Surface Diffusion

Since millennia heat treatment of minerals has been part of human history. Application of this on siliceous rocks were found in the Indus Valley and Mesopotamia. And the oldest bibliographical reference of heat treatment in corundum can be found in a book: Al-Jawāhir wa ma Shabhala (Gems and the Likes), about 850-870 AD, written by al-Kindi (Abū Yūsuf Yāqūb bin Ishāq ibn as-Sabbah ibn Ōmran ibn Ismāil al-Kindi3 , a great Hellenizing Arab philosopher and scientist (801 - 873 AD) (Fig 1). (Notari F. et al., 2018, The corundum's heat treatment at moderate temperature-Incolor-InColor, Spring. 42, pp 14-23)

Heat treatment in sapphires have been known in Ceylon (Serandib at that time) for at least 1,500 to 1,800 years ago. (Notari F. et al., 2018, The corundum's heat treatment at moderate temperature-Incolor-InColor, Spring. 42, pp 14-23)

Heating corundum in oxidizing environment referred as Batā kubalā or reducing environment in Ceylon, a good enough process to eliminate the unwanted sub-hue or enhance or reduce the color of an otherwise dark toned sapphire. In corundum, heating has been used to induce an optical phenomena or removal of the same, for color transformation process and to enhance clarity, which is generally acceptable by the trade.

Pure corundum is composed of aluminum and oxygen and hence it is colorless. The different colors seen in corundum is a direct result of impurities/trace elements or color center in the crystal structure. (Heat Treating Rock Creek Sapphires, By John L. Emmett and Troy R. Douthit, G&G Winter 1993). Heating may not always produce desired colors and in the absence of impurities/trace elements or color center, color can be induced by diffusion of aliovalent ions.

Surface diffusion around the 1980’s became prevalent and trade witnessed a large quantity of diffusion treated sapphires with vibrant blue color. Colorless sapphires that failed the traditional heating process were commonly used as the starting material. The first experiments started with titanium, one of the color causing trace element in natural sapphire that is blue. Titanium is a heavy metal with atomic weight of 47.867u, and when diffused with titanium the penetration of color was less than 0.5mm into the surface and hence this treatment was carried out on faceted gems. Although the treatment is stable, but the cutters had to be careful during repolishing or recutting since it could remove the shallow blue color on the facets and reveals the true color of Fig 2: Color concentrations along facets junctions under immersion. the stone. Identification was pretty straight forward when diffused with titanium. One could observe color concentrations along facet junctions, girdle edge or the keel. (Fig 2) Diffusion treatment of corundum: a history, current practices, and how to test for it. Maria Katharina Schenk Oslo, September 2020

The new millennium witnessed new technique of corundum treatment that was first reported by American Gem Trade Association (AGTA) on January 8, 2002. Gemological laboratories worldwide researched and found that this new process used beryllium (Be) a lighter element with atomic weight of 9.012182u and produced a wide variety of corundum colors such as yellow, orange, or brown color. The effectiveness of color shift observed were dramatic. When these stones were immersed in methylene iodide, many displayed unusual yellow to orange rims surrounding pink cores (color of the gems being orange to pinkish orange). Be-diffusion analyses have indicated concentration in the diffused region of about 10-35 parts per million and is the most broadly applicable artificial coloration ever achieved. BERYLLIUM DIFFUSION OF RUBY AND SAPPHIRE, John L. Emmett, Kenneth Scarratt, Shane F. McClure, Thomas Moses, Troy R. Douthit, Richard Hughes, Steven Novak, James E. Shigley,WuyiWang, Owen Bordelon, and Robert E. Kane, GEMS & GEMOLOGY SUMMER 2003

Discussion

Reports published on internet created a worldwide controversy on the corundum treatments, which ranged from simple heat treatment to high-heat treatments such as beryllium diffusion. A lot uncertainty about the nature of treatment created not just controversy but lack of confidence among the traders, investors or even consumers. One of the most common and important question since its inception has been, “are these sapphires treated and can they be identified?”

