Colour Theory in Gemmology
The physics and chemistry of gem colour: crystal field theory, charge transfer, colour centres, allochromatic and idiochromatic minerals, and band-gap colouration.
Overview – The Four Main Causes of Colour
Kurt Nassau's classification (Nassau, The Physics and Chemistry of Color, 2nd ed., 2001)
identifies fifteen mechanisms of colour in minerals, of which four are central to gemmology:
- Crystal field theory (d–d transitions) – the dominant mechanism for chromophore transition
metals in most coloured gem species (Cr³⁺ in ruby and emerald, Fe²⁺ in peridot). - Charge transfer – electron transfer between adjacent ions; responsible for blue sapphire
and aquamarine colour. - Colour centres – defects in the crystal lattice that trap electrons or holes and absorb
visible light; responsible for smoky quartz, amethyst, blue topaz, and coloured diamond. - Band-gap (semiconductor) absorption – the entire conduction band absorbs photons above
the gap energy; explains the pure colourlessness of type IIa diamond and the vivid red of
cinnabar.
Understanding which mechanism operates in a given stone allows the gemmologist to interpret
spectroscope data, Chelsea Colour Filter reactions, fluorescence, and treatment susceptibility.
Crystal Field Theory and d–d Transitions
Crystal field theory (CFT) describes how the electrostatic field of surrounding ligand anions
(usually O²⁻) splits the five degenerate d-orbitals of a transition-metal cation into two
energy sub-sets. When an electron absorbs a photon to jump between sub-sets, the remaining
transmitted wavelengths constitute the body colour.
Mechanism:
- In an octahedral coordination site (6 O²⁻ ligands), the d-orbitals split into a lower set
(t₂g) and a higher set (eɡ), separated by the crystal field splitting energy Δ_oct. - In a tetrahedral site (4 ligands), the splitting Δ_tet ≈ 4/9 Δ_oct – smaller, producing
weaker colour. - The energy Δ determines the absorbed wavelength (E = hc/λ); a larger Δ absorbs shorter,
higher-energy wavelengths. - The same chromophore ion produces different colours in different host structures, because
Δ depends on coordination number, bond length, and the nature of the coordinating anion. - Cr³⁺ has strong absorption in the blue–green and a weaker band in the yellow; what
survives is deep red to pink. The exact energies are host-dependent.
Key diagnostic character: d–d transitions produce broad, smoothly curved absorption
bands in the visible, distinguishable from the sharp lines of colour centres and the steep
asymmetric edges of charge transfer.
Source: Nassau, pp. 101–153 [VERIFIED]; Read 3rd ed., DOI: 10.4324/9780080507224 [VERIFIED]
Ruby – Cr³⁺ in corundum (Al₂O₃)
Al³⁺ octahedral sites are comparatively small; the strong field around Cr³⁺ gives Δ_oct
large enough to produce broad absorption at ~410 nm (violet–blue) and ~560 nm (yellow–green),
leaving deep red transmitted.
Cr³⁺ also produces a characteristic sharp emission (fluorescence) doublet at ~692–694 nm
(the "R lines"), visible in the absorption spectrum as the ruby doublet and responsible for
the internal red "fire" of Mogok rubies when illuminated with a tungsten source.
Chelsea Colour Filter reaction: strongly red (Cr³⁺ transmits deep red and absorbs green).
Emerald – Cr³⁺ in beryl (Be₃Al₂Si₆O₁₈)
The same Cr³⁺ ion sits in a slightly larger Al³⁺ octahedral site in beryl compared with
corundum, shifting Δ to smaller values. The absorption bands move to slightly longer
wavelengths (~430 nm and ~610 nm), broadening the transmission window into green and
reducing fluorescence efficiency relative to ruby.
V³⁺ in some Brazilian emeralds produces a similar but slightly weaker green via an
identical d–d mechanism. Chrome emeralds appear red through the Chelsea Colour Filter;
vanadium emeralds may appear red–orange or inert depending on V concentration.
Alexandrite – dual transmission window
Alexandrite (Cr³⁺ in chrysoberyl, BeAl₂O₄) is the classic example of the alexandrite
effect. The Cr³⁺ crystal field in chrysoberyl produces two transmission windows:
one in the red (680 nm) and a narrower one in the blue–green (580 nm region excluded
and ~470–490 nm transmitted).
