Part XXV Overview: Platinum and Palladium

For four parts you have been running the same reaction with a different second reagent. Light reduces iron(III) to iron(II); the iron(II) then makes the picture out of whatever was coated beside it — Prussian blue in Part XXI, silver in Part XXIV. This part changes that second reagent one last time, to a noble metal, and everything that is different about the part follows from what noble metals cost and how inert they are.

Two things make it unlike the four before it. The first is that the consumable is the expense: about 40 mg of palladium on an 8 × 10 print, which is a quantity you count in drops and a cost the course cannot currently quote you. The second is that the image substance will outlive the paper it is in, which turns “permanence” from a slogan into an argument about which part of the object you are talking about.

The third variation on one reaction

The pattern is Part XXI’s and it does not change here.

UV + Fe₂(C₂O₄)₃ → 2 FeC₂O₄ + 2 CO₂

Act one, unchanged since Part XXI: ultraviolet reduces the iron and the ligand pays for it with carbon dioxide

2 [Fe(C₂O₄)₂]²⁻ + [PdCl₄]²⁻ → 2 [Fe(C₂O₄)₂]⁻ + Pd + 4 Cl⁻

Act two, this part's own: iron(II) hands two electrons to palladium(II) and metal is left where the light fell

But one change is not cosmetic and it is why this is a separate part. The iron salt is an oxalate rather than a citrate, and the reason is a redox potential: the citrato iron(III)/iron(II) couple of a cyanotype sensitiser sits at +0.372 V, which reaches silver at +0.80 V and gold at +1.00 V but does not reach the tetrachloropalladate couple at +0.62 V or the tetrachloroplatinate at +0.73 V. The oxalato couple, at +0.02 V, reaches all four. Change the ligand on the iron and the noble metals become available; that is the whole of the step from Part XXI to here.

And one thing about the reaction above is genuinely surprising, so it is worth meeting before the first page: in the traditional process the iron(II) that light makes is insoluble and cannot move. The developer’s job is not to reduce anything — it is to dissolve the photoproduct so it can reach the metal sitting a few micrometres away. The chemistry lesson is built around that fact, because most of what is odd about this process comes out of it.

What is different in practice

Four things, and each of them is a page. They are not four difficulties bolted onto a familiar process: each one follows from a property of the chemistry you have just met, and knowing which property is what makes them learnable rather than a list of fussy rules.

The drop counting follows from the metal’s price. The humidity sensitivity follows from the fact that the photoproduct has to travel before anything happens, so how much water the paper holds decides how far it gets. The developer’s temperature is a colour control and not a speed control, because the reaction stops when the iron(II) runs out rather than when the bath tires. And the three-bath clearing follows from the one thing this process has that no silver process does: an image that will outlast its own residue, so the residue has to go.

What this part asks of you that the last four did not

in the order you meet them

1. Every drop is counted, and a drop is not a unitA whole 8 × 10 sheet takes about two millilitres of sensitiser, mixed for that sheet and no other. The published literature counts drops — and Photographers' Formulary assume 20 drops to the millilitre while Bostick and Sullivan state that 10 mL of their palladium solution is 255 drops, which is 25.5. A 25 per cent disagreement between two suppliers of the same process is why the first thing you do in this part is calibrate your own dropper
2. The coating is humidity sensitive, and you cannot see itWare's own figures put palladium's relative speed at 0.5 at 32 per cent relative humidity and 2.5 at 80 — a factor of five from a variable a hygrometer reads and an eye does not. Below about 30 per cent the print can come out with its shadows reversed. Humidity is a reagent in this part rather than a condition
3. The developer may be worked warm, and it reduces nothingA concentrated oxalate solution, sometimes warmed, in which the image is complete in seconds. The trade worked it at 60 to 77 °C before 1892 and Willis's own announcement of cold development names the problem that solved — workers scalding their fingers. This course caps the bath at the 38 °C the modern makers publish and says why
4. The clearing is three baths and it decides everythingNot an afterthought and not one bath. Chelation alone cannot detach all the iron from the cellulose, so the sequence is chelate, reduce, chelate again at a higher pH. Accelerated ageing shows badly cleared prints looking perfectly acceptable on the day and staining later in proportion to the iron left in them

What the process gives

A long tonal scale with a shallow slope. Ware’s own densitometry gives palladium an exposure scale of about 2.4 with a mid-tone slope of about 0.78, against platinum’s 1.9 and 0.96, and attributes palladium’s extra length to a substantial non-linear toe. Read as a photographer, that says the process spends its gradation on the light tones.