Different literatures have mentioned about presence of color zoning confined to the facets or surface-conformal color layers but for understanding color causing mechanism is the use of special analytical methods, such as UV-VIS-NIR (origin of color analyses), ED-XRF (e.g. Iron (Fe), Chromium (Cr) and Titanium (Ti) trace element analyses), SEM-EDS (chemical analyses of contaminated substances at the gemstones' surface), visual and spectroscopic cathodoluminescence investigations (analyses of internal growth structures), color-stability measurements (long-term colorstability to UV and visible light), and, most importantly, Laser Ablation Inductively Coupled Plasma Mass Spectrometry (LA-ICP-MS).LA-ICP-MS is used to characterize the content and variation of all trace elements including lighter elements such as Lithium (Li), Boron (B) and Beryllium (Be) on a micro scale. Since LA-ICP-MS was not available, this report contributes to understand if ED-XRF, UV-Vis and FTIR can be helpful in identifying “Beryllium–Diffusion Corundum”. Contributions to gemology- No.4, December 2005, Beryllium Treatment-GRS- Dr. A. Peretti, FGG, FGA, EuroGeol, GRS Gemresearch Swisslab Diffusion

This report will discuss briefly:

• Types Of Diffusion Treatments
• The Identification Techniques (basic & advanced)
• Summary Conclusion

Diffusion: the penetration of certain elements into the atomic lattice of a gemstone during heat treatment, with the objective of changing or accentuating its color.

Types Of Diffusion Treatments:

The main two types of diffusion treatments are: Titanium Diffusion and Beryllium Diffusion.

Around the 1980’s diffusion treatment became ubiquitous in the gem industry (Nassau, 1994) and the market was flooded with large quantities of diffusion treated sapphires with vibrant blue colors. Colorless sapphires that could not be improved by traditional heat-treating methods, were used as starting material (Griffith, 2020). In the beginning treaters used titanium, a heavy metallic color causing element for blue with color penetration less than 0.5mm into the surface (Griffith, 2020). Due to the lower penetration levels, it was carried out only on faceted gems. Though stable treatment but repolishing or recutting removes the blue on the facets and reveals the true color of the stone. Experiments to enhance the red in rubies by adding diffusing chromium was not successful since chromium does not penetrate easily into the stone. (Griffith, 2020).

Visual Observation (Basic Gemology)

The study sample on which this article is based included rough and faceted stones ranging from 0.41 to 5.67 ct. The 8 stones in the core sample ranged from medium to dark in tone and included bluish green-yellowish green and orangy yellow hues. Faceup, cut and polished stones appeared uniform in color, while roughs displayed desaturated color. 4 samples were bG-yG before treatment and 4 samples were oY post Be-diffusion (Fig 4).

Standard Gemological Testing: Routine gem testing was performed on a portion of the samples examined for this study. The tests included determination of refractive index, specific gravity by the hydrostatic method, and pleochroism, as well as observing absorption spectra from desk-model spectroscopes and the reaction to ultraviolet radiation. Since color distribution was anticipated to be the most important identification criterion, we examined virtually all samples with a gemological microscope, using various types of lighting and immersion techniques.

Magnification: No specific or classic features of beryllium diffusion treated stones were observed under standard magnification. The best way to examine beryllium diffused gem is under immersion microscope (Fig 5). The faceted samples showed a slightly darker core in the centre with an almost near-colorless rim.

Beryllium diffused gems do not show color concentration on facets edges as seen in titanium diffused gems, as the treated color penetrates deeper into the stone and hence the distinctive rim-core pattern is not visible due to complete color penetration.

(Diffusion treatment of corundum: a history, current practices, and how to test for it. Maria Katharina Schenk Oslo, September 2020)

LWUV (Long Wave Ultraviolet): Under LWUV the untreated natural samples had no reaction whereas the beryllium treated samples showed deep orangish belly with whitish culet. In one of the samples a horizontal whitish band was observed. (Fig 6)

Infrared Spectroscopy (FTIR): Detectable quantities of hydrogen in several forms such as hydrogen bonding with oxygen forming O-H or hydrogen can also exist as H2O, are usually found in corundum. Natural corundum with O-H bond can show sharp absorption bands at 3367, 3310, 3233, and 3185 cm−1 and hydrogen in H2O can show strong absorption at 3400 cm-1. While studying the samples, IR spectra of 8 samples were collected. Four samples were Be-diffused while the remaining four were sold as natural untreated.

Two natural untreated samples showed strong absorption at 3310 cm-1 related to hydrogen oxygen bonding whereas the beryllium treated samples lacked this absorption, which was expected, since hydrogen rapidly diffuses out of corundum at high temperatures in a hydrogen-free atmosphere or an oxygen-rich atmosphere. (Fig 7) shows the O-H absorption peak and absence of O-H peaks in beryllium treated stones. Also seen was a hump at 3050 cm-1 in two of the treated samples which in the recent past is one of the diagnostic IR feature observed in beryllium diffused gems.

