Noble Metal Chemistry and the Platinotype

Every developer you have used since Part IV was a reducing agent. It found silver that light had already begun to reduce and finished the job, multiplying a speck of a few atoms into a whole developed grain. The bath a platinum print goes into contains no reducing agent at all. It is a concentrated solution of a salt you could put in a rust remover, and what it does to the sheet is dissolve something.

That single fact is the key to the whole part. Get it and the rest follows: why the image appears in seconds rather than minutes, why development time is not a contrast control, why the developer slowly fills with metal, why a variant of the process needs no developer at all provided the paper is damp enough, and why the two metals in the part’s title behave differently enough that the course prints one of them and only reads about the other.

The two salts, named properly

A platinum or palladium sensitiser has exactly two active components and they do completely different jobs.

The light-sensitive one is an iron(III) salt of an organic acid, which is the siderotype pattern Part XXI established and Part XXIV repeated with silver. In the traditional process it is ferric oxalate, and Ware — a chartered chemist writing about a substance he has used for forty years — calls it “a chemists’ nightmare”: ill-characterised, polymorphic, apparently uncrystallisable, sold in states of hydration that different sources give as the tetrahydrate, pentahydrate and hexahydrate, with formula weights running from 375.76 to 483.84. Twenty grams of it is a weight of an ill-defined solid rather than a known quantity of iron, and that uncertainty runs through every published sensitiser strength in this part.

The image-forming one is a noble-metal chloro-complex, and it is not light-sensitive. Coat a sheet with the metal salt alone and it will sit in the dark or in the sun and do nothing. The names matter here because the literature uses several for each substance and they are not interchangeable:

Modern name	Formula	Older names you will meet	Formula weight
Potassium tetrachloroplatinate(II)	K₂PtCl₄	potassium chloroplatinite, platinous chloride of potash	415.11
Sodium tetrachloropalladate(II)	Na₂PdCl₄	sodium chloropalladite, palladium #3	294.19 anhydrous
Potassium tetrachloropalladate(II)	K₂PdCl₄	potassium chloropalladite	326.4
Palladium(II) chloride	PdCl₂	palladium dichloride	177.31

Three traps sit in that table. The -ite ending of the period names marks the metal(II) oxidation state: chloroplatinite is platinum(II) and chloroplatinate is platinum(IV), and the difference is the whole of Willis’s first three years of failure. The cation is not always a spectator — a gram of the potassium palladium salt carries about a tenth less palladium than a gram of the sodium salt, and Ware’s print-out process deliberately uses the ammonium salt because ammonium ions are part of its mechanism. And the hydrate is frequently unstated: the sodium salt is usually sold as roughly the trihydrate, which differs from the anhydrous form by about a sixth of its palladium content, and the kit sheets give a percentage without saying which.

⚗Ware's two ways to the same bottle

The palladium solution the course’s practical pages use is 0.5 molar in Na₂PdCl₄, and Ware publishes two routes to it that he presents as equivalent, choosing between them on price and availability.

Either dissolve 7.35 g of the anhydrous double salt, or 8.7 g of the trihydrate, and make up to 50 cc. Or dissolve 2.92 g of sodium chloride in 30 cc of water, heat to about 70 °C, add 4.43 g of powdered palladium(II) chloride in small portions with stirring until the dark orange-brown solution is clear, and make up to 50 cc.

The second route is chemistry rather than measuring: two extra chloride ions per palladium convert the sparingly soluble dichloride into the soluble tetrachloropalladate(II) anion. It is also the cheaper route, and it is the one that carries a dust hazard, because it starts by weighing a powder. Both are in the formulary with the arithmetic.

The reduction, and what is actually established about it

Light does one thing to the coating: it reduces iron(III) to iron(II) and destroys an oxalate ligand as carbon dioxide. Ware writes the solid-state case as one equation.

UV + Fe₂(C₂O₄)₃ → 2 FeC₂O₄ + 2 CO₂

Ware's equation for the photolysis of solid iron(III) oxalate

The iron(II) then hands electrons to the metal, two per metal atom because each metal(II) needs two and each iron supplies one.

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

The reduction of palladium(II) by the iron(II) oxalato complex — the reaction the course's prints are made of

2 [Fe(C₂O₄)₂]²⁻ + [PtCl₄]²⁻ → 2 [Fe(C₂O₄)₂]⁻ + Pt + 4 Cl⁻

The same reaction with platinum, which the course teaches and does not run

Why it goes is a matter of redox potentials, and Ware’s Table 11.1 is worth having in front of you, because it explains as much by what it excludes as by what it permits.

