How Paints Work - Part 1
Knowing what is supposed to happen can help if you are making your own.

Pigments

Ever since people started painting, they have, in effect, wandered around the surface of this planet, looking for and collecting interestingly coloured rocks, taking them home and smashing them up into a fine powder and using them. There are plenty of naturally occurring minerals giving rise to many colours in the artist's palette.

Naturally occurring minerals include;

• cinnabar (red);
• minium (orangey red);
• realgar (orange);
• litharge (peachy yellow);
• orpiment (yellow);
• malachite (green);
• azurite (greenish blue);
• lapis lazuli (a very rich blue);
• hydrocerussite (white); and,
• so on.

It might be worth noting at this stage that realgar and orpiment are both arsenic compounds,

There are also a number of 'earth' colours, one group of which contain amounts of manganese dioxide and various hydration states of iron oxide ranging from a purplish red, through red and orange to yellow, then there are other earths that are shades of green (containing chromium compounds) and blue (copper compounds).

Man-made pigments include:

• vermilion (chemically identical to cinnabar but has been made as far back as Egyptian times);
• Egyptian Blue (a copper compound);
• lead white (chemically identical to hydrocerussite);
• lead-tin yellow I;
• lead-tin yellow II;
• carbon black (collected from lamps where the smokey flame deposits carbon);
• verdigris (a cyan colour made from treating copper with acetic acid in the presence of water); and,
• so on.

Since the 15th century, there have been many pigments synthesised and with the discovery of new elements such as cadmium and chromium; and the production of titanium as a white pigment as opposed to a buff-coloured pigment, there is a lot more choice. Examples include:

• french ultramarine (chemically identical to lapis lazuli);
• lead chromate;
• prussian blue; and many more.

Dyes

Dyes are normally thought of as dying fibres or fabric but they can be used in painting as well. However, where, with a normal pigment, you would just add it to, say, linseed oil or acacia gum solution, you cannot get away with that. The dye would not be stable in the medium and the dye could migrate within it and leach or bleed into the ground, other paints layers, nearby areas of paint or varnish - this action being labelled 'fugitive'. Clearly, dyes need to be stabilised with something that is not going to move around within the paint.

Normally, a fine, inert, colourless compound is used as the target of the dye so it gets dyed and then is used as a normal pigment. The compound normally used for this is alum. The dyed alum 'pigments' are called 'lakes' which comes from the Indian word 'lakh' which means 100,000 and reflects the large quantity of insects needed to dye a useful quantity of cloth.

Woad or Indigo has been used since Egyptian times and the modern chemical industry equivalent is the colour used to dye denim. You can obtain woad lake pigment from suppliers today.

Rose madder is another lake pigment that has been around since Egyptian times. One of the properties of rose madder is that it will fluoresce a 'day-glow orange' if you shine ultraviolet light on it. Inside Egyptian tombs, you can see the rose madder on the walls by shining ultraviolet light on them. (This was noted once by an explorer, I don't suggest that it is something that you should try for yourself.)

This dye was used to dye the red army tops although over the years, it fades if left in sunlight - clearly, if it is not, it will last literally thousands of years. Note that fading in a very short period of time is also called 'Fugitive.'

On the right, you can see two images of the Romy 2017 painting - the top in normal, visible light and the lower, illuminated by ultraviolet light. You can tell from the 'day-glow' orange areas where the rose madder is in the larger harlequin top and where it isn't in the cat's toy, even though the reds look very similar - look at the orange-red in the toy's headdress and at the very similar colour right next to it in the harlequin sleeve right next to it.

There are a number of pigments that fluoresce in a distinctive manner - lead white fluoresces white and curiously, Egyptian blue fluoresces in the infrared when illuminated by visible light.

Carmine is another lake pigment and comes from a beetle from South America. As South America wasn't discovered by Europeans until a long time after the beginning of the fifteenth century, its inclusion in paintings from before the continent's European discovery indicates that either the inclusion of the lake is from a more recent repair to the painting or that the painting was made after that time.

