Optical properties of Corundum

 

Effects of Light

Refraction

Anthony et al. (Handbook of Mineralogy, 1997, the most reliable source) assign the values 1.767–1.772 to the ordinary ray and 1.759–1.763 to the extraordinary ray. Corundum is uniaxial negative. Gem materials for the

 

most part accord with these values and show a birefringence of 0.008. Both ruby and sapphire show distinct pleochroism: blue sapphire nor- mally shows blue-green to yellow-green in the extraordinary ray and pale to deep blue in the ordinary ray, ruby showing orange-red in the extraordinary ray and bluish red in the ordinary ray. When possible, both rubies and sapphires should be cut with the c-axis at right angles to the table facet, to display the best colour.

 

Lustre

The lustre of ruby and sapphire varies from sub-adamantine to vitreous and is pearly on parting surfaces.

 

Dispersion

The dispersion of the stones is only 0.018 so little fire is perceptible.

 

Absorption Spectra

In describing absorption spectra throughout the book details have been checked against B.W. Anderson and C.J. Payne, The Spectroscope and Gemmology, edited by R Keith Mitchell, London, NAG Press, 1998; ISBN 071980261X. This is a compilation of papers by the authors in the journal The Gemmologist between 1954 and 1957, with additional notes by Mitchell who also re-drew some of the diagrams. The book should be available to all engaged in gem testing. Details and techniques of absorption spectrum investigation are also found in this book and in Gemmology (Read), 3rd edition, 2005 (ISBN 0750664495).

Ruby shows the classic chromium absorption spectrum though corun- dum of any colour which contains significant chromium may show some of its elements. Two lines close together in the red form a doublet which, depending upon the lighting conditions, may be coloured (emission dou- blet) or dark (absorption doublet). Emission lines occurring elsewhere in the spectrum will be found to have emanated from the room lighting, never from the specimen under examination. I have not recorded the fainter absorption lines but only those which are prominent in normal conditions of observation or which are significant in diagnosis.

The lines are situated at 694 nm and are followed by absorption lines in the orange at 668 and 659.2 nm. A broad absorption band extending from 610 to 500 nm obscures the green (this band, present though with slight differences of position in all materials coloured by chromium is the major cause of the residual colour (the colour seen by the unaided observer).

 

Three narrow lines in the blue, two of them equally strong, at 476.5 and 468.5 nm are accompanied by another weaker line at 475 nm and by a general absorption of the violet. Red spinel, easily confused with ruby when testing is undertaken in haste or under unsatisfactory conditions, does not show the strong, close lines in the blue. The absorption spectrum of synthetic ruby is identical though all elements may be deceptively easy to see. A strong emission doublet and other elements suggest that further tests may well be needed. Readers should also remember that elements of the absorption/emission spectrum may well be stronger in different directions and specimens should always be tested in several directions (back-to-front will give the same results as front-to-back!).

Anderson and Payne (1998) remind us that a chrome-rich ruby need not always be synthetic as some Myanmar and East African stones have a higher-than-usual chromium content. Furthermore, it has been shown that in deep red rubies the absorption beginning in the violet extends, in fact, not only close to the absorption line at 468.5 nm but also extends, after more transmission, into the near UV. The final absorption region begins in natural rubies between 300 and 290 nm: in synthetic rubies this region does not begin until about 270 nm (or not before).

 

Fluorescence

A red colour, seen through a red filter when a ruby is illuminated only by monochromatic blue light (the crossed filter effect), proves the presence of chromium. As might be expected, the effect is particularly spectacular with synthetic rubies since they contain no iron to weaken the effect. Iron is usually included in natural ruby and if present in sufficient quantities inhibits fluorescence, which is the transmission of energies of certain wavelengths when a specimen is irradiated by energies of a lower wave- length (but higher energy).

Despite the occasional claim, the response of ruby to UV and X-radiation should not be relied on either for identification, place or type of origin of a suspected ruby, though any effects seen may make very useful confirmatory tests. In ruby the excitation of a chromium ion to a higher energy level produces red light from the emission doublet described above. The effect produced by the crossed filter experiment, ruby’s response to the ultraviolet radiations present in daylight and to other radiations of higher energy than these and the mechanism operat- ing are all lucidly described by Kurt Nassau in the two editions of The Physics and Chemistry of Colour (second edition, 2000; ISBN 0471391069). Ruby fluoresces most effectively if no iron is present; this can only be

 

guaranteed when the specimen is man-made so that a strong crimson response to UV or to X-rays requires the investigator to make additional tests. Rubies, whatever the origin, can be separated from red garnet or tourmaline if a UV source is available. While most Myanmar rubies commonly fluoresce more vividly than most (relatively iron-rich) rubies from Thailand, place of origin determination needs a study of inclusions as well as or more than fluorescence. Purple and mauve sapphires will usually respond with something of a red colour to activating sources.

 

Absorption Spectra of Other Colours of Corundum

Iron causes the colours of natural non-ruby corundum occurring, like chromium, as a replacement for aluminium. This is trivalent iron and less than 1% may be enough to cause colour. What might be thought of as the classic iron spectrum in corundum is that shown by green sapphire in which no other element influences the colour. Here there are three bands, at 471, 460 and 450 nm in the blue region of the spec- trum. The band at 450 nm is the strongest and nearly coalesces with its neighbour at 460 nm, the next strongest. The 471 nm band is distinctly separate from the other two and is the weakest of the three. The same spectrum can be seen in natural but not synthetic flame-fusion blue sapphires but specimens from different locations show variations. In blue sapphires from Australia all three bands are strong and easily seen; on the other hand, Sri Lanka blue sapphires show only a faint 450 nm band. Anderson and Payne tell us that blue sapphires from Myanmar, Kashmir, Thailand and Montana show a clear 450 nm band accompanied by a smudge on the long-wave side – the only trace of the other two bands.

These bands belong to the ordinary ray and observations may be made

more easily if the specimen is illuminated by light which has passed through a blue filter (copper sulphate solution is still used). If a Polaroid filter is available it can be useful in finding the strongest effect direction. Some deep blue natural sapphires show a rather vague broad absorption band near 585 nm.

Yellow sapphires may also show the three ferric iron bands. Sapphires from Australia, Montana and Thailand show them most clearly, while yellow sapphire from Sri Lanka may show only a faint 450 nm band or none at all (though they do give a characteristic apricot-yellow fluores- cence under both types of UV). Generally speaking, synthetic flame- fusion sapphires neither fluoresce nor show any absorption bands – but exceptions have been recorded.

In some cases absorption bands in the UV may give a clue not only to the identity of an unknown yellow stone but also to its possible origin.

 

Photography needs to be used (a photograph will often show otherwise undetectable features in a synthetic stone, in particular) to find absorp- tion bands in the near-UV at approximately 379 and 364 nm. These bands have been noted in Sri Lankan yellow sapphires when the 450 nm band is scarcely pereptible. The UV bands are usually obscured in sap- phires which show the ‘450 complex’ strongly. Synthetic sapphires in general, certainly the nearly ubiquitous flame-fusion products, show absorption neither in the visible nor in the UV regions. Anderson and Payne remark on the oddness of this since iron (with titanium) needs to be present in any sapphire to give a blue colour. Some observers have noted a weak 451 nm absorption band in some Verneuil and Chatham flux-grown sapphires.