Precipitation, Nucleation and Crystal Growth

Pour a silver nitrate solution into a bromide solution and you get silver bromide. Pour the same solution into the same bromide solution with gelatin in it, at a controlled temperature, over a controlled time, and you get a photographic material whose speed can differ by a factor of seventy depending on how long the pouring took. The chemistry is identical. What differs is the physics of how many crystals started and how large they were allowed to become, and that is decided in the first few seconds and the following hour.

Duffin sets out the whole of emulsion making as five stages, and the first two are this page:

1. the precipitation of very small silver halide crystals, which he calls emulsification;
2. the growth of those crystals to the appropriate size, which is ripening;
3. the removal of the by-products, which is washing;
4. the sensitisation of the crystals, which is digestion;
5. the preparation of the emulsion in its final form for coating.

Stages 3 and 4 belong to washing, digestion and sensitisation, and stage 5 to coating, drying and hardening.

The first seconds: supersaturation and the nucleation burst

AgNO₃ + KBr → AgBr + KNO₃

Emulsification, in molecular form: the silver halide, and a soluble nitrate that will have to be washed out later

Duffin’s account of what that reaction looks like is worth having exactly. The solutions used are around one gram molecule per litre, and because silver halide is so nearly insoluble, “very high supersaturation occurs as soon as the mixing process commences. This results in an exceedingly rapid formation of small silver-halide crystals which may be observed by the immediate clouding of the solution where mixing occurs, and the whole bulk of the mixture will soon become an opaque milky liquid, its colour depending on which halide is precipitated.”

That clouding is the nucleation burst, and it is the most consequential event in the make. A supersaturated solution is a solution holding more dissolved material than it can keep; the relief is to throw some out as solid, and the first solid has to appear as tiny clusters that then act as sites for everything else. Once enough of them exist, the supersaturation collapses, because it is being consumed as fast as it is created. The number of crystals is fixed almost immediately, and nothing later in the make will change it. Everything after that redistributes the same silver between the same crystals.

🔬How many crystals, and how few atoms decide their fate

Duffin gives an order of magnitude that makes the rest of this part comprehensible. A typical emulsion carries about 4 × 10¹⁴ silver halide grains per gram molecule of silver, and each grain holds something like 10⁹ silver ions. So one mole of silver — 108 grams — is divided among four hundred million million separate crystals, and the sensitisation lesson will show that a few dozen atoms at the surface of each one decide whether it develops.

That is why the nucleation burst matters so much. You are not making a substance. You are making a population, and the moment of birth sets its size.

Growth: why slow addition grows crystals instead of making new ones

If nucleation is a burst that ends, then silver arriving afterwards has two possible fates: it can make new crystals, or it can deposit on the ones that exist. Which one happens depends on how much supersaturation is present when it arrives. Add the silver quickly and you keep the supersaturation high, which keeps making new nuclei. Add it slowly and the supersaturation stays low — high enough to deposit on an existing surface, too low to start a new crystal from nothing — so the same silver goes into growth.

That separation of nucleation from growth is the picture usually attributed to LaMer, and this course names it without leaning on it: it has not read LaMer’s papers, nor a source applying his treatment to silver halide, so it publishes no diagram of his and quotes no threshold from him. What it can publish is the operational statement from a Tier 1 emulsion chemist, and a table of numbers.

Duffin: “In a single-jet emulsification very rapid mixing of the halide and silver solution produces a large number of small nuclei and the longer the time of the silver addition, the larger the grains will grow.” His explanation is the same one in the maker’s own terms — nuclei produced early in the emulsification, with the excess halide acting on them as a ripening agent, grow a little and then act as nuclei for the silver halide formed later to grow on; a very rapid addition instead gives a large number of similar-sized nuclei which then do not ripen so rapidly.

Duffin’s Table 4.1, which is the whole argument in seven rows

Effect of emulsification time on grain size and on speed

• Relative speed against addition time (Duffin, Table 4.1)

Show the numbers behind this plot

Relative speed rises steeply and continuously with the time taken to add the silver solution, from 12 at an addition time of about half a minute, to 52 at 4.4 minutes, 150 at 10.2 minutes, 250 at 19.5 minutes, 450 at 42.7 minutes, 630 at 54.8 minutes and 910 at 85.2 minutes. Mean grain area rises with it over the same range, from 0.14 to 2.56 square micrometres, and gamma falls from 2.5 to 0.65, so the emulsion becomes faster, coarser and softer together. A marker at the half-minute point is labelled fast addition, many small nuclei, and a marker at the 85-minute point is labelled slow addition, few large crystals. The seventy-fivefold difference in speed is produced by the addition time alone, with the same chemicals in the same quantities.

Series	Time of silver addition (minutes)	Relative speed
Relative speed against addition time (Duffin, Table 4.1)	0.52	12.00
Relative speed against addition time (Duffin, Table 4.1)	4.37	52.00
Relative speed against addition time (Duffin, Table 4.1)	10.20	150.00
Relative speed against addition time (Duffin, Table 4.1)	19.50	250.00
Relative speed against addition time (Duffin, Table 4.1)	42.67	450.00
Relative speed against addition time (Duffin, Table 4.1)	54.83	630.00
Relative speed against addition time (Duffin, Table 4.1)	85.17	910.00

Published data, read from the printed table in Duffin's Table 4.1 (page 62); the axis values are his, the plotting is this course's. The same table gives mean grain area rising from 0.14 to 2.56 µm² and gamma falling from 2.5 to 0.65 across these seven rows.

