Glycerol

Every other liquid in this encyclopaedia earns its place by what it does to silver, or to iron, or to the pH of a bath. Glycerol does none of those things. It is not a developing agent, not a silver salt, not an alkali and not a restrainer in any chemical sense, and it takes no part in the formation of an image. What it does is hold water in a coating that would otherwise dry out, and thicken a solution so that things stop moving through it — and those two physical properties are enough to decide the colour of an argyrotype and to have made local development of a platinum print possible at all. It is the clearest case in the course of a substance whose photographic effect is entirely mechanical.

In photography

It is the humectant of Mike Ware’s argyrotype, and therefore its colour control. Ware’s sensitiser carries 1 cc of glycerol in 100 cc, added after the iron salt has dissolved and before the solution is made up to volume, and he marks it optional. His own note explains what the option buys: the colour of the silver print is determined by the size of the metal particles, that size is affected by the humidity of the coating during the printing-out exposure, and higher humidity is promoted by the presence of a hygroscopic substance — glycerol, “included in the sensitizer formulation as a humectant to ensure a pleasing purplish-brown colour”. Without it, he writes, the image may be a more yellowish-brown. The chain runs molecule to picture without a gap: a hygroscopic solute keeps water in the dried coating, the water lets ions move during the exposure, ion mobility governs how the silver particles nucleate and grow, and particle size governs the colour. The argyrotype page works that chain through in full; this page is about the first link.

The same trick, and the same substance, runs through the print-out iron processes generally. Ware lists the concentration of humectants in the sensitiser among the parameters that push a palladium print towards neutral grey, naming glycerol alongside ammonium nitrate and dimethyl sulfoxide as hygroscopic substances whose presence “may promote an increased amount of water to be absorbed by the cellulose fibres”. Pizzighelli’s nineteenth-century variations on the print-out platinum sensitiser contained glycerol for the same reason, and Ware’s gloss on them is worth keeping — he describes the humectant and the gum arabic in those formulas as undesirable additional variables that only complicate testing the process. That is a fair summary of what a humectant is: a way of decoupling a coating from the room it is drying in, bought at the price of one more thing that can be wrong.

⚗Why a non-volatile solute holds water, and why it is the same fact as antifreeze

A humectant is not a sponge and it does not attract water in any special way. It lowers the vapour pressure of the water it is dissolved in, and that is a property of solutions in general rather than of this molecule in particular.

OpenStax’s Chemistry 2e states the rule and, as it happens, uses this very substance as its worked example of a non-volatile solute: dissolving a non-volatile substance in a volatile liquid lowers that liquid’s vapour pressure, because solute present at the surface reduces the area available to solvent molecules and so the rate of vaporisation, while the rate of condensation is unaffected. Equilibrium is reached with fewer solvent molecules in the vapour. Glycerol qualifies as non-volatile by a wide margin: Chemical Safety Card 0624 gives its vapour pressure as 0.01 Pa at 25 °C and states that evaporation at 20 °C is negligible. It stays in the coating; the water is what comes and goes.

So a sheet coated with a sensitiser containing glycerol, left to equilibrate with a room at a given relative humidity, settles at a higher water content than the same sheet without it. The room has not changed. What has changed is the water content at which the coating stops losing water to the room.

The freezing points of glycerol solutions are the same fact seen from the other side, and the numbers are on the record: a 10 per cent solution freezes at −1.6 °C, a 30 per cent at −9.5 °C, a 50 per cent at −23.0 °C and a 66.7 per cent at −46.5 °C. That is the bottom of the curve rather than the end of it: the same record gives 80 per cent at −20.3 °C and 90 per cent at −1.6 °C, so beyond about two-thirds glycerol the mixture behaves increasingly like glycerol with a little water in it and the freezing point climbs again. Chemistry 2e draws the two effects together on one phase diagram — the vapour-pressure lowering and the freezing-point depression are the same displacement of the same curves. Water that will not freeze out of a glycerol solution at −20 °C is water that will not evaporate out of one in a dry room either.

