Detecting synthetic CVD-diamond with GemmoRaman-532SG™

Mikko Åström, Alberto Scarani, Marco Torelli

Synthetic diamonds grown by CVD-method (Chemical Vapor Deposition) have recently gained a lot of publicity in jewelry trade publications and triggered many alerts about undisclosed stones submitted to major gemological laboratories. Development of pure CVD- diamond production has been fast and dynamic because the material has great potential for semiconductor industry. Detecting CVD-diamonds polished as gems has become extremely difficult if not impossible to deal with by using traditional gemological tools.

CVD synthetic diamonds currently on the market can be unambigously identified only by combination of two advanced test methods; UV-fluorescence microscopy and photoluminescence spectroscopy. UV-fluorescence microscope (such as DeBeers DiamondView) is used for revealing curved growth striations of CVD-diamond also known as “terraces” (Fig 1.) Extended range and enhanced signal to noise ratio of scientific grade GemmoRaman-532SG™ spectrometer addresses the latter requirement by detecting subtle photoluminescense peaks arising from silicon impurity in these diamonds.

While silicon impurities have been found from very limited set of natural samples, it is still considered as very strong indication of CVD- origin of a diamond. The crystallographic SiV–defect, consisting of negatively charged Silicon atom next to a vacancy (a missing carbon atom) is very stable and can not be removed in commercial HPHT- treatment temperatures.

In some CVD-stones it is also possible to detect SiV-peak by UV-Vis-NIR spectroscopy, but many scientific studies have shown that its sensitivity is not sufficient enough in all cases. Therefore, for studying CVD diamonds, the more sensitive PL-spectroscopy technique is preferred.

Fig1. Curved growth striations of CVD-diamond photographed with MAGI in-house build deep ultraviolet (<225nm) fluorescence microscope. 15x, Marco Torrelli

Screening procedure for colorless CVD- diamonds

Colorless or near colorless (D-J) CVD-diamonds to can be classified as type IIa, containing no significant amount of aggregated nitrogen defects. There has been indications this may change in the future when CVD- processes develop towards adding nitrogen in the reaction chamber on purpose (for speeding up the growth) and subsequent HPHT- treatment which has tendency to aggregate portion of the nitrogen atoms in the form of A or even B- defects. However, the following screening procedure works for all the colorless to near colorless CVD-diamonds manufactured to date.

1.) Raman screening of diamond and it’s simulants

First crucial step of the screening procedure is to make sure all the stones under the study are actually diamonds. This can be achieved with a Fast Diamond Screening- feature built inside all the GemmoRaman-532™ -series spectrometers. It will separate colorless to near colorless diamond from all of it’s simulants based on diamond’s unique Raman fingerprint – there is absolutely no room for mistakes!

This tool has been designed to be as fast as possible, typical test time being a fraction of a second. The tool is especially efficient when working with large parcel of stones. Basically, testing time is limited to operator’s skills in handling diamonds with gem tweezers.

2.) Diamond type screening

The vast majority (about 98%) of natural colorless to near colorless diamonds belongs to type Ia – also known as Cape Series. These diamonds contain aggregated nitrogen atoms as minor impurity, leading to yellowish tinge. The same impurities causes another distinct feature: opacity to short wave ultraviolet (SWUV) radiation. On the contrary, type II diamonds (less than 2% in nature) which do not contain measurable amount of nitrogen and pure type IaB diamonds (0.2%) do transmit SWUV radiation.

UV- transparency of diamonds can be conveniently and safely tested with MAGI DiaGuard an add-on product for GemmoRaman-series spectrometers. If the test result indicates the sample is transparent to UV radiation (Type II, IaB or IaB with very low A component) there is a chance that the stone is of CVD-synthetic origin.

3.) Photoluminescence spectroscopy with GemmoRaman-532SG™ spectrometer

GemmoRaman-532SG™ is based on state of the art Ocean Optics TEC cooled scientific grade spectrometer capable for detecting smallest Raman and photoluminescence signals of the materials with very high signal to noise ratio.

The spectrometer’s extended spectral range combined to MAGI Liquid Nitrogen Kit– accessory set is extremely useful for diamond studies, for example allowing clear detection of Silicon-Vacancy photoluminescence peaks in CVD-diamonds.

Typical test time including the sample set-up and cooling is about 2 minutes, which can be considered as a very fast method. It is also often possible to detect somewhat broadened negatively charged Silicon-Vacancy peak of CVD diamond even at room temperature at about 738 nm. Room temperature measurement can be used as a first – even faster step – for verifying a stone which is already known or suspected as to be CVD grown.

The spectra above illustrates the difference between detected Raman & photolumisescence peaks of CVD synthetic diamond measured at Room Temperature (RT) and Liquid Nitrogen Temperature (LNT) with GemmoRaman-532SG™. Cooling the sample near to liquid nitrogen temperature enhances the PL- peaks, narrows down their spectral width and shifts them towards the shorter wavelengths. Strong Silicon-Vacancy peak located at about 737 nm (actually a doublet at 736.6/736.9 nm) is a very good indication of CVD origin. However, the operator must be careful with the interpretation – many natural type IIa diamonds exhibit a weak to moderate GR1 (general radiation) – peak at 741 nm.

HPHT-grown synthetic diamonds do not usually contain detectable amount of SiV-defects, unless purposely doped with silicon. This kind of stones may be expected to appear on the market in the future, as recent research has shown that using excess silicon in CVD- process with possibly other dopants turns the stones to attractive blue color. These blue CVD-stones are naturally extremely easy to identify as their Silicon-Vacancy peak is pronounced and their PL spectrum does not resemble anything found in the natural blue stones.

References and further reading

CVD Synthetic Diamonds from Gemesis Corp, Gems & Gemology, Summer 2012, vol. 48
Wuyi Wang, Ulrika F. S. D’Haenens-Johansson, Paul Johnson, Kyaw Soe Moe, Erica Emerson, Mark E. Newton, and Thomas M. Moses

Identification of Synthetic Diamond Grown Using Chemical Vapor Deposition (CVD) Gems & Gemology, Spring 2004, vol. 42
Philip M. Martineau, Simon C. Lawson, Andy J. Taylor, Samantha J. Quinn, and David J. F. Evans, and Michael J. Crowder

New Generation of Synthetic Diamonds Reaches the Market (Part A):  CVD-grown Blue Diamonds
Contributions to Gemology, Nov 2013, No. 14, p.1-14
Adolf Peretti, Franz Herzog, Willy Bieri, Matthias Alessandri, Detlef Günther, Daniel A. Frick, Ed Cleveland, Alexandre M. Zaitsev, Branko Deljanin