Kodak D-23

Two ingredients. Not two ingredients plus water and a restrainer and a trace of something — two. Kodak Limited’s 1949 handbook calls it an “Elon fine-grain developer for plates and films; a soft-working developer of simple formula giving normal emulsion speed”, and the formula underneath that description is complete in three lines.

Ingredient	Quantity	Form the source specifies
Metol	7.5 g	Kodak calls it Elon
Sodium sulfite	100.0 g	anhydrous — or 200.0 g of the crystalline salt
Water to make	1000 mL

Everything interesting about this formula is in what is absent. There is no second developing agent, no added alkali, no restrainer, and no buffer pair beyond whatever the sulfite provides on its own.

Purpose

To develop a soft, fine-grained negative at full film speed, slowly, with the smallest possible number of variables. That last clause is why it belongs early in a formulary rather than as a curiosity: a developer with two ingredients has two things that can be wrong with it, which makes it the right bath to mix first and the right control against which to judge everything more complicated.

Recommended uses

Sheet film and roll film where highlights need holding back. A soft-working developer suits a contrasty subject — a bright interior with a window in it, a landscape in hard sun — because the gradient climbs slowly and the highlights do not run away while the shadows are still arriving.

As the reference bath in a teaching darkroom. It is the developer Part VIII asks you to mix first, and the one Part IX’s experiments build on, because a two-ingredient solution can be varied one ingredient at a time without ambiguity.

Where metol on its own is wanted. Any negative developed here was developed by metol alone, which makes it the natural control for any experiment about the second agent.

When another formula is preferable

• For normal contrast in a third of the time, D-76 adds hydroquinone and a weak alkali. Eighteen minutes against eight or nine is the price of simplicity here.
• For finer grain still, D-25 is this formula with 15 g/L of sodium bisulfite added, and Kodak states the cost plainly: a 50 to 100 per cent increase in exposure.
• For a fine-grain developer that does not cost time, DK-20 reaches extra-fine grain in fifteen minutes with a thiocyanate solvent and a proper alkali.
• For high contrast, nothing in this direction will do; go to D-19.

Mixing

Metol first, sulfite second, which is the same rule as every other Kodak metol formula and matters more here than anywhere else in this reference, because the metol dose is nearly four times D-76’s. Kodak’s 1949 note is explicit: Elon is readily soluble in warm water but only slightly soluble in sulfite solutions without alkali. Put 100 g of sulfite in the litre first and a substantial part of 7.5 g of metol will simply refuse to dissolve, and there is no alkali here to rescue it later.

So: warm water, well short of the final volume; the metol in and completely dissolved; then the sulfite; then cold water to 1000 mL.

!The one formula where getting the order wrong is visible

In D-76 a little undissolved metol shows up as a slightly slow bath. Here the metol is the developer, and there is no borax to bring it back into solution. If you can see undissolved powder, or a faint haze that will not clear, the bath is not D-23 and no development time will fix it. Warm it, stir it, and if it will not go, start again.

Behaviour

Slow, and Kodak says so: about 18 minutes at 18 °C, undiluted. The handbook offers no dilution and no second time, which is unusual — most of its developers carry at least a note about diluting for more or less contrast. That silence is consistent with what the formula is: there is nothing to dilute except the thing doing all the work.

Its keeping is as good as D-76’s — 24 hours in a dish, a month in a tank, six months in a full stoppered bottle and two months in a half-filled one — which tells you the 100 g of sulfite is doing its preservative job whatever else it is up to.

Its capacity is the same as D-76’s on the 1949 table in a dish and lower in a deep tank: 24 sheets of 8 × 10 per 160 fluid ounces either way, but 36 in a deep tank against D-76’s 36 with a footnote saying replenishment greatly increases it. The replenisher for this bath is D-25R, which serves both D-23 and D-25.

It was never sold as a packed powder. Kodak Limited marked the formulas available as prepared powders with an asterisk, and D-23 has none. Whatever else it is, it is a formula Kodak expected you to weigh yourself.

Image characteristics

Grain: fine, on the same mechanism as D-76 and for the same reason — 100 g of sulfite per litre is a silver solvent, not just a preservative. Speed: normal, and this is Kodak’s own claim. There is no bromide raising the threshold and no high pH to bring on fog, so the film gives what it has. Contrast: low. “Soft-working” is Kodak’s word, and it is the defining property. Tonality: a long, gently rising scale in which highlights are restrained relative to shadows. Sharpness: modest — a solvent developer at full sulfite strength etches the grain boundaries that edge effects need.

⁇Compensating, and who says so

D-23 is very widely described as a compensating developer: one whose highlight development slows as local exhaustion sets in, so that a long-scale subject fits on the film.

The mechanism is plausible from what is on this page — very low activity, an alkali with no reserve behind it, and a bath that must produce acid as it develops. But no manufacturer publication in this course’s corpus makes the claim for D-23, and the course does not make it either. Part VIII covers compensating development as a mechanism and this page notes that the widely repeated attribution is a photographic convention rather than a sourced fact. It is a good candidate for measurement rather than for repetition.

