EDTA and sodium sulfite clearing sequence

An iron-silver print comes out of its developer with an image in it and a problem in it. The image is silver. The problem is iron — most of it soluble and on its way out in the wash, and a stubborn remainder chemically bound to the cellulose of the paper, invisible on the day, and quite capable of turning the highlights yellow in ten years and oxidising the silver image away in fifty. This is the sequence Mike Ware devised to get that remainder out. It is three trays, two rinses and half an hour, and the reason it is three trays rather than one is that chelation on its own gets only about half of the iron off the paper.

Clearing bath 1: disodium EDTA, ca. 5 per cent w/v — the acid chelating bath, which takes out most of the iron(III)

Ingredient	Quantity	Form the source specifies
EDTA disodium salt dihydrate	50 g	Ware writes the formula of the dihydrate in his chemicals list, and gives the quantity as "ca. 50 g" for a litre, or 100 g for 2 litres
Water	1000 mL, added	Ware's instruction is to dissolve the solid in about a litre of tap water at room temperature, which is water added rather than a make-up volume, and is why he calls the result "ca. 5 per cent w/v" rather than 5 per cent. Tap water, not distilled: he specifies it, and a hard supply is discussed under Interactions.

Clearing bath 2: sodium sulphite, ca. 2.5 per cent w/v — the reducing bath, which turns the iron(III) still bound to the paper into iron(II)

Ingredient	Quantity	Form the source specifies
Sodium sulfite (anhydrous)	25 g or 25 g (sodium metabisulphite, which Ware numbers as the solution in his own list and measures as a level tablespoonful to the litre)	Anhydrous, the salt sold for photography. Ware gives this bath as sodium sulphite or disulphite at about 2.5 per cent w/v, and names sodium bisulphite and Kodak Hypo Clearing Agent as further alternatives
Water	1000 mL, added	Again water added at room temperature, not a make-up volume. This bath is made fresh for a day's printing and discarded: it oxidises in air to sulphate, and a bath that has become sulphate reduces nothing.

Clearing bath 3: tetrasodium EDTA, ca. 5 per cent w/v — the alkaline chelating bath, which takes out the iron(II) and leaves the paper alkaline

Ingredient	Quantity	Form the source specifies
Sodium hydroxide	9.6 g	Dissolved in the water first. Ware also prints, for the same conversion, sodium carbonate anhydrous 12.7 g, sodium carbonate decahydrate 34.3 g and sodium bicarbonate 20.2 g
EDTA disodium salt dihydrate	45 g	Added to the alkali, not the other way about. Ware's own line for this bath is 50 g of tetrasodium EDTA in a litre; this is his published route to it from the disodium salt, and a reader holding the tetrasodium salt should weigh 50 g of that and leave the alkali out
Water	1000 mL, added	Ware: dissolve the alkali in a litre of water, then add the EDTA. The order is not interchangeable, and it is the one step on this page that needs its own precautions.

Used in this order — clearing an iron-silver or noble-metal print of residual iron

1. Clearing bath 1: disodium EDTA, ca. 5 per cent w/v — 10 minutes — Tray-processed at room temperature with intermittent agitation. Ware: this first bath must be acidic, pH 3 to 4, and tetrasodium EDTA must not be used for it.
2. Water — half a minute — A rinse, not a wash. Its job is to stop bath 1 travelling into bath 2.
3. Clearing bath 2: sodium sulphite, ca. 2.5 per cent w/v — 10 minutes — Made fresh for the session and discarded after it.
4. Water — half a minute
5. Clearing bath 3: tetrasodium EDTA, ca. 5 per cent w/v — 10 minutes — Ware: examine the print for a yellow stain of residual iron in the borders of unexposed sensitiser, under a bluish light, and prolong this bath if the stain is there.
6. Water — a minimum of 30 minutes in running water — Ware allows at least three fresh static baths where water is short. On an iron-silver print the toning and fixing baths follow, so this wash is the course's ordering rather than Ware's; see Recommended uses.

Tray-process the exposed paper with intermittent agitation at room temperature in disodium EDTA 5 per cent w/v for 10 minutes, rinse in water half a minute, the sulphite bath for 10 minutes, rinse in water half a minute, tetrasodium EDTA 5 per cent w/v for 10 minutes, then wash in running water for a minimum of 30 minutes.

Ware's step 2 is a develop-and-clear bath, because in the print-out platinum-palladium process the disodium EDTA does both jobs at once. An iron-silver print has been developed already, in citrate, tartrate or borax, or has printed out and been washed, so on those papers the same bath is a clearing bath only. The times are unchanged; what changes is what the bath finds.

Purpose

To remove residual iron from a print made by an iron-based process, completely, while the print is still wet, and before any of it can turn into something that cannot be removed at all.

The iron in question is not the iron that dissolves. A kallitype leaves most of its ferric oxalate and ferrous oxalate in the developer; a Van Dyke Brown leaves most of its ferric ammonium citrate in the first wash, which is why Bostick & Sullivan’s Van Dyke sheet asks for three to five minutes of plain tap water and nothing else. Photographers’ Formulary say it exactly for their New Kallitype kit: “most of these have dissolved into the developer but some remain in the paper’s fibers.”

That remainder is the subject of this page, and it is a different chemical problem from the one the wash solves. Ware’s account is that iron(III) binds to the hydroxyl groups of cellulose strongly enough that it is chemisorbed rather than merely wetted — strongly enough, he notes, that the same binding has been developed as a way of permanently dyeing cellulose fibres. Water does not lift it. Worse, above about pH 4 iron(III) hydrolyses to a colloidal iron(III) hydroxide that lodges in the fibres, and if that is left it ages irreversibly into iron(III) oxyhydroxide, FeO(OH) — the mineral goethite — which is insoluble even in dilute acid. Ware’s conclusion is the sentence to carry away from this page:

!Important

It is essential to remove all the iron(III) at the wet processing stage, before the print dries. After that the chemistry has moved on and the ordinary baths will not touch it.

The three baths do three different jobs, and none of them can do another’s. Bath 1 chelates iron(III) in acid, where it is complexed rather than hydrolysed. Bath 2 reduces whatever iron(III) is still gripping the cellulose down to iron(II), which grips it less. Bath 3 chelates that iron(II) in alkali, where the iron(II) complex is at its strongest, and leaves the paper faintly alkaline, which is what a sheet of paper wants for its own sake.

