Sapphire colour
Discovering what had to be done to the crystal corundum (Al_2 O_3) to make the deep blue colour of sapphire was not an easy task. The addition of a percent or so of chromium (Cr^3+ replacing Al^3+) was known to create the purple-red of ruby; but what could be added to make blue?. Professor Auguste Victor Louis Verneuil (1856-1913), the French chemist who had grown ruby successfully in his laboratory before he announced his success in 1902, set out to make sapphire. To him, the obvious additive was cobalt - the origin of blue in many materials. This attempt failed as did his following trials with a whole list of impurities. He was forced to analyse the chemical composition of a number of natural sapphires to find that all contained very small fractions of iron and titanium oxides: less than 0.1%. However, the addition of neither of these oxides on their own produced the desired colour. Adding them together though did the trick! But it took another sixty years to find out how this worked.
Both iron and titanium can replace aluminium in the Al_2 O_3 structure of corundum. If Fe^2+ and Ti^4+ find themselves in neighbouring sites. A photon of the right energy (2.11 eV ~ 588nm) can cause the iron to donate an electron to titanium to produce Fe^3+ and Ti^3+. Since the crystal stucture allows adjacent Al sites in a different direction with a slighly larger spacing, there is another (weaker) absorption at a somewhat longer wavelength. These two absorptions can be separated by looking though a polarizing filter in different orientations to reveal the blue/blue-green dichroism of sapphire. These charge transfer transitions are hundreds or thousands of times stronger than the ligand field transitions in ruby, which explains why such small quantities of impurity iron and titanium are needed for sapphire colouration.
The narrower absorption band at 452nm is due to a transition in trivalent iron (Fe^3+) which, similar to transitions in ruby, is a result of the energy-level splitting by the crystal (ligand) field: the regular electic field produced by the charges on the surrounding ions. This can be accompanied by weaker lines at 460 and 470nm - as seen in the lower two of these spectra. Fe^3+ also produces a strong double absorption in the deep blue at 376 and 389nm which are only marginally visible in these spectra due to the low blue signal from the lamp I use. The very broad, diffuse absorption in the red and far-red is thought to be due to divalent iron (Fe^2+ , see Lehmann & Harder, 1970, Am. Min. 55, 98): these are strong in blue sapphires.
The three sapphires shown here were acquired by me in very different ways. The rectangular cut stone on the left was bought in a junk shop in Eastbourne, UK for 50 pence in around 1970. Thinking it was synthetic or a glass imitation (as did the seller), I was surprised to find when I looked at it with my visual spectroscope that it appeared to be a natural, iron-rich stone of the type often found in Australia. The beautiful and perfect hexagonal crystal (10.3 mm high) in the middle came from a small-holding in Northern New South Wales in 1991. We had called in after seeing a sign by the road advertising gemstones. The farmer had, apparently, used his tractor do dig and prepare a melon patch by the house in order to bring in some extra cash. Amongst the debris, he and his family found sapphires and have made a business out of it. He showed me this crystal as one of his prize findings but refused to sell until I produced my Zeiss loupe to examine it. Taking me to be an expert (!) he offered it for 50 Oz dollars. I took it! The synthetic on the right was supplied by the Gemmological Association of Great Britain when I first developed an interest in the spectroscopy of coloured gems in the late 1960s.
Sapphire colour
Discovering what had to be done to the crystal corundum (Al_2 O_3) to make the deep blue colour of sapphire was not an easy task. The addition of a percent or so of chromium (Cr^3+ replacing Al^3+) was known to create the purple-red of ruby; but what could be added to make blue?. Professor Auguste Victor Louis Verneuil (1856-1913), the French chemist who had grown ruby successfully in his laboratory before he announced his success in 1902, set out to make sapphire. To him, the obvious additive was cobalt - the origin of blue in many materials. This attempt failed as did his following trials with a whole list of impurities. He was forced to analyse the chemical composition of a number of natural sapphires to find that all contained very small fractions of iron and titanium oxides: less than 0.1%. However, the addition of neither of these oxides on their own produced the desired colour. Adding them together though did the trick! But it took another sixty years to find out how this worked.
Both iron and titanium can replace aluminium in the Al_2 O_3 structure of corundum. If Fe^2+ and Ti^4+ find themselves in neighbouring sites. A photon of the right energy (2.11 eV ~ 588nm) can cause the iron to donate an electron to titanium to produce Fe^3+ and Ti^3+. Since the crystal stucture allows adjacent Al sites in a different direction with a slighly larger spacing, there is another (weaker) absorption at a somewhat longer wavelength. These two absorptions can be separated by looking though a polarizing filter in different orientations to reveal the blue/blue-green dichroism of sapphire. These charge transfer transitions are hundreds or thousands of times stronger than the ligand field transitions in ruby, which explains why such small quantities of impurity iron and titanium are needed for sapphire colouration.
The narrower absorption band at 452nm is due to a transition in trivalent iron (Fe^3+) which, similar to transitions in ruby, is a result of the energy-level splitting by the crystal (ligand) field: the regular electic field produced by the charges on the surrounding ions. This can be accompanied by weaker lines at 460 and 470nm - as seen in the lower two of these spectra. Fe^3+ also produces a strong double absorption in the deep blue at 376 and 389nm which are only marginally visible in these spectra due to the low blue signal from the lamp I use. The very broad, diffuse absorption in the red and far-red is thought to be due to divalent iron (Fe^2+ , see Lehmann & Harder, 1970, Am. Min. 55, 98): these are strong in blue sapphires.
The three sapphires shown here were acquired by me in very different ways. The rectangular cut stone on the left was bought in a junk shop in Eastbourne, UK for 50 pence in around 1970. Thinking it was synthetic or a glass imitation (as did the seller), I was surprised to find when I looked at it with my visual spectroscope that it appeared to be a natural, iron-rich stone of the type often found in Australia. The beautiful and perfect hexagonal crystal (10.3 mm high) in the middle came from a small-holding in Northern New South Wales in 1991. We had called in after seeing a sign by the road advertising gemstones. The farmer had, apparently, used his tractor do dig and prepare a melon patch by the house in order to bring in some extra cash. Amongst the debris, he and his family found sapphires and have made a business out of it. He showed me this crystal as one of his prize findings but refused to sell until I produced my Zeiss loupe to examine it. Taking me to be an expert (!) he offered it for 50 Oz dollars. I took it! The synthetic on the right was supplied by the Gemmological Association of Great Britain when I first developed an interest in the spectroscopy of coloured gems in the late 1960s.