Couple	Potential	What the course does with it
Fe(III)/Fe(II) oxalato	+0.02 V	The reductant this part is built on
Fe(III)/Fe(II) citrato	+0.372 V	The cyanotype sensitiser’s iron — too weak for platinum or palladium
[PdCl₄]²⁻ / Pd	+0.62 V	Reducible by the oxalato couple
[PtCl₄]²⁻ / Pt	+0.73 V	Reducible by the oxalato couple
Ag⁺ / Ag	+0.80 V	The kallitype and Van Dyke of Part XXIV
[AuCl₄]⁻ / Au	+1.00 V	The chrysotype, studied in the atlas only
Fe(III)/Fe(II) EDTA	−0.12 V	Not a sensitiser — the clearing bath, and a stronger reductant than either

Read the first two rows together and a fact from Part XXI falls out with a mechanism attached. A cyanotype sensitiser makes Prussian blue and, with an added metal salt, gold or silver — but it will never make platinum, because +0.372 V is not enough. The change from citrate to oxalate is what makes the noble metals reachable, and it is the reason this part exists as a separate part rather than as a variant of Part XXIV.

⚗Deeper: what the equations above simplify, and what nobody has established

Three things in the account above are the course’s simplification of what the source actually says, and the difference is worth knowing before you quote any of it.

The complex the electron actually reaches is aquated, not the one written. Ware’s account has the reduction proceed not through [MCl₄]²⁻ but through the aquatrichloro anion in which one chloride has been displaced by water:

H₂O + [PdCl₄]²⁻ ⇌ [Pd(H₂O)Cl₃]⁻ + Cl⁻

The aquation equilibrium that governs both metals, and separates them

At 20 °C the equilibrium constant is 0.17 for palladium and 0.015 for platinum. At sensitiser strength that puts more than half the palladium in the reactive form and fewer than a fifth of the platinum. This one number explains most of the difference between the two metals, and it is why the course could teach palladium and would struggle to teach platinum even if the hazard allowed.

The electron-transfer pathway is a proposal. Ware states plainly that the pathway “has not been elucidated”, argues from the lability of iron complexes that an inner-sphere mechanism is unlikely, and offers an outer-sphere route through a hydrogen-bond bridge with hydrogen-atom transfer in one direction and proton transfer in the other, giving a transient platinum(I) species. He writes the overall result as a stoicheiometry rather than as a mechanism, and this page does the same.

The number of iron ions is a design constraint, not a coincidence. Ware’s rule for his own sensitiser is that the volume of iron solution must equal the combined volume of metal solution, and at 1.40 M against 0.67 M that supplies about 2.1 iron(III) ions per metal(II) ion — just over the two the reduction consumes. Mixing “equal volumes” is not a convenience; it is the stoicheiometry written as a bench instruction.

Why the classical process needs a developer at all

This is the mechanism students most often get wrong, and getting it right changes what you expect from every step of the two labs that follow.

In a dry ferric oxalate coating the photoproduct is ferrous oxalate, and ferrous oxalate is almost insoluble — Ware gives 0.022 g per 100 cc of water. So at the end of the exposure the sheet holds a perfect distribution of reducing agent that cannot move, sitting beside a metal salt that also cannot move. The Getty atlas describes what you actually see at that moment as “a faint brownish image made from the photochemically generated ferrous oxalate”: a picture of the iron, not a picture in metal.

The developer’s job is transport. Saturated potassium oxalate floods the sheet with free oxalate anions, which take the immobile iron(II) into solution as a mobile bisoxalatoferrate(II) complex.

FeC₂O₄ + C₂O₄²⁻ → [Fe(C₂O₄)₂]²⁻

The whole of what a platinum developer does: it dissolves the photoproduct

Now the iron(II) can reach the palladium sitting a few micrometres away, and the reduction that light had already paid for finishes itself in the liquid film at the image site. That is physical development in the strict sense — the image substance is deposited from solution rather than made in place — and it is worth naming, because the phrase gets used loosely elsewhere.

Three acts, and the developer is the second

1. 1Ultraviolet reduces iron(III) to iron(II) and takes an oxalate off as carbon dioxideThe only photochemical step in the process. The palladium salt beside it is untouched by light and stays as palladium(II). What you can see at the end of this step is a faint brown image of ferrous oxalate, which is not the picture
2. 2The developer dissolves the insoluble photoproductConcentrated oxalate ions convert immobile ferrous oxalate into a mobile bisoxalatoferrate(II) anion. Nothing is reduced by the bath: it contains no developing agent, and this is the step every other developer in the course does not have
3. 3The mobile iron(II) reduces palladium(II) to metal at the image siteTwo iron(II) per palladium atom, in solution, in seconds. The iron returns to iron(III) and stays in the sheet as the residue the clearing baths exist to remove
4. 4Three clearing baths take the iron out, and a wash takes the baths outThe step that decides whether the print survives. Residual iron is the failure mode of this whole family, and it is invisible on the day

The pattern is Ware's; the fourth step is where Part XXV puts most of its practical attention, for reasons the permanence section below sets out.

Four consequences follow from that middle step and every one of them is testable at the tray.

The image appears in seconds. Both surviving kit suppliers say development is complete within seconds; Bostick and Sullivan instruct that the developer be poured over the print fast enough to break air bubbles, because an area the liquid reaches a second late develops differently. Wall’s 1924 handling instruction for hand-coated paper is a one-pass answer to the same problem: take both ends of the sheet, immerse one end face down, draw the paper right through and turn it face up.