A maximum density of about 1.45, which is well under a glossy silver gelatin print’s and is the half of the claim that never gets made. This process sells separation in the middle and high values, not impact.

An image with no binder at all. The AIC put it in one line: the image layer is embedded within the top fibre structure of the support. Electron microscopy reports the metal as single ellipsoidal nanoparticles 15 to 25 nm across, distributed through the body of the surface fibres. There is nothing on the sheet, so there is no gloss, nothing to delaminate, and the paper’s own surface is part of the picture.

And the most chemically inert image substance in this course. Platinum and palladium do not fade, do not oxidise and do not sulfide under any condition a print will meet. Two qualifications belong with that, and both are on the record: the paper is what fails — Ware calls acidic embrittlement of the cellulose the chief besetting problem historic platinotypes present to a conservator — and Ware also notes that palladium is not as resistant to chemical attack as platinum and may therefore be less permanent archivally, especially if a print is chemically treated later. The cost and permanence lesson takes both apart properly.

Where this part sits in the cluster

Five parts, five second reagents, one photochemical step. Set them side by side and the shape of the decision this part asks you to make is visible before you spend anything.

	Image substance	Sensitiser’s iron salt	Exposure scale	What decides permanence
Cyanotype, Part XXI	Prussian blue, a pigment	Ammonium iron(III) citrate	about 0.9 for the classic formulation; about 2.2 for the New	Alkali. The pigment is destroyed by a buffered mount
Salted paper, Part XXII	Printed-out silver, gold-toned	none — a silver process	long, and matched to the same negative as the New Cyanotype	Residual thiosulfate, sulfiding, and whether it was gold-toned
Albumen, Part XXIII	Printed-out silver in an albumen layer	none — a silver process	long	The same, plus the binder’s own yellowing
Van Dyke and kallitype, Part XXIV	Iron-reduced silver	Ferric oxalate or ferric ammonium citrate	long	Residual iron and the silver’s own vulnerability to sulfur
Palladium, this part	Palladium metal	Ferric oxalate	about 2.4	Residual iron, and then the paper. Nothing attacks the image

Read the last column downwards and the argument of this part appears. Every process above the last line has an image substance that can be attacked: a pigment that alkali decolourises, or silver that sulfur converts. This one does not. What is left when you remove that failure mode is the paper — and the paper is a failure mode this process makes worse, because its own processing leaves the sheet acidic unless the clearing sequence is chosen to leave it alkaline.

And read the exposure-scale column and the practical problem appears. The classic cyanotype at about 0.9 and palladium at about 2.4 differ by more than a stop and a half in what the negative must carry. One negative cannot suit both, which is the difficulty the closing assignment is built around rather than the difficulty it avoids.

What it costs, before you spend anything

This part puts the arithmetic first because the alternative is discovering it after buying the bottle.

🧮What one print takes, and what the course can and cannot tell you

Ware’s standard palladium solution is 0.5 M in Na₂PdCl₄. It is mixed with an equal volume of iron solution, which is stoicheiometry rather than convenience — the reduction consumes two iron(II) per palladium(II), and equal volumes at his strengths supply about 2.1. So the sensitiser reaching the paper is 0.25 M.

His specific coating volume for this material is 24 to 36 cm³/m², with a guide of about 1.5 cm³ for a 10 × 8 inch print coated by glass rod. Brush coating uses more; he calls it expensively wasteful.

1.5 mL × 0.25 mmol/mL × 106.4 mg/mmol = 40 mg

Palladium applied to one 8 × 10 print, by rod

His own independent figure agrees: coating weights of 6 to 9 millimoles of metal per square metre, on about 0.055 m², is 35 to 53 mg. Two routes to the same answer.

What the course cannot give you is the money. src/data/prices.json holds no price for any platinum or palladium salt, none for ferric oxalate, potassium oxalate or EDTA, and none for unbuffered cotton paper in the sizes this part prints at. All five are gaps the file names, and the entry for the palladium salt carries this part’s own coating figure inside it: searched again on 7 September 2026, the only UK routes found were a kit page that would not resolve and industrial suppliers who quote rather than list. This part names the most consequential missing figure in the course, gives you the arithmetic complete in every other term, and asks you to finish it yourself with a dated listing in front of you.

