Lab Notes Gems & Gemology, Summer 2019, Vol. 55, No. 2

Separation of Kornerupine and Prismatine

Stones identified as kornerupine (left), prismatine (center), and undetermined (right) based on their chemistry.
Figure 1. The two stones on the far left, corresponding to the blue analytical points in figure 2, were classified as kornerupine. The center 20 stones, corresponding to the yellow analytical points in figure 2, were classified as prismatine. For the eight stones on the right, corresponding to the red analytical points, the species could not be determined. The species classifications of all the stones were based on chemistry acquired by LA-ICP-MS. The largest stone is the 11.9 ct brown oval at the top of the center group. The smallest is the light blue cushion stone at the bottom of the center group, weighing 0.93 ct. Photo by Diego Sanchez.

Kornerupine and prismatine form a solid solution series with a general chemical formula of X(□,Mg,Fe)M(Al,Mg,Fe)9T(Si,Al,B)5(O,OH,F)22 (X: cubic site; M: octahedral sites; T: tetrahedral sites). Chemically, they differ mainly in magnesium, iron, aluminum, and fluorine content, with boron content that ranges from 0 to 1 apfu (atoms per formula unit) in one of the tetrahedral sites. When the boron apfu is less than 0.5, the mineral is classified as kornerupine; above 0.5, it is classified as prismatine (E.S. Grew et al., “Prismatine: revalidation for boron-rich compositions in the kornerupine group,” Mineralogical Magazine, Vol. 60, No. 400, 1996, pp. 483–491).

GIA laboratories have recently implemented a method using laser ablation–inductively coupled plasma–mass spectrometry (LA-ICP-MS) to quantify boron concentration in kornerupine and prismatine in order to distinguish the two. The 30 stones shown in figure 1—all thought to be kornerupine—were borrowed from the GIA Museum and tested using a Thermo Fisher iCAP Qc ICP-MS, coupled with an Elemental Scientific Lasers NWR213 laser ablation system. GSD-1G and GSE-1G (U.S. Geological Survey) and NIST 610 were used as external standards. 29Si was used as an internal standard. Three 55 μm circular spots were ablated on the girdle of each stone. The data reduction was a modified version of that used at GIA for tourmaline (Z. Sun et al., “A new method for determining gem tourmaline species by LA-ICP-MS,” Spring 2019 G&G, pp. 2–17). As a result, we have slightly modified the criteria for separating the two to account for analytical error. If the boron apfu is between 0 and 0.45, the stone is classified as kornerupine. Between 0.45 and 0.55 apfu, the species cannot be determined by this method. If the boron apfu is between 0.55 and 1, the stone is classified as prismatine. For three spot analyses, if one spot has a boron apfu between 0.45 and 0.55, the species is undetermined.


Aluminum-boron plot showing the apfu ranges used to identify kornerupine from prismatine.
Figure 2. The Al-B plot shows three sections. When the B (apfu) is between 0 and 0.45, stones are classified as kornerupine (blue spots). When the B (apfu) is between 0.45 and 0.55, the species for the stones (red spots) cannot be determined. When the B (apfu) is between 0.55 and 1.00, stones (yellow spots) are classified as prismatine. For three spot analyses, if one spot has a B (apfu) between 0.45 and 0.55, the species is considered undetermined.

All analytical points were plotted in figure 2 (full chemical results are in appendix 1). There were 2 kornerupine, 20 prismatine, and 8 undetermined species, for a total of 30 tested stones. This result suggests that, contrary to popular belief, prismatine may be a far more common gem species than kornerupine, which is also supported by the analyzed boron concentration of many production stones submitted to GIA laboratories. Interestingly, for the five green stones tested here the species could not be determined, although they slightly favor prismatine chemistry (some of the red spots in figure 2). Aluminum apfu is also a good indicator for separating kornerupine from prismatine. Boron-rich prismatine has an Al apfu around 5.8 to 6.0, compared to boron-poor kornerupine with an Al apfu around 6.6 to 6.8. Extra Al in kornerupine functions as a substitute for B in one of the tetrahedral sites in the crystal lattice.

Kornerupine is an appropriate group name for kornerupine-structure minerals of an unknown boron content or for cases where the species cannot be determined. For extreme cases, such as when boron apfu is close to 0 or 1, Raman spectroscopy could be sufficient to separate kornerupine from prismatine (B. Wopenka et al., “Raman spectroscopic identification of B-free and B-rich kornerupine (prismatine),” American Mineralogist, Vol. 84, No. 4, 1999, pp. 550–554). However, the more accurate separation for the two is determined by LA-ICP-MS analysis because the species definition is based on concentration of a chemical element, boron.

Ziyin Sun is a research associate, Jonathan Muyal is a staff gemologist, Nathan D. Renfro is identification manager, Shane F. McClure is global director of colored stone services, and Aaron C. Palke is a senior research scientist, at GIA in Carlsbad, California.