Development time is not a contrast control. The reaction stops when the iron(II) runs out, not when the developer tires, which is why the modern sheets can say “you cannot overdevelop a platinum print” and Wall could write in 1912 that “with correct exposure, over-development is impossible”. The AIC’s conservators put the general case: contrast in this family is achieved mostly by exposure rather than by the developer.

The developer accumulates iron and metal, and does not exhaust in the ordinary way. Nothing in the bath is consumed by developing; what changes it is what the prints put in. Bostick and Sullivan describe a bath with “an indefinite life”, replenished for evaporation, its colour darkening as dissolved metal accumulates, with the sludge filtered out periodically through a coffee filter. Ware adds the warning that follows from the same fact: a re-used platinotype developer always contains sensitiser reagents in solution and is therefore still photosensitive, and if it is not protected from actinic light it will steadily precipitate particulate platinum metal.

And the developer is a colour control, which is the one thing it genuinely is. Hotter is warmer; the potassium oxalate entry records the historic bath worked at 60 to 77 °C before 1892 and the modern kits at 32 to 38 °C, and the sodium citrate entry records Wall’s opposite instruction for palladium black papers, 7 to 16 °C, the coldest working temperature in the whole formulary.

Print-out, where humidity replaces the developer

Change the iron salt and the developer becomes unnecessary. That is the single most important variation in this part and the reason it needs its own name.

Ammonium iron(III) oxalate supplies the trisoxalatoferrate(III) anion, and its photoproduct is soluble. If the paper holds enough water, the iron(II) can migrate a short distance without any bath at all and reduce the metal during the exposure. Ware’s arithmetic for “enough water” is precise and surprising: paper conditioned at about 70 per cent relative humidity carries about 8 per cent water by weight, which works out at roughly ten water molecules to each trisoxalatoferrate(III) ion in the amorphous regions of the cellulose — enough for short-range ion diffusion, while the sheet still rattles and is to all appearances dry.

Three practical consequences follow, and they change the way the work feels rather than just the chemistry.

Print-out is self-masking. Darkening shadows inhibit their own further darkening, so a longer negative density range is accommodated by exposing longer rather than by changing the chemistry. The contrast lever moves from the bottle to the lamp.

You judge the print by looking at it. A hinged-back frame and an eye replace the test strip, which is a different craft from the develop-out route where the sheet looks like almost nothing until it goes in the tray.

And the failure mode changes. Ware reports the print-out route as apparently immune to the “plague of black spots” that sometimes afflicts ferric oxalate development. He offers it as an observation without a mechanism, and this page does the same.

🕰Pizzighelli 1887, and what the Ziatype is

Giuseppe Pizzighelli, by then working alone in Bosnia, discovered the print-out route in 1887 using sodium ferric oxalate rather than ferric oxalate, and reported testing the potassium and ammonium salts too: the sodium salt gave the most satisfactory result, the potassium salt less sensitivity, and the ammonium salt greater sensitivity but less brilliant images with a tendency to a cold bluish tone. Note that this is five years after the 1882 drop-counting publication he and Hübl are better known for, and it is a different discovery.

Commercial “direct printing platinotype” papers followed and did not last, which Ware attributes tentatively to limited storage life and the need to control humidity in use — the same two problems the modern practice solves with a desiccant tin and a humidifying box.

The Ziatype is the living commercial descendant, sold by Bostick and Sullivan and built around a lithium palladium solution rather than an ammonium one. Its formulary entry records what the maker publishes and what it does not. The cation choice is not arbitrary and it is a nice piece of physical chemistry: lithium chloride dissolves exothermically because the small, highly charged lithium ion binds water tightly, ammonium chloride dissolves endothermically because the large diffuse ammonium ion binds it weakly, and a cation that holds water differently changes how much water a coating must take up before its ions become mobile. Lithium and ammonium are not interchangeable, even though both are monovalent and both leave no residue.

What actually sets the image colour

Four things, and one of them is not a bath.

The metal. Platinum runs neutral to cool black; palladium runs warm, brown, softer and lower in contrast from the same negative. The AIC’s identification notes describe platinum as “a matte, deep, rich, yet warm, black” and palladium as “warmer, yellower brown”, and add that mixtures modulate between them — while recording that the mixture “was not commonly employed by historic practitioners”, so the Pt/Pd print of a modern portfolio is a modern object.

The humidity. This is the control most printers underestimate, and Ware’s Table 7.1 puts numbers on it for his print-out sensitisers. Palladium at 32 per cent relative humidity gives sepia; at 55 per cent Van Dyke brown; at 80 per cent a warm black. Over the same span platinum barely moves in colour and moves the other way in speed.

The developer and its temperature. Hotter and stronger both warm the image, and the Getty atlas supplies the physical reason: hotter development gives smaller metal particles and browner prints. The ammonium citrate bath is chosen by modern palladium printers specifically for the opposite effect — its formulary entry collects three independent statements, from the supplier, from Jill Enfield and from a master printer Ware quotes, all saying it gives cooler and more neutral tones than potassium oxalate, and notes that none of the three is a measurement.

And any additive. Mercury(II) in a platinum sensitiser makes an image “resembling that of palladium” — lower in contrast, brown and very smooth — which is how the sepia platinotype was made before palladium was available. That chemistry is named here and treated properly, without a procedure, in the Level D lesson at the end of this part.