In platinum printing it did the opposite job, by being thick rather than by being wet. From the early 1890s until the platinotype papers disappeared, a platinum print could be developed with a brush instead of a tray, and glycerol is what made that controllable. Ware’s account is precise about the mechanism and about the misnomer: glycerine “somewhat belies its naming as a ‘developer’”, because the presence of the viscous, inert, but water-miscible liquid actually inhibits and slows the development process, allowing time for a selective local application of developing agents by brush. Elsewhere he puts it as a diffusion problem — the premature washing out of undeveloped sensitiser may be inhibited by increasing the viscosity of the developer with an inert water-miscible liquid such as glycerol, to diminish the rates of diffusion.

The numbers behind that are startling. PubChem’s viscosity entry gives 954 centipoise at 25 °C for the pure liquid and 17 centipoise at 25 °C for a 70 per cent solution. Diluting by three parts in ten drops the viscosity by a factor of about fifty-six — the course’s arithmetic on the record’s two figures — and that steepness is exactly why the period instructions work by mixing glycerine and developer in graded cups rather than by timing anything.

🕰What the worker actually did, from the 1900 paper and from Wall

Keiley and Stieglitz’s Camera Notes article of April 1900 sets out the procedure. The paper is printed somewhat deeper than for ordinary development, aiming at the delicate sky tones, because the foreground can be held back later. It is laid face up on a heavy sheet of glass that has itself been glycerined — which stops it sliding — coated evenly with pure glycerine by a broad brush, blotted so that the glycerine is worked into the sensitised surface, then thinly recoated, and coated heavily wherever development is to be restrained. Two working solutions stand ready: equal parts developer and glycerine, and pure developer. The print is built up with brushes, blotting and re-glycerining any part that has gone far enough. Parts coated with glycerine will not develop, which is the whole of the control. One print takes fifteen to fifty minutes, often longer. Clearing is one part muriatic acid to sixty of water, three fresh baths, then five minutes’ wash.

Wall’s 1912 dictionary gives the same method with four cups and a costing — “new brushes, varying in size from a fine spotting one to a 2-inch flat dusting-brush, a pound of glycerine, four cups, a plentiful supply of good fine clean blotting paper, and a sheet of glass, form the whole outfit” — and his 1924 formulary gives the four cups as quantities: pure glycerine; glycerine 10 parts to saturated potassium oxalate 1 part; glycerine 10 parts to saturated oxalate 5 parts; and saturated oxalate alone. Wall’s summary of what the glycerine is for is a single clause and it is the right one: it “merely prevents any lines of demarcation showing”.

The 1900 paper also gives mercury(II) chloride solutions for warm brown and flesh tones, with its own warnings that the mercury “is a very uncertain quantity and rarely acts in the same way twice”. The course teaches that substance as history and gives no home procedure for it; the glycerine is not the problem in that formula and cannot make it safe.

It is a plasticiser wherever gelatin has to stay flexible. The Getty Conservation Institute’s carbon atlas records that preparing canvas, silk or linen to receive a carbon print took three or four coats of a solution of gelatin, glycerin, sugar and chrome alum, sanded smooth between coats. Wall’s 1924 baryta paper is the same idea under an emulsion — gelatine 100 g, barium sulfate 30 g, glycerine 5 ccm and water 940 ccm, the baryta rubbed to a cream with the glycerine before it goes in, and chrome alum added drop by drop afterwards. A dried gelatin layer with no plasticiser in it is brittle; the glycerol holds enough water in the layer to keep it from cracking when the support flexes.

And the carbon tissue itself carries the course’s most quotable statement of when a humectant is wanted. Wall’s instruction for the tissue is sugar candy as standard, “in very hot dry climates a little glycerine”, and then, flatly, “the latter addition is not advisable under ordinary circumstances”. Ware says the same of Pizzighelli’s platinum variants — a humectant for use in very dry environments. Two authorities a century apart, on two unrelated processes, reach the identical rule: a humectant is a correction for a dry room, not a default ingredient. The reason is the one Ware’s argyrotype sheet gives from the other end — a coating carrying too much water may not lie flat and can damage a silver-gelatin negative pressed against it in the frame.