The mechanism

With no alkali in the bottle, the sulfite has to be the alkali. Sodium sulfite is the salt of a strong base and a weak acid, so its solution is alkaline on hydrolysis, and at 100 g/L there is a great deal of it:

100 g/L ÷ 126.05 g/mol = 0.79 mol/L

Sulfite

7.5 g/L ÷ 344.39 g/mol = 0.022 mol/L

Metol, as the hemisulfate

What no source in this course gives is the pH that produces. The sodium sulfite page records the same gap: no sourced figure for a working-strength solution. It is certainly alkaline and certainly far below D-76’s borax region, because metol at three and a half times the concentration takes more than twice as long to do the same job. The course states that as a comparison of two published development times, not as a pH.

The metol dose is the compensation for everything that is missing. Kodak’s 1928 primer sets out the general rule that the energy of a developer follows the quantity of alkali, and that agents of higher reduction potential need less of it. Metol sits at the top of the primer’s potential ranking, which is precisely why it is the agent that can work in a bath like this at all — a hydroquinone developer with no added alkali would do essentially nothing. Even so, metol working alone at low pH takes eighteen minutes — and that is with 0.022 mol/L of the hemisulfate salt, which on Part VIII’s convention of two developing molecules per formula unit is 0.044 mol/L of agent, very nearly four times D-76’s total agent load.

There is no restrainer and no fog problem, and the two facts are connected. Fog is a rate ratio between exposed and unexposed grains, and it is a low-pH bath’s reward for being slow. Adding bromide here would cost shadow speed and buy nothing.

The bath has no buffer pair when it is fresh, and acquires one as it works. Development produces acid; acid meets sulfite and makes bisulfite; sulfite and bisulfite are a conjugate pair. So a seasoned D-23 is buffered in a way a fresh one is not, and its pH has fallen to get there. That is also, exactly, the design of D-25 — which is D-23 with the bisulfite put in at the start.

Function of every ingredient

Metol, 7.5 g. The whole developer. It supplies every electron that reduces silver, sets the speed, sets the contrast and sets the time. At 7.5 g/L it is close to a quarter of metol’s solubility in cold water, which is why it must be dissolved warm and first. More metol speeds the bath up and raises the gradient somewhat, but it cannot substitute for alkali: past a point the limit is the pH, not the agent. Less and the eighteen minutes becomes half an hour. Because there is no hydroquinone to carry on when the metol is used up, this bath’s exhaustion curve is steeper than a metol-hydroquinone developer’s — there is no reserve behind the front line. Metol is also the ingredient that makes this a Level B page: it is a skin sensitiser, and sensitisation is permanent.

Sodium sulfite, 100.0 g anhydrous. Four jobs, one more than it does in D-76.

• Preservative. It scavenges dissolved oxygen and intercepts the oxidation products of metol, and the published keeping figures show it succeeding.
• Alkali. With nothing else in the bottle, the alkalinity of the sulfite solution is what allows metol to work at all.
• Buffer, once the bath has run. Bisulfite formed from the acid of development pairs with the remaining sulfite.
• Silver solvent. The same job it does in D-76: dissolving a little of every grain during development, which is where the fine grain comes from.

More sulfite would raise the pH slightly, dissolve more silver and give finer grain and less speed; less would lower the pH, slow an already slow bath and coarsen the grain. Unlike D-76, you cannot change the sulfite here without changing the alkalinity, because they are the same ingredient. That is the single most important thing to hold on to about this formula, and it is why every one of the course’s sulfite experiments uses D-76 rather than D-23 as its base.

Water to 1000 mL. Warm to dissolve the metol, cold to make up. The 1949 handbook’s warning about a cold bottle throwing down a precipitate applies with force here: what comes out of a chilled D-23 is metol, and it is half the formula.

Interactions

Agent and alkali collapse into one interaction, because there is only one other ingredient. In every other developer in this reference you can move the pH without moving the solvent action. Here you cannot. The sulfite concentration is simultaneously the preservative dose, the solvent dose and the alkalinity, and any change to it moves grain, speed, contrast and time together.

Agent and agent: absent. There is no superadditivity here, which makes D-23 the natural control for measuring it. Part IX’s agent comparison is built on exactly this idea.

Agent and restrainer: absent by design, and acquired by use. The bromide released by the film accumulates, and because metol is much less sensitive to bromide than hydroquinone is, a seasoned D-23 loses activity more gently than a seasoned metol-hydroquinone bath does. That is an inference from Kodak’s own statement that bromide affects hydroquinone strongly and metol far less, and it is offered as an inference.

Variants

D-25 is Kodak’s own variant: this formula plus 15 g/L of sodium bisulfite. One added acid salt, and Kodak prints both what it buys — minimum grain — and what it costs: a 50 to 100 per cent increase in exposure, and a development temperature raised to 25 °C to keep the time at eighteen minutes.