Impurities

On the right, you can see some malachite as it occurs naturally. Around a quarter of the way up and roughly 40 per cent of the way across, you can see a small inclusion of rock. When the malachite is ground up, this impurity is also ground up and unless it is removed from the pigment by some means or other, it will end up in the paint.

Sometimes, the level of impurities in the pigment is low enough for it not to be a problem. However, when the level is too high, something needs to be done about it.

The easiest way to end up with pigment that has a low enough level of impurities in it is to select sources that are naturally low in impurities and, of course, the people who sell these know this and will part you from more of your money for the privilege but sometimes even these sources still need purifying.

So, how do we purify pigments?

Separation methods break down into two main areas - chemical and physical.

Chemical Methods

You can use chemical methods if you are not particularly bothered about the state of the particles of pigment that you end up with. Chemical methods include dissolving the pigment or impurities in water, an acid, alkali, some organic solvent or other.

Examples:

• If the impurities dissolved in acid but the pigment didn't, you could use that to remove the impurities and, you would retain the structural integrity of the pigment - removal of the copper impurity in stack lead white utilises this.
• If both the impurities and the pigment dissolve in acid and say, the pigment was more soluble than the impurity, once they have dissolved, you can add alkali to precipitate the impurities out first, filter out the precipitate and then add the rest of the alkali to precipitate out the pigment. This would have the disadvantage that you would lose the particle structure of the pigment.
  The analogous process would exist for pigments that dissolved in alkalies where you would add acid to precipitate out the components differentially.
• If the impurities dissolve in an organic solvent but the pigment does not, you can 'wash' out the impurities with the solvent - again, this preserves the structure of the pigment.
• If both the pigment and the impurities dissolve in an organic solvent that is immiscible with water but the impurities are acidic (for example, an organic acid), dissolve them in the organic solvent and then add some alkali in water and shake up then let the two layers separate, take the organic solvent layer and evaporate off the solvent.

There are of course many permutations of the above but that should give you an idea of what can be done.

Physical Methods

These methods rely upon differentiating between the pigment and its impurities by physical properties such as density, whether they are hydrophylic or hydrophobic and so on.

In the photograph on the right, you can see powdered malachite in water. The malachite has tended to settle towards the bottom of the Erlenmeyer flask and the impurities have tended to stay suspended in the water. Here, the malachite has a greater density so it sinks first. If the water is tipped off and this procedure repeated, reasonably pure malachite can be obtained.

This separation method depends upon the densities of the pigment and the impurities remembering that if they both sink, you can make the water denser using a soluble salt such as sodium chloride until you get the best separation. When they are very similar to each other, the rate at which the particles fall through the liquid becomes a factor. That can be determined by the shape of the particles when they are ground up insofar as if, say the pigment particles tend to form roundish particles but the impurities tend to be flat, they will sink at different rates and then it is a matter of selecting the best time to separate them.

However, there is a limit to this because some malachite stays in suspension with the impurities and some of the impurities sink with the malachite. Over several purification steps, you can decide how much malachite you are prepared to discard for the sake of getting an appropriate level of purity in the finished pigment.

You don't have to tip the impurity-rich layer down the sink, you can keep hold of it and in the case of malachite, you can separate off the water and then dissolve the copper carbonate in acetic acid, filter it and then either, let it dry out to make verdigris or, split the solution into two equal amounts and to one part, add an excess of sodium carbonate which will precipitate out copper carbonate and to the other, add an excess of sodium hydroxide solution which will precipitate out copper hydroxide. Filter these out and you have a very fine particle-sized artificial malachite.

