Header: Footer:

Gemstone Identification Lab

Identifying Gemstones with Raman Spectroscopy

Learning Objectives:

  1. Identify and determine authenticity of gemstones using Raman spectroscopy

  2. Understand importance of differentiating gemstones

 

Background:

Gemstones can be found in a large variety of colors, shapes, and sizes. For example, quartz, a very common crystal, has many varieties including transparent forms such as translucent amethyst (purple), citrine (yellow), milky (white), rose (pink), and smoky (gray). There are also opaque forms such as agate, jasper, and onyx. Due to the large variety of mechanisms through which minerals can form and their similarity in appearance, it is imperative to be able to identify and distinguish them.

 

This is especially important when it comes to the sale of these valuable crystals. Many synthetic or 'lab-grown' gems are on the market. Such gems can be sold to unsuspecting customers and never be identified as imitations as they have almost exactly the same physical and chemical characteristics as their authentic counterparts. For example: Cubic zirconia can be pawned off as diamond, colored glass as peridot, and citrine quartz as topaz. Even an experienced collector would never know the difference at first glance. 

 

The method of Raman spectroscopy is a definitive way to ascertain the true identity of a crystal. Each mineral has its own unique Raman “fingerprint”. Once the Raman signal of a sample is taken, its spectrum can be superimposed with that of an authentic crystal to see if they match. Raman spectroscopy distinguishes materials based on the stretching of their constituent bonds. For instance, a mineral composed of only carbon (i.e. a diamond) will only show one peak on a Raman spectrum as all of its bonds are the same and will stretch identically. 

When taking data with Raman, it is crucial that your signal is accurate. If a signal is 'saturated', the instrument is unable to process all of it and the corresponding peak will  plateau at a certain point. If the signal is too weak, the corresponding peak will be 'fuzzy' or poorly defined.  The easiest way to rectify that issue is to increase the integration time (the time spent calculating the area between the graph and the x-axis).