🔬Why smaller particles are browner, and why palladium reverses

The colour of a nanoparticle metal image is a size effect, not a chemical one. Ware invokes the von Weimarn coefficient: a faster precipitation gives smaller particles. Palladium’s complexes are more labile, so its reduction is faster, so its particles are smaller, so its images are warmer — one causal chain from an equilibrium constant to a hue.

The same speed has a cost. Ware attributes tonal reversal — the partial “solarization” seen in some palladium prints, where heavy shadow exposures come back lighter — to nanoparticles formed too rapidly in the shadows, and his own tests name five conditions that promote it: a coating dried below about 30 per cent relative humidity, an oxalate rather than a citrate developer, a paper of lower absorptivity, a contrasty negative allowing heavy shadow exposure, and a degree of partial print-out. Platinum images are not known to suffer it. The reversal disappears entirely in the print-out process, and it is the clearest evidence in the part that humidity is a reagent here rather than a condition.

What the image is, physically

Not a layer. There is no gelatin, no baryta, no supercoat and no binder of any kind: the sensitiser soaks into the cellulose and the metal is precipitated among and inside the surface fibres. The AIC states it in one line — the image layer is embedded within the top fibre structure of the support — and the electron microscopy Ware reports puts a size on it: single ellipsoidal platinum nanoparticles, 15 to 25 nm across, distributed through the body of the surface fibres rather than lying on them.

A palladium print, in section, beside the two structures you already know

Drawn to show the arrangement, not the depth. The particle size is Ware's electron microscopy of a platinum print; the paper weights are his own recommendation for this process. Layer depths are drawn to be readable, not to scale: on real film the base is roughly a hundred times the emulsion, and drawn honestly the emulsion would vanish. Any thickness given in the labels is the real one.

Three properties follow directly from that arrangement, and they are what people are buying when they buy a platinum print.

A matte surface that is the paper’s own. There is no gloss because there is nothing on top to be glossy. The paper’s texture is part of the picture rather than a substrate under it.

An image that appears to be in the paper rather than on it, because it is. The AIC records the odd corollary: sometimes the image material darkens the support enough to be visible from the verso.

And a long, smooth tonal scale. Ware’s densitometry of his own print-out sensitisers 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, which is the “delicacy in the high values” the process is admired for. The maximum density in most of those tests was about 1.45, which is well below a glossy silver gelatin print’s, and that pairing is the honest statement of what the process offers: separation in the middle values, not impact.

Where the noble-metal scale sits, drawn to teach

• Palladium, exposure scale about 2.4
• Platinum, exposure scale about 1.9

Show the numbers behind this plot

Two curves on the same axes. The palladium curve rises from base density through a long gently curved toe, straightens through the middle values at a shallow slope, and reaches a maximum density of about 1.45 after roughly 2.4 log units of exposure. The platinum curve rises more steeply from a shorter toe, reaches a similar maximum density, and does so within about 1.9 log units. The palladium curve therefore sits below and to the right of the platinum one through the high values, which is the extra tonal delicacy the process is valued for, and reaches the same maximum. Both curves are drawn to show the published shape and were not measured by the course.

Series	log relative exposure	Reflection density
Palladium, exposure scale about 2.4	0.00	0.06
Palladium, exposure scale about 2.4	0.30	0.09
Palladium, exposure scale about 2.4	0.60	0.16
Palladium, exposure scale about 2.4	0.90	0.29
Palladium, exposure scale about 2.4	1.20	0.48
Palladium, exposure scale about 2.4	1.50	0.72
Palladium, exposure scale about 2.4	1.80	0.98
Palladium, exposure scale about 2.4	2.10	1.24
Palladium, exposure scale about 2.4	2.40	1.42
Palladium, exposure scale about 2.4	2.70	1.45
Platinum, exposure scale about 1.9	0.00	0.06
Platinum, exposure scale about 1.9	0.30	0.13
Platinum, exposure scale about 1.9	0.60	0.29
Platinum, exposure scale about 1.9	0.90	0.55
Platinum, exposure scale about 1.9	1.20	0.86
Platinum, exposure scale about 1.9	1.50	1.15
Platinum, exposure scale about 1.9	1.80	1.38
Platinum, exposure scale about 1.9	2.10	1.45
Platinum, exposure scale about 1.9	2.40	1.46
Platinum, exposure scale about 1.9	2.70	1.46

The shape is drawn to teach and was not measured from a print. The exposure scales, the mid-tone slopes and the maximum density are Ware's own densitometry of his print-out sensitisers; the curve between those points is the course's drawing. Your own measured curve is the business of the experiment two pages on. The curve is drawn to show the shape, not measured from a real material. Your own materials will differ, and measuring them is what the sensitometry part of the course is for.

Permanence: the metal is not the problem

Platinum and palladium are chemically inert under any condition a print will meet. There is no equivalent of silver’s sulfiding, no oxidation, no tarnish and no fading of the image substance. Ware’s own summary of palladium prints is that they are highly lightfast, may be displayed under good lighting, and are robustly resistant to all contaminants likely to arise in a normal environment.

