Cross-contamination between alternative processes sharing a bench

Owned by Break/Fix — the iron-silver print that would not clear, which is where the cluster has accumulated enough processes for this to become the commonest inexplicable result, and which reduces the whole thing to three rules: a brush belongs to one process, a rod may be shared because it can be got genuinely clean, and nothing metal touches anything. Part XXI’s break/fix page states the principle this entry opens with — in this cluster the contaminants are each other’s reagents — and adds the order rule for a print carrying two processes. Both name this entry as carrying the full route list. It was written earlier from Part IX’s contamination work.

What you see

Four appearances, each pointing at a different route:

What you see	What it points at
Blue stains or blue tone on a print that should have none	Ferricyanide or ferrocyanide residue meeting an iron sensitiser
Small white spots, sharp, scattered, with no image	Crystals of a sparingly soluble salt formed where two sensitisers met
An image that bleached away in its own developer	The developer of one process used on the print of another
Fog or unexplained density with no light	A metal image acting on a sensitiser catalytically

Likely causes

1. A shared tool. Bostick & Sullivan’s instruction is blunt: if you coat with a brush, use it only for one process, and avoid a brush that has coated other alternative-process chemistry. A glass rod is the shareable alternative — if it is properly washed between uses.
2. Two sensitisers mixed. Ware calls mixing New cyanotype sensitiser with Simple or Classic “a chemical nonsense”, because the whole point of the New preparation is to eliminate potassium ions, and putting them back allows crystals of the sparingly soluble potassium ferric oxalate to form on the paper and wreck the print.
3. The wrong developer. Ware sets out the incompatibility exactly: New cyanotype is best with a mineral acid developer — 1 per cent nitric or hydrochloric, or 5 per cent sulfamic — while Simple and Classic can only tolerate a very weak acid, under 1 per cent citric, because stronger acids cause blue chemical fog in the highlights.
4. Splashes, dry chemical and shared vessels, which are Kodak’s four generic routes and apply here unchanged.

What is happening: the chemistry and physics

These processes share ions, and the ones they do not share are the ones that ruin each other.

Iron plus hexacyanoferrate makes Prussian blue wherever they meet, which is the whole cyanotype and is a disaster anywhere else. The ammonium iron(III) citrate page records the practical rule that follows: a stray drop, a shared pipette or a shared brush leaves a blue stain that does not wash out, and the instruction is a dedicated syringe or pipette for each solution and a third for the mixture.

Potassium plus ferric oxalate makes a crystal. Ware’s account of the New cyanotype is a whole preparation built around removing potassium ions for this reason, and the visible consequence of putting them back is white spots where a crystal sat — and possibly, he suggests, brush marks, if friction is promoting microcrystal formation in the coating itself.

Oxalate destroys Prussian blue. Ware is explicit that a cyanotype cannot be made first and platinum printed over it, because traditional platinotype developer will rapidly bleach and dissolve any Prussian blue — the same oxalate reaction the Prussian blue page gives as one of the three ways to undo the pigment.

And a finished image can act on a fresh sensitiser without any light at all. Ware reports that simply coating a processed platinum-palladium print with New cyanotype sensitiser has been observed to cause immediate deposition of Prussian blue on the graduated tones, without any exposure, and suggests the Pt/Pd image may be acting catalytically. That is a contamination fault in which neither party is a contaminant: it is two finished chemistries meeting.

!The confirming test: coat a blank and look for a process that is not there

Coat a sheet with each sensitiser separately, using the tools you actually use, and process each by its own route with no exposure at all.

• Blue appears on a sheet that has no ferricyanide in its formula: a tool or a tray is carrying it.
• White spots appear on an otherwise clean coating: a crystal is forming, and two sensitisers have met.
• Everything is clean: the tools are not the route, and the fault is in a bath or a bench surface.

Then test the developer separately. Process one exposed print of each type in its own developer and one in the neighbouring process’s. Ware’s incompatibilities predict what should happen, and a print that behaves as predicted has identified the route.

And keep an uncoated margin on every sheet. Ware’s fogging algorithm ends with exactly this question — stain in the uncoated areas of the paper means the wet chemistry is contaminated — and it is the one test that distinguishes a contaminated bath from a contaminated coating.

Diagnostic questions

• Is any tool shared? Brushes, rods, pipettes, trays, tongs, measuring cylinders, drying racks.
• Which sensitisers have been on this bench, and when? These solutions stain wood, metal and many plastics, which means they stay.
• Which developer went with which print? The commonest single error when two processes run in an evening.
• Are you printing one process over another? Ware’s order rule is specific: platinum first, then cyanotype — never the reverse.
• Was dry chemical weighed near open solutions? Kodak’s airborne route applies to a coating bench as much as to a darkroom.
• Are the bottles distinguishable? Several of these solutions are pale yellow-green liquids in brown bottles.

Corrective action

Discard contaminated stock. A sensitiser that has met another process’s ions is not recoverable by dilution, and the crystals it will form are a coating fault waiting to happen.

Wash the tools properly or retire them. A glass rod can be washed; a brush’s ferrule cannot, which is why the suppliers’ rule is one brush per process.

Re-coat and re-print. As with every coating fault, the sheet is the cheap part.

And do not adjudicate Ware’s argument about mixing processes by experiment on a picture. His position is that mixing two published formulae and naming the result is not inventing a process, and that it has yet to be demonstrated — with step tablet tests, not pictures — how New cyanotype is improved by mixing it with Classic. If you want to try it, the step tablet is the instrument.

Prevention

One brush per process, one rod washed between processes, one tray per chemistry.

Separate the benches, or separate the sessions. Running one process an evening removes most of the routes.

Label every bottle and every tray, per the labelling SOP.

Cover the bench. Bostick & Sullivan put newspaper or blotter under the sheet because these solutions stain the table — and a stained table is a reservoir.