Read the table as three columns moving together. Addition time up, grain size up (0.14 to 2.56 µm² of mean grain area, an eighteenfold increase), speed up (12 to 910, seventy-fivefold), and gamma down (2.5 to 0.65). That is the fundamental trade of emulsion making and it is not a manufacturer’s compromise, it is geometry: a bigger crystal is a bigger target for a photon and holds more silver to be developed, and a population of bigger crystals covers a wider range of exposures before it saturates, which is what a lower gamma means. Part IV’s grain, speed and resolution owns that relationship; this page shows you the tap that controls it.

🧮What this means for a make you are going to do

Ross’s plain silver iodobromide recipe adds 45 ml of silver solution at one teaspoon — 5 ml — per minute, so the addition takes about nine minutes, and then holds the emulsion in the bath for ten minutes more without stirring. Baker’s Trumm bromide paper runs the silver through a funnel fitted with about 4 mm bore tubing so that it enters as a fine jet near one edge of the vessel, leaving room for stirring.

Both of those are the same control: the rate of arrival. Duffin’s table says that if you halve the addition time you are not making the same emulsion faster, you are making a different, slower, harder emulsion. Time the addition with a clock, write the time in the batch record, and if a batch surprises you, look at that number first.

Single jet and double jet

Duffin’s Figure 4.1 shows the two classical arrangements, and this course reproduces the arrangement rather than his drawing.

Single-jet and double-jet precipitation, and why only one of them is done in this course

1. 1Single jet — silver into all the halide; the halide excess falls throughout the addition, so the first and last crystals grow under different conditions
2. 2Double jet — silver and halide together into a small halide charge; the excess stays constant, so every crystal grows under the same conditions
3. 3The measurement that makes a double jet worth doing — a silver electrode against a calomel reference, reading pAg continuously; without it the two rates are matched by hope

Drawn as an arrangement, not as apparatus to be copied. This course runs the single jet in every project; the double jet is designed and costed in the fifth project and is not performed, because no published home procedure for it exists.

In the single-jet method the silver nitrate is run at a controlled rate into a stirred solution that already holds all the halide, the gelatin and everything else. In the double-jet method a silver solution and a halide solution equivalent to it are added simultaneously through two jets into stirred gelatin. Duffin’s summary of the difference is one sentence: “In broad outline, the double-jet method gives a more even grain distribution at the end of emulsification.” The reason he gives is the one that matters — with a double jet “the concentration of ripening agents remains fixed throughout, no initial large excess is present to cause the more rapid growth at the beginning of the process.”

Duffin names a third route as well, which the chloride and chlorobromide projects in this part are closer to than either of the classical two: reversed precipitation, in which either the whole of the halide is added to the silver, or, more usually, the silver is added to insufficient halide and the extra halide is then added afterwards.

!This course single-jets, and says why

Every make in Part V is single-jet, and that is a limitation rather than a preference.

No source at any tier, in or reachable from this course’s corpus, publishes a simultaneous double-jet procedure at domestic scale. Ross, the only practitioner in the corpus with tested domestic makes, adds silver by single jet at 5 ml per minute; her nearest approach to a double addition is sequential rather than simultaneous — add B with a syringe, then follow with C in another — and she says plainly that each technique yields its own characteristic results and invites experiment. That is not a double jet.

Rule 6 does not let this course assemble a procedure out of an industrial patent and a domestic recipe and print the result as though somebody had run it. So the double jet is taught here, designed on paper in the fifth project (design-an-emulsion-make-an-orthochromatic-one), and not performed. That project is where you cost the two burettes, the two matched drive rates and the silver electrode, decide what precision each would need, and write down what error a constant-rate addition with no electrode introduces — which is a better exercise than following a recipe nobody has tested.

Excess halide, complex ions, and why crystals eat each other

An emulsion is never made with silver and halide in equivalent amounts. Duffin: “Except with very special emulsions, the quantities of halide and silver are never equivalent; in the vast majority of photographic emulsions, particularly those of a negative type, excess halide is always present.”

The reason is not stoichiometric caution. It is that the excess halide is the solvent that lets the crystals grow at all.

AgBr + Br⁻ ⇌ [AgBr₂]⁻

Silver bromide dissolving in excess bromide, first step

[AgBr₂]⁻ + Br⁻ ⇌ [AgBr₃]²⁻

Second step: the complex Duffin reports as the one that matters for crystal growth

[AgBr₃]²⁻ + Br⁻ ⇌ [AgBr₄]³⁻

Third step: the most abundant complex at the excess-bromide concentrations emulsions actually use

Silver bromide is almost insoluble in water — Part III’s solubility and precipitation gives the solubility product, and Part III’s complex formation gives the machinery. Add bromide ions and a genuinely different equilibrium opens: the solid dissolves as a series of bromide complexes, and Duffin reports that the most predominant of these at the concentrations photographic emulsions use is the tetrabromoargentate, while “it has been suggested that the important complex from the point of view of crystal growth is AgBr₃²⁻”, which is also present in appreciable quantity.

