Gem News International Gems & Gemology, Spring 2016, Vol. 52, No. 1

Red Cordierite from Madagascar


Faceted Cordierite
Figure 1. These three gems, weighing (from left) 4.17, 1.28, and 1.28 ct, were fashioned from the same piece of cordierite rough; red is a new color for this gem. The specimen originated from a recently explored deposit in southeast Madagascar. Photo by Benjamin Rondeau.

At the Mineral Expo show in Paris in early December 2015, we procured a large piece of dark red cobbed rough presented as cordierite. This would be the first documented occurrence of red cordierite. The slightly fractured 3 cm rough offered a magnificent deep red color when examined using transmitted light from a Maglite illuminator. The identity of the rough as cordierite was confirmed using Fourier-transform Raman (FTR) spectrometry. The spectrum obtained showed excellent agreement with a reference spectrum for cordierite from Madagascar in the RRUFF database (http://rruff.info).

Author TP confirmed that the piece we examined was from the Iakora district, Fianarantsoa province, in southeast Madagascar. Cordierite has long been known in this general area (A. Lacroix, Minéralogie de Madagascar, A. Challamel, Paris, 1922). There is no mechanized mining. According to a miner’s sketch obtained locally by one of the authors, red cordierite occurs in a “vein” (probably a layer) associated with orange and blue cordierite, parallel to other layers containing kornerupine or blue spinel. The cordierite vein is about 1.50 m thick and wide and is apparently contained in a kornerupine layer. Cordierite is usually found in a metamorphic environment, but the layered appearance, with each band containing specific minerals in thicknesses of about 1 m, is reminiscent of metasomatic deposits. The association of cordierite with kornerupine and spinel is already known; blue gem-quality cordierite is found in such a metasomatic reaction zone in Kenya (C. Simonet, pers. comm., 2015).

To determine the material’s gemological properties, three stones were fashioned by Jacques Le Quéré (Auray, France): a 2.69 ct modified rectangle; a 1.28 ct flat cabochon, cut to obtain a lighter color; and a 4.17 ct parallelepiped, faceted based on the optical directions to best show its trichroism (figure 1). The RIs, measured with a Rayner LED sodium-equivalent lamp, ranged from 1.527 to 1.541 (nx = 1.528–1.530, ny = 1.532–1.533, and nz = 1.538–1.540). Interestingly, this crystal presented a biaxial positive character, whereas gem cordierite is typically biaxial negative.

The trichroism was very strong, as expected with cordierite. With a standard dichroscope, the colors ranged from deep red to orange to grayish brown with a hint of blue or purple (figure 2), this last color becoming black at a thickness of about 1 cm (figure 3). Hydrostatic SG measured 2.548–2.554. These properties are consistent with cordierite, though SG was at the lower end of the range.

Pleochroism in Cordierite
Figure 2. Pleochroism of the oval cabochon observed in polarized light in the two orthogonal directions. On this thinner sample (about 3 mm thick), the desaturated grayish brown color with a hint of blue or purple is best observed. The stone measures 9.85 × 6.75 mm. Photos by Emmanuel Fritsch.
Trichroism in cordierite
Figure 3. Trichroism of the 9.45 × 6.95 × 5.28 mm cordierite parallelepiped observed in polarized light in the three orthogonal directions to represent the three “extreme” colors of the material’s pleochroism: deep red, black (both seen with a thickness of 9.45 mm), and orange (seen with a thickness of 6.95 mm). Photos by Emmanuel Fritsch.

As the rough was sawed, we noticed that the vivid red transparent color of the thick crystal turned to orange with reduced thickness. This is the definition of dichromatism, also known in gemology as the Usambara effect, a variation of hue that is affected by the optical path length.

The UV-visible absorption spectrum was measured on a Unicam UV4. When the material was a few millimeters to a centimeter thick, in the direction where red is observed there is an extremely strong absorption in the violet and blue areas; this band leads to the red color. The absorption decreases abruptly from about 550 to 685 nm, with only a weak broad band from about 710 to 800 nm. This leaves a steep transmission window in the red, which is consistent with the color observed. The shape of the absorption and a rapid decrease in the orange-yellow also explains the Usambara effect. Through a few millimeters, the perceived color is orange; through a centimeter or more, it is deep red.

Emmanuel Fritsch (CNRS Team 6502) is a professor of physics at the Institut des Matériaux Jean Rouxel, at the University of Nantes, France. Benjamin Rondeau (CNRS Team 6112) is an assistant professor at the Laboratory of Planetology and Geodynamics in Nantes, France. Thierry Peclet is a gem dealer from Manakara, Madagascar. Patrick Lefebvre is an independent gemologist from Aix en Provence, France.Yves Lulzac is a retired geological engineer from Nantes, France.