Lapis lazuli is usually mixed in with other rock and predictably, the purest is the most expensive. However, if you grind it up into a fine powder then mix it with a mixture of wax and resin, let it cool down then each day, warm it up and massage it, for about a week or so. Finally, warm it up and then massage it in water and the pure lapis lazuli powder will fall out and sink to the bottom. This starts off fairly pure but as the massaging process continues, an increasingly large amount of the impurities start to fall out as well. The best way to do this extraction process is to change the water at regular intervals so the first batch is the purest then you have another batch and so on. Then, you can combine the batches that are of a certain grade so that you are not mixing the purest with the most contaminated to make your paint.

The lapis lazuli relies on the fact that the lapis lazuli is more hydrophylic (water loving) than the impurities and so will tend to fall out of the resinous wax mixture first. This differential of physical properties between the pigment and its impurities is what all of these processes have in common.

Effects of Particle Size

The diagram on the right represents what happens to a ray of white light as it interacts with a pigment particle in a transparent medium. At each interface between: air and medium; medium and particle; and, medium and ground, there is an opportunity for reflection and refraction - within the pigment particle, there is also an amount of absorption going on as well.

The incident rays starts off in the upper-right corner (as ray 1) and you can follow it around as it makes its way into the medium (where it becomes ray 2), losing some light as a reflection off the surface of the paint, into the particle (where it becomes ray 3), and where that ray starts to lose some of its blue content, then is internally reflected, (becoming ray 4), passing out of the particle into the medium (5) and finally out into the outside world whereupon, it can be observed by the viewer (6).

At each interaction with an interface, some of the original beam of light is lost and you can work out how much by inserting values for the refractive indices of the two materials at each interface as well as the angle of the incident ray. There is a page on Wikipedia about Snell's Law, Fresnel's equations here. Basically, the greater the difference in refractive index between the pigment and the medium, the greater the reflections at the surfaces that the beam encounters.

Hopefully, the medium is colourless enough so that it doesn't absorb any appreciable amount of light as the beam passes through it. The Predominant colouring process happens as the beam passes through the particles of pigment.

Note that the absorption of the light is a process that happens gradually within the pigment particle and that each time light enters a particle, some of it is reflected, back into the medium where it has the opportunity either to return to the outside world or end up encountering another pigment article.

Effect of Particle Size on Depth

The paint sample on the right (pigments in acacia gum) shows both coarsely- and finely-ground pigment samples for: Lapis Lazuli; and, Smalt.

These pigments are transparent when they are a single mass - smalt is alkaline (potassium) cobalt glass - and as such, light is coloured as it travels through the pigment.

The coarse samples show what the pigment is capable of and what an artist would expect of the colours. However, both smalt and lapis lazuli can be rather coarse, especially if you are painting fine detail, and they are applied more easily of the particle sizes are finer. However, you can see what happens to them when they are finer.

Let's see what happens when the particle size is reduced.

This is a closeup of what happens when light falls onto a fairly coarsely ground pigment.

The light is reflected uncoloured from the surface of the pigment particles when it enters them, reflects internally and then exits the particle having been coloured on its travels through the pigment, leaving the paint layer and being observed by the viewer.

When the particles are smaller, the light can travel through more of them in the same volume of paint but there are more surfaces to be reflected from and straight back out to the viewer before it has a chance to be coloured.

There are two main processes going on here so let us have a look at them in more detail...

Effect of Particle Size on Hue

This is natural cinnabar in its granular form. Its colour is quite intense and as a result of the large particle size, the light incident upon it can travel a long way and be modified to a great extent before it encounters the other end of the particle along with the opportunity to find its way back to the viewer.

These particles are around the size of granulated sugar so there is no way that you would paint with them - they need to be ground - normally by sliding the muller over the slab so that the particles end up with an even size. However, there is an interesting artefact of not doing that so, to illustrate a point...

This is the same sample of cinnabar but it has been ground by rotating the muller - not something that you would normally do. Grains at the centre of the pile of pigment have nowhere to go so they are ground more finely whereas pigment at the edge will spread out and end up as larger particles. You can see in the photograph that the pigment in the middle is ground so finely that it is almost an orange colour (hue value of 13) whereas that around the edge is a deeper red (hue value of 0).