So the excess bromide has put a small amount of dissolved silver into the water around the crystals, and that changes everything.

Ostwald ripening, in Duffin’s word: cannibalism

Ostwald ripening is what happens next, and Duffin’s explanation is unusually direct. Growth, he says, “normally only takes place by ‘cannibalism’ … some of the crystals grow at the expense of others produced in the same precipitation process”. Surface ions dissolve off one crystal under the action of the ripening agent, migrate, and reprecipitate on a neighbour.

Why one and not the other? Because of surface energy. “In a silver-halide system, the larger the surface area of a crystal, relative to its volume, the higher will be the energy state of that crystal.” A small crystal is nearly all surface; a large one is mostly interior. A system of many small crystals therefore sits at higher energy than the same silver in a few large ones, and — like water running downhill, which is the comparison Duffin uses — it moves that way if it can. The excess halide is what lets it: without a solvent the crystals cannot exchange material and the population is frozen as precipitated.

The three stages of a make, and what each one settles

1. 1Mixing begins — supersaturation very high; nothing solid yet
2. 2Nucleation burst — the number of crystals is fixed here, in seconds, and never changes again
3. 3Growth during the addition — silver arriving late deposits on crystals that already exist; the earliest-formed run ahead
4. 4Ripening — small crystals dissolve through the bromide complexes and their material is redeposited on large ones

The same silver in all four panels. Drawn to show the mechanism: neither the sizes nor the numbers are to scale, and a real emulsion carries about 4 × 10¹⁴ crystals per mole of silver rather than twenty.

How much excess, in numbers

Duffin gives the working ranges, expressed as the percentage excess of halide over the quantity equivalent to the silver:

• Iodobromide emulsions: 20 to 250 per cent excess bromide during precipitation and ripening. Negative emulsions commonly carry the highest amounts; X-ray emulsions the lowest.
• Chlorobromide and chloride emulsions: rarely more than about 10 per cent, and as low as 1 or 2 per cent where very little growth is wanted.

He also notes a subtlety worth carrying: some authors express the excess as a normality rather than a percentage, and since it is the concentration of excess halide ions that controls complex formation, that is arguably the more logical measure. The two agree in practice only because most emulsions are made at roughly the same concentration, around 2,000 ml per gram mole of silver.

Compare those ranges with the makes in this part, computed from the published formulas. Ross’s iodobromide plate emulsion runs about 45 mol per cent excess halide, comfortably inside Duffin’s iodobromide band. Baker’s Trumm bromide paper runs about 9.5 per cent, and Wall’s soft-working chlorobromide paper about 23 per cent. Baker’s chlorobromide transparency emulsion runs about 2.4 per cent, which matches his own remark that these emulsions use a very small excess only of soluble halide over the combining weight. The pattern is exactly Duffin’s: the plate emulsions that need growth carry a large excess, the paper and transparency emulsions that need fine grain carry almost none.

pAg, and the variable behind the variable

pAg is the way an emulsion chemist writes the concentration of free silver ion in the liquid phase.

pAg = −log₁₀[Ag⁺]

pAg

[Ag⁺] is the concentration of free silver ion in mol/L, and the logarithm is taken to base ten and negated, exactly as pH is. A high pAg means very little free silver, which means halide is in excess. A low pAg means silver in excess. It is the same variable as “how much excess halide”, read from the silver’s side instead of the halide’s, and it is the one an instrument can actually measure: Duffin’s method is a silver electrode against a standard calomel reference.

Two consequences follow, and Part IV has already met the second.

pAg decides crystal habit. The silver halides gives the industrial statement: ordinary precipitation gives octahedral (1,1,1) faces, and Kodak’s double-jet patent gets cubic-regular grains by holding pAg between 8.6 and 9.2 at pH 4.0 or below. Duffin adds the mechanism, from Moisar and Klein’s work: at low pAg — that is, low excess bromide — the (111) faces are where growth happens, so they grow themselves out of existence and leave a 100-faced crystal; at higher bromide concentration, bromide ions adsorb strongly on those faces and hinder growth there, so growth is forced onto the (100) faces and the finished crystal shows (111) faces. He adds a fact worth remembering when you make the chloride paper: in silver chloride, no (111) planes are observed even at very high chloride-ion concentration.

pAg decides ripening rate, because it is the same number as the excess-halide concentration that drives the complexes. This is why “excess halide”, “solvent”, “ripening agent” and “pAg” are four names for one control, and why an emulsion chemist reaches for a silver electrode rather than a stopwatch.

!What this course does not have

Nothing in Part V measures pAg. A silver electrode, a calomel reference and a meter able to read them are not in the Part II laboratory, and no source in this course’s corpus gives a domestic substitute. The consequence is honest and worth stating: the makes in this part control the excess halide by weighing it, and then hope. That is exactly how Wall’s and Baker’s commercial formulas worked too, which is why they specify quantities and temperatures to the gram and the degree — with no electrode, the recipe is the control.

Physical ripening: heat, time and what it costs

Physical ripening is the deliberate heat treatment after mixing, and Duffin names it carefully to keep it apart from the chemical ripening that comes after washing: “This is called ripening, or physical ripening to distinguish it from chemical ripening, or Ostwald ripening after Ostwald who first recorded the change of fine precipitates of silver iodide into coarser ones on heat treatment where a proportion of the crystals grew at the expense of the rest.”

