|
Special Issue
A new corundum treatment has recently emerged that, in a very short time, has created a storm of controversy. In December 2001, these authors were told of a new type of heat treatment being done in Thailand that could take the abundant light pink sapphire from Madagascar and change it to a beautiful “padparadscha” color. The only other information available from the treaters at the time was that the stones were being heated in an oxygen atmosphere and they believed the process was new and revolutionary.
We were subsequently told that in addition to the pink Madagascar material, green sapphire from Songea, Tanzania, and even old stocks of Thai ruby were involved. When, in early January, we had our first opportunity to examine some of these stones, it was clear that more than one type of corundum was being treated by this process. By that time, gemologists had noted that many of these stones were different from the heat-treated stones we typically examine, which gave rise to the belief that there was more to the “heat” treatment than was being disclosed. The following is a preliminary report of GIA’s examination of 48 stones treated by this purportedly new process.
Examination of Treated Samples
Pala International of Fallbrook, California, kindly loaned GIA 27 of these treated sapphires, all faceted, which they had recently obtained in Thailand. Shortly afterward, we received from D. Swarovski and Co. of Austria an additional 16 treated sapphires—10 faceted and 6 preforms. They reported that the 10 faceted samples were provided by Metee Jungsanguansith of World Sapphire in Chanthaburi, Thailand, who (according to Swarovski) is one of the key people involved in this process. At the same time, five corundums that appeared to have been treated in a similar manner were submitted to our New York laboratory for identification. This provided us with a total of 48 samples on which to collect data for this preliminary study; 6 were orange, 35 were pinkish orange to orangy pink, and 7 were orangy red to red-orange (see, e.g., figure 1). The samples ranged from 0.34 to 3.53 ct, with most between 1 and 2 ct. Four of the six orange stones were purported to originate from Songea, while the origin of the orangy red to red-orange stones is unclear at this time. All of the other samples were purported to be from Madagascar.
The gemological properties of these stones were, for the most part, typical for corundums of these colors. There were only minor variances from the norm. For example, the orangy red stones had slightly elevated refractive indices (1.766-1.774) and an abnormally strong iron absorption in the desk-model spectroscope. The R.I. measurements were consistent when measured on several facets of the same stone. With magnification, most of the stones showed evidence of high-temperature heat treatment in the form of altered and heat-damaged inclusions.
Soon after reports of this new treatment appeared, it was noted that the orange color did not go all the way through many of these stones. In fact, Richard Hughes posted a photo on Pala International’s Web site that clearly showed an orange layer of color following the faceted shape of the stone. Similar photos were posted on the AGTA Gemological Testing Center website. Our samples were extremely variable in this regard, but most showed evidence of a surface-related color when they were immersed in methylene iodide and examined over diffused light with a microscope. In fact, 36 of the stones showed a distinct orange color layer over a pink central core (see, e.g., figure 2). In many of these, the layer was only near the surface, penetrating just 10%-20% of the stone, with an appearance similar to that produced by diffusion treatment in blue sapphires (see, e.g., R. E. Kane et al., “The identification of blue diffusion-treated sapphires,” Gems & Gemology, Summer 1990, p. 117, figure 2). However, the color layer in some of the stones extended deeper than any diffusion treatment we have seen in the past. In fact, in a few samples the layer penetrated more than 50% of the distance to the center of the stone (figure 3). The extent of the orange color penetration did not appear to be related to the size of the samples. All of the samples we examined were treated after they had been cut at least into preforms, since the color zoning followed the shape of the facet arrangement.
The coloration in three of the orangy red to red-orange stones appeared dramatically different from the other samples when viewed with immersion. One was almost entirely orange (approximately 80% penetration of the orange zone), with only a small pink core and an adjacent red hexagonal area in the center. This pink core had small blue spots within it. Both of the other two stones showed a deep orangy red color layer that extended from the table past the girdle plane, encompassing the entire crown area. A light orange layer followed the surface of the pavilion, and the central core was purplish pink, again with small blue spots (figure 4).
Five of the pinkish orange to orangy pink samples displayed uneven coloration, but either did not have a surface-related color zone or the zone was very indistinct. One showed alternating bands of orange and pink in a hexagonal pattern (figure 5). Another was mostly orange with small areas of yellow. These stones were so different from the other samples that it would be logical to assume that the treatment method was also somewhat different.
Only four of the 48 stones, all orange, had the same color throughout. These four were represented as being from Songea, with the starting material said to be green.
