The following are several criteria to consider when selecting an ATR crystal material for a specific sample:

• Refractive Index
  The crystal should have a higher index of refraction than the sample. The majority of organic materials have refractive indices in the area of 1.4. Refractive indices of standard ATR crystals span from 2.4 to 4.0 – which in most cases provides sufficient sample to crystal differentiation. Inappropriate refractive index ratios may cause distortion of spectral features. These may be manifested by diminished peak symmetries, sharp baseline/peak shoulder transitions, and in extreme cases, presence of derivative-like features in the spectrum.
• Spectral Range
  All ATR crystals have different spectral ranges. Specifically, in Mid-IR, the ATR range cut-off at low wavenumber varies from approximately 1500 cm-1 for Si to 400 cm-1 for KRS-5. To a certain extent, the cut-off values are also affected by the length (thickness) of the crystal. In light of these facts, it is important to determine whether the spectral features of the sample correspond with the spectral range of the ATR crystal selected. Spectral ranges of selected ATR materials are listed below.
• Chemical and Physical Properties
  For obvious reasons, the ATR crystal must be chemically and physically compatible with the sample. Some crystal materials may react with samples. This will typically damage the crystal surface and may produce unpleasant side effects (e.g. acidic solutions, pH<5, may etch the ZnSe crystal while strong acids may generate toxic hydrogen selenide) or Alkalies, pH>9 can be harmful to ZnSe or AMTIR. Physical considerations are equally important since some crystals are more susceptible to pressure and temperature changes than others.
• Sensitivity
  The effective pathlength of the infrared beam in the sample must be sufficient to produce an adequate spectrum. This parameter is affected by the number of reflections (more reflections yield higher absorbance) and the depth of penetration – which is a function the refractive indices and the angle of incident beam. For high absorbing samples, Ge is a good choice due to relatively low depth of penetration.
• Optical Design
  The overall optical design of an HATR accessory – its optical path, mirrors, quality and throughput has great effects on analytical results. Placing a good quality HATR accessory with a 45-degree ZnSe crystal in the sample compartment of an FTIR spectrometer should result in an energy throughput of >20 % T for multi-reflection systems and > 35 % T for single reflection systems.

The following is a basic review of common ATR crystal materials:

• AMTIR produced as a glass from selenium, arsenic and germanium. It is highly toxic during the manufacturing process. However, the brittle nature of this material and its total insolubility in water makes it safe for use as an internal reflectance element. It has a similar refractive index to zinc selenide and can be used in measurements that involve strong acids.
• Diamond (Raw or Synthetic) offers excellent chemical and physical properties when used as a sampling interface in infrared experiments. This hard, scratch-resistant material is suitable for applications involving a wide range of chemicals. It withstands highly acidic and basic samples very well. It does not react with strong oxidizers or complexing agents. Diamond ATRs can be used to analyse hard powders and other difficult to analyse solid samples. The main disadvantage of the diamond is its relatively high absorbance (when used as an ATR element) in the 2,500 cm-1 to 1,650 cm-1 region. It also is the most expensive ATR material, which is less practical for multi-reflection applications.
• Germanium has been used extensively as a higher refractive index material for samples that produce strong absorptions (e.g. rubber). The crystal is also used when analysing samples that have a high refractive index, such as in passivation studies on silicon. Although slightly higher than ZnSe, it is fairly low cost.
• KRS-5 was the most widely used material for ATR elements prior to the common availability of Zinc Selenide. Although it has a wide spectral range, KRS-5 is very soft and is easily damaged. Like the Zinc based compounds, the thallium in KRS-5 is readily complexed by ammonium compounds and amino-based chelates. The main advantage of KRS-5 is its wide spectral range. It is moderately expensive.
• Silicon has a relatively high refractive index and it is useful for analysing highly absorbing samples. Silicon is scratch resistant and is insoluble in water and organic solvents. However, it is affected by strong acids and is soluble in alkalis. Another limitation of a silicon crystal is its relatively narrow infrared spectral range. Although slightly higher than Ge, Si is still priced moderately.
• ZnSe is the preferred replacement for KRS-5 for all routine applications. Its useful spectral range is less at the low frequency end than that of KRS-5, but the mechanical strength of this rigid, hard polycrystalline material is superior. Although it is a general-purpose material, it has limited use with strong acids and alkalies due to the surface becoming etched during prolonged exposure to extremes of pH. Complexing agents, such as ammonia and EDTA, will also erode the surface because of the formation of complexes with the zinc. It is the lowest cost ATR material available today.