Three knobs, and each one is in the recipes you will follow.

Temperature. Duffin gives the emulsification range as 40 to 70 °C, and states the direction plainly: “generally, increasing the emulsification temperature causes faster crystal growth and therefore gives a coarser grained emulsion”, while warning that this is not always the case and that in some situations altering the temperature changes the grain pattern considerably. Ross ripens her iodobromide make at 55 °C for 45 minutes, and drops the bath to 50 °C when an active gelatin is used.

Time. Wall’s chlorobromide formulas digest for ten minutes, thirty minutes, an hour or seventy-five minutes at temperature, depending on what the emulsion is for. Kodak’s 1928 primer puts the whole control in one sentence: ripening is by temperature and duration of heating.

Solvent. The excess halide is the default one. Ammonia is the classical alternative and this course excludes it, for reasons set out below. Duffin also names thiocyanate in place of bromide, and amines such as morpholine and N-hydroxyethylpiperidine in place of ammonia; none of those is a domestic reagent and this course gives no procedure for any of them.

What ripening buys and what it costs

Speed, and contrast. That is Duffin’s Table 4.1 again, read the other way: as ripening proceeds the crystals get larger, the emulsion gets faster, and gamma falls. Take it too far and two things go wrong. The size distribution broadens rather than narrowing, because cannibalism accelerates as the spread widens; and — the practical limit — fog rises, because a long hot hold in gelatin is also a chemical sensitisation, whether you meant one or not.

Size distribution, which one number cannot describe

An emulsion is not “0.5 micrometres”. It is a population, and two emulsions with the same mean can behave completely differently: a narrow spread gives high contrast because most crystals need much the same exposure to become developable, and a broad spread gives a long scale because different crystals come in at different exposures.

Two grain-size distributions with the same silver in them

• A narrow distribution: higher contrast
• A broad distribution: longer scale, lower contrast

Show the numbers behind this plot

Two distribution curves plotted against grain size from zero to one point six square micrometres of surface area, with the number of crystals per thousand on the vertical axis. The first curve is narrow and tall: it rises steeply from almost nothing at 0.05, peaks at about 240 crystals per thousand at a grain size near 0.30, and falls away to almost nothing by 1.0, so most of the crystals are close to one size. The second curve is broad and low: it rises gently, reaches a shallow maximum of only about 125 crystals per thousand somewhere near 0.5, and is still carrying twenty or more crystals per thousand at 1.6, so the population runs from very small crystals to crystals several times larger. Both curves enclose the same total quantity of silver. The narrow distribution gives high contrast because most crystals need much the same exposure to become developable; the broad one gives a long tonal scale because different crystals come in at different exposures. A note records that ripening in a large excess of bromide is what produces the broad kind.

Series	Grain size (µm² of surface area)	Crystals per thousand
A narrow distribution: higher contrast	0.05	10.00
A narrow distribution: higher contrast	0.15	60.00
A narrow distribution: higher contrast	0.25	190.00
A narrow distribution: higher contrast	0.30	240.00
A narrow distribution: higher contrast	0.40	150.00
A narrow distribution: higher contrast	0.55	50.00
A narrow distribution: higher contrast	0.75	14.00
A narrow distribution: higher contrast	1.00	4.00
A narrow distribution: higher contrast	1.30	1.00
A narrow distribution: higher contrast	1.60	0.00
A broad distribution: longer scale, lower contrast	0.05	30.00
A broad distribution: longer scale, lower contrast	0.15	80.00
A broad distribution: longer scale, lower contrast	0.30	115.00
A broad distribution: longer scale, lower contrast	0.50	125.00
A broad distribution: longer scale, lower contrast	0.70	110.00
A broad distribution: longer scale, lower contrast	0.90	85.00
A broad distribution: longer scale, lower contrast	1.10	60.00
A broad distribution: longer scale, lower contrast	1.30	40.00
A broad distribution: longer scale, lower contrast	1.60	22.00

Drawn to teach the shape; neither curve is measured from a material. Duffin's Figure 4.2 gives real distributions for one emulsion made at 0, 2.56 and 6.15 molar per cent iodide, and his own octahedral example, ripened in a 244 per cent excess of bromide, came out at 0.87 µm mean diameter with what he calls a fairly broad spread. 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.

Duffin’s teaching example is the anchor worth remembering, because it names the cause. His octahedral emulsion is ripened in 244 per cent excess bromide at 1.0 molar, and the result is a mean grain diameter of 0.87 µm with “a fairly broad spread”. His comment on it is the general rule: “The broad spread of grain sizes in this emulsion is characteristic of emulsions ripened in the presence of a large excess of bromide. The contrast of emulsion of this type is fairly low and this kind of grain pattern is most suitable for negative and colour emulsions when a wide range of light intensities must be recorded.”

So the excess halide that lets you ripen at all is also what broadens the population, and the double jet exists because holding it constant is the only way to ripen without broadening.

🧪Four ripening stages from one batch, without four batches

The Bureau of Standards researchers faced the same problem you will: comparing ripening times means comparing batches, and batches differ. Their solution was to draw samples from one batch at about 30, 60 and 120 minutes from the end of the mix, giving four stages of ripening — the three samples plus the remainder — with everything else identical by construction.

