The use of Fourier-Transform Infrared Spectroscopy (FTIR) in gem identification, with examples of sapphire & jadeite.
Introduction to Infrared Spectroscopy
Most gemologists are familiar with the direct vision spectroscope. This instrument uses a series of prisms or a diffraction grating to separate white light into its spectral components. If white light is passed through a gem, it will reveal any areas of absorption or emission.
The main limitation of the direct vision spectroscope is that it can only analyze the visible region (400–700 nm), but a small portion of the electromagnetic spectrum.
What's up with these wavenumbers?
For many gemologists, the only spectral tool they have used is the direct-vision spectroscope. Spectral features for that instrument are described by wavelength using nanometers (1 nm = 1 millionth of a millimeter).
The main region of infrared interest lies in the range of 2,500 to 16,000 nm. But chemists prefer to describe wavelenths in the infrared by reciprocal centimeters (cm-1) or "wavenumbers." Conversion between nm and cm-1 is complicated, but a handy conversion tool can be found online here.
Key Infrared Regions
The infrared is a huge spectral region, many times larger than the visible region. But most FTIR units sample the region from 7500 to 370 cm-1. Within this range, two key regions exist, as follows:
- Group Frequency (Functional Group) Range: Extending from about 4000-1000 cm-1, this is the key region for identifying features useful for gemology. Features important for separating certain heated from unheated corundums, natural vs. treated jadeite, natural vs. synthetic quartz, etc.
- Fingerprint Range: Extending from < 1000 cm-1 and below, this region is useful for separating one gem material from another.
Today there are three major techniques used with FTIR in gem testing, as follows:
- Beam condenser (narrow beam) method
- Diffuse reflectance (DRIFTS) method
- Specular reflectance method (used for 'fingerprinting' materials)
Each of these techniques requires a separate accessory for the sampling chamber. In many cases these accessories are made by third parties.
Narrow Beam vs. DRIFTS
The Fourier-Transform Infrared Spectrophotometer (FTIR) is used to help determine whether a sapphire has been thermally enhanced. The heights of the various peaks give the gemologist an indication of whether the sapphire being tested has been thermally enhanced. For example, when a strong peak at 3160 wavenumbers (cm-1) is found, this is an excellent indication that the sapphire being tested has not been thermally enhanced.
When a strong peak is found at 3309 in non-basaltic sapphire, this gives an indication that the gem may have been heat treated. Interestingly, a weak-to-medium 3309 peak can be found in both heated and unheated sapphire. Thus great care must be given to interpreting FTIR data.
Two different FTIR methods are commonly used for testing gems. One involves a transmission scan, where a relatively narrow infrared beam is passed through the specimen. As a result of the restricted path, only a portion of the gem is sampled. Another technique, the Diffuse Reflectance Infrared Fourier Transform Spectroscopy (DRIFTS) method, diffuses the beam throughout the gem and thus samples a greater portion of the piece.
In 2007, the authors tested a 5.72-ct blue sapphire with both techniques. As one can see from the pictured spectra, the narrow-beam method produced a small peak at 3309. However when the DRIFTS method was utilized, a much stronger 3309 peak was found, suggesting that the gem was probably heat treated. This was later confirmed by examination of the inclusions.
Why the two different results? By their very nature crystals possess directional properties and this can impact spectra of all sorts, including infrared. But more important, crystals are composed of layers of atoms stacked one on top of another. As the growth takes place, the conditions constantly change, from start to finish. Thus one layer is never identical to another. We can see this in the arrangement of inclusions within a crystal and it is most obvious in the growth and color zoning that many crystals display.
Differences in composition and structure from one layer to another create growth zoning and are also probably responsible for the different FTIR spectra we obtained in this specimen. In this particular gem, the DRIFTS method was better able to capture the full amplitude of the 3309 peak because it sampled a greater portion of the specimen.
This Fourier-Transform Infrared Spectrophotometer is one of the most important tools in our arsenal. It is useful in helping to separate a number of natural gemstones from their treated counterparts, including various sapphires, emerald and jadeite, to name but a few. It is also of use in identifying some synthetics, including synthetic emerald. The instrument is shown with the beam condenser transmission attachment.