Recommended uses

Platinum and palladium prints, which is what Ware published it for. It is the wet processing of the Malde-Ware print-out platino-palladiotype, and in that process the first bath is doing two jobs at once: it develops the print as well as clearing it. Ware reports that Matthew Clarke and Dana Hemmenway, testing clearing procedures by X-ray fluorescence at the National Gallery of Art, found this one the most effective of all they tried, leaving less residual iron than any other combination — comparable with, or even less than, the iron already present in the uncoated paper — and that accelerated ageing of those sheets produced no perceptible yellow stain. That is the strongest evidence in the literature for any clearing procedure the course has read, and it is why this page exists rather than a page about citric acid.

Chrysotype, in the variant form described under Variants, where both chelating baths are the tetrasodium salt.

Kallitype and Van Dyke Brown, as the course’s extension of it. This is a reading and it is marked as one. Ware does not publish this sequence for the iron-silver papers; his own argyrotype clearing bath is de-chlorinated water acidified with about 25 g of citric acid to ten litres, at pH about 4. Three things justify the extension.

• Sandy King says so directly. Kallitype and platinotype share ferric oxalate as the light-sensitive substance and almost identical processing, and therefore “the developers and clearing agents used for platinum can be used for kallitype”. King is describing his own practice of a process he has published in detail, and he is describing exactly this substitution.
• A kit maker already ships it. Bostick & Sullivan’s traditional kallitype kit contains 250 g of EDTA clearing agent, and its instruction is to clear for three to five minutes in tetrasodium EDTA at two tablespoons to the litre. EDTA clearing of a kallitype is not a proposal; it is what is in the box.
• The residue is the same residue. The iron chemisorbed to the cellulose of a kallitype is the same iron in the same cellulose as the iron chemisorbed to a palladiotype. Ware’s own outline for non-chemists puts kallitype and argentotype in the same iron-based family as platinotype, “with appropriate changes in the chemistry, although the principles are the same”.

Where it goes in an iron-silver workflow. In the place the citric acid or potassium oxalate bath occupies: after development and the first rinse, before toning, and well before fixing. That ordering is King’s and Bostick & Sullivan’s, not Ware’s, and both are explicit about it — King’s fourteen steps run develop, rinse, clear, rinse, tone, rinse, fix; Bostick & Sullivan tell you to tone “between your clearing bath and your fixing bath”. Two constraints tie it down at both ends. Before it, King’s rule that the first rinse must be neutral or slightly acidic, because an alkaline rinse forms iron(II) hydroxide compounds in the paper that make complete clearing “difficult or impossible”. After it, the fact that the sequence ends alkaline, at pH 9 to 10, and the thiosulfate fixer and any noble-metal toner come next; the course rinses between, and says under Interactions why that is a precaution rather than a published instruction.

Where the yellow highlight will not come out of a print you already made. The troubleshooting entry is the place to start, and the answer there is usually this sequence run properly and promptly. If the print has dried and yellowed, see the last paragraph of Variants: the answer is a conservation treatment, not a darkroom one.

When another formula is preferable

• For a Van Dyke Brown that clears in water, water. The Van Dyke’s sensitiser iron is ferric ammonium citrate, which is freely soluble, and neither Van Dyke kit sheet asks for a clearing agent at all: Bostick & Sullivan give three to five minutes of plain tap water “until most of the yellow stain has left the highlights”, and Photographers’ Formulary about one minute in running soft water at 20 °C, with three separate one-minute trays of distilled or demineralised water where the supply is a problem. Their reason for insisting on the water quality is the whole of this page in two sentences: “if your wash water is slightly alkaline, the iron salts will not be removed. Hard water is not satisfactory; it usually contains dissolved iron salts, which will contaminate the print.” If the highlights come out white and stay white on the paper you use, there is nothing for a chelating bath to do that the wash has not already done, and three extra trays buy nothing. The case for the sequence on a Van Dyke is a case about the invisible remainder and about the print in twenty years, not about the print today.
• For a kallitype, 3 per cent citric acid, 30 g to the litre, which is King’s own recommendation and Photographers’ Formulary’s for the New Kallitype kit. King uses one bath and renews it often, because the chemical is cheap and proper clearing is “absolutely vital to print stability”; the Formulary uses two, or three, rotated up so that the freshest is always the last the print meets. It is simpler, it is one substance, and it is acid, which is where iron(III) is best removed. What it does not do is the second half of the job — the iron that chelation and acid together leave behind, which is what bath 2 exists for.
• For the Photographers’ Formulary kallitype kit, the kit’s own potassium oxalate bath, 60 g in 500 mL, five minutes at 20 °C. It is the developer chemistry used as a clearing bath, and oxalate is a good enough ligand for iron(III) to work. Note the kit’s own warning attached to it: potassium oxalate is toxic, this is the bath your hands are most likely to meet, and gloves are not optional.
• For platinum and palladium in the traditional developed-out method, the EDTA, citric acid and hydrochloric acid sequence, which the course keeps on its own page. That is where the historical hydrochloric acid clearing belongs, and it is a different argument — Ware’s objection to it is that dilute hydrochloric acid dissolves palladium and weakens the cellulose.
• For a stain that has already appeared in a dry print, none of these. Once iron(III) hydroxide has aged to goethite it is out of reach of a clearing bath, and the treatment is reductive dissolution with sodium dithionite, which is conservation practice rather than darkroom practice. The course has no encyclopaedia entry for that substance and gives no procedure for it. See Variants.
• Never as a substitute for fixing. This sequence removes iron. It does not remove the unreduced silver, which is what the dilute alkaline thiosulfate fixer is for, and it is not a washing aid either — see Interactions, where the two sulfite baths of an iron-silver workflow are told apart.

Mixing

Three jugs, three labels, and the labels matter more than usual. Two of the three baths are the same ligand at opposite ends of the pH scale, they look identical, and using the wrong one first is the single documented way to make this sequence worse than no sequence at all. Follow the labelling SOP and write the pH on the bottle as well as the name.

Bath 1. About 50 g of disodium EDTA into about a litre of tap water at room temperature, stirred in a large plastic jug. Ware specifies tap water, not distilled; a hard supply costs the bath some of its capacity, which is discussed under Interactions. Nothing is heated and nothing evolves.

Bath 2. About 25 g of sodium sulfite into about a litre of tap water. Ware measures the equivalent metabisulphite as “a level tablespoonful”, which tells you how much precision this bath needs. Mix it on the day you print and pour it away at the end; it is not a bath you keep.