Copy the design, not their apparatus (they washed by centrifuge, and their formula is ammoniacal). Draw three equal samples into chilled containers at your chosen times, let the rest run on, wash and coat all four, and expose one step wedge through each. You will have a controlled ripening series and a single gelatin, a single silver charge and a single temperature behind all four points.

Their other observation is worth borrowing too: they could follow the progress of ripening by texture, the silver bromide on the centrifuge bowl wall becoming distinctly rougher with each successive sample. You have no centrifuge, but you do have the transmitted-light check below.

Reading the emulsion by transmitted light

Two different things can be read from the colour of a wet emulsion held up to a safelight, and the course keeps them apart because they are answers to different questions.

Composition. Ross’s ladder: a chloride emulsion is white; a pure bromide emulsion is pale yellow; adding iodide deepens the yellow. That is the absorption edge moving, exactly as the silver halides describes, and it tells you what you made rather than how coarse it is.

Size. Abney’s rule, from 1885, is about scattering rather than absorption: “it will be found that the emulsions which remain of an orange tint by transmitted light after boiling are those which possess most covering power. The reason of this is not far to seek, as the orange emulsion is in finer particles than the blue or violet.” An emulsion that has gone blue or violet by transmitted light has coarsened.

Both are practitioner-scale observations rather than measurements, and neither replaces a step wedge. Take them as the check you make during the make, when there is still time to change something.

Ammonia: the fastest route, and why it is not on this course

Duffin explains the historical shape of the literature in one passage, and it answers the question every reader of Wall and Baker eventually asks — why do so many of these formulas call for ammonia?

“Ammonia emulsions, particularly those using converted silver, ripen rapidly and frequently a temperature of 40 °C or 45 °C is quite high enough to give a satisfactory ripening rate. In the absence of ammonia, when the only ripening agent is excess bromide ions, then much higher temperatures are needed, frequently up to 70 °C and in earlier days even higher, and these are therefore termed neutral or boiled emulsions.”

The chemistry is a second silver complex, formed with a much better ligand than bromide:

2 AgNO₃ + 2 NH₄OH → Ag₂O + 2 NH₄NO₃ + H₂O

What the first drops of ammonia do to a silver nitrate solution: precipitate silver oxide

Ag⁺ + 2 NH₃ ⇌ [Ag(NH₃)₂]⁺

What further ammonia does: redissolve it as the diammine complex, which is 'converted' silver

An ammoniacal emulsion is one in which that conversion has been done. It ripens fast and cool, which is why the nineteenth-century plate makers reached for it and why so many published formulas — Baker’s negative emulsion, Baker’s orthochromatic emulsion, Wall’s bromide paper, the Bureau of Standards research formula, and Ross’s own two fastest dry-plate emulsions — are ammoniacal.

Part V excludes it, and the exclusion has a price this page states rather than hides. The course’s hazard framing keeps ammonia out at Level B, so every make here is a neutral or boiled emulsion, which means working hotter and slower for less speed. Ross reaches the same conclusion from the practitioner’s side and for a different reason — the stuff stinks, and with so much darkroom work now being done in shared and improvised spaces the smell of ammonia is a barrier to deeper technique — and she went looking for a plain-silver route because of it. This part follows her.

Two further notes belong with that decision. Duffin’s ammonia raises the pH of the emulsion, which makes it worse for the digestion that follows and for coating. And there is one hazard a reader should be told about explicitly even though the course gives no procedure that could produce it: a solution of silver in ammonia must never be stored, and must never be allowed to dry, because of the possibility of forming fulminating silver. That rule is set out in silver nitrate handling and it is not restated in a formula because no formula here contains it.

Iodide, and how it gets in

Iodide is a small addition with an outsized effect, and Duffin devotes a section to it.

Ag⁺ + I⁻ → AgI

Silver iodide, precipitated first because it is the least soluble of the three

Experimental work shows that when silver nitrate is added to mixed bromide and iodide ions, the silver iodide precipitates first, because it is so much less soluble. Ostwald ripening then dissolves some of those first-formed grains and redistributes their iodide — but the coarser grains of a broadly dispersed iodobromide emulsion still contain the most iodide, which suggests the grains containing most iodide were precipitated first and grow more quickly. Some crystals have been shown to carry a core with quite large concentrations of iodide, which stays there because silver iodide is so insoluble.

Iodide’s other function is to introduce defects into the silver bromide lattice, and Duffin lists the consequences: increased photographic speed, especially after chemical sensitisation, with much less tendency to fog than a low-iodide emulsion; and improved spectral sensitisation. Above about 10 per cent iodide the sensitivity falls again, which he attributes to the formation of more internal image.

How it is introduced in practice is a weighing problem before it is a chemical one. Ross puts it squarely: the original Kodak formula uses 32 g of potassium bromide to 0.8 g of potassium iodide, and her domestic version divides everything by about eight, which would make the iodide 0.1 g — a weight she says is almost impossible to weigh accurately on any scale a home darkroom is likely to have. Her answer is the percentage solution: dissolve 10 g of potassium iodide in 100 g of water, and measure 1 ml, which is 20 drops from a standard eye dropper, delivering 0.1 g. That single technique is how every milligram-scale addition in this part is made, dye and restrainer included, and it is worth practising with water before you practise with silver.