Last, it is a solvent chosen because it will not evaporate. Wall’s 1924 formulary gives dye solutions for liquid safelight filters and then adds that the water in them is apt to evaporate, “and this trouble may be overcome by using glycerine instead”; for a filter standing in the heat of a lamp he says the same again, that the evaporation can be obviated by using glycerine as the solvent instead of water. Lowy’s gelatine safelight coating in the same book carries 50 ccm of glycerine in a litre of 8 per cent gelatine, where it is doing the plasticiser’s job in a layer that must not craze. The same non-volatility explains the use Wall’s 1912 dictionary records for the developing bench: glycerine as a preservative of pyro. Edwards’s Pyro-glycerine stock, printed in the 1924 formulary, dissolves 40 g of pyrogallol in 40 ccm of glycerine and 250 ccm of methylated spirit, with a second solution of bromide, ammonia and a further 40 ccm of glycerine, the two mixed with thirty parts of water at the moment of use. The glycerine is not protecting the pyro chemically; it is keeping a concentrated stock a syrup rather than a crust, and keeping air out of it by being viscous.

⚗The one place glycerol does take part in the chemistry

Everything above depends on glycerol being unreactive. It is not unreactive; it is slow. Put a strong enough oxidant and a catalyst in the same place and the three alcohol groups become three places to take electrons from.

Ware records a gold toning of platinotypes due to A. W. Dollond, in which a coating of glycerol is laid over the print and a solution of gold(III) chloride brushed onto it. The platinum image catalyses the reaction, glycerol is oxidised to dihydroxyacetone or glyceraldehyde, and the gold(III) is reduced to nanoparticle gold, which gives the image a bluish cast:

3 C₃H₈O₃ + 2 [AuCl₄]⁻ → 3 C₃H₆O₃ + 6 H⁺ + 2 Au + 8 Cl⁻

Dollond's gold toning: platinum catalyses the oxidation of glycerol, and gold(III) is the oxidant

Three glycerol molecules give up two hydrogens each; the six electrons are exactly what two gold(III) ions need to reach the metal. The Getty’s platinotype atlas describes the same process, first used about 1894, spelling the name Dolland, and adds a step the equation does not show — the toned print was afterwards treated in an alkaline developer to complete the reduction of the gold chloride.

Read that reaction next to the incompatibility list in this page’s Handling section and they are the same chemistry at two temperatures. A noble-metal salt on a catalytic surface oxidises glycerol quietly enough to tone a print. Potassium permanganate or a chlorate ground together with it does the same thing all at once.

Properties

Identity. Propane-1,2,3-triol, C₃H₈O₃, relative molecular mass 92.09, CAS 56-81-5, PubChem CID 753, EC 200-289-5. Three carbons, each carrying one hydroxyl group. PubChem’s computed properties count three hydrogen-bond donors and three hydrogen-bond acceptors on a molecule of six heavy atoms, with a topological polar surface area of 60.7 Ų and an XLogP3 of −1.8; the experimental partition coefficient agrees at log P_ow −1.76. There is no more compact way to say what this substance is. It is almost entirely hydroxyl, which is why it mixes with water in all proportions, why it will not go into petroleum ether, and why it holds onto water so tenaciously when a coating dries.

Physical constants, from the ILO-WHO card unless another source is named. Boiling point 290 °C, with decomposition. Melting point 18 °C. Relative density 1.26. Vapour pressure 0.01 Pa at 25 °C. Relative vapour density 3.2. Flash point 176 °C closed cup, auto-ignition 393 °C, explosive limits 2.6 to 11.3 vol% in air. NIOSH classes it a Class IIIB combustible liquid, and although a flash point of 176 °C is far above anything a darkroom reaches, the word “combustible” belongs on the shelf label. Viscosity, from PubChem, is 954 cP at 25 °C for the pure liquid and 17 cP for a 70 per cent solution. Refractive index 1.4746 at 20 °C. Heat of combustion −4310 cal/g.