Under daylight (photopic peak ~555 nm), the blue–green window dominates → green.
Under incandescent light (spectral peak shifted toward 600 nm), the red window
dominates → red–purple. The colour change depends on which window the illuminant's
spectral peak falls nearest.
Source: Nassau 2001, pp. 112–115 [PARTIALLY_SUPPORTED]; F-009 confirmed in principle
Peridot – Fe²⁺ in olivine
Fe²⁺ in the M1 and M2 octahedral sites of olivine (Mg₂SiO₄) produces three distinct
absorption bands at approximately 493 nm, 473 nm, and 453 nm – the classic three-banded
iron spectrum visible in the spectroscope. The yellow-green colour results from this
triplet absorbing blue and leaving yellow-green transmitted.
The relatively weak crystal field around Fe²⁺ in olivine results in smaller Δ values
compared with Cr³⁺ in corundum, which is why the colour is yellow-green rather than red.
Peridot is idiochromatic: Fe²⁺ is an essential constituent of gem-quality olivine.
Source: Nassau, pp. 108–112 [VERIFIED]
Charge-Transfer Colouration
Charge transfer (CT) colour arises when an electron is transferred between two adjacent ions,
either between two metal ions (intervalence charge transfer, IVCT) or from oxygen to a
metal (ligand-to-metal charge transfer, LMCT), rather than remaining on a single ion as
in d–d transitions.
Mechanism:
- Heteronuclear IVCT: Fe²⁺–Ti⁴⁺ pairs on alternating face-sharing octahedral sites
along the c-axis of corundum. Under photon absorption:
Fe²⁺ + Ti⁴⁺ + hν → Fe³⁺ + Ti³⁺. Absorbs strongly from blue through yellow–green,
leaving blue–violet transmitted. - Homonuclear IVCT: Fe²⁺ + Fe³⁺ on adjacent sites:
Fe²⁺ + Fe³⁺ + hν → Fe³⁺ + Fe²⁺. Absorbs in yellow–red, leaving blue transmitted. - LMCT: Electron transfer from O²⁻ to an empty d-orbital of Fe³⁺ absorbs in the UV
and tails into the visible, producing yellow or orange. - CT transitions have much higher extinction coefficients than d–d transitions; trace
concentrations of CT-capable pairs (~0.01% Ti in sapphire) suffice for intense colour. - CT bands are broad and asymmetric, without the sharp emission lines of Cr³⁺.
Source: Nassau, pp. 184–210 [VERIFIED]; Dubinsky 2020, DOI: 10.5741/gems.56.1.2 [VERIFIED]
Blue sapphire – Fe²⁺–Ti⁴⁺ IVCT in corundum
Nassau's canonical IVCT example. The Fe–Ti pair on alternating face-sharing octahedral
sites absorbs strongly in the yellow–red (peaking ~550–700 nm), leaving blue–violet
transmitted. Spectroscope shows three broad, unresolved bands at approximately 450 nm,
460 nm, and 470 nm – the diagnostic blue sapphire triplet.
No sharp Cr fluorescence lines; no red CCF reaction (no Cr³⁺ unless the sapphire is
deliberately Cr-bearing, as in some Sri Lanka padparadscha).
Source: Dubinsky et al. 2020, DOI: 10.5741/gems.56.1.2 [VERIFIED]
Aquamarine – Fe²⁺–Fe³⁺ IVCT in beryl
Fe²⁺ and Fe³⁺ on adjacent sites in the beryl channel structure produce a broad
absorption band in the yellow–red, leaving blue transmitted. The spectroscope shows
a broad unresolved band without sharp lines.
Indicolite tourmaline (blue) carries a similar Fe²⁺–Fe³⁺ IVCT colour. Kyanite blue
is also attributed to this IVCT mechanism.
Yellow sapphire – Fe³⁺ LMCT
Fe³⁺ alone (without Ti⁴⁺ to form IVCT pairs) absorbs in the UV, with the absorption
tail cutting into the blue end of the visible, leaving yellow–orange transmitted.
Some orange sapphires combine Fe³⁺ LMCT with Cr³⁺ d–d bands.