The inversion to carry from Part XXI. In cyanotype the paper costs more than the chemistry — about 96p a sheet against about sixpence of sensitiser. Here the metal is comparable with the paper or larger, and the whole discipline of the part follows from that one ratio.

The accessibility route

This is part of the design and not an apology for it. A student who cannot afford the metal is not excluded from this part, and the substitutions are specific rather than gestural.

Three routes through this part

1. 1The full route — palladium, at 8 × 10About 40 mg of palladium a sheet by rod coating. Eight pages, two labs, one experiment, one assignment
2. 2The small route — palladium, at 4 × 5About a quarter of the area is about a quarter of the metal, and Bostick and Sullivan's own remark is that small negatives make beautiful jewel-like prints. Every measurement in this part works at that size, provided you say what size you worked at
3. 3The substitute route — a palladium-toned kallitypeThe same iron chemistry with silver in the metal's place, then plated with palladium from a bath Bostick and Sullivan build from 7 to 15 drops of palladium solution in a litre of 1 per cent citric acid. Since 10 mL of that solution is 255 drops, a whole litre of toner uses less palladium than one coated sheet. What you keep is the surface, the scale, the craft and a noble metal on the image; what you give up is an image made of palladium, and an equal-permanence claim the course cannot support

Six of the eight pages need no palladium at all. The chemistry lesson, the cost lesson, the Level D lesson and this overview are reading; the experiment can be run in full on a kallitype, where humidity, developer, oxidiser dose and coating weight are the same four variables on the same iron photochemistry; and the closing assignment is completable with one row filled from the course’s published data and marked as such. Two pages — the two labs — are the ones that need the metal in some form, and even there the coating technique is rehearsable with a cyanotype sensitiser at pennies a sheet and the entire wet sequence of the second lab runs unchanged on a kallitype.

What you actually lose is stated on the pages that lose it, rather than glossed here: you will not have measured this process’s own exposure scale, you will not have seen what a second coating does to maximum density, and you will not own an object made of an inert noble metal.

The map

Eight pages, and the order is the reasoning: the chemistry before the coating, the coating before the processing, the processing before the measurement, the measurement before the money — and the comparison last, because it needs every process the cluster has taught.

Page	What it settles	Level
Noble metal chemistry and the platinotype	The two salts, the reduction and its potentials, why the developer dissolves rather than reduces, print-out, what sets the colour, the permanence question and the history	—
Lab: preparing the sensitiser and coating for palladium	Two millilitres mixed for one sheet, a calibrated drop, a paper chosen on chemistry, a coating at a measured humidity	B
Lab: exposing, developing and clearing a palladium print	A measured dose, a developer that reduces nothing, and the three baths that decide whether the print survives	B
Experiment: tonal scale, contrast and image colour	Eight conditions on three sheets, and which of metal load, humidity, developer and oxidiser is worth using as a control	B
Cost, permanence and when a platinum print earns its price	The metal per print, the permanence claim taken apart into four, and the sentences you may honestly write on the back	—
The additives we study and do not reproduce	Mercury, lead, uranium, thallium and a dichromate inside the processes this cluster prints, and what the course does instead	D
Assignment: the process comparison atlas	One negative in seven processes, eight fields, and the measurement the published literature does not contain	B

Behind them sits a formulary that is already ahead of the lessons. The three-solution drop system, the 1886 Pizzighelli and Hübl formulations, Ware’s print-out platino-palladiotype, Willis’s potassium oxalate developer, the sodium citrate and ammonium citrate developers, the historic clearing sequence and the modern EDTA and sulfite one, the two bought kits, and the Level D entries for the mercuric sepia platinotype and the lead additions. No quantity is repeated in the lesson pages. A lesson explains why a formulation is shaped the way it is and sends you to the formulary for what to weigh.

What you have already done, and what changes

You have coated a sheet by rod, dried it, contact-printed it under ultraviolet, developed it and cleared it of iron. That is the kallitype of Part XXIV, and every one of those operations reappears here unchanged. Nothing in this part’s craft is new.

Four things change, and none of them is a technique.

The cost of a mistake changes by an order of magnitude. A spoiled kallitype costs a sheet of paper and some silver; a spoiled palladium sheet costs about 40 mg of palladium as well. That single fact is why the coating lab makes you rehearse with a cheap sensitiser before the metal comes out of the bottle, why the experiment is designed on three sheets rather than eight, and why every page in this part counts what was left unused in the shot glass.