It has no acid-base behaviour a photographic bath will notice. PubChem gives the pH as “neutral to litmus” and the dissociation constant as pK_a 14.4. An alcohol that weak is un-ionised across the entire pH range this course works in, from a 2 per cent acetic stop bath to a carbonate developer, so glycerol neither buffers a solution nor moves its pH. Adding 1 cc of it to 100 cc of an argyrotype sensitiser held at about pH 3.5 changes the acidity by nothing at all, and that is a designed property: the humectant had to be something that would not interfere with a formulation whose contrast is controlled by acid.

And it does not oxidise on standing. PubChem’s decomposition entry states that pure glycerin is not prone to oxidation by the atmosphere under ordinary conditions, but decomposes on heating with the evolution of toxic acrolein. That first clause is why a bottle keeps and why a stock solution carrying it keeps; the second is in the Handling section, where it belongs.

C₃H₈O₃ → C₃H₄O + 2 H₂O

On heating: glycerol dehydrates to acrolein, the corrosive fume the safety card warns about

!Three names one letter apart, and only one of them is this

This is the substance in the encyclopaedia most likely to be confused by name, and the confusions are not subtle in chemistry — only in typography.

• Glycerol, also spelled glycerin and glycerine, CAS 56-81-5, is this page: a triol, a humectant, no photographic chemistry of its own.
• Glycin, CAS 122-87-2, is a developing agent — the encyclopaedia files it as 2-(4-hydroxyanilino)acetic acid, and it is also written N-(4-hydroxyphenyl)glycine. In a period manual set in a small serif face, “glycin” and “glycerin” differ by two letters, and both appear in developer formulas.
• Glycerol trinitrate, nitroglycerine, CAS 55-63-0, is what glycerol becomes in a mixture of nitric and sulfuric acid. It is a different substance with a different record — HSE’s EH40 lists it separately from glycerol, with a short-term limit and an Sk skin-absorption notation that the glycerol entry does not carry. It is a hazard mentioned on this page and never a synonym for it.

The register carries glycerin and glycerine as aliases of this entry, and deliberately does not carry glycine, so that a page naming it fails and somebody decides which of the three was meant. Buy and verify on 56-81-5, not on a word.

Handling

What the hazard record says, and what it does not. The aggregated GHS entry is Not Classified, from 5,617 reports across 16 notifications to the ECHA C&L Inventory, with 98.3 per cent of those reports saying the substance does not meet GHS criteria. Ninety-five reports, in thirteen notifications, do lodge hazard statement codes; the aggregation does not show which, because none of them is held by more than 10 per cent of the companies that classify. And the substance’s own ECHA CHEM record carries an empty index number, so there is no harmonised classification — nothing here is agreed across the European Union, and everything here is a notifier’s own view. The ILO-WHO card leaves its rows for effects of short-term exposure and effects of long-term or repeated exposure blank, listing only dry skin from skin contact and diarrhoea from ingestion.

That is an unusually quiet record, and the course states it as a record rather than as a verdict. Three hazards survive it, and every one is about what glycerol is put next to rather than about touching it.

☠Glycerol and the darkroom's oxidisers

This is the reason a bottle of syrup gets a shelf position rather than a corner. The ILO-WHO card’s whole storage instruction is four words — separated from strong oxidants — and PubChem states that an explosion may occur if glycerol is triturated with chromium trioxide, potassium chlorate or potassium permanganate. Ground together, not merely mixed in water.

The record explains why glycerol is worse in this respect than its molecular weight suggests, and the explanation is worth reading twice: the violent or explosive reactions with solid oxidants are due to glycerol having three centres of reactivity, to its being a liquid, which ensures good contact with the solid, and to its high boiling point and viscosity, which prevent the heat of the reaction from dissipating. Those are the same three properties every photographic use on this page depends on. A substance that is chosen because it wets everything, stays put and does not boil away is, for exactly those reasons, the wrong thing to spill into a permanganate.