Diagnostic contrast – CT vs d–d
A gemmologist can distinguish CT-coloured stones from d–d-coloured stones by:
- Spectroscope: CT gives broad, asymmetric bands; d–d gives narrower, more
symmetrical bands; Cr³⁺ gives narrow sharp emission doublet at 692–694 nm. - CCF: CT blue sapphire appears inert to green (no Cr); Cr³⁺ stones appear red.
- Fluorescence: Cr³⁺ ruby fluoresces strongly red under LWUV; Fe²⁺–Ti⁴⁺
sapphire is typically inert or shows weak orange.
Colour Centres
A colour centre is a localised lattice defect (typically an electron F-centre or a hole
h-centre) trapped at a structural imperfection such as an anion vacancy or adjacent to a
substitutional impurity, that absorbs visible light and creates colour in an otherwise
colourless material.
Mechanism:
- F-centre (Farbe = colour): an anion vacancy trapping one or more electrons. The
trapped electron absorbs photons to move between energy levels of the defect potential. - Hole centre: an unpaired electron hole trapped adjacent to a substitutional cation.
In quartz, Al³⁺ substituting for Si⁴⁺ creates a charge imbalance; irradiation leaves an
AlO₄ unit with a trapped hole: the [AlO₄]⁰ centre. - Irradiation (natural γ/α/β or artificial neutron/electron beam) creates centres by
displacing atoms and generating vacancies. The stability of the centre determines whether
colour is permanent or fades. - Heat bleaching: raising temperature allows trapped electrons/holes to escape and
recombine, destroying the centre. Smoky quartz bleaches at ~300–400 °C; amethyst converts
to citrine at ~450–500 °C; irradiated blue topaz would bleach at high heat.
Source: Nassau, pp. 211–250 [VERIFIED]; Nassau & Prescott 1977,
DOI: 10.1180/minmag.1977.041.319.01 [VERIFIED]
Smoky quartz – Al–O hole centre
Al³⁺ substituting for Si⁴⁺ in quartz, combined with natural or artificial irradiation,
produces a hole trapped at the AlO₄ unit (the [AlO₄]⁰ centre). Nassau & Prescott (1977)
demonstrated that the smoky colour arises from the A₃ absorption band at 2.90 eV (428 nm).
Heating above 300–380 °C bleaches the centre, converting smoky quartz to colourless.
Source: Nassau & Prescott 1977, DOI: 10.1180/minmag.1977.041.319.01 [VERIFIED]
Amethyst – Fe⁴⁺ hole centre
Fe³⁺ substituting for Si⁴⁺ in quartz, under irradiation, produces an [FeO₄]⁰ centre
(Fe⁴⁺, a hole localised on iron). This centre absorbs at ~520 nm (green), transmitting
purple-violet.
Heating above ~450 °C converts Fe⁴⁺ back to Fe³⁺, producing yellow citrine. This
transformation is exploited commercially to produce heat-treated citrine from amethyst.
Blue topaz – irradiation-induced colour centres
Natural blue topaz is rare. Commercial blue topaz is produced by neutron or electron
irradiation of colourless topaz, creating colour centres that absorb in the yellow–red
region. The colour is stable at room temperature but bleaches at high heat.
The irradiated origin must be disclosed in trade (treated material). The colour mechanism
distinguishes blue topaz (colour centre, heat-sensitive) from aquamarine (Fe IVCT, stable).
Diamond colour centres – N3, NV, H3
Nitrogen defects in diamond create several named optical centres:
- N3 centre (three N atoms + vacancy): responsible for Cape blue-white fluorescence
under LWUV and the 415 nm absorption line (Cape series). Characteristic of type Ia
diamond – the most common natural diamond type. Spectroscopic literature gives 415.5 nm
for the zero-phonon line. - NV⁻ centre (nitrogen–vacancy pair, negative charge): zero-phonon line at 637 nm;
produces pink to red photoluminescence. Found in pink diamonds and exploited in
quantum sensing applications. - NV⁰ centre (neutral charge state): zero-phonon line at 575 nm.
- H3 centre (two N atoms + vacancy): zero-phonon line at 503 nm (503.2 nm in
spectroscopic precision); produces green–yellow fluorescence; found in HPHT-treated
diamonds. Detected by photoluminescence at 77 K.
HPHT treatment of brown type IIa diamond anneals out vacancy-related brown centres,
destroying N3 characteristic fluorescence, detectable by FTIR and DiamondView.