Bath 3, and the one step on this page that needs care. Ware’s own instruction is 50 g of tetrasodium EDTA into a litre of water — one solid, no alkali, nothing to go wrong. The course’s chemical encyclopaedia has no entry for the tetrasodium salt, so this entry records his published route from the salt the encyclopaedia does cover: 9.6 g of sodium hydroxide dissolved in a litre of water first, then about 45 g of disodium EDTA added to it in portions, with stirring. That is a route, not an improvement.

⚠If you are making bath 3 from the disodium salt

Sodium hydroxide is corrosive to skin and eyes, and the course’s own sodium hydroxide page records, from CAMEO’s datasheet, that dissolving it can liberate enough heat to cause steaming and spattering. Alkali into water, never water onto alkali. Splash goggles rather than spectacles, gloves, a bench you can flood, and an eyewash within reach; the course’s alkali spill SOP is the procedure if it goes wrong. This single step meets the Level B criteria and the rest of the page does not; the classification rubric allows exactly that, provided the step is named, which it is here. The whole sequence is Level A if you buy tetrasodium EDTA and weigh out 50 g of it. That is what the course recommends, and the only reason this page does anything else is the state of its own reference shelf.

Ware’s “ca.” is not modesty, it is arithmetic. Fifty grams dissolved in a litre of water is not a 5 per cent w/v solution, because the solid adds volume: the finished bath is somewhat more than a litre and somewhat less than 5 per cent. The course’s formulary distinguishes a make-up volume (“water to make 1000 mL”, which fixes the concentration) from water added (“dissolve in a litre of water”, which does not), and Ware is doing the second throughout. That is legitimate here in a way it would not be in a developer: a clearing bath is judged by whether the margin is white, not by its molarity, and Bostick & Sullivan say so out loud about their own — “this is not a critical measurement; less will just take longer to clear, more will clear faster.”

🧮Why 45 g and 9.6 g make the same bath as 50 g

Disodium EDTA dihydrate, Na₂H₂EDTA·2H₂O, has a formula weight of 372.24. Sodium hydroxide is 40.00.

45 g ÷ 372.24 g/mol = 0.121 mol

The EDTA

9.6 g ÷ 40.00 g/mol = 0.240 mol, which is 1.99 equivalents per mole of EDTA

The alkali

Two protons come off, which is exactly what turns the disodium salt into the tetrasodium salt:

Na₂H₂EDTA + 2 NaOH → Na₄EDTA + 2 H₂O

Ware's conversion, written out

0.121 mol × 416.20 g/mol (Na₄EDTA·2H₂O) = 50.3 g of tetrasodium EDTA

What the tray now holds

Ware’s two figures reproduce his own 50 g to within a gram. The arithmetic is the course’s, from Ware’s published masses and standard formula weights; the numbers are his.

The same table gives the other alkalis for the same conversion, and they are the equivalents you would expect: 12.7 g of anhydrous sodium carbonate, 34.3 g of the decahydrate, or 20.2 g of sodium bicarbonate — 0.12 mole of carbonate, which is two equivalents, or 0.24 mole of bicarbonate, which is one apiece. The carbonate route keeps the whole page at Level A, at the cost of one step of reading: Ware writes “9.6 g of NaOH etc., or alkali as above”, and the alkali as above is that column.

Do not try to convert between the two EDTA salts by weight. 50 g of the tetrasodium salt and 45 g of the disodium salt hold the same amount of ligand, but they are six units of pH apart, and pH is the whole design of this sequence. Ware’s third conversion runs the other way: 31 g of citric acid added to each 100 g of tetrasodium EDTA makes the disodium salt, and he notes that the citrate left behind “can only assist the clearing”, since citrate is itself a chelator of iron.

Behaviour

It works by end point, not by clock. The published times are ten minutes a bath, and they are generous on purpose. What tells you the sequence has worked is the margin: Ware’s instruction is to examine the print for a yellow stain of residual iron in the borders of unexposed sensitiser, under a bluish light, and to prolong the final bath if the stain is there. This is why practitioners mask the negative — King tapes round the image area with red lithographer’s tape, or cuts a frame in construction paper — and it is that margin, “those areas that were masked during exposure”, that reports on the clearing. King’s own instruction is to clear “until there is absolutely no stain left in the sensitized but unexposed areas”.

The paper answers as loudly as the chemistry. Two independent limits are published, and they disagree by a factor of two. King’s is that a kallitype paper taking more than about four minutes to clear should be replaced with a better one. Photographers’ Formulary’s is that a print taking longer than ten minutes to clear indicates a paper too absorbent to use without sizing. Either way, a clearing time that keeps growing is a fact about the sheet, not about the bath — and Bostick & Sullivan add a third: “water hardness will sometimes affect clearing time.”

The image lightens while it clears, and it comes back. This is the observation that frightens people the first time, and King’s reassurance is worth quoting because it is specific about when the density returns: “the image will lighten considerably during clearing, but don’t worry because all the lost density will return during toning and fixing.” On a platinum print there is nothing to lose in the first place, which is one of the differences between clearing a noble-metal print and clearing a silver one.

Bath 1 does most of the work and empties first. Ware: “there is so little iron left in the prints by the time they reach [bath 3]; most comes out in the first bath.” That asymmetry is why the published capacities differ by a factor of two — about fifty 10 × 8 prints per litre for the disodium bath against about a hundred for the tetrasodium — and why Ware’s chrysotype notes recommend rotating the exhausted first bath out and the third bath into its place, then making a fresh third. It is the two-bath fixing discipline applied to a chelating bath, for the same reason: the last bath a print meets should be the cleanest one.

The two published capacity figures for bath 1 do not quite agree, and the course records both. Section 7.9 gives about fifty 10 × 8 prints per litre; section 7.22, discussing the two-litre bath of the published procedure, gives about sixty. Two litres at fifty per litre would be a hundred. The likeliest reading is that the second figure is the whole session’s figure — section 7.3 says the quantities listed there “suffice to process ca. 60 10x8 in. Pt/Pd prints” — rather than a contradiction of the first, but the course cannot demonstrate that and does not pick.