A variable you have been ignoring becomes the dominant one. Relative humidity has been a footnote since Part XXI. Here it changes the speed by a factor of five across Ware’s own measured range, moves the image colour from sepia to warm black, and — below about 30 per cent — can reverse the shadows of a print you have already paid for. The hygrometer stops being laboratory furniture and becomes an instrument you read and record.

A word you thought you understood changes meaning. You have used developers that reduce since Part IV, and one that strips ligands in Part XXI. The bath in this part dissolves, and reduces nothing. Three baths called developers doing three different things is the distinction the closing assignment tests directly.

And a step you have been treating as housekeeping becomes the one that decides the outcome. Clearing. Accelerated ageing shows prints with raised residual iron looking entirely acceptable on the evening they were made and staining later in proportion to the iron in them. The permanence of your print is decided by a difference you cannot see at the time you decide it, which is a genuinely uncomfortable thing to learn and is the most transferable lesson in the part.

Safety framing

The part’s level, 4, is difficulty. The safety letters are assigned page by page from the classification rubric and are not the same scale.

⚠Four things govern every practical page here

Platinum is taught and never handled, and the reason is one hazard statement. Potassium tetrachloroplatinate(II) carries H334 — may cause allergy or asthma symptoms or breathing difficulties if inhaled — in 99.5 per cent of 212 notified reports, and HSE’s EH40 sets halogeno-platinum compounds at 0.002 mg/m³ as platinum with the Sen notation, against 5 mg/m³ for platinum metal. A limit 2,500 times tighter for the compound than for the element is the clearest possible statement that the hazard belongs to the compound. HSE’s own paragraphs add that once the airways are hyper-responsive, further exposure to tiny quantities may provoke symptoms, and that it is impossible to identify in advance who will become hyper-responsive. Ware records that platinum allergy was first identified in 1911 as an occupational disease of photographic factory workers handling platinotype paper — the people who coated it. The course ruled on 4 September 2026 that students perform palladium and never handle platinum, and the Level D policy carries the argument.

Palladium is permitted because its record is different, not because it is benign. Sodium tetrachloropalladate(II) is Danger with H302, H317, H318 and both aquatic statements — and no block of its notified record gives respiratory sensitisation. That absence is the whole difference: a glove is a real control against a skin sensitiser, and no glove is a control against an asthmagen. EH40 lists no palladium compound at all, and its own introduction says absence from the list does not indicate a substance is without risk.

No noble-metal salt is ever weighed as a powder in this course. Solutions are bought ready made. The operation sets the level, not the substance, and weighing a fine salt is the dust-generating step — the one Ware attaches a mask warning to on the palladium side and the one HSE’s control regime is built around on the platinum side.

The oxalates are systemic toxins and the developer is a concentrated one. About 280 g of soluble oxalate in every litre in the tray. The notified classification is Warning with H302, H312, H315 and H319; CAMEO goes further, recording systemic poisoning from inhalation of the dust of the potassium salt, and the course handles the substance against that description. EH40 gives oxalic acid 1 mg/m³ long-term. Gloves, goggles, tongs, nothing eaten at the bench — and oxalate never meets silver, because silver oxalate is explosive when dry and this cluster’s other benches are silver benches.

And the period’s contrast agents are not on any bench here. Potassium chlorate is Level D on the availability of a better route; a dichromate is excluded by the chromium policy at any level anywhere; mercury and lead are Level D. The part’s own Level D lesson teaches all of them as chemistry and gives no procedure for any.

Two levels appear in this part. The two labs, the experiment and the assignment are Level B, on concentrated solutions of a skin sensitiser, a concentrated soluble oxalate, an ultraviolet source and, in one declared raised step, a warmed developer tray. The additives lesson is Level D, which is the level at which this course gives no procedure to anybody.