The darkroom neighbours that matter, all of which have pages here: potassium permanganate, potassium dichromate and ammonium dichromate as chromium(VI), potassium chlorate, ammonium persulfate and potassium persulfate, hydrogen peroxide, nitric acid and sulfuric acid. PubChem names the nitric-plus-sulfuric mixture as forming the explosive glyceryl nitrate, and names chlorine as exploding with glycerol in a confined mixture heated to 70 to 80 °C. Store it away from all of them; wipe a spill up completely rather than leaving a film on a bench that will later take a permanganate weighing; and do not let the spatula or the measuring cylinder used for it be the one that goes into an oxidiser next.

Heat is the second hazard, and it needs no oxidiser at all. Chemical Safety Card 0624 states that glycerol decomposes on heating and that this produces corrosive fumes of acrolein. Nothing in this course heats it — the argyrotype sensitiser has cooled before the glycerol goes in, and every other use here is at room temperature — but a bottle stored above a radiator, or a spill onto a hotplate used for dissolving gelatin, is the situation the sentence is written for. The card’s fire advice is that the substance is combustible, gives off irritating or toxic fumes in a fire, and that there should be no open flames.

The exposure limits disagree by a factor of twenty, and the course reports the spread rather than picking a winner. HSE’s EH40 Table 1 lists Glycerol, mist, CAS 56-81-5, at 10 mg/m³ as a long-term limit over an eight-hour reference period, with no ppm figure, no short-term limit and none of the Carc, Sen or Sk annotations. The ILO-WHO card gives a German MAK of 200 mg/m³ for the inhalable fraction, with peak limitation category I(2) and pregnancy risk group C. NIOSH records the OSHA permissible limit as 15 mg/m³ total and 5 mg/m³ respirable, and sets no recommended limit of its own: glycerine (mist) appears in its Appendix D among substances for which NIOSH concluded that the documentation cited by OSHA was inadequate to support the proposed 10 mg/m³ limit. So one regulator says 10, another says 200, a third says 15, and a fourth declined to say. What all four are measuring is a mist, produced by spraying or by heating, and no operation in this course produces one. EH40’s own introduction states that absence from its list does not indicate that a substance is without risk; the same caution applies to a limit that exists but describes a form of exposure the darkroom never creates.

Why Level A. Against the course rubric, glycerol as a material meets no Level B criterion. There is no classification at all, let alone one reaching the rubric’s ceiling of irritant, harmful if swallowed or corrosive; the quantities are one to a few millilitres; nothing is heated above 50 °C; there is no mains-voltage work; and the waste from the substance itself is dilute rinse water. The ordinary Level A controls — nitrile gloves, eye protection, a ventilated room, dedicated labelled containers — are what the ILO-WHO card asks for anyway.

The level belongs to the operation, not to the jar, and this substance is a clean illustration of the distinction. Glycerol is Level A. The argyrotype sensitiser it goes into contains silver oxide reacted with sulfamic acid and is assessed on its own page, with its own controls, none of which this ingredient changes in either direction. The historical platinum developer it thickened contained mercury(II) chloride, and no quantity of glycerol makes that a home procedure.

Waste

Dilute glycerol in rinse water is biodegradable organic load with no classification of its own, and on its own it would be an uninteresting waste stream. It is never on its own. In every use on this page it leaves the darkroom dissolved in something that decides the routing for it: the wash water from a sensitised argyrotype sheet carries silver and belongs with the silver stream; a spent siderotype clearing bath carries iron and citrate; a historical platinum developer carries oxalate and, in the formulas that used it, mercury. Route the stream by its worst constituent. Kodak’s J-300 guidance for amateur photographers sets out what may and may not go to a sewer system, and ILFORD’s route for domestic users in the United Kingdom is a household waste and recycling centre’s chemical cupboard, with each stream bottled, labelled with what it contains and dated. A spill of the neat liquid is collected in covered containers as far as possible and the remainder absorbed in sand or an inert absorbent, which is what Chemical Safety Card 0624 specifies, and then stored and disposed of according to local regulations. Check your local regulations; they govern.