Bath 2 fails on the calendar, not on the print count. No capacity is published for it because it is not exhausted by iron; it is exhausted by air. Sulfite oxidises to sulfate on standing, and a bath that has become sulfate reduces nothing while still looking, smelling and pouring exactly like one that works. Ware’s instruction is unambiguous in three separate documents: made up fresh for a day’s printing, not stored, not re-used.

Image characteristics

A clearing bath is not supposed to have image characteristics, and most of what this sequence does to a print is remove something invisible. Four things it does are visible.

The margin goes white. That is the whole of the visible result and the whole of the test. Under ordinary tungsten light a faint iron yellow in an unexposed border is easy to miss, which is why Ware specifies a bluish light for the inspection: yellow is what a blue-rich source shows up.

The image lightens, then recovers. Discussed under Behaviour. On an untoned, unfixed kallitype the loss during clearing can look severe, and it is not.

The paper is left alkaline. Bath 3 finishes at pH 9 to 10 and, in Ware’s words, “leaves the paper sheet in an alkaline condition, which is desirable for its preservation”. A sheet of cellulose ages better slightly alkaline than slightly acid; and on an iron-silver print there is a second reason to welcome it, because Ware’s own argument elsewhere is that molecular oxygen is a more potent oxidising agent under acidic than under alkaline conditions, so the silver image is a little safer in an alkaline sheet than an acid one.

Nothing changes in ten years. That is the characteristic worth having, and it is the one you cannot see. Ware’s report of the National Gallery of Art work is that accelerated ageing of papers cleared this way produced no perceptible yellow stain, while the same test on prints given “short cuts” in clearing produced yellow stains that correlated with the residual iron measured by X-ray fluorescence — and, notably, produced them on sheets that “would have been judged acceptable as artists’ prints” on the day they were made.

The mechanism

What is in the paper when the print goes into bath 1

A sensitised sheet that has been exposed and developed holds far more than the image. On a kallitype: unreduced silver nitrate, silver already reduced to metal, iron(II) and iron(III) oxalato-complexes, free oxalate, the developer’s citrate or tartrate, and whatever the paper’s sizing and buffering brought to the party. Almost all of it is soluble and almost all of it leaves in the developer and the rinses. What does not leave is iron(III) bound to the cellulose itself.

Ware’s account of that binding is careful about what is known and what is inferred. The evidence that iron(III) binds cellulose strongly is good — strong enough that it has been developed as a method of permanently dyeing cellulose fibres — and the geometry of the binding, a cis-bidentate or facially terdentate coordination through the ring ether oxygen and the hydroxyls, is offered as conjecture. He is equally careful about the colour: a monomeric [FeO₆] centre is only weakly absorbing, but two such centres on adjacent chains can lose water and cross-link through a bridging oxide, and the resulting metal-to-metal charge transfer absorbs strongly in the visible. That, in his hypothesis, is why the yellow stain of an under-cleared print appears slowly, over years, rather than on the day.

Why iron(III) must not be allowed to hydrolyse

Above about pH 4, iron(III) hydrolyses. The chain is short and it ends badly:

[Fe(H₂O)₆]³⁺ → Fe(OH)₃ + 3 H₂O + 3 H⁺ → FeO(OH) + 4 H₂O + 3 H⁺

Hydrolysis, then irreversible ageing to goethite

Freshly formed iron(III) hydroxide will redissolve in dilute acid. Left alone it transforms slowly and irreversibly into the polymeric oxyhydroxide, which is the mineral goethite and is, in Ware’s words, “quite insoluble in dilute acids”. A print that dries with hydrolysed iron in it has locked the problem in.

There is a third party in this, and it is the tap. Calcium — from a hard water supply, or from a chalk buffer in the paper — precipitates insoluble calcium oxalate, which strips oxalate off the iron complex and starts the hydrolysis:

Ca²⁺ + [Fe(C₂O₄)₃]³⁻ + 2 H₂O → CaC₂O₄ + [Fe(C₂O₄)₂(H₂O)₂]⁻

Ware: calcium drives the iron complex apart

Bath 1: chelation, in acid

EDTA is a hexadentate ligand — four carboxylate arms and two tertiary nitrogens — that wraps around the six coordination positions of an octahedral metal without much strain. Its four acid dissociation constants are pK₁ 2.0, pK₂ 2.7, pK₃ 6.2 and pK₄ 10.3, and everything about this page follows from them: the mono- and disodium salts give solutions at pH 3 to 4, the trisodium salt about neutral, and the tetrasodium salt pH 9 to 10. Same molecule, six units of pH apart.

At pH 3 to 4 the iron(III) is complexed rather than hydrolysed, and the complex is enormously stable: the formation constant is 10²⁵.

Na₂H₂EDTA + Fe³⁺ → [Fe(EDTA)]⁻ + 2 Na⁺ + 2 H⁺

Bath 1: iron(III) chelated in acid

⁇The first bath must be the acid one, and one manufacturer disagrees

Ware is unusually direct about this, and he names the disagreement: “It is important not to use tetrasodium EDTA for the first processing bath — in spite of the fact that it has been recommended by Bostick & Sullivan — it has a high pH ~10 at which the iron(III) is not effectively complexed, but rather hydrolysed, with the eventual result of yellow or brown staining.” His New Chrysotype instructions say the same of the developer bath in fewer words: “Do not use tetrasodium Edta, which is alkaline, and will cause iron stains.”

Bostick & Sullivan’s kallitype sheet does specify tetrasodium EDTA, as a single bath, and their platinum-palladium sheet pairs an EDTA clearing agent with sodium bisulfite in one tray. So this is a real disagreement between a Tier 1 manufacturer sheet and a Tier 2 specialist, and the course does not pretend to resolve it by authority. What can be said is that Ware states a mechanism and a prediction, that the prediction is about a stain appearing years later rather than on the day, and that the accelerated-ageing evidence he reports is consistent with it. What cannot be said is that anyone has published a head-to-head test of the two orderings on an iron-silver print. That is an experiment, not a settled question.

There is a wrinkle worth knowing before deciding either way: Ware’s own chrysotype sequence uses tetrasodium EDTA for the first bath as well as the third. It is set out under Variants, and the difference is that in the chrysotype the acid job is done by a separate developer bath before it.

Bath 2: reduction, because chelation is not enough

Here is the fact that turns a two-bath idea into a three-bath sequence. Chelating agents alone do not strip iron(III) off cellulose. Ware cites measurements showing that only about 50 per cent of the iron(III) comes off suspensions of cellulose pulp even with ligands considerably more powerful than EDTA — CDTA and DTPA, whose formation constants are orders of magnitude larger. A stronger chelator is not the answer. Changing the oxidation state is.