The paper is the problem, and Ware puts it in a sentence that is worth memorising: although the metal constituting a platinotype image is quite invulnerable, its paper substrate is certainly not, and acidic embrittlement of the cellulose is the chief besetting problem that historic platinotypes present to the conservator. The AIC’s treatment notes agree from the bench: paper supports can become brittle from the acidic processing and yellow, and cracks, breaks and broken corners are common.

⁇Does the platinum itself destroy the paper it sits in?

This is the single claim in this part most likely to be overstated, in either direction, and the evidence has a specific shape.

The mechanism proposed. Finely divided platinum — “platinum black” — is the classic heterogeneous catalyst; Ware’s own chapter 2 traces catalysis itself from Davy in 1817 through Döbereiner’s 1823 lighter to Ostwald’s 1902 oxidation of ammonia. The proposal is that the image metal catalyses the oxidation of atmospheric sulfur dioxide to sulfur trioxide, which then forms sulfuric acid that concentrates in the fibres and cuts the cellulose chains.

2 SO₂ + O₂ → 2 SO₃

The catalysed step, with platinum black as the catalyst

H₂O + SO₃ → H₂SO₄

And the irreversible one that follows, giving a non-volatile acid in the fibres

The evidence for it is one unpublished experiment from 1989, and Ware presents it as such. A sheet half-coated with platinotype sensitiser was exposed and processed to maximum density, cut in four, and two pieces were held in about one atmosphere of sulfur dioxide for three days. Pulped and measured, the controls read pH 6.3 for the white paper and 6.1 for the platinum; the gassed pieces read 4.5 for the white paper and 2.6 for the platinum. Ware’s own verdict on his figures is the honest one: they “seem to provide preliminary evidence” and “the topic deserves full investigation with control of parameters such as relative humidity.”

And he immediately argues against over-reading them. Assuming a concentration–time reciprocity he says he has no evidence for, one atmosphere for three days is roughly 100 ppm for eighty years — against outdoor sulfur dioxide of about 1 ppm, a peak of 1.34 ppm in the Great London Smog of 1952, and a fall by a factor of ten in parts of the UK since the Clean Air Acts.

Who holds what. Ware proposes the mechanism, supplies the only measurement, and marks it preliminary. The AIC records the brittleness as a fact of the objects and attributes it to the acidic processing rather than to catalysis. What the course states is that the embrittlement is real and documented, that platinum catalysis is a plausible and partially tested contributor, that the sulfur dioxide exposures in the experiment are far above anything a print has met, and that the acidity the process itself leaves in the sheet is the better-established cause. A modern print cleared by the EDTA and sulfite sequence leaves the paper alkaline rather than acid, which addresses the documented cause and not the conjectural one.

The related phenomenon is worth naming because you may meet it. Platinum prints in books and albums sometimes generate a faint brown mirror image on the facing sheet. Ware proposes the name autoplatinography, lists five non-exclusive hypotheses for it, and records that a team at the US National Archives investigated it in 2013–14 and reached mostly negative results: no analytical difference between light and dark areas of the ghost, absolutely no evidence for transfer of image metal, ghosts produced by accelerated ageing on four of five cellulose papers but not on the lignin-containing one, ghosting through an interleaving tissue implying a gas-phase route, and — the finding that unsettles the simple story — prints with no history of ghosting giving negative results under the same test. He also notes rare instances of untoned silver prints doing the same, so platinum-catalysed degradation “cannot be the sole explanation”.

The history, compressed

Fifty years elapsed between the dawn of photography and a commercially viable platinum process, and about forty more before it had practically gone. Ware frames the whole of his history around that paradox, and the answer to the first half is chemical while the answer to the second is economic.

Platinotype and palladiotype, from the first patent to the modern revival

Willis

1873 patent, platinum with silver and lead1878 silver-free; Platinotype Company founded1892 cold development, never published1913 Satista1917 Palladiotype

Published formulae

1882 Pizzighelli & Hübl, drop-counted chlorate1887 Pizzighelli print-out1895 Abney & Clark; von Hübl Der Platindruck

The market

1879–80 first commercial papers1911: fifteen kinds listed1916 embargo1920 price peak; 1937 Company wound up; 1941 British manufacture ends

Revival

1970s–80s: kits, not papers1982– Malde–Ware print-out palladium

Spans are drawn to place events in order and are not to a linear scale of years. Every date is from Ware's history except the commercial arc, which is the Getty atlas's.

Willis’s three problems. Ware names the three chemical choices the process needed and Willis could not easily get: the right platinum salt, the right iron salt and — the one people forget — the use of potassium oxalate as a developer. Willis began in 1872 with platinic chloride, platinum(IV), and by his own later account “all my early experiments were naturally made with platinic chloride”. They failed, and Ware explains why in terms this page has already given: platinum(IV) is thermodynamically reducible at +0.68 V but far too inert kinetically to make an image in the minutes available. In 1873 he turned to the little-known platinous salts and, “after a troublesome operation, made some potassic chloro-platinite”. He also admitted that his attention had been directed to potassium oxalate by “a note by a French chemist” who has never been identified, and that he had to prepare the salt himself because he could not obtain it in London.

