One of the most difficult gemological problems is the detection of beryllium diffused sapphire. The treatment process was discovered by accident by a Thai treater in 2001 when he was heating low quality sapphires from Madagascar contaminated with chrysoberyls. By the end of 2002 the practice was widespread in Thailand and not disclosed to buyers.

Not all rubies and sapphires can be treated with Be successfully. The treated corundum must have the proper chemistry to react to the Be, and according to the leading expert on this subject, Ted Themelis, “...most of the sapphires suited for the beryllium treatment are nearly depleted. However, the beryllium process continues to be used in treating sapphires previously heated by other processes.”

Be treated stones are ubiquitous in the marketplace and tend to be considerably less expensive than their natural counterparts. The depth of the diffusion process (e.g. depth of coloration) is directly predicated on the magnitude of the soaking temperature and the time of treatment. While complete coloration through the stone is possible, in a sizeable stone the process could take weeks. Most Be diffusion treated stones seen on the market tend to be surface diffused only, and in some cases the treatment can be seen when the stone is immersed in diiodomethane (methylene iodide) and observed under a microscope with backlighting. There are a number of articles in gemological journals and trade publications that deal with the various detection methods and for a description of the treatment process, Ted Themelis’ newly published book Heat Treatment of Ruby and Sapphire is the definitive information source available from Ted directly (http://themelis.com)

The concentrations of Be necessary to color a stone can be very, very low, ranging from 3-4 ppm up to 30-40 ppm. And, because Be is a light element with an atomic number of 4 and a mass just over 9 amu, it is difficult to detect by common analytical methods. Moreover, the windows of both wavelength dispersive and energy dispersive x-ray spectrometers are comprised of beryllium and so the element is undetectable with most electron microprobes and conventional energy dispersive x-ray fluorescence equipment.

The very low concentrations of Be can be detected by three analytical methods, LIBS (laser induced breakdown spectroscopy), LA-ICP-MS (laser ablation inductively coupled mass spectroscopy) and SIMS (secondary ion mass spectroscopy). Of the three techniques, LIBS is the least expensive as far as equipment acquisition is concerned, with the typical set-up costing about $70,000, compared to at least $400,000 for LA-ICP-MS and perhaps $1 million for SIMS. Both LA-ICP-MS and SIMS have quantitative capability, while LIBS is typically only semi-quantitative.

A typical LIBS set-up consists of several components, the first being a relatively high powered laser that ablates or vaporizes a tiny volume of the surface of the sample. The ablated area is usually just barely visible with a microscope under 10x and is usually about 50 micrometers in diameter. Typically a pulsed UV laser of 266 nm is used as higher wavelength lasers cause additional heat that could damage nearby inclusions in the stone. A microscope with video camera is often used as an aid in positioning the sample.

When the laser impacts the sample it vaporizes a minute quantity into a high temperature plasma cloud that reached 50,000° C in the center and is about 5,000°C at the edges. The excited elements in the sample emit light at specific wavelengths that are in turn measured by a series of six or seven sensitive spectrometers that measure the wavelengths from 200-900nm. The results are graphically displayed by computer.

LIBS equipment does not require a vacuum, however the analytical sensitivity of the process is increased approximately four-fold in an argon atmosphere, so argon gas is typically injected into the sample chamber during analysis.

The limits of detection for Be are nominally about 3 ppm under ideal circumstances, which is fairly close to the minimum required for coloration in corundum. And again, the limitation of the equipment and process is that it will tell you Be is present, but not how much - a situation that makes LA-ICP-MS and SIMS preferable for this type of analysis. At this low level of detection care must be taken to prevent Be contamination and in some cases chromium in ruby can partially mask the beryllium spectra.

Overall, the equipment is not complex to operate and its relative cost effectiveness makes it the most sensible proposition for a lab that would like to be able to definitively detect Be treated corundum.