Where colour in stone comes from

Colour is the first thing anyone notices about a stone tool and the last thing that should be trusted to identify it. A mineral’s colour usually comes from impurities present in tiny quantity, which means the same species can appear in half a dozen colours and different species can appear in the same one. Understanding the four mechanisms that produce colour is what makes that intelligible rather than arbitrary.

Colour that belongs to the mineral

A small number of minerals are idiochromatic: an element essential to the composition is itself a colouring element, so every specimen shows the colour. Malachite is green because copper is part of what malachite is. Turquoise is likewise copper-coloured by definition.

Most gem minerals are the other case — allochromatic. The pure compound is colourless, and colour arrives with trace impurities that vary from deposit to deposit and from crystal to crystal. Pure quartz is colourless. Pure corundum is colourless; add a little chromium and it is ruby.

This is the root of the identification problem. If colour comes from an impurity present in trace quantity, colour is close to uninformative about the host mineral.

The four mechanisms

Transition metal impurities. Iron, chromium, manganese, copper, titanium, vanadium, nickel and cobalt absorb particular wavelengths when they sit in a crystal lattice, and what passes through is what you see. The same element gives different colours in different structures, and in different oxidation states, which is why iron accounts for green, yellow, brown and black across various minerals. In nephrite, the position along the tremolite–actinolite series — how magnesium-rich or iron-rich the material is — accounts for most of its range from near-white to deep green.

Colour centres. A defect in the lattice can absorb light on its own: a missing atom, a trapped electron, an atom in the wrong place. These are often produced by natural radiation over geological time and can frequently be undone by heat, which is exactly why some colours are unstable. Smoky quartz and amethyst owe their colour to this class of mechanism, and both can be altered in a domestic oven’s temperature range or a little above it.

Inclusions. The host is colourless or pale and something else inside it supplies the appearance. Green aventurine is a quartz aggregate full of platy chromium mica; its colour and its glitter are both the inclusions rather than the quartz. The pink of massive rose quartz is generally attributed to microscopic fibrous inclusions rather than to anything dissolved in the quartz lattice.

Charge transfer is worth separating out as a variant of the first mechanism. Where two neighbouring ions of different elements or oxidation states sit close together in a lattice, an electron can move between them under illumination, and that process absorbs light strongly. It produces some of the most saturated colours in mineralogy from very small quantities of impurity, and it is why “trace” and “faint” are not the same thing.

Structural colour. No pigment at all — light interacts with fine-scale features. Thin-film interference, diffraction from an ordered array, and scattering all produce colour this way. Opal’s play of colour and the sheen in some obsidian belong here. So does the iridescence of a vapour-deposited metallic coating, which is a manufactured version of the same physics.

What that means for a stone in your hand

WHAT YOU CAN TELL — reading colour

  · Colour follows banding, mottling or visible
      inclusions → likely natural distribution
  · Colour pooled in a web along fine cracks
      → dye. Strong indicator
  · Uniform saturated colour, no internal variation
      → suspect dye, glass or resin
  · Glitter from flat reflective flakes
      → inclusions, natural or manufactured
  · Iridescent film that stops at a scratch
      → surface coating, not body colour
  · Which element or defect causes a natural colour
      → NOT DETERMINABLE at home. Needs
        spectroscopy

The distinction that actually matters is body colour versus surface colour versus fracture-filled colour. Natural body colour is generated throughout the material and varies with the material’s own structure. Dye is introduced from outside and follows the paths available to a liquid, which in a fine-grained aggregate means grain boundaries and cracks. A coating sits on top and is only as thick as it was applied.

How the trade uses colour

Colour is where most of the money is, and therefore where most of the intervention is.

The premium is for colour, so colour is what gets manufactured. A pale, cloudy stone and an intense even one may be the same species and grade in every respect except appearance. Bringing the first to look like the second is the commonest treatment in the trade, and it is cheap relative to the price difference it creates.

Colour names get used as species names. A green stone becomes “jade”, a pink one becomes “rose quartz”, and the label describes appearance while implying composition. This is the ordinary mechanism of a trade name.

Some natural colours are unstable, and this is disclosed inconsistently. Where a colour comes from a colour centre, prolonged strong light or heat can fade it. That is a genuine property of the material, and a faded stone has not necessarily been treated.

One observational caution

Colour is not a fixed attribute of an object; it is the product of the material and the light falling on it. Daylight, incandescent light and the various spectra sold as white LED contain very different proportions of wavelengths, and a mineral whose absorption happens to fall where two sources differ will look markedly different under each.

The practical consequence is that any judgement about colour — whether it is even, whether it is too saturated, whether it matches a listing photograph — should be made in one consistent light, and daylight is the conventional choice for exactly this reason. A stone that looks wrong under a kitchen bulb may look entirely ordinary at a window, and a photograph taken under one source cannot be compared with an object seen under another.

The limit

You can often tell how colour is distributed, which is a real forensic observation and the basis of the strongest home indicator there is. You cannot tell what causes a natural colour without a spectrometer, and you frequently cannot tell a stable natural colour from a well-executed dye job that did not leave pooling — some modern dyes penetrate evenly, and a stone with no visible pooling is not thereby established as undyed.

Colour, in other words, is useful in exactly one direction. It can raise suspicion. It cannot confirm a species, and the things it cannot confirm are most of what people want to know.