The patents, and one date correction. The first patent is 5 June 1873, for a process still carrying silver, gold or lead. On 17 March 1878 he made the first silver-free platinum print, annotated and witnessed on the verso, and patented it later that year. The Platinotype Company followed. Ware notes that all the histories give 1879 for its founding, and that 1878 is on the company’s own headed notepaper and embossed on its paper tins. The first commercial papers went on sale in Britain in 1879 at a shilling for a 17¾ × 22¾ inch demy sheet.

1892 is the year the process was perfected and the year the knowledge was lost. Willis announced before the Camera Club that he could make fine prints by development at room temperature rather than at the temperatures that “had caused so many workers previously to scald their fingers”, and then said in terms: “For reasons which will be understood I cannot make this method public.” He never patented it, no technical records of the Platinotype Company are known to have survived the London Blitz, and the manufacture of the most successful commercial platinum papers ever made is therefore a lost secret. Ware infers from Willis’s own wording that the answer lay in the sizing or coating rather than in the sensitiser — possibly an acidic alum–rosin size or a clay filler that slowed the dissolution of the exposed sensitiser during processing and so allowed the image to develop fully.

The trade Willis did not have. Pizzighelli and Hübl, two Austrian army captains in the photographic department of the military technical administration, published detailed recipes in 1882 — Ware notes that British writers seem to have been careful not to, probably for fear of Willis’s patents. Their publication won the Vienna Photographic Society’s Voigtländer Prize, reached English in 1883, and carried the drop-counted chlorate contrast system into every kit sold today. The 1886 formulations and the drop system itself are in the formulary, both at Level D and both with no procedure.

The economics, then and now

Platinum stopped being a photographic material because it became a strategic one, and the mechanism was the same catalysis this page has already discussed.

Ostwald discovered in 1902 that platinum catalyses the oxidation of ammonia to nitric acid. Nitric acid makes explosives. Germany banned platinum’s use in photography as early as 1901, on account of lamp filaments; by 1916 the British government had embargoed it for anything but munitions, which stopped Willis’s manufacture and export. Ware’s price history has the metal peaking in 1920 at 800 shillings — £40 — per troy ounce and falling in the Depression to 93 shillings in 1931. Willis’s answer was palladium, launched in 1917: a metal discovered by Wollaston in 1803, proposed for photography by Burnett in 1856 and by Willis himself in 1878, and passed over at the time because it was then more expensive than platinum — 6 pence a grain, about £12 an ounce, against about £2 for platinum.

🕰The commercial papers that imitated platinum with silver

Two of them are Willis’s own, and they are the reason the word “platinotype” on a mount is not a statement about the image substance.

Satista, patented in 1913, replaced most of the platinum with silver. It is the ancestor of a whole class the Getty and the AIC call “faux platinum” papers — silver papers platinum-toned to imitate the look, of which Aristo Platino is the example Ware illustrates. They are a conservation problem in their own right: Ware’s chapter records that autoplatinography has been seen from a platinum-toned Aristo Platino print as well as from true platinotypes.

The other imitation is Nicol’s kallitype of 1889, and it is the honest one, because it never claimed to be platinum. Nicol’s stated motive, in a letter Ware quotes, was “a silver printing process, which in point of simplicity and artistic effect would rival the expensive platinotype”. Part XXIV teaches it, and the cost lesson argues whether a palladium-toned kallitype is a substitute for what this part makes.

And the economics have not gone away. Ware’s own summary of contemporary practice is that it is very much easier and less expensive to print in palladium than in platinum, that very few practitioners today even attempt pure platinum, and that most use palladium alone or a mixture in which palladium predominates. He adds a chemical sting to that: because palladium prints out about 2.5 times faster than platinum in the middle tones, a sensitiser mixed 1:1 does not give a 1:1 image. X-ray fluorescence on a gelatin-sized paper found the deposited ratio starting near Pt:Pd = 1:3 at low exposures and levelling near 3:4 at maximum; on a non-gelatin paper, about 1:2 in the middle tones reaching 1:1 only at maximum density. His conclusion is a recommendation nobody wants: to get equal molar amounts of the two metals in the image, coat with a sensitiser at Pt:Pd = 2:1 or more. Many prints sold as platinum-palladium must therefore consist chiefly of palladium.

What you should expect this process to do

Stated as testable claims, because the experiment two pages on exists to measure them against your own prints rather than to admire them.

A long exposure scale. Ware puts the unmodified traditional platinotype sensitiser at about 2.0 and his print-out palladium at 2.4, against the classic cyanotype’s 0.9 that Part XXI had you measure. The suppliers converge from the other side: Bostick and Sullivan ask for a negative of density range 1.35 to 1.50, which is not the same claim and is discussed at length in the negatives lesson.

A modest maximum density. About 1.45 in Ware’s tests, well under a glossy silver gelatin print. Anyone expecting a deeper black from an expensive process has misread what it sells.

