fills: soak an untreated cyanotype in plain water for the same length of time as the nickel soak, dry it, and read the mid-tone density. If it loses as much as the nickel-treated strip did, the density cost belongs to the water. If it loses less, it belongs to the nickel. Either answer is worth more than the paragraph the question currently sits in.

Find the document nobody in this chain has read. The lead toner is attributed to Oscar Bolle about 1900 on the strength of one citation: George E. Brown’s Ferric & Heliographic Processes, page 16, a book whose own date Ware’s endnote gives as undated and probably 1902. This course has read Ware and not Brown, and says so. Tracking down that page would settle whether Bolle’s bath was the strong one the specific gravity implies, whether Brown names a pH, and whether the attribution is Brown’s or somebody else’s before him. The same exercise is waiting on Holtzman’s 1945 paper.

Do the specific-gravity arithmetic yourself and then attack it. The Mixing section argues that Bolle’s S.G. 1.24 bath cannot be the 5 per cent solution Ware works at. Redo it with the anhydrous density instead of the trihydrate’s, then look up how far volumes really do add in a concentrated salt solution, and decide how much of the conclusion survives. The lower bound of 24 per cent should survive everything; the estimate of 40 per cent should not survive much.

Write the three hazard assessments and compare them. Take the three chemical pages, the classification rubric and nothing else, and place each substance at a level with the criteria named. You should find lead failing on an accumulating dose, thallium on an acute classification and a skin route, and nickel on a hazard its acute classification hides completely. Then read Safety above and see whether you reached the same level by the same route — and, more interestingly, whether you would have put all three at the same level at all.

Argue the exception. This entry publishes concentrations that the Level D policy says a Level D page does not publish, and gives a three-part reason. The uranium toner makes a related argument in different words, and Herschel’s hydrargyro-cyanotype faces the opposite problem — a bath whose strength nobody ever published. Read the three together and decide whether the line this page draws between a composition and a procedure is the right one, or whether it is a distinction that only looks principled from the inside.