2 Fe³⁺ + SO₃²⁻ + H₂O → 2 Fe²⁺ + SO₄²⁻ + 2 H⁺

Bath 2: sulfite reduces iron(III) to iron(II)

Where the bath is made from metabisulphite instead, the disulphite anion hydrates first and the reduction is written through hydrogen sulphite:

S₂O₅²⁻ + H₂O → 2 HSO₃⁻

Metabisulphite in water

HSO₃⁻ + 2 Fe³⁺ + H₂O → HSO₄⁻ + 2 Fe²⁺ + 2 H⁺

The same reduction, by hydrogen sulphite

Iron(II) is both less strongly bound to cellulose and far less prone to hydrolysis than iron(III), so after this bath the residue is in a form the next chelating bath can actually reach. The thermodynamics are comfortable in either direction — the iron(III)/iron(II) couple sits at +0.771 V, and the sulphate/sulphite couple at −0.93 V in alkali, +0.17 V in acid — so the sulphite reduces the iron easily, and more easily in alkali than in acid.

There is a small piece of history attached to this tray. Irving Penn, working on platinum in the mid-twentieth century, introduced a 5 per cent sodium bisulphite bath after his hydrochloric acid clearing, having noticed only that “the paper seems to whiten and generally clear”. Ware’s comment is that this “would have had the effect of reducing any residual iron(III) to iron(II), which is more easily removed from the cellulose; it is now a key step in the Malde-Ware processing sequence”. A printer’s observation, correctly explained forty years later.

Bath 3: chelation again, in alkali

The iron is now iron(II), and the optimum pH for chelating iron(II) is not the optimum pH for chelating iron(III). The formation constant of the iron(II) EDTA complex peaks at 2 × 10¹⁴ at about pH 10, which is precisely where a tetrasodium EDTA solution sits.

Na₄EDTA + Fe²⁺ → [Fe(EDTA)]²⁻ + 4 Na⁺

Bath 3: iron(II) chelated in alkali

The bath removes the last of the iron and leaves the sheet alkaline. Both halves are wanted, and Ware says so: the pH is “optimum for complexation of iron(II) and leaves the paper in a beneficial alkaline condition”.

What is different when the image is silver

Everything above is about iron and paper, and it is identical whether the image is platinum, gold or silver. Three things are not identical, and an honest page has to say which of them are established and which are the course’s reasoning.

Established: acid clears iron better, and acid is harder on a silver image. This is Ware’s own statement of the central difficulty of the iron-silver processes — to clear the print of iron salts “without dissolving the image silver in the presence of the oxidising nitrate ion”, for which alkaline treatments were recommended but “are not very effective at removing excess iron(III), which is better done in acid”. He gives the practical corollary for the argyrotype: a little ammonia in the clearing bath, taking it to pH 9 to 10, “inhibits the dissolution of silver, but may raise the level of residual iron in the image”. The sequence on this page is interesting for iron-silver work precisely because it refuses that trade: it takes the iron out in acid at pH 3 to 4, where acid works, and then finishes in alkali at pH 9 to 10, where the silver is safest and the iron(II) is chelated best.

Established: this silver is unusually vulnerable. Ware puts the particles of a brown plain-paper silver image at around 20 nm, far smaller than the wavelength of the light they scatter, and sitting in the paper fibres with no binder over them. A reaction that attacks the surface removes a large fraction of the metal, which is the same geometry that makes highlights bleach first in the fixer.

The course’s reasoning, not a published finding: a chelator lowers the silver ion activity, and that makes metallic silver easier to oxidise. Ware’s general principle, stated for thiosulfate, is that “any involvement of highly insoluble products or complex ions in the redox equilibrium greatly increases the ease of oxidising metallic silver” — the silver couple sits at +0.7991 V and anything that removes Ag⁺ from solution pulls it down. EDTA does complex silver(I). The course therefore expects a small effect in the same direction, and found no measurement of it on an iron-silver print in any source it read, no formation constant for the silver EDTA complex in a source it read, and no report of image loss in this sequence from anyone. Two things bound the concern: EDTA is a far weaker ligand for silver(I) than for iron(III), and the ten-minute baths are at room temperature and open to the air for no longer than a fixer is. Treat it as a reason to watch a first print carefully, not as a reason to avoid the sequence.

The course’s reasoning again, and the sharper of the two: sulfite is a reducing agent, and an unfixed iron-silver print still holds silver nitrate. The alkaline sulphate/sulphite couple at −0.93 V sits far below the silver couple at +0.7991 V, so on the published potentials sulfite should reduce dissolved Ag⁺ to metallic silver. In a platinum print bath 2 meets no soluble silver; in an unfixed kallitype or Van Dyke it may. The predicted symptom would be a faint overall veil or a highlight fog rather than a stain, and again the course has found no report of it happening and has not tested it. What follows practically is the ordering the page already recommends for its own reasons — develop, rinse thoroughly, then clear — because the rinse is what carries the free silver nitrate away before bath 2. It is written up as the second experiment below, and it is the most useful piece of work a reader of this page could do.

Function of every ingredient

Disodium EDTA, 50 g in bath 1. What it is: the disodium salt of ethylenediaminetetraacetic acid, sold as the dihydrate, Na₂H₂EDTA·2H₂O, formula weight 372.24, an unremarkable white solid that all six companies reporting to the European C&L inventory record as not meeting GHS hazard criteria. Why it is here: it is the only substance in bath 1, and its job is to take iron(III) off the cellulose and hold it in solution. What it does: it wraps six donor atoms — four carboxylate oxygens and two amine nitrogens — around the iron, forming a complex with a formation constant of 10²⁵. And, just as important, what it does not do: it does not raise the pH. With two acid protons still on the molecule the bath sits at pH 3 to 4, which is below the threshold at which iron(III) hydrolyses, so the iron is complexed rather than turned into colloidal hydroxide. More of it: Ware’s own strength is “ca. 5 per cent w/v” and Bostick & Sullivan say a stronger clearing bath simply clears faster; the honest limit is that iron(III) chelation caps out around half the bound iron however much ligand you add, which is the finding that makes bath 2 necessary. Less of it: longer clearing, and a bath that reaches its capacity — about fifty 10 × 8 prints per litre — sooner. What it interacts with: calcium, which it will chelate perfectly happily instead of iron, so a hard water supply or a chalk-buffered paper spends the bath faster.