A shallow mid-tone slope, about 0.78 for palladium against 0.96 for platinum, which is the numerical form of “lower contrast and greater delicacy in the high values”.

A speed that depends on humidity, in opposite directions for the two metals. Ware’s Table 7.1 has palladium’s relative speed rising from 0.5 at 32 per cent relative humidity to 2.5 at 80 per cent, while platinum’s falls from 1.8 to 1.0 across the same span. That is why he recommends a 3:1 platinum-palladium mixture to a printer who does not want to control humidity: the two effects nearly cancel.

And a speed that is not, in absolute terms, as hopeless as the rest of the family. The Image Permanence Institute’s atlas records that platinum was about three times faster than silver chloride printing-out papers and could be printed in diffused light — which is a comparison against another contact-printing process rather than against a camera film, and is the right comparison to make.

⚠Two metals, two hazard records, and the ruling that follows

The course ruled on 4 September 2026 that students perform palladium and never handle platinum, and the reason is in the two classifications rather than in any judgement about difficulty.

Platinum. The notified GHS classification for potassium tetrachloroplatinate(II) carries H334 — may cause allergy or asthma symptoms or breathing difficulties if inhaled — in 99.5 per cent of 212 reports. 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, and the tightest figure anywhere in this course. Ware records that platinum allergy was first identified in 1911 as an occupational disease of photographic factory workers handling platinotype paper, and his own instruction is blunt — never touch the surface of platinum-sensitised paper, never put ungloved fingers in the processing solutions, dust inhalation is the worst hazard, and do not store large amounts of dried sensitised paper.

Palladium. Sodium tetrachloropalladate(II) is classified Danger with H290, H302, H317, H318 and both aquatic statements, aggregated from 176 reports across 10 notifications. No block of that record gives respiratory sensitisation, and that absence is the whole of the difference: skin sensitisation is controllable by a glove, and an asthmagen is not controllable by anything a page can assume.

Where the two sources appear to disagree. Ware writes that if you develop the platinum allergy, “platinum printing is not for you — but you could still use palladium, which is non-allergenic”. That is a clinical statement about occupational asthma from the platinum salts, and it is compatible with the notified H317 on the palladium salt, which is a different endpoint. The course states both and takes the more cautious reading on skin: gloves are not optional on either metal.

And EH40 lists no palladium compound at all, whose introduction says in terms that absence from the list does not indicate a substance is without risk. Palladium being permitted here is a judgement about controls, not a claim of an absent hazard.

Recap

A platinum or palladium print is made by two reactions in sequence, and only the first involves light. Ultraviolet reduces iron(III) to iron(II) in an oxalate coating; the iron(II) then reduces a platinum(II) or palladium(II) chloro-complex to metal, two iron per metal atom. In the traditional process the photoproduct is insoluble, so a concentrated oxalate developer must dissolve it before it can reach the metal — the developer reduces nothing, and that is why the image builds in seconds, why development time is not a contrast control, and why the bath fills with iron and metal instead of exhausting. Change the iron salt for ammonium iron(III) oxalate and the photoproduct is soluble, so a damp sheet completes the reduction under the lamp and needs no developer at all.

Palladium is the faster, warmer, softer, cheaper and — decisively — the less hazardous of the two metals, because more of it is aquated at sensitiser strength and because its notified classification carries no respiratory sensitisation. The image is metal among bare paper fibres, matte, long in scale, modest in maximum density and chemically inert. What can fail is the paper.

Check your understanding

Question 1. Why does a traditional platinum print need a developer while a Ziatype does not, in terms of where the iron(II) has to go?

• The Ziatype sensitiser contains its own reducing agent, so no external one is needed
• In dry ferric oxalate the photoproduct is insoluble ferrous oxalate and cannot reach the metal salt, so an oxalate bath must dissolve it; in the print-out sensitiser the photoproduct is a soluble complex, so water held in the paper lets it migrate during the exposure
• The Ziatype is exposed for far longer, so the reaction has time to complete without help
• Potassium oxalate reduces the palladium directly, and the Ziatype uses light to do the same job

Show the answer and why

Answer: In dry ferric oxalate the photoproduct is insoluble ferrous oxalate and cannot reach the metal salt, so an oxalate bath must dissolve it; in the print-out sensitiser the photoproduct is a soluble complex, so water held in the paper lets it migrate during the exposure

The variable is the solubility of the iron(II) photoproduct, not the presence of a reductant. Ferric oxalate photolyses to ferrous oxalate, which Ware gives as 0.022 g per 100 cc — effectively immobile in a dry sheet. Ammonium iron(III) oxalate photolyses to a soluble bisoxalatoferrate(II) anion, and at about 70 per cent relative humidity the paper holds roughly eight per cent water by weight, or about ten water molecules per sensitiser ion, which is enough for short-range diffusion. Note what option four gets wrong: no platinum developer reduces anything. It dissolves.

Question 2. You mix a sensitiser three parts palladium to one part platinum by volume of the standard solutions. Predict the image colour, and say what else has to be held constant for the prediction to mean anything.