The history, in one paragraph, because the economics repeat

William Willis began in 1872 with the wrong platinum salt, turned to the platinous salts in 1873, made the first silver-free platinum print on 17 March 1878, and founded the Platinotype Company that year. The first commercial papers went on sale in Britain in 1879 at a shilling for a demy sheet. In 1892 he announced a paper that developed at room temperature rather than at the temperature that had been scalding printers’ fingers — and refused to publish how, so that the manufacture of the finest platinum papers ever made is a lost secret. Then in 1902 Ostwald discovered that platinum catalyses the oxidation of ammonia to nitric acid, nitric acid makes explosives, and photography lost its claim on the metal: Germany banned its photographic use in 1901, Britain embargoed it in 1916 for anything but munitions, and the price peaked in 1920 at £40 a troy ounce. Willis’s answer, in 1917, was palladium — a metal that had been proposed for photography as early as 1856 and passed over because it was then dearer than platinum. Commercial manufacture ended in Britain in 1941, and what revived the process was not a paper but a kit, which is why the hazard that used to belong to a factory now belongs to whoever opens the bottle. The chemistry lesson has the whole of it with the evidence attached.

What you produce

A palladium print, made from a sensitiser you mixed for that sheet, coated at a humidity you measured, cleared through three baths and checked at a masked margin under a bluish light.

A measured curve for the process as you run it — exposure scale, maximum density and mid-tone slope from your own step wedge, with your instrument’s precision stated — and a family of curves around it showing which of four controls actually moves it.

A completed permanence argument, in which every sentence you are prepared to write about the print is marked as measured, published or inferred.

And the finished process comparison atlas: one negative printed in every process you can reach, measured the same way each time, eight fields per row. The course’s own printing processes matrix says in its scope that nobody has made that comparison and that no column in it is a substitute for the one this part’s assignment owns. That is the thing to keep.

What you need first

Part XXIV’s iron-silver chemistry, which is this reaction with silver in the metal’s place, and printing a kallitype, which is where the coating and clearing craft is learned at a tenth of the cost. If you have not coated an even sheet there, do not practise here.

Part XXI’s negatives and papers lesson, for the unbuffered-paper argument, the exposure-scale-against-density-range problem and the standard negative the closing assignment depends on. This part will not re-teach any of it.

Part XIII’s sensitometry and Part XV’s densitometer, because the experiment measures an exposure scale and a maximum density and teaches neither from scratch.

Part XVI’s ultraviolet chain — the source, the unit and the contact frame glazed in glass.

And Part II’s laboratory discipline, because the first thing this part asks you to do is measure the volume of a drop and report it with its spread.

Alternative route

No ultraviolet unit. Sunlight, for everything in this part, on the terms Part XXI’s alternative route sets out: you lose repeatability, you gain the sharpest contact print available because the sun is nearly a point source, and what you must supply instead of a dose is a written record. For the experiment and the assignment, expose the whole comparison in one session under an unchanging sky so that the strips remain comparable with each other even though the absolute dose is unknown, and say so in the analysis.

No mains. Only two things in this part use it: the exposure unit, for which sunlight substitutes, and the warmed developer, which is optional — the default of both labs is room temperature, which is what Ware’s own tests used.

No darkroom. None is needed. Every page of this part is worked under dim tungsten or in a shaded room; what you have escaped, as in Part XXI, is the dark rather than the sun. The only page that wants a darkroom is the closing assignment’s silver gelatin row, and that row can be filled from Part XIII’s published data and marked as published.

No palladium. The accessibility section above is the answer, and each practical page carries its own version of it in detail.

And one thing has no substitute. There is no version of this part that makes a palladium print without palladium, and no cheaper metal behaves like it. The substitutions give you the chemistry, the craft, the measurement and the record; the metal is the metal, and the course would rather say that than pretend otherwise.

The pages of this part

Part25 of 28Level4 — SpecialistPages7Estimated time17.0 hoursHighest safety levelLevel D

✱3 of 7 pages in this part need a darkroom, a UV source or mains-powered equipment, marked below. Each says what can be improvised and, where one exists, gives an alternative route.

0 / 7 lessons in this part completed

• LessonNoble Metal Chemistry and the Platinotype60 min
• LabLab: Preparing the Sensitiser and Coating for PalladiumB180 min
• LabLab: Exposing, Developing and Clearing a Palladium PrintB✱ UV · Mains210 min
• ExperimentExperiment: Tonal Scale, Contrast and Image Colour in PalladiumB✱ UV · Mains180 min
• LessonCost, Permanence and When a Platinum Print Earns Its Price45 min
• LessonThe Additives We Study and Do Not ReproduceD45 min
• AssignmentAssignment: The Process Comparison AtlasB✱ Darkroom · UV · Mains300 min