Sodium sulfite, 25 g in bath 2. What it is: the anhydrous salt sold for photography, and the same substance the course meets as a developer preservative and as a washing aid. Here it is neither. Why it is here: because chelation alone removes only about half the iron(III) from cellulose, and the way past that ceiling is to change the oxidation state rather than to find a stronger ligand. What it does: it reduces iron(III) to iron(II). Iron(II) is bound less strongly to the cellulose and hydrolyses far less readily, so it can be lifted off by the bath that follows. More of it: Ware publishes 2.5 per cent and his chrysotype step sheet prints 2 per cent for the same bath, which tells you the tolerance; the practical cost of more is only chemical wasted. Less of it: incomplete reduction, and the residue that bath 3 leaves behind is exactly the residue that stains a print years later. What happens with time rather than use: it oxidises in air to sulfate and stops working, which is why this bath alone in the sequence is made fresh each session. What it interacts with: acid, which liberates sulfur dioxide — see Incompatibilities — and, potentially, any silver nitrate still in an unfixed print, for which see the last paragraph of The mechanism. Ware’s alternatives, at the same 2.5 per cent: sodium metabisulphite — which is the salt he numbers first in his own list, and which is a corrosive acid when moist — sodium bisulphite, or the working solution of Kodak Hypo Clearing Agent, whose own principal component is sulfite.

Disodium EDTA again, 45 g in bath 3, with 9.6 g of sodium hydroxide. What the pair is: Ware’s published route to tetrasodium EDTA from the salt most darkrooms have. What the alkali does: it is not a component of the finished bath in any useful sense — it is two equivalents of base that strip the two remaining acid protons off the ligand and are consumed doing it. The product is the tetrasodium salt and water, and the arithmetic is in the maths box under Mixing. What the tetrasodium salt then does: the same chelation as bath 1, on a different oxidation state, at a different pH. With no acid protons left the bath sits at pH 9 to 10, which is where the iron(II) EDTA complex reaches its peak formation constant of 2 × 10¹⁴. More alkali: a higher pH than Ware specifies, no benefit the sources describe, and a correspondingly more caustic tray. Less alkali: an incompletely converted ligand, a bath somewhere between disodium and trisodium, and the wrong pH for the iron(II) you have just made. Sodium hydroxide is the substance to respect on this page, and the alternatives Ware prints for the same conversion — 12.7 g of anhydrous sodium carbonate, 34.3 g of the decahydrate, 20.2 g of sodium bicarbonate — do the same job without the caustic. And the simplest route of all is to buy the tetrasodium salt and weigh out 50 g, which is what Ware publishes and what the course would record if its encyclopaedia had a page to point at.

Water, about a litre per bath, and Ware says tap. Not a filler. Tap water carries the calcium that this sequence’s chelator has to compete for, and the Formulary’s Van Dyke sheet warns that hard water “usually carries dissolved iron salts which contaminate the print” — an iron problem arriving in the solution meant to solve one. In a hard district, or where the paper is chalk-buffered, distilled water in bath 1 buys capacity. Ware’s own instruction is tap water; the exception is his.

Interactions

The two sulfite baths of an iron-silver workflow are not the same bath and must not be confused. This sequence’s bath 2 is 25 g/L of sulfite, it comes before fixing, and its job is to reduce iron. King’s hypo-clear is 10 g/L of the same substance, it comes after fixing, and its job is to displace thiosulfate so the wash can be shorter; the course covers that role at the sodium sulfite washing aid and at Kodak Hypo Clearing Agent. The reason the confusion is easy is that Ware himself names Kodak’s packaged hypo clearing agent as a perfectly good bath 2 — which it is, because its principal component is sulfite and sulfite is what the job needs. One product, two completely different reasons to reach for it, at two different points in the same afternoon.

The rinse before clearing must not be alkaline. King is explicit for the kallitype: if the first rinse is alkaline, “ferrous hydroxide compounds may be formed in the paper, making complete clearing difficult or impossible”. Photographers’ Formulary say the same for the New Kallitype kit — “water with a base ph will make the print difficult to clear” — and add that a neutral or slightly acid rinse also extends the life of the clearing baths. Neutral or slightly acid, then bath 1.

Hard water and chalk-buffered paper both spend the chelator. Calcium is a metal ion and EDTA does not care that it is the wrong one. Ware’s calcium equation also shows the second harm: precipitating calcium oxalate strips oxalate off the iron complex and pushes the iron towards hydrolysis, which is the reaction the sequence exists to forestall. Bostick & Sullivan report the practical symptom, that “water hardness will sometimes affect clearing time”.

The developer carries over, and an exhausted developer makes clearing harder. King’s observation is that unreplenished kallitype developer accumulates iron(II) and that this “will make it increasingly difficult to clear the print during processing”. The bath your print goes into is only half the story; the bath it came out of is the other half.

What follows bath 3 is alkaline paper, and the course rinses before the next step. Ware’s sequence ends in a wash because his print is finished. An iron-silver print is not: it still has to be toned and fixed. No source the course has read states what a sheet left at pH 9 to 10 does to a gold, platinum or palladium toner, or to a dilute alkaline thiosulfate fixer, and the course will not guess. What it does is put a rinse between them, on the general principle that carrying one bath into another is how baths are ruined, and note that the fixer that follows is alkaline anyway, so the direction of the carry-over is at least not a contradiction.

Tone before fixing, as always in this family. Both King and Bostick & Sullivan place toning between clearing and fixing, and both give reasons that belong to the fixer rather than to the clearing bath — a toned print does not bleach in the hypo. The clearing sequence does not change that ordering; it occupies the same slot the citric acid bath occupied.

Variants

The all-tetrasodium sequence, which is Ware’s own for chrysotype. In the New Chrysotype instructions the sequence is: a separate acid developer — disodium EDTA at about 1 per cent, or tartaric, citric or oxalic acid — then clearing bath #I of tetrasodium EDTA at 5 per cent, a rinse, sodium metabisulphite or sulphite, a rinse, clearing bath #III of tetrasodium EDTA at 5 per cent, and a wash of 30 to 60 minutes. Both chelating baths are the alkaline salt. This is not a contradiction of the “do not use tetrasodium EDTA first” rule — that rule is about the bath that first meets the iron(III), and in the chrysotype that bath is the acid developer. It is worth noticing for the iron-silver printer, because an iron-silver print also arrives with its acid stage already done: a kallitype has been through citrate, tartrate or borax and an acid rinse. Whether the disodium or the tetrasodium salt is the better first bath for a kallitype is therefore genuinely open, and the course does not claim to know.

