Violet-Colored Diopside from 
Southern Baffin Island, Nunavut, Canada

Christopher D.K. Herd and Ronald C. Peterson
Department of Geological Sciences, Queen's University
Kingston, Ontario, K7L 3N6

George R. Rossman
Division of Geological and Planetary Sciences
California Institute of Technology
Pasadena, California, 91125, U.S.A.

ABSTRACT: Diopside with an unusual violet color in hand specimen from south Baffin Island, Nunavut, Canada, is found in calc-silicate lenses, associated with mirialite, pargasite, phlogopite, calcite, apatite, titanite, talc, chlorite, plagioclase and quartz. It occurs as massive aggregates of roughly equant grains. Basal parting is evident and pyroxene cleavage is subtle. Indices of refraction are na=1.670(1), nb=1.675(1), and ng=1.695(1), 2VZ is equal to 57.6(5)° at 589 nm. Pleochroism is weak to nonexistent. Dcalc = 3.300 g/cm3. Cell dimensions determined from powder X-ray diffraction are a = 9.730(4) Å, b = 8.873(3) Å, c = 5.275(2) Å, b 109.95(3)°. A single-crystal X-ray structural refinement was performed to determine bond lengths and angles. The empirical formula, based on microprobe analysis and absorption spectroscopy indicating 0.30% H2O in the structure, is (Ca0.96Na0.04)(Mg0.86Al0.06 Fe2+0.05Ti4+0.02) (Si1.89Al0.11)O5.93OH0.07. The unusual colour is due to intervalence charge transfer between Fe2+ and Ti4+ at the M1 site, and is observed due to the low overall concentration of Fe in the diopside. Cathodoluminescence indicates that Mn2+ is present in the M1 site. Absorption spectroscopy demonstrates that Mn3+ does not contribute to the violet colour.

The field location

The diopside occurs in meter size calc-silicate lenses within a marble (a grey irregular band in the upper center of the photo) that occurs about 1 km west of Crooks' Inlet, on the south coast of Baffin Island (red cross on map).

Image of the violet diopside absorption spectrum

The violet diopside spectrum (4K); presented normalized to 1.0 mm thickness.

These spectra are taken in the directions of the alpha, beta and gamma indices of refraction: gamma-to-c = 42 degrees, beta = b-axis. Two slabs were used: (010) for E||a and E||g; and a slab perpendicular to c for the b = E||b-axis.

The violet color is due to the slightly higher transmission (less absorbance) near 700 nm (red) than at 450 nm (blue). The band near 1050 nm in the b-direction is due to Fe2+ in the M2 site. The band in a near 2300 nm is also from Fe2+ in the M2 site. In the a and g directions, there is a pair of M1-site Fe2+ bands near 950 and 1100 nm.
 

Is the color due to Mn3+?

Manganese minerals commonly have a similar color. However, the 480 nm band does not come from the Mn. Here is a comparison of the spectrum (5K) of the Baffin Island sample with a synthetic diopside doped with Mn3+.  It is obvious that the standard diopside pattern does not match that of the violet diopsidee.
 

Is the color due to Cr3+?

Does chromium contribute to the color? [remember, rubies are red.]
This comparison (4K) of the Baffin Island diopside to a chromian diopside from Russia (both in the E||b direction) shows that Cr3+ is not a factor.  Furthmore, the chemical analysis indicates that the Cr content in the diopside is below the detection limits. If chromium were present in minor concentrations we would expect to see an absorption band from chromium centered near 650 nm in the spectrum of the Baffin Island sample.  This clearly is not present.
 

What is the color due to?

The comparison to other pyroxenes with the Fe2+ - Ti4+  intervalence charge transfer band is more interesting.  Here is a comparison (4K) of the proxene (fassaite) from the Angra dos Reis meteorite to the Baffin Island sample.  Here, the match is much closer.  There is an appreciable amount of Ti in the microprobe analysis: 0.83 TiO2. This, together with the iron (which we know is, in part, Fe2+ based on the optical spectrum) can cause bands in this region.

In the a* direction (5K) (E vector vibrating perpendicular to both the b- and c-axes) there is essentially no aborption from the 480 nm band.  This is an important point of comparison to the ADOR meteorite and speaks to the intervalence charge transfer.  In this direction, there are not any adjacent cation sites, so IVCT can not occur.  It should reach a maximum in the E||c direction if it occurs along the M1-M1 chains, and in the E||b direction if it is along the M1-M2 cation pairs.

Data Files as ASCII x,y data   alpha 0.404 mm sample (77K file);   beta 1.101 mm sample (79K file);   gamma 0.404 mm sample (76K file);   a* direction 1.101 mm (21K file).