Source: Hainschwang 2012, DOI: 10.5741/gems.48.4.252 [VERIFIED]; Nassau 2001,
pp. 227–235 [VERIFIED]
Maxixe beryl – unstable colour centre
Deep blue beryl from the Maxixe mine, Minas Gerais, Brazil, owes its colour to an
irradiation-induced colour centre (involving NO₃⁻ or CO₃²⁻ radical anions substituting
in the beryl channel; the precise identity remains under study).
The centre is unstable in ambient light and fades to pale pink or colourless over months
of daylight exposure, making it the standard gemmological example of light-bleachable colour.
This distinguishes maxixe beryl from aquamarine (Fe IVCT, stable) and irradiated blue
topaz (different centre, stable at room temperature).
Duty to advise: a gemmologist identifying maxixe beryl must advise the client that
the colour will fade in sunlight, as this is a material fact affecting value.
Source: Adamo 2008, DOI: 10.5741/gems.44.3.214 [VERIFIED]
Fluorite colour centres
Purple, golden, and blue fluorite owes its colour to electrons and holes trapped at Ca²⁺
vacancies and interstitial F⁻ defects; natural irradiation from uranium-bearing inclusions
is often responsible. Colour is bleachable by heat. Fluorite's own very low dispersion
(0.007) gives it a "dead" optical appearance despite sometimes vivid colour.
Source: Nassau, pp. 215–218 [VERIFIED]
Allochromatic, Idiochromatic, and Pseudochromatic
Three categories describe whether colour is essential to a mineral species or incidental.
- Allochromatic: the pure end-member is colourless; colour depends on trace chromophore
impurities. The same mineral can appear in many colours. Removal of the impurity would
restore colourlessness. - Idiochromatic: the colouring element is a required component of the mineral formula.
Even an ideal, pure crystal will be coloured. The colour range is narrow. - Pseudochromatic: colour-like effects produced by light interference, diffraction, or
scattering, not electronic absorption. No chromophore is involved; the apparent colour
changes with viewing angle or disappears when the specimen is oriented edge-on.
Source: Nassau, pp. 100–104 [VERIFIED]
Allochromatic examples
| Mineral | Pure form | Chromophore → colour |
|---|---|---|
| Corundum (Al₂O₃) | White sapphire (colourless) | Cr³⁺ → ruby; Fe²⁺+Ti⁴⁺ → blue sapphire; Fe³⁺ → yellow; Cr+Fe → padparadscha |
| Beryl (Be₃Al₂Si₆O₁₈) | Goshenite (colourless) | Cr/V → emerald; Fe²⁺ → aquamarine; Mn → morganite; Fe³⁺ → heliodor |
| Quartz (SiO₂) | Rock crystal (colourless) | Al³⁺ + irradiation → smoky; Fe⁴⁺ + irradiation → amethyst; Fe³⁺ thermal → citrine |
Idiochromatic examples
| Mineral | Essential chromophore | Colour |
|---|---|---|
| Peridot / gem olivine ((Mg,Fe)₂SiO₄) | Fe²⁺ essential to composition | Yellow-green; cannot exist colourlessly |
| Malachite (Cu₂(CO₃)(OH)₂) | Cu²⁺ essential | Always green from Cu²⁺ d–d absorption |
| Turquoise (CuAl₆(PO₄)₄(OH)₈·4H₂O) | Cu²⁺ essential | Always blue-green; more Fe → greener |
Pseudochromatic examples
| Mineral/Phenomenon | Cause | Notes |
|---|---|---|
| Precious opal (play-of-colour) | Bragg diffraction from ordered silica sphere arrays (~150–300 nm) | Colours shift with angle; no chromophore |
| Labradorite (labradorescence) | Thin-film interference in lamellar feldspar intergrowth (Bøggild intergrowth) | Structural colour; changes with angle |
| Moonstone (adularescence) | Scattering + interference at thin albite lamellae within orthoclase | Billowing blue sheen; not absorbed colour |
Band-Gap Colouration
In crystalline solids with a band structure, colour arises when the band gap (Eg, the energy
difference between the valence band and the conduction band) falls within or near the visible
range. Photons with energy > Eg are absorbed; those with energy < Eg are transmitted.
