From the GIA Laboratory: Beryllium in Corundum – The Consequences for Blue Sapphire
January 26, 2007
Figure 1. The smaller stones in this group of sapphires (1.24–8.00 ct) were known to have been diffusion treated with beryllium. The four larger stones were identified as Be-treated during the investigations. Photo by C. D. Mengason.
In 2001−2002, the gem industry witnessed the introduction of beryllium (Be)-diffused corundum in “padparadscha,” yellow, and orange colors. The treated “padparadscha” sapphires could be identified by their yellow-to-orange surface-related color zoning, which was easily seen with diffused light in immersion. The yellow and orange Be-diffused sapphires proved more challenging to identify, often requiring sophisticated means of chemical analysis such as secondary ion mass spectroscopy (SIMS) or laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS) to detect the trace levels of Be that are indicative of this treatment; Be is a light element that is not detectable by typical analytical methods used in gemological laboratories.As both LA-ICP-MS and SIMS require expensive and complicated instrumentation, more cost-effective alternatives were explored. Another technique that has shown promise is laser-induced breakdown spectroscopy (LIBS). Although this technique is not as complicated to use and can detect trace levels of Be, it is not as sensitive or accurate as the other two methods, and it is still relatively expensive compared to standard gemological instruments such as infrared and UV-visible spectrometers.
Figure 2. These unusually shaped and patterned clouds were seen in some of the Be-treated blue sapphires. Photomicrograph by Shane F. McClure; magnified 25×.
Blue Be Diffusion-Treated Sapphire. With the increasing availability of analytical methods that can detect Be, and the analysis of corundum for this element becoming more routine, gemological laboratories are identifying ever-greater numbers of blue sapphires in the international marketplace that have been treated by this technique. In mid-2006, the Gemological Association of All Japan reported that they had identified several blue sapphires that were treated with beryllium (click here for full report). Subsequently, a group of more than 50 Be-diffused blue sapphires were obtained by Vincent Pardieu, then of the AIGS laboratory in Bangkok, and sent to several laboratories for study.The concentration of Be measured in these and other known Be-treated blue sapphires has varied from approximately 4 parts per million (ppm) to over 20 ppm.
The possibility of using Be diffusion to treat blue sapphires was initially investigated in 2002 by a team of GIA and other researchers working with John Emmett (Crystal Chemistry, Brush Prairie, Washington). As illustrated by Emmett et al. (“Beryllium diffusion of ruby and sapphire,” Summer 2003 Gems & Gemology, pp. 84–135), Be diffusion and subsequent heating can be used to lighten the color of dark blue sapphires.
Figure 3. Some Be-treated blue sapphires show rows and strings of circular inclusions that may resemble corkscrews. Photomicrograph by Shane F. McClure; magnified 25×.
In light of the emergence of so many Be-treated blue sapphires, GIA initiated a broad-ranging project to determine the extent of Be-treated blue sapphire in the colored stone market. Laboratory staff used the GIA Laboratory's LA-ICP-MS and standard gemological techniques to analyze more than 500 sapphires from the New York trade; the majority were blue sapphires that had been heat treated, but the group also included some green, yellow, orange, and pink sapphires. Analysis of these 500 samples revealed four blue sapphires that showed consistently high levels of Be (over 3 ppm) throughout, proving they were diffusion treated (figure 1). Be diffusion−treated blue sapphires have indeed entered the marketplace in the U.S.Certain inclusion scenes noted in the Be-diffused blue sapphires appear to be indicative of this treatment. These include unusually shaped and patterned clouds (figure 2), rows and strings of circular inclusions that sometimes resemble corkscrews (figure 3), and spherical clouds surrounding highly altered crystals (figure 4). However, not all of the Be-diffused blue sapphires contained these inclusion scenes, and some contained no inclusions at all. In addition, Be was detected in some blue sapphires that did not appear to have been diffusion treated.
Figure 4. Also characteristic of Be-treated sapphires are these spherical clouds surrounding highly altered crystals. Photomicrograph by Shane F. McClure; magnified 20×.
Beryllium in Natural Sapphire? Pure corundum consists only of aluminum oxide (Al2O3), but in nature this mineral always contains minor-to-trace amounts of other elements as impurities. For example, natural untreated blue sapphires may contain roughly 1,000–10,000 ppm of iron and titanium, depending on the type of deposit in which they formed. Yellow sapphires normally contain tens of ppm of magnesium. However, beryllium has not been previously documented in untreated corundum. Therefore, it was assumed that the presence of any Be in corundum was indicative of beryllium treatment.During its analysis of the 500 stones mentioned above, the Laboratory encountered 17 blue sapphires with Be concentrations that were highly variable from one area of the stone to the next. This is not consistent with the results obtained on known treated stones. One of these anomalies was a 2.63 ct. round brilliant-cut blue sapphire that showed clear evidence of heat treatment.
Initial LA-ICP-MS analyses of two spots on the girdle of this sapphire showed 3.5 ppm Be in one that coincided with the edge of an angular milky cloud within the stone, whereas the other spot (with no Be detected) was well away from any inclusions. Further analyses revealed up to 13.3 ppm Be in the cloud, and essentially no Be elsewhere in the stone. The cloud also showed consistently elevated concentrations of niobium (Nb) and tantalum (Ta).
Figure 5. The clouds that are associated with traces of Be often are extremely fine grained and appear milky white in darkfield illumination (left). However, they usually have a brownish appearance when viewed in transmitted light (right). The planar structure of this cloud can also be seen in the image on the right. Photomicrographs by Shane F. McClure; magnified 20×.
The cloud in question had a distinctive appearance: a very fine particle size (so small that the individual particles could not be resolved with a standard gemological microscope), planar structures, a milky white appearance in darkfield illumination, and a brownish appearance in transmitted light (figure 5). It is important to note, however, that whereas clouds with a similar appearance were seen in other samples investigated, not all of these clouds contained Be. Due to the extremely fine size of the particles in these clouds, the phase(s) present have not yet been conclusively identified.In addition to these 17 stones, the Laboratory has detected traces of Be (along with elevated levels of Ta and Nb) associated with such clouds in blue sapphires submitted to the Laboratory for identification reports. A few of these sapphires showed no signs of heat treatment. This was also the case for some samples collected by GIA researchers – and others obtained by AGTA’s Richard Hughes – at the deposits in Ilakaka, Madagascar.
For the heated samples examined, there does not appear to be a likely scenario in which Be diffusion treatment could produce such a dramatically uneven distribution of elements. This, and the detection of Be in sapphires with no evidence of heating, conclusively shows that trace amounts of Be do occur naturally in corundum, at least in conjunction with foreign mineral inclusions (forming the clouds). In addition to ongoing research to better understand this phenomenon and the nature of the clouds, GIA is collaborating on a project headed by Vincent Pardieu (now with the Gübelin Gem Lab in Lucerne, Switzerland) to investigate the effect of heat treatment on the Be that is naturally present in such sapphires. Careful chemical mapping by LA-ICP-MS is being conducted at the GIA Research facility in Bangkok as part of this project.
About the Authors
This report was prepared by Andy Shen, Shane McClure, and Mike Breeding of the GIA Laboratory in Carlsbad; Ken Scarratt of GIA Research in Bangkok; Wuyi Wang and Christopher Smith of the GIA Laboratory in New York (the latter now with the American Gemological Laboratories in New York); and James Shigley of GIA Research in Carlsbad.