UV-Vis-NIR Spectroscopy: (Fig 8) shows the UV-Vis spectra of natural and Be diffused sapphire samples. These spectra show several distinct features. Each spectrum has three sharp absorption peaks centered at 377, 388, and 450 nm, and an additional broad absorption band at 570 nm in rough samples. The peaks at 377 and 450 nm are assigned to Fe3+ - Fe3+ ion pairs, whereas the peak at 388 nm is assigned to isolated Fe3+ ions. The broad absorption band at 570 nm is due to Fe2+ - Ti4+ ion pairs. The absorption peak at 450 nm is attributed to the Fe3+ ion pair which is the main cause of yellow color in iron-containing sapphires. Similarly, the broad absorption centered at 570 nm is attributed to the Fe2+ - Ti4+ ion pair which is the cause of blue color in iron and titanium-containing sapphires. The strength of these bands decides the final color of the stone. The relative strength of 570 nm in faceted sapphire is lower than that of rough stone, explaining the color transition from green to yellow. However, as per the literature, the Fe3+ ion causing absorption band at 450 nm is not a strong chromophore, it cannot produce such a vivid yellow color which is observed in faceted sapphire samples. Therefore, in faceted orange-yellow sapphire, the color causing chromophores or defects is not iron but some anonymous chromophores which is to be explored.

UV-Vis-NIR spectroscopy is typically used to confirm the trapped-hole nature of color in these Be-diffused stones, but we could not find it useful as an identification method.

(Beryllium Diffusion of Ruby and Sapphire Gems & Gemology Summer 2003, John L. Emmett, Kenneth Scarratt, Shane F. McClure, Thomas Moses, Troy R. Douthit, Richard Hughes, Steven Novak, James E. Shigley,WuyiWang, Owen Bordelon, and Robert E. Kane)

Trace element analysis ED XRF: The data revealed considerable amount of Iron (Fe) and least amount of Chromium (Cr) and Titanium (Ti). Traces of Vanadium (V) and Gallium (Ga) % was too low for any consideration. (Table 1)

Comparing the quantitative data, (e.g. focusing on the concentration of trace elements between beryllium and untreated natural sapphires, and the intensity of the absorption lines, the following differences were noticed:

Sapphires that were heat-treated in the presence of Beryllium, it was found that chemical concentrations (Fe) were nearly the same whereas and presence of Cr were not seen. (Contributions To Gemology, No.4 December 2005- Beryllium-Treatment- Dr. A. Peretti)

In such stones, we have the condition that [Ti4+ + Si4+] ≥ [Mg2+] (that is, donor impurities are slightly greater than or about equal to acceptor impurities).

The addition of beryllium as an acceptor changes the concentration relationship to [Mg2+ + Be2+] [Ti4+ + Si4+]—so there is an excess of acceptor impurities.

In a reducing atmosphere, the excess acceptor impurities (Mg2+, Be2+) will be charge compensated by oxygen vacancies and thus produce no color. However, in an oxygen atmosphere, excess Mg2+ and Be2+ will trap holes, producing a strong yellow to orangy yellow coloration. (2003 Be di R & S)

ED-XRF cannot access origin of color concentrated in certain levels within the gemstone since ED-XRF cannot see these layers in the gemstone due to restricted analytical sampling volume and hence the color seen faceup is difficult to explain by ED-XRF.

Extensive measurements by LA-ICP-MS will be required to understand the orangy yellow coloration seen in our samples. (Contributions To Gemology, No.4 December 2005- Beryllium-Treatment- Dr. A. Peretti)

Summary And Conclusion:

In this study basic and advanced gemological test were performed on samples “before and after” Be diffusion to find out the identification features of Be diffusion. It was observed that in both before and after treated stone there is no significant change in optical properties of stones i.e. RI, SG, etc. as an identification feature of Be diffusion. FTIR spectra of before treatment samples, 3310 cm-1 peaks were observed which were absent in after treatment samples. This was confirming heat treatment on aftertreatment sample but not indicative of Be diffusion. The after-treatment samples showed a broad peak centered at 3050 cm-1 which can be considered as an identification feature of Be treatment. (GTL, Lab Information Circular -July 2010- Vol 58- Beryllium Treated Yellow Sapphires - yet another challenge; Atichat et al (2009, The 2nd International Gem & Jewelry Conference, Bangkok, Thailand; proceedings volume). The origin of this special feature is anonymous. The typical UV-Vis spectra and Ed-XRF trace element analysis explained the green coloration in the before-treatment samples but in after-treatment samples, this data only confirmed the presence of Fe. Since Fe is not a strong chromophore and cannot cause such a strong orangy-yellow coloration, the data procured from UV-Vis and ED-XRF were inconclusive to the cause of such a strong color.