Mechanism:
- Pure diamond has Eg ≈ 5.47 eV, corresponding to ~227 nm (deep UV). All visible photons
pass through → pure type IIa diamond is colourless. - When nitrogen or boron substitutes for carbon, new energy levels appear inside the gap,
allowing selective visible absorption. - In semiconductors such as cinnabar, the gap itself falls in the visible range, giving a
sharp absorption edge.
Source: Nassau, pp. 253–275 [VERIFIED]; Read 3rd ed., DOI: 10.4324/9780080507224 [VERIFIED]
Diamond type classification and colour
| Type | Nitrogen/Boron configuration | Body colour | Notes |
|---|---|---|---|
| Type Ia (most common) | N in aggregated A and B centres | Colourless to Cape yellow–brown | N3 centre at 415 nm absorbs blue → yellow |
| Type Ib (rare natural; common synthetic) | Single substitutional N | Canary yellow (intense) | N donor level absorbs blue ~430 nm |
| Type IIa (rarest natural) | Essentially no N or B | Colourless | Largest diamonds (Cullinan); HPHT treatment target |
| Type IIb (rare) | B acceptors; no N | Blue (e.g. Hope Diamond) | P-type semiconductor; electrically conductive |
Band-gap materials beyond diamond
| Material | Band gap / absorption | Body colour |
|---|---|---|
| Cinnabar (HgS) | Eg ≈ 2.0 eV (~620 nm); blue–green absorbed | Vivid red; SG 8.0–8.2 |
| Realgar (As₄S₄) | Eg absorbs violet and blue | Orange-yellow |
| Orpiment (As₂S₃) | Eg absorbs blue-violet | Lemon yellow |
Diagnostic Relevance
A gemmologist uses colour-cause knowledge at multiple practical points:
- Chelsea Colour Filter interpretation: Cr³⁺ (d–d, CFT) → red; Fe²⁺–Ti⁴⁺ CT blue
sapphire → inert; S₃⁻ colour centre in hauyne → inert. - Treatment detection: Colour-centre stones (smoky quartz, blue topaz, maxixe beryl,
irradiated diamond) are heat-sensitive. Knowing the mechanism identifies which treatments
could destroy or alter the colour. - Advising clients: Maxixe beryl fades in light; irradiated blue topaz is stable at
room temperature but not at jewellery-repair temperatures. These are material disclosure
facts. - Spectroscope interpretation: d–d transitions → broad bands; CT → asymmetric bands;
colour centres → sharp zero-phonon lines (in diamond photoluminescence); all contrasted
with sharp Cr³⁺ doublet at 692–694 nm. - Simulant/treatment separation: diamond type classification by FTIR (N content),
optic character, and UV response is a direct application of band-gap/colour-centre theory.
Sources
- Nassau, K. The Physics and Chemistry of Color, 2nd ed. Wiley-Interscience, 2001.
ISBN 978-0-471-39106-7. Chapter DOI: 10.1016/b978-044451251-2/50008-8. [VERIFIED] - Read, P. Gemmology, 3rd ed. Routledge, 2012. DOI: 10.4324/9780080507224. [VERIFIED]
- Nassau, K.; Prescott, B.E. "Smoky, blue, greenish yellow, and other irradiation-related
colors in quartz." Mineralogical Magazine, 41(319), 1977.
DOI: 10.1180/minmag.1977.041.319.01. [VERIFIED] - Hainschwang, T. et al. "Photoluminescence at 77K in treated diamond." Gems & Gemology,
48(4), 2012. DOI: 10.5741/gems.48.4.252. [VERIFIED] - Dubinsky, E.V. et al. "Fe²⁺–Ti⁴⁺ charge transfer in blue sapphire." Gems & Gemology,
56(1), 2020. DOI: 10.5741/gems.56.1.2. [VERIFIED] - Adamo, I. et al. "Maxixe-type color in beryl." Gems & Gemology, 44(3), 2008.
DOI: 10.5741/gems.44.3.214. [VERIFIED] - Dubey, S.; Rai, A.K. et al. "Mineralogical application of LIBS: idiochromatic,
allochromatic and pseudochromatic stones." Journal of Optics, 2022.
DOI: 10.1007/s12596-022-00870-8. [PARTIALLY_SUPPORTED]