Acetone is sometimes used to clean specimens prior to taking an FTIR spectrum, the idea being that this will remove surface contaminants. But one must take great care that the acetone itself is not contaminated, as the following two spectra show.
At the Tucson 2007 show, I performed a brief experiment that shows the potential perils of acetone cleaning as it relates to the FTIR. First, the FTIR spectrum was run on an orange sapphire specimen without any sample preparation other than wiping with a gem cloth. Then, the gem was dipped into a bottle of acetone that had been previously used to clean many specimens. Following a quick dip, the excess acetone was removed with a paper towel (without touching the gem with the fingers) and the spectrum re-run. The difference in the two spectra was remarkable, with large oil peaks found following the acetone bath.
This simple experiment makes clear that acetone dipping has many perils with the FTIR. As a powerful solvent, it will dissolve oils and other dirt that might be on the specimen, but this then leaves a residue as the acetone evaporates. Even if the gem is carefully wiped clean following this dipping, one can suppose that residue would still remain in pits and fissures that escape the touch of the wiping medium.
The lesson is that acetone "cleaning" may add contamination, rather than remove it. If acetone is to be used at all (and the above suggests it should not), one should use clean acetone for every dipping.
Dr. John Emmett had the following comments on the idea of cleaning a specimen in acetone prior to testing with the FTIR:
Washing a stone in anything without knowing the spectra of it is not advisable. First, use semiconductor-grade solvents and only use them once. Second, know the spectra of the pure solvent so you know where in the spectrum to look for interferences and for solvent in cracks. Third, you can measure the spectra of even volatile solvents by putting a drop between two pieces of polished synthetic sapphire plates. Run the spectra of the plates without solvent first. Using paint store solvents for multiple cleanings is worse than no cleaning.
A piece of jadeite jade brought in for testing recently was intriguing in a couple respects. First were the inclusions. While one does not normally think of distinctive inclusions in jadeite, this gemstone possessed a clear cluster of crystals, including many which broke the surface and displayed a much higher luster than the surrounding jadeite. Micro-Raman analysis identified these as zircon, which had been previously identified in jadeite by John Koivula and Tom Moses in 1998 (Gems & Gemology, Vol. 34, No. 1, p. 45).
In addition to the zircon inclusions, the specimen displayed a network of fine cracks that suggested it had previously been bleached. Bleaching is the first stage in the common B-jade process where jadeite is bleached to remove foreign matter from its pores and then the tiny micro-fissures are impregnated with either a wax or polymer, greatly improving both color and clarity.
Such treatment is typically unmasked by reference to FTIR spectra. When we first checked the FTIR spectrum via the DRIFTS method, large polymer peaks were found. A recheck with the beam condenser method revealed the same polymer peaks, but at reduced levels.
- Bradley, M. (n.d.) Gemstone analysis by FT-IR: Identifying treated jades. Thermo Scientific, Application Note: 50833, 2 pp.
- Hainschwang, T. and Notari, F. (2008) Specular reflectance infrared spectroscopy – a review and update of a little exploited method for gem identification. Journal of Gemmology, Vol. 31, No. 1/2, pp. 23–29.
- Koivula, J.I. and Moses, T. (1998) Lab Notes: Jadeite with inclusions of zircon. Gems & Gemology, Vol. 34, No. 1, Spring, p. 45.
- Lowry, S. (n.d.) Analysis of emeralds by FT-IR spectroscopy: Identifying treated and synthetic emeralds. Thermo Scientific, Application Note: 51123, 2 pp.
- Martin, F., Mérigoux, H. et al. (1989) Reflectance infrared spectroscopy in gemology. Gems & Gemology, Vol. 25, No. 4, Winter, pp. 226–231.
- Williams, B. (2012) Infrared spectroscopy in gemstone testing. The Valuer, May–August, pp. 2–4.
First published in 2007 in three parts, while RWH was at the AGTA GTC. This version features significant updates.
About the author
Richard W. Hughes is the author of the classic Ruby & Sapphire and over 170 articles on various aspects of gemology. Many of his writings can be found at www.lotusgemology.com and www.ruby-sapphire.com. His latest book is Ruby & Sapphire: A Collector's Guide (2014).