Normally, you wouldn't grind paint in that way as it is better to have a fairly consistent particle size. If you do have a wide range of particle sizes and the medium that you are using has a low viscosity, the larger particles will tend to sink faster than the smaller ones - if they are really small, they might even stay in suspension because of Brownian motion. The result of this would be that the deeper coloured pigment would settle to the bottom and as you paint your way through the preparation, the hue would change.

This spectrum shows what happens to the wavelengths of light when they fall onto cinnabar that has been ground to different particle sizes.

The deeper red is the larger particles, where all of the blue and quite a lot of green light has been absorbed; and, the lighter, more orangey spectrum is from the smaller particles where the colour of the light is not changed very much within each particle and there are more interfaces where light can be reflected back to the viewer, unaltered.

This effect has been used in paintings where a single pigment has simply been ground differently in order to produce two different shades of the same colour. This has been done with azurite and also with malachite.

The reduction in saturation and shorter pathlength effects that you see here happen the most when there is a greater difference in refractive index between the pigment and the particles - the amount of reflected light increases with this difference. So, if the refractive index is increasingly similar, you would expect this effect to diminish and instead, the depth of colour to increase or, as a minimum, stay very similar.

Masstone

Ideally, the refractive index of the medium will be around the same value as the pigment. In this case, the light has minimal chance to be lost in reflection before it has its colour changed by the pigment.

Another effect is that the light travels further into the paint layer before it is either scattered back to the viewer or reflected back by the ground or underlying paint layers.

The consequence of this is that the amount of colour that is absorbed becomes more a function of the thickness of the paint - with opaque paints, it doesn't matter how thick they are but with transparent layers, they get deeper, the thicker they are.

This would not be a problem if the pigments were perfect - the magenta absorbed only green light, the cyan absorbed only red light and so on. However, these are real-world pigments and magentas tend to absorb a bit of blue light and cyans tend to absorb a bit of green light.

Many of us, as children, have wondered at the orange-coloured liquid in our mother's pantries that is labeled 'yellow food colouring' because it is almost invariably a, orangey-red colour that miraculously turns yellow when it is diluted. It does this for the same reason.

On the right, you can see some mango flavouring which is coloured orangey-yellow. You can see from the image on the top-right that it is the same liquid in two identical jars.

The image on the right is the same two jars with the same liquid in (you can tell from the shadows) but on the right, where the light path is not as long, there is a shift in hue - a value of 0 on the left (pure red) to 46 on the right (40 is orange and 60 is yellow). This is the same liquid so why does this happen?

On the right, you can see the spectra of various sample points from the image. They show that whilst the food colouring absorbs almost all of the blue light and transmits almost all of the red light, it also absorbs some green light.

As the path length (could be concentration instead) increases, the change in the amount of green light absorbed is greatest so the most noticeable thing that happens is that with roughly the same amounts of red and blue light being seen, is a hue change in an analogous way that as it did with the particle size in the cinnabar.

With regards to the real-life cyan paint that absorbs a little green light, when there is a lot of it in a paint, there is enough green light absorbed for people to think that they are looking at blue paint. Similarly, the real-life magenta paint that absorbs a bit of blue light will convince most people that they are looking at red paint.

The problem with this is that if you ask people what colours the three primary colours are in painting (subtractive primaries) and they will reply; 'Red, yellow and blue' when they are really talking about what colours the masstones of those colours are. In painting, like anywhere else, the three subtractive primary colours are yellow, magenta and cyan, regardless of what people want to call them because, at the dilutions that you use them at, they are the colours.

If people are going to insist that they call them red, yellow and blue because they are the colours that they are in the tubes, then you might as well call French Ultramarine, 'Black.'

---

All images and original artwork Copyright ©2020 Paul Alan Grosse.

---