D-25R replenishes both, and its own composition is a third variation on the same two ingredients, with sodium metaborate as the alkali the working baths do not have.

Two-bath and divided versions of this formula, in which the film is developed in a sulfite-metol solution and then transferred to a plain alkali, circulate widely and are not published here: the course has no manufacturer or primary source for any of them, and under Rule 6 a formula without traceable provenance is not printed however sensible it looks.

A warning about the number. Kodak’s 1924 and 1928 primers both print a “Formula D-23” that is a motion-picture developer with hydroquinone, carbonate, bromide, citric acid and metabisulfite in it. It is not this formula, and a citation to “Kodak D-23” without a date can mean either.

Safety

Level B, on the metol alone: skin sensitisation, and 7.5 g of a fine powder per litre is the largest metol dose in this batch of formulas. Sodium sulfite is a Level A material. There is no hydroquinone here and no borate, so this is the mildest developer in the film-developer section by classification — but “mildest” and “safe” are different words, and the sensitiser is the one that matters.

The classification rubric sets what Level B assumes. In practice, for this page: weigh the metol with extraction or in an enclosure, nitrile gloves and eye protection for mixing and for the tank, and treat any sign of skin trouble as the end of your exposure to metol rather than something to work through.

What is not a hazard here. Nothing evolves a gas. Sulfite releases sulfur dioxide when it meets an acid, but there is no acid in this formula — that hazard belongs to D-25, which has bisulfite in it, and to any tray where a stop bath could be splashed into the developer. Nothing is corrosive at these concentrations and nothing is heated beyond the warm water used to dissolve the metol.

Storage

A tightly corked bottle, filled nearly full, small bottles rather than one large one. Six months full and two months half-filled at 18 to 21 °C on Kodak’s table.

Two things are specific to this formula. First, a chilled bottle is a real risk, because 7.5 g/L of metol is a substantial fraction of its cold solubility; warm and redissolve rather than decanting. Second, because there is no alkali, a stock solution that has oxidised will look less obviously wrong than an oxidised D-76 — there is less to go brown. Date the bottle and believe the date.

Incompatibilities

Acid, and here it is a gas question as well as a chemistry one. Sulfite meeting acid releases sulfur dioxide. At stop-bath dilutions in a tray this is not the hazard, but a splash of concentrated acetic acid into a litre carrying 0.79 mol/L of sulfite is a different proposition, and the incompatibilities page covers the pair.

Fixer, in either direction: thiosulfate carried back into a developer already full of silver solvent is the classic route to dichroic fog.

Oxidising agents — ferricyanide, dichromate, permanganate, persulfate — never in the same bottle or the same drain as a developing agent.

Waste

A dilute alkaline solution whose load is metol and sulfite. Metol carries aquatic-toxicity classifications, and sulfite exerts a real biochemical oxygen demand, which is the constituent a sewer code is most likely to care about at darkroom volumes. Its own labelled bottle, kept out of the fixer so the fixer can go for silver recovery. The disposal caveat governs and the general chemical waste SOP gives the procedure. Local regulation decides.

Troubleshooting

Negatives thin and flat at the published time. Expected, if you are judging by D-76 standards — this is a soft developer and eighteen minutes is a starting point for 1949 materials. Extend the time before you change the formula, and keep the strip.

Negatives thin and the bath was mixed today. Suspect undissolved metol. There is no other candidate.

Grain coarser than expected. Check the sulfite: 100 g of the anhydrous salt, not 100 g of the crystalline, which would be half the sulfite the formula calls for.

A slow drift to lower contrast over a working session. The bath has no alkali reserve, so its pH falls as it develops. That is the formula behaving as designed, and it is the reason a tank of D-23 wants replenishing with D-25R rather than topping up.

A yellow-green cast on the film base. Metol oxidation products. Usually an old bottle; check the date and the fill level.

Experiments

Add the alkali and watch the time collapse. Mix a litre of D-23 and a second litre with 2 g/L of borax added, develop matched strips at the same time, and compare. You have converted D-23 into something close to a hydroquinone-free D-76 and isolated the alkali as a single variable. This is Part IX’s activity series in its simplest form.

Test the compensating claim rather than repeating it. Expose a step wedge with more range than the film can hold, develop one strip in D-23 and one in D-76 to the same mid-tone density, and compare what happened at the top of the scale. If D-23 is compensating, its highlight steps will be closer together. Whatever you find, it is your measurement and it belongs in the laboratory report as such.

Halve the metol. One ingredient, one arrow. The prediction is a longer time at similar contrast, because you have changed the amount of agent and not the pH.

Measure the pH the course cannot cite. Calibrate a meter against buffers using the pH meter SOP, read a fresh litre, then read it again after four films. Two numbers nobody has published for you, and the second one is the more interesting.