The alternative route, conversion — precipitating silver bromide and then adding iodide to convert part of it — is real chemistry and it is what happens inside the crystal anyway:

AgBr + I⁻ → AgI + Br⁻

Conversion: the less soluble halide displaces the more soluble one

No home make in this course’s corpus does it deliberately. It is named here so that you recognise it when you meet it in the literature, and so that you understand why an iodide contamination in a bromide emulsion is not a small matter.

Why chloride emulsions are slow, and stay slow

Wall divides the emulsion world at the halide, and the practical consequence for this part is that Project 1’s chloride paper is a genuinely different operation from Project 3’s iodobromide plate.

Silver chloride is the most soluble of the three halides by a wide margin — Part IV gives the solubility products — so a chloride emulsion is intrinsically the least stable against ripening and would coarsen readily if pushed. It is also intrinsically the slowest and the least sensitive spectrally, responding only to ultraviolet and the near edge of the violet. Both facts push the same way: a chloride emulsion is made with very little halide excess (Duffin’s 1 to 10 per cent, against 20 to 250 for iodobromide), given little or no ripening, and used for printing papers exposed under a lamp rather than in a camera.

=A correction to what this course previously said about Project 1

Part V’s own manifest described the first project’s chloride emulsion as “unwashed and unripened”, and the second word is wrong. The recommended formula does ripen — about thirty minutes in a hot water bath, which is the time it takes to stage the coating — and the course’s research pass corrected the description to “unwashed, with a short thermal ripening”.

The wash is genuinely absent, and that is well sourced: Wall states the division explicitly, that bromide papers “are always washed emulsions, the gaslight being usually unwashed”, and that the gaslight or development papers “are as a rule unwashed emulsions containing varying ratios of bromide and chloride of silver”. A paper base absorbs the potassium nitrate that a glass plate would let crystallise on its surface, which is why the same emulsion coated on glass would need washing and the same emulsion coated on paper does not.

The correction matters because the ripening is not incidental: it is where the emulsion gets whatever speed it has.

Crystal irregularity, and the defect this creates

Duffin’s Table 4.1 and his ripening account both assume clean growth. Two things spoil it.

Coarse grain from too little gelatin. More gelatin gives slower grain growth, and Duffin is careful to say this is not merely a viscosity effect — it is a function of the peptide structure and of the ripening restrainers the gelatin brings with it. Wall makes the same point as an instruction: two of his paper emulsions “contain a fairly high ratio of gelatine in mixing, and are then diluted down to the required bulk. This is to prevent the formation of a coarse grain.”

Pepper. Wall names a specific defect of chloride emulsions: “a characteristic coarse grain, technically known as pepper, which is distributed throughout the emulsion and is reduced to the metallic state without exposure to light. This is sometimes so fine that it cannot be seen by mere visual examination, therefore, test plates should always be examined with an eyepiece, particularly in those parts protected from light action.” Three sources give three different avoidances, which is unusual enough to be worth listing: Wall gives two — reduce the quantity of water during the mix, or add the silver nitrate dry to the chlorized gelatine and stir until dissolved — and Ross gives a third, tempering the silver solution warm immediately before precipitation, for the same stated reason. Ross separately warns that undissolved potassium bromide pellets give black slugs or sunspots on a negative, and that the bromide should be ground to a fine powder and stored airtight. The break/fix page emulsion-coating-defects owns the diagnosis; the point here is that these are precipitation faults, made in the first minutes and invisible until development.

The make as a written schedule

Everything above collapses, in practice, into a table you write before you start and fill in as you go. Duffin lists ten variables that a comparatively simple emulsion formula can be manipulated through; the ones you control at domestic scale are these.

Variable	Where it appears in a recipe	What it moves
Mode of addition	single jet, at a stated rate	size distribution; the double jet is not available here
Addition time	“5 ml per minute”, “in a fine stream”, “over 10 minutes”	number of nuclei, and therefore grain size and speed
Emulsification temperature	the water-bath figure, 40 to 70 °C	growth rate, and so coarseness
Halide excess	the weights of the salts against the silver	ripening rate, habit, and how far ripening can go
Ripening time and temperature	“45 minutes at 55 °C”	grain size, speed, gamma, and eventually fog
Gelatin quantity	how much in the make, how much after washing	growth restraint, viscosity, coating weight
Halide proportions	chloride to bromide, and the iodide percentage	speed, the shape of the scale, spectral reach
Cation of the halide	potassium, sodium or ammonium	solubility of the salt, and the pH it leaves behind
pH	rarely stated in the historical formulas	growth and, later, whether chrome alum will harden

Write those nine down for every batch, with the actual figures rather than the intended ones. A photographic emulsion cannot be debugged after the fact from the coated sheet alone: the sheet tells you what happened, and only the batch record tells you why.

How heat ripening was found out

The dry plate arrived in 1871, when Richard Leach Maddox published a gelatine emulsion, and Abney’s 1885 account calls it what it is: “a gelatine emulsion, as it is somewhat crudely called, as first made by Dr. Maddox in 1871, is in reality silver bromide, &c., emulsified in a gelatine solution, with which plates are coated.” What Maddox had was a binder that could be manufactured, stored and sold. What he did not have was speed.