• Neutral black, because platinum dominates the colour even in small proportions; nothing else matters
• Warm brown, and the prediction is meaningless unless the relative humidity of the coating at exposure is stated, because humidity moves palladium from sepia at 32 per cent to warm black at 80 per cent
• Exactly three quarters of the way from platinum’s colour to palladium’s, because the metals mix linearly
• Unpredictable, because the ratio of the two metals in the sensitiser has no relationship to the image

Show the answer and why

Answer: Warm brown, and the prediction is meaningless unless the relative humidity of the coating at exposure is stated, because humidity moves palladium from sepia at 32 per cent to warm black at 80 per cent

Palladium-dominant means warm, but the honest answer is the qualification. Ware's Table 7.1 gives palladium sepia at 32 per cent relative humidity, Van Dyke brown at 55 and warm black at 80: humidity moves the colour further than a change of metal ratio would. Option three is wrong for a second, independent reason worth knowing — palladium prints out about 2.5 times faster than platinum in the middle tones, so the image ratio is not the sensitiser ratio, and Ware recommends coating at Pt:Pd of 2:1 or more to obtain equal molar amounts in the print.

Question 3. Ware's standard palladium solution is 0.5 molar in Na₂PdCl₄ (formula weight 294.19 anhydrous) and is mixed with an equal volume of iron solution. His specific coating volume is about 25 cm³ of mixed sensitiser per square metre of paper. Roughly what mass of palladium metal is deposited in a fully reduced square decimetre?

• About 0.13 mg
• About 1.3 mg
• About 13 mg
• About 130 mg

Show the answer and why

Answer: About 13 mg

Work it in three steps. The mixed sensitiser is 0.25 M in palladium, because equal volumes halve the 0.5 M stock. A square decimetre is a hundredth of a square metre, so it carries 0.25 cm³ of sensitiser, which is 0.25 mL × 0.25 mmol/mL = 0.0625 mmol of palladium. At 106.4 g/mol that is about 6.6 mg — and if you coat closer to Ware's upper figure of 36 cm³/m², about 9.6 mg. The answer is therefore the 13 mg band rather than the one above or below it, and Ware's own independent statement of coating weights, 6 to 9 millimoles of metal per square metre, agrees: 6 to 9 mmol/m² is 0.06 to 0.09 mmol/dm², or 6.4 to 9.6 mg of palladium. Note that this is the metal applied; complete reduction is an assumption, and unreacted salt washes out.

Question 4. A dealer describes a print as a platinum print made in 1925, and offers as evidence that platinum is permanent and the print is in perfect condition. Which part of that argument is weakest?

• Nothing; platinum images do not deteriorate, so condition is good evidence of the medium
• The date; commercial platinum papers were unavailable in 1925
• The inference from condition to composition: prints of that period may be palladium, a platinum-palladium mixture, or a platinum-toned silver "faux platinum" paper, and the permanence claim has been transferred between metals without anyone checking which is in the sheet
• The permanence claim itself, since palladium prints fade in ordinary display conditions

Show the answer and why

Answer: The inference from condition to composition: prints of that period may be palladium, a platinum-palladium mixture, or a platinum-toned silver "faux platinum" paper, and the permanence claim has been transferred between metals without anyone checking which is in the sheet

Condition is evidence about an object, not about its composition, and this family is where that distinction bites hardest. The Getty records that only after 1917 did photographers begin to use palladium and then platinum-palladium mixtures on paper they coated themselves; Ware shows that a 1:1 sensitiser deposits far more palladium than platinum; and Willis's own Satista of 1913 and the platinum-toned silver papers imitated the look with silver. Option two is wrong on the facts — supplies had returned by 1920, at prices that had photographers asking whether the process could survive. What settles the question is X-ray fluorescence, and what an honest label says is which metal made the image.

Question 5. Why does an iron(III) citrate sensitiser make Prussian blue, silver and gold, but never platinum?

• Citrate ions form an insoluble compound with platinum(II) that cannot be reduced
• The citrato iron(III)/iron(II) couple sits at +0.372 V, which is below the tetrachloroplatinate couple at +0.73 V and the tetrachloropalladate at +0.62 V, but above the couples for silver and gold; the oxalato couple at +0.02 V is a strong enough reductant for all four
• Citrate sensitisers are not sensitive enough to ultraviolet to produce useful quantities of iron(II)
• Platinum requires a developer and citrate processes are all print-out, so the mechanism cannot complete

Show the answer and why

Answer: The citrato iron(III)/iron(II) couple sits at +0.372 V, which is below the tetrachloroplatinate couple at +0.73 V and the tetrachloropalladate at +0.62 V, but above the couples for silver and gold; the oxalato couple at +0.02 V is a strong enough reductant for all four

It is a thermodynamic exclusion and it is why Part XXV is a separate part rather than a variant of Part XXI. Ware's Table 11.1 lays the couples side by side, and the citrato iron couple simply does not reach the two noble metals of this part while it comfortably reaches silver at +0.80 V and gold at +1.00 V. Option four inverts the argument: whether a process prints out depends on the solubility of the photoproduct, which is a separate variable, and there are develop-out and print-out versions of both chemistries.

Check my answersClear