Bath 2 by any of four reagents. Ware prints sodium sulphite, sodium disulphite (metabisulphite), sodium hydrogen sulphite (bisulphite) and Kodak Hypoclear powder as interchangeable at about 2.5 per cent. This page records the sulphite because it is the one the course’s own encyclopaedia treats as the darkroom staple and because the sequence is named for it; Ware’s own numbered solution is the metabisulphite, at the same strength, measured as a level tablespoonful.

Bostick & Sullivan’s combined bath, which is one tray rather than three: two tablespoons each of EDTA clearing agent and sodium bisulfite to a quart of water, one or two successive baths of five minutes, up to a dozen 8 × 10 prints. It is faster, it is simpler, and it puts the chelator and the reductant in the same tray at the same pH instead of separating them by pH and by order. Ware’s objection is specifically to their tetrasodium-first recommendation rather than to combining the reagents, and the course has found no test of the combined bath against the sequenced one.

Ware’s three conversions between the EDTA salts, all published in the same section and all recorded in the maths box under Mixing: tetrasodium to disodium with 31 g of citric acid per 100 g; either salt from about 35 g of the free acid H₄EDTA with the appropriate alkali; and disodium to tetrasodium with two equivalents of base. The course has no encyclopaedia entry for the free acid or for the tetrasodium salt, so those two routes are recorded here and not as ingredients.

A stain that has already appeared is a different problem with a different treatment, and this page does not give it. Once iron(III) hydroxide in a dried print has aged to goethite, no clearing bath reaches it. The conservation answer is reductive dissolution with a stronger reductant than sulfite — sodium dithionite, used with tetrasodium EDTA in the same bath at about pH 9 — and Ware traces that treatment from Helen Burgess’s work on chelating agents in conservation through Jacquelyn Rees and Megan Gent’s removal of iron stains from historic platinotypes at the Victoria and Albert Museum, to Clarke and Hemmenway’s finding at the National Gallery of Art that 0.01 M DTPA with 2 per cent dithionite at pH 8 outperformed EDTA. The course has no encyclopaedia entry for sodium dithionite and publishes no procedure for it. It is named here so that a reader with a stained print knows what to ask a conservator for, and because it is the clearest demonstration that a darkroom tray and a conservation bench are doing the same chemistry with the same reagents.

Safety

Level A for the sequence as Ware sells it, with one Level B step if you make bath 3 yourself. Three trays of dilute salts at room temperature, nothing heated, nothing volatile, no reaction between the baths. The classification rubric admits substances that are at most irritant or harmful if swallowed at the concentrations actually handled, and that is where two of the three baths sit. The exception is named in the callout under Mixing and is repeated here because it is the only real hazard on the page: dissolving sodium hydroxide is a Level B step — corrosive to skin and eyes, exothermic on dissolution, alkali into water and never the reverse, splash goggles, gloves, eyewash within reach; the sodium hydroxide page carries the sourced detail. Buy tetrasodium EDTA and the step disappears.

The powders are the bigger risk, not the baths. Ware’s own safety note for this process says it plainly: “dry powdery solids present a greater risk than solutions, because the dust can be inadvertently inhaled, so it should be well-contained and a mask worn.” Weigh over a tray, do not tip from a height, follow the weighing SOP.

What the hazard classifications actually say. Disodium EDTA dihydrate: all six companies reporting to the ECHA inventory record it as not meeting GHS hazard criteria, and the course’s EDTA page records that the HSE’s list of workplace exposure limits holds no entry for EDTA or edetate. Sodium sulfite: the aggregated notifications are split — about half report H314, causes severe skin burns and eye damage, about a fifth H315 and about a fifth H319, about a fifth H302 harmful if swallowed, and about a quarter of reports record no GHS hazard at all. Sodium metabisulfite is more consistently classified, with H302 and H318, causes serious eye damage, in nearly every report, and PubChem’s description adds that it is toxic by inhalation and “a corrosive acid when moist”. Sodium hydroxide is corrosive in essentially every report there is.

Gloves or tongs, throughout. Ware’s instruction for this sequence is unambiguous: “do not allow the processing solutions, especially bath 2, to come in contact with your skin: use print tongs or gloves.” Note which bath he singles out. See glove selection for what the published permeation data support.

Ordinary room light is fine. Nothing here is light-sensitive; the print has been exposed and developed, and the clearing trays can be worked under normal lighting.

Storage

Baths 1 and 3 keep and are re-used. Ware says of the tetrasodium bath that it may be stored, and that in the platinum sequence it has a long life “because there is so little iron left in the prints by the time they reach it”. It is changed “when discoloration begins to be apparent”. No source publishes a period in months for either EDTA bath, and the course will not invent one; what is published is a print count, and it is under Capacity above.

Bath 2 is not stored. Three separate documents say so. It is oxidised by air to sulfate, and the failure is silent: a dead sulfite bath is indistinguishable from a live one until a print fails to clear. Mix it on the day.

Label the two EDTA salts apart and keep them apart. They are both white solids, they are used at the same 5 per cent, they are six units of pH apart, and Ware’s warning about using the wrong one first is the mildest consequence of confusing them. If you make bath 3 by conversion, label the bottle with what is in it and how it was made, not with the name of the salt you weighed — see the container labelling SOP and the troubleshooting entry on an unlabelled or undated container.

Keep the solids dry. Metabisulphite in particular is described as a corrosive acid when moist, and a damp jar of it is both a hazard and a bath that will not reduce anything.

Incompatibilities

Sulfite and acid. Acidifying a sulfite or metabisulphite solution liberates sulfur dioxide. Bath 2 must never meet an acid stop bath, an acid fixer, a citric acid clearing bath or the acid contents of bath 1 in a shared container. Between bath 1 and bath 2 there is a rinse, and the rinse is there for this reason as much as for cleanliness.

Sodium hydroxide and acid. If you make bath 3 from the disodium salt, the caustic solution neutralises violently and exothermically with any acid it meets. Keep the alkali step away from the acid trays entirely; make it on a different bench or at a different time.