Speed came from an accident of patience. Abney again: “Mr. C. Bennett first showed how extremely sensitive plates could be prepared by keeping the gelatine solution liquid at a temperature of about 90° for six or seven days” — his degrees are Fahrenheit, so about 32 °C — and Bennett’s process, published in 1878, “was the first process published which gives extreme rapidity”. Abney’s own summary of the mechanism is exactly right and predates any of the chemistry on this page: “Sensitiveness is attained by slow digestion at a low temperature instead of by boiling.”

🕰What Bennett was actually doing, and what it cost him

Read Bennett’s week through this page and two things become clear. He was performing an extremely long physical ripening, growing his crystals by cannibalism at a temperature low enough not to wreck the gelatin. And he was simultaneously performing a chemical sensitisation he could not have known about, because the sulfur impurities in his gelatin were making silver sulfide specks the whole time — Sheppard would not identify them for another forty-six years.

Abney names the cost in the same paragraph: “in certain states of the weather this long emulsification was attended with enormous risks of decomposing the gelatine, and when gelatine decomposes, the products are apt to reduce the” silver. A week at 32 °C in a warm summer is also a week of bacterial growth in a protein broth. That is the whole argument for the modern arrangement in one sentence: separate the two ripenings, wash between them, and control each with a measured addition rather than with a week of luck.

Recap

• Five stages: emulsification, ripening, washing, digestion, and preparation for coating. This page owns the first two.
• The nucleation burst fixes the number of crystals in the first seconds, and nothing later changes it. About 4 × 10¹⁴ crystals per mole of silver, each holding about 10⁹ silver ions.
• Addition time is the master control. Duffin’s Table 4.1: from half a minute to eighty-five minutes, mean grain area rises eighteenfold, relative speed seventy-fivefold, and gamma falls from 2.5 to 0.65 — with identical chemicals.
• Double jet gives a more even distribution because the halide excess stays constant, and this course does not perform one, because no domestic procedure for it is published anywhere it can reach.
• Excess halide is the solvent, working through the bromide complexes, and it is the same variable as pAg. Iodobromide emulsions carry 20 to 250 per cent excess; chloride and chlorobromide rarely more than 10.
• Ripening is cannibalism driven by surface energy: small crystals dissolve and large ones grow.
• pAg decides habit as well as rate: low pAg gives cubic crystals, higher bromide gives octahedral, and silver chloride shows no (111) faces at any chloride concentration.
• Ammonia ripens fast and cool, which is why the historical formulas use it; excluding it means working hotter and slower, and the page says so rather than pretending the choice is free.
• Iodide precipitates first, ends up concentrated in the coarser grains, buys speed and clean highlights up to about 10 per cent, and is delivered from a percentage solution because you cannot weigh it.
• A chloride paper emulsion is unwashed but it is ripened. The manifest’s “unripened” was wrong and is corrected here.

Check your understanding

Question 1. Two makers use identical quantities of the same chemicals at the same temperature. One adds the silver solution in thirty seconds, the other over eighty-five minutes. Using Duffin's Table 4.1, what differs in the finished emulsions, and why?

• Nothing measurable; the chemistry is identical, so the emulsions are identical
• The fast addition gives finer grain, lower speed and higher contrast; the slow addition gives coarser grain, much higher speed and lower contrast, because a rapid addition holds supersaturation high and keeps making new nuclei while a slow one lets arriving silver deposit on crystals that already exist
• The fast addition gives coarser grain because the crystals collide more often
• The slow addition gives a fogged emulsion because of the extra time at temperature

Show the answer and why

Answer: The fast addition gives finer grain, lower speed and higher contrast; the slow addition gives coarser grain, much higher speed and lower contrast, because a rapid addition holds supersaturation high and keeps making new nuclei while a slow one lets arriving silver deposit on crystals that already exist

Duffin's seven rows put numbers on it: mean grain area 0.14 against 2.56 µm², relative speed 12 against 910, gamma 2.5 against 0.65. The mechanism is the separation of nucleation from growth. A rapid addition keeps the supersaturation high enough to keep nucleating, so the same silver is divided among a very large number of crystals; a slow addition lets the supersaturation fall to a level that will deposit on an existing surface but not start a new crystal, so the silver goes into growth. Fog is a real risk of long hot holds, but it is not what this table is showing.

Question 2. Why does an emulsion recipe deliberately include more halide than the silver can combine with, and what would happen with an exactly equivalent quantity?

• To make sure all the silver is used up; with an equivalent quantity some silver nitrate would remain and fog the emulsion
• The excess halide forms soluble bromide complexes that dissolve silver halide, which is what allows crystals to exchange material and grow; with no excess, ripening would be very slow and the crystals would stay as precipitated
• To keep the pH low during precipitation
• To harden the gelatin before coating

Show the answer and why

Answer: The excess halide forms soluble bromide complexes that dissolve silver halide, which is what allows crystals to exchange material and grow; with no excess, ripening would be very slow and the crystals would stay as precipitated

Silver halide is almost insoluble in water, so on its own a precipitate cannot ripen — there is no route for material to move from one crystal to another. Excess bromide opens that route by forming AgBr₂⁻, AgBr₃²⁻ and AgBr₄³⁻; Duffin reports the tetrabromo complex as the most abundant at emulsion concentrations, with the tribromo complex suggested as the one that matters for growth. That is why the excess is called the ripening agent and why the same number, expressed as pAg, is what an industrial maker measures. The first option is the common misconception and it is backwards: it is the halide that is in excess, not the silver.