Although the near-surface, facet-related orange layer in many of these stones is very similar to the type of color zoning seen previously in diffusion-treated stones, we observed none of the other identifying characteristics of diffusion: There was no higher relief in immersion, patchy surface color, or concentration of color at facet junctions.
We performed a quick color stability test on one pinkish orange sample by heating the stone in the flame of an alcohol lamp and then comparing it to a control sample of the same original color. We observed no change in the color appearance of the test sample.
Advanced Testing
All surface diffusion-treated stones we have studied to date showed an abnormal concentration of the color-producing metallic ion used in the diffusion process, such as titanium or chromium. Qualitative chemical analysis of the surface of five of the samples with energy-dispersive X-ray fluorescence spectroscopy revealed only a slightly elevated – but not extraordinary – iron content. Among the trace elements detected were titanium, iron, chromium, and gallium.
To further investigate the composition of the treated corundums, we sawed in half three preforms that represented varying depths of the orange layer (figure 6). The sawn surfaces were then polished and analyzed by electron microprobe at Rutgers University, New Brunswick, New Jersey. For each sample, five to eight spots were quantitatively analyzed along a line from center to rim. In all analyses, the trace elements were close to or slightly above the detection limits of the instrument (about 0.015 wt.% for MgO, TiO2, and Cr2O3; and about 0.005 wt.% for SiO2, CaO, NiO, CoO, and Ga2O3). Iron varied from 0.20 to 0.28 wt.% FeO, but no systematic variations from center to rim were detected in any sample. However, it is possible that the color differences in the samples may be due to very low concentrations of trace elements that are not measurable by this technique.
Discussion and Conclusion
Although we believe that the color in these stones is being produced by a form of high-temperature heat treatment, we do not currently know the exact treatment method(s) being used. However, the fact that the coloration may not extend through the entire stone is a significant issue for the gem trade. When a component of a stone’s color is confined to a relatively shallow surface layer, there is always a danger that the color of the stone may change if it has to be recut. For example, according to William Larson of Pala International, a 2.69 ct emerald cut examined for this study originally had a large “window.” When the stone was recut in an attempt to reduce the window, the color became more pink than orange. As can be seen figure 7, most of the orange color layer was removed from the pavilion during recutting.
According to the Academic Press Dictionary of Science and Technology (1992), one definition of diffusion is simply “the movement of individual atoms through a crystal lattice.” We cannot dispute that some kind of diffusion is taking place in these stones, since we know of no other mechanism that can create these distinct, surface-related color layers in corundum. At this time, however, we question whether this new treatment should be classified in the same category as the surface diffusion-treated stones we reported on in the past. If only for the sake of the numerous dissimilarities between the two, it may be necessary to separate them in some manner. We do not believe we have enough information at this time to make that judgment.
John Emmett, president of Crystal Chemistry in Brush Prairie, Washington, has suggested that the diffusion of certain light elements, such as magnesium, into the structure of corundum could be the cause of these color layers. Magnesium can react with traces of titanium to create color centers that give rise to a yellow to orange color in corundum. According to Dr. Emmett, magnesium diffuses into corundum easily at high temperatures, and even extremely low concentrations (such as 20-30 ppm) can have a significant effect on color. Such small amounts of Mg could be present as contamination in the powders used to surround these stones during heat treatment. Unfortunately, such low concentrations of magnesium are below the detection limits of the instrumentation commonly used by gemological laboratories. Therefore, we will continue to pursue other, more sensitive techniques.
Regardless of the actual mechanism involved, proper disclosure of this treatment is essential. Such disclosure is complicated in this case by the extreme variability in the depth of penetration of the orange color, which in many cases would not be considered near surface. Until more information is available, GIA Gem Trade Laboratory identification reports will note in the comments section the nature of the color zoning in these stones, using notations such as “The orange color of this stone is confined to a near-surface layer” when appropriate. As we learn more information on this new treatment process, we will pass it along to you, both in the GIA Insider and Gems & Gemology.
Editor’s Note: This report was prepared by Shane McClure, Director of West Coast Identification Services, GIA Gem Trade Laboratory; Tom Moses, Vice President of Identification Services, GIA Gem Trade Laboratory; John I. Koivula, Chief Research Gemologist, GIA Gem Trade Laboratory; and Wuyi Wang, Research Scientist, GIA Gem Trade Laboratory.
|