The two EDTA salts, with each other. Not a chemical incompatibility — a procedural one, and the one this sequence is most likely to fall to. Tetrasodium in bath 1 hydrolyses the iron and stains the print; disodium in bath 3 puts the wrong pH on the iron(II).

Calcium. A chalk-buffered paper and a hard water supply are both, in effect, competing ligands for the chelator. See Interactions.

Chlorinated water. Ware acidifies the argyrotype’s own clearing bath with citric acid partly to scavenge chlorine, and warns elsewhere against highly chlorinated supplies for iron-silver work. The course records the caution and has no measurement of its size for these baths.

Nothing in this sequence is compatible with a food container or a kitchen utensil. The general rule, and it applies to a bath of chelated heavy metal as much as to anything.

Waste

Bath 1 is the valuable stream on a noble-metal print and should not go down the drain. Ware’s instruction is to save the spent disodium EDTA bath for recovery of precious metals: it carries whatever platinum or palladium did not become image. On a kallitype or a Van Dyke there is no platinum in it, but there may be silver, and the same logic applies — see the silver-bearing waste SOP.

The iron is the environmental question rather than the economic one. All three baths leave carrying iron held as a soluble chelate, and that is what makes the stream different from a rinse: the chelate keeps the metal in solution instead of letting it settle out, and EDTA is poorly degraded in conventional treatment. Bottle the clearing baths rather than pouring them.

Bath 2 is a spent reducing agent and leaves as sulfate and sulfite. Bath 3, if you made it with sodium hydroxide, leaves alkaline; check the pH before it goes anywhere.

Jurisdiction governs. Kodak’s guidance for amateurs sends developers, stop baths, fixers after silver recovery and wash water to a household sewer where there is one, while naming what must not go; ILFORD tells domestic users in the United Kingdom to bottle wastes separately, label them and take them to a household waste and recycling centre. The course publishes no jurisdiction-specific instruction anywhere and makes no exception here. Follow the general chemical waste SOP and the course’s disposal ruling, and check your local regulations.

Troubleshooting

Yellow in the highlights or the masked margin, wet. Residual iron, and the sequence has not finished. Ware’s own instruction is to prolong bath 3 and to look under a bluish light. If prolonging bath 3 does not shift it, the fault is upstream: an exhausted bath 1, an alkaline first rinse, or a paper too absorbent to clear. The full entry works through the diagnosis.

Yellow that appeared months or years after the print dried. Not a bath that failed on the day — a residue that was there all along and has now cross-linked and darkened. This is the exact failure Ware’s accelerated-ageing account describes, on prints “judged acceptable as artists’ prints” when they were made. It is out of reach of a clearing bath; see the last paragraph of Variants, and the platinum equivalent for the mechanism written out.

Clearing got slower as the session went on. In order of likelihood: bath 1 has reached its capacity — about fifty 10 × 8 prints per litre — and should be replaced, or rotated out and the tetrasodium bath moved into its place with a fresh third bath made up; the developer has accumulated iron(II) and needs replenishing, which is King’s diagnosis; or the water is hard and has been spending the chelator on calcium all along.

The print lightened badly in bath 1. Expected on a silver image, and King’s assurance is that the density returns during toning and fixing. If it has gone further than that, suspect a print that was under-exposed rather than a bath that was too strong: this family is printed dark deliberately, and the entry on a print bleached in the fixer covers the same loss one bath later.

A veil or fog over the whole print after bath 2. The course has no report of this happening and records it as a predicted symptom only, for the reason set out at the end of The mechanism: sulfite is a reducing agent and an unfixed print may still hold silver nitrate. If it occurs, lengthen and freshen the rinse between the developer and bath 1, which is what carries the free silver away.

Nothing happened at all in bath 2. The likeliest cause is a sulfite bath that was kept. It oxidises to sulfate on standing and gives no sign of having done so.

The bath went brown, or a sludge appeared. Iron, which is the bath doing its job, and in a platinum print also dissolved metal — Bostick & Sullivan’s advice for a developer in the same state is to filter it through a coffee filter. Discoloration is Ware’s own criterion for changing bath 3.

You cannot remember which jar is which. Stop, and do not guess. Test a spoonful in a little water with pH paper: about 4 is the disodium salt, about 9 to 10 the tetrasodium. The consequence of guessing wrong is the one failure on this page that shows up years later.

Experiments

The clearing audit. Coat and mask a sheet, expose it, develop it, then process it through bath 1 only. Process a second the same way through all three baths. Process a third through the three baths with each time cut to one minute. Dry all three, keep a strip of each in the dark, and put a strip of each in a south-facing window or a warm dark place for six months. Compare the masked margins under a blue-rich light against a strip of the uncoated paper. This is Clarke and Hemmenway’s accelerated ageing done with a windowsill instead of a humidity oven and a pair of eyes instead of an X-ray fluorescence spectrometer, and it will answer the only question that matters about a clearing procedure. That it takes six months to answer is the point.

The sulfite-and-silver question, which is the useful one. Take two identical unfixed Van Dyke or kallitype prints. Give the first a thorough rinse before bath 1 and put it through the sequence. Give the second a deliberately brief rinse. Compare the highlights and the unexposed margin after both are fixed and dried. If the prediction at the end of The mechanism is right, the second will carry a faint veil the first does not; if it is wrong, they will match, and the course will have learned something it could not find in any source. Record it in the lab notebook either way.

Acid first or alkali first, on a kallitype. The disagreement between Ware and Bostick & Sullivan is on the record and has never, so far as the course can find, been tested on an iron-silver print. Run the full sequence on one print with disodium EDTA as bath 1 and on a second with tetrasodium EDTA as bath 1, everything else identical, and age both alongside a control. This is the same experiment as the clearing audit with a different variable, and it needs the same patience.

Where the density goes and whether it comes back. Read a step wedge printed on your paper before clearing, after bath 1, after bath 3 and after fixing, using a densitometer or a visual comparison against a reference strip. King says the lost density returns during toning and fixing. Find out by how much, on your paper, with your developer, and whether the loss is even across the scale or concentrated in the highlights — the geometry argument in The mechanism predicts the highlights.

Capacity, honestly measured. Ware’s fifty prints per litre is for 10 × 8 platinum prints on his papers. Clear prints through a measured litre of bath 1, one at a time, checking each masked margin, until a margin stays yellow after ten minutes. That number is your capacity, and it will be a different number from anyone else’s.