Question 3. You want a fast, coarse-grained negative emulsion. Which combination of choices moves you towards it, and which one of them is not available to you in this course?

• Fast addition, low temperature, low halide excess, ammonia; the ammonia is unavailable
• Slow addition, high emulsification temperature, large halide excess, and ammonia as a ripening agent; the ammonia is the one this course excludes
• Slow addition, low temperature, no halide excess, and a double jet; the double jet is unavailable
• Fast addition, high temperature, ammonia and a large gelatin quantity; the gelatin is unavailable

Show the answer and why

Answer: Slow addition, high emulsification temperature, large halide excess, and ammonia as a ripening agent; the ammonia is the one this course excludes

Each of the first three pushes the same way: a long addition time grows the crystals rather than multiplying them, a higher emulsification temperature speeds growth and coarsens the emulsion, and a large excess of bromide is the solvent that lets ripening proceed. Ammonia adds a much stronger silver complex and, in Duffin's words, lets an emulsion ripen satisfactorily at 40 to 45 °C where a neutral emulsion needs up to 70. Excluding ammonia at Level B therefore means this course works hotter and slower for less speed, and the page states that cost rather than absorbing it. More gelatin would move you the wrong way: it restrains growth.

Question 4. A published home iodobromide recipe calls for 0.1 g of potassium iodide in a batch carrying 5 g of silver nitrate. Your balance reads to 0.01 g. What is the right way to make that addition, and what general problem does the method solve?

• Weigh 0.1 g directly; the balance is adequate
• Make a 10 per cent solution — 10 g in 100 g of water — and measure 1 ml, which is about 20 drops; the method turns any milligram-scale addition into a volume you can measure
• Weigh 1 g and add a tenth of the batch by eye
• Leave the iodide out; at that quantity it cannot matter

Show the answer and why

Answer: Make a 10 per cent solution — 10 g in 100 g of water — and measure 1 ml, which is about 20 drops; the method turns any milligram-scale addition into a volume you can measure

A balance reading to 0.01 g does not weigh 0.1 g to a useful tolerance, and iodide is one of the most consequential small additions in an emulsion: it precipitates first, concentrates in the coarser grains, buys speed and clean highlights, and improves spectral sensitisation. Ross's percentage-solution method is the standard answer and it generalises to every milligram-scale addition in this part — the dye, the bromide restrainer, and the sensitiser doses in the design exercise. Note the second half of her advice, which is what makes it repeatable: label the stock with its concentration, solvent and date.

Question 5. Bennett kept a gelatine emulsion liquid at about 32 °C for six or seven days in 1878 and got an unprecedentedly fast plate. Which two processes was he running at once, and which one could he not have known about?

• Washing and coating; he could not have known about coating weight
• Physical ripening, which grew his crystals by cannibalism at a temperature the gelatin would tolerate, and chemical sensitisation by the sulfur impurities in the gelatin, which he could not have known about because Sheppard did not identify them until the 1920s
• Spectral sensitisation and hardening; he could not have known about hardening
• Nucleation and washing; he could not have known about nucleation

Show the answer and why

Answer: Physical ripening, which grew his crystals by cannibalism at a temperature the gelatin would tolerate, and chemical sensitisation by the sulfur impurities in the gelatin, which he could not have known about because Sheppard did not identify them until the 1920s

The week at 32 °C was a very long physical ripening, growing crystals by Ostwald ripening at a temperature low enough not to destroy the gelatin — Abney's own summary is that sensitiveness is attained by slow digestion at a low temperature instead of by boiling. It was also, unavoidably, a long chemical sensitisation, because the sulfur compounds in his gelatin were forming silver sulfide specks throughout. Abney names the cost in the same passage: in warm weather the long emulsification risked decomposing the gelatine, whose decomposition products reduce silver. Separating the two ripenings, with a wash between them, is what the modern sequence is for.

Question 6. Your iodobromide emulsion, held up to a safelight while still wet, has changed from pale yellow to a bluish cast during a long ripening. What does that suggest, and what is the limit of the observation?

• The iodide has been lost; the observation is quantitative and can replace a step wedge
• The crystals have coarsened, following Abney's rule that orange emulsions are in finer particles than blue or violet ones; the observation is a practitioner's check made during the make, not a measurement, and it does not replace a step wedge
• The emulsion has fogged; the observation is definitive
• The gelatin has denatured; nothing can be inferred about the crystals

Show the answer and why

Answer: The crystals have coarsened, following Abney's rule that orange emulsions are in finer particles than blue or violet ones; the observation is a practitioner's check made during the make, not a measurement, and it does not replace a step wedge

Abney's 1885 rule concerns scattering rather than absorption: emulsions that stay orange by transmitted light after boiling have the most covering power, "as the orange emulsion is in finer particles than the blue or violet". Keep that apart from the other transmitted-light rule on this page, Ross's composition ladder — white for chloride, pale yellow for bromide, deeper yellow with iodide — which reads the absorption edge and tells you what you made rather than how coarse it is. Both are checks you can make while there is still time to act; neither is a number, and the step wedge after coating is still what settles speed and contrast.

Check my answersClear