History

It arrived in photography as a by-product of the soap trade. Wall’s 1912 dictionary describes it as “a peculiar, sweet, viscid liquid, obtained from oils and fats as a bye-product in saponification”, specific gravity 1.260, and then states its photographic character in one sentence that has not needed correcting since: “It is extremely hygroscopic, and its non-drying properties are taken advantage of in photography to prevent the too rapid drying of some substances, and it is also used as a preservative of pyro.” He adds that it “has also been suggested as a restrainer in developing, its action being probably rather physical in this respect than chemical”. That hedged final clause is the finding of this whole page, written in 1912. Ware reaches the same conclusion in 2017 with the vocabulary of diffusion rates and viscosity; Wall got there with “probably rather physical”.

It was a stocked photographic chemical rather than an improvisation. Kodak Limited’s 1949 Chemicals and Formulae lists Glycerine among “Some ‘Kodak’ Tested Chemicals”, in the same short column as Elon, hydroquinone, silver nitrate and Kodalk.

⁇Who invented glycerine development, and in what volume it was published

Four accounts of the origin are in front of the course, and they do not agree.

• The Getty Conservation Institute’s platinotype atlas states that the glycerin-developed platinotype process was invented by Alfred Maskell (1864–1931) in 1892.
• Ware reports that Stieglitz is credited with inventing it around 1900 in collaboration with Joseph Keiley, but that Adrienne Lundgren has pointed out that Stieglitz himself first described the use of a glycerine developer as early as 1892, as a remedy for a poor-quality negative, while discussing the newly marketed cold-bath papers.
• Ware also records that William Willis recommended the use of glycerine to the Camera Club in 1893, and his chronology carries “1893 William Willis — ‘glycerine development’ of platinotype”.
• Keiley and Stieglitz themselves, in 1900, disclaim it. Their opening paragraph states that the use of glycerine with the platinum developer “is not new”, that it had been tried at different times by various experimenters with records in the photographic year books, and that it had been regarded first as a means of obtaining clear high lights and modifying heavy shadows, then as “a valuable and useful means of mechanically restraining or retarding the development”, and most recently as an eliminator of obtrusive detail. Their claim is not for the idea but for its corrective and interpretive use — for bringing it within the power of the photographer to correct on the print the falsely rendered values of the negative.

The course does not settle it. What can be said with the documents in hand is that the technique was in circulation by 1892 or 1893 from more than one direction, that the 1900 paper is the first full published method rather than the invention, and that its own authors said so.

One detail the course can settle. The Getty’s bibliography cites the paper as Camera Notes 4: 221–26; Ware cites it as volume 3, April 1900. The volume header at the head of the number reads “Vol. III. APRIL, 1900. No. 4.”, so the article is in volume III, number 4, and the Getty’s entry has the volume and the number transposed.

What the prints look like a century later. The Getty’s atlas keeps glycerin-developed platinotypes as one of the named variants of the process and describes what they look like a century on: a sketchy character with background or distracting detail eliminated, vignetting more asymmetrical than any mechanical tool would give, and — the detail that gives the material away — a darker halo, in old prints, over the area the glycerin covered. Alice Boughton, Joseph Keiley, Alfred Stieglitz and a trade of commercial portraitists worked this way. It ended when platinum paper did.

This page was written last, and for the ordinary reason. The chemical encyclopaedia was planned by asking what each substance does to silver, and a substance that does nothing to silver is exactly what such a list misses. Glycerol entered the register on 6 September 2026 because the argyrotype page named it as an ingredient and could not link it, taking the planned count from 136 to 137 — the fifth time in two days that a written formula page, rather than the plan, told the register a substance existed.