FTIR spectrometers use a system of moving and stationary mirrors (called an interferometer) to split the infrared radiation transmitted through a gem into two halves (figure 1). The two waves travel different distances before they are recombined, which creates interference. When the two halves of the transmitted radiation are rejoined, they add together either constructively or destructively. For example, if the peaks of the two waves coincide, then they will add together to twice the original amplitude, or displacement. However, if the peak and trough coincide, then the two waves will sum to zero. The combination of the two waves produces a pattern called an interferogram (figure 2, left), which is then translated into an absorption curve (figure 2, right) using a mathematical operator known as a Fourier transform (thus the term for this type of spectroscopy is Fourier Transform Infra-Red). IR spectra are usually expressed in wavenumbers (number of waves per centimeter); this unit is convenient since the wavelength of infrared light varies from about 14,000 cm^-1 to 10 cm^-1.

For additional information, see A. Beran, E. Libowitzky, Eds. (2004) Spectroscopic Methods in Mineralogy, Eötvös University Press, Budapest.