Infrared (IR) diffuse reflectance spectroscopy is an emerging technology that relies on statistical relationships between IR soil spectra and results obtained from ‘conventional’ measurements, in this case for gypsum. This non-destructive analytical option derives from the knowledge (Beech et al. 2003) that mid infrared reflectance (MIR) can be used to assess (for regulatory purposes) the purity of samples of commercial gypsum. Earlier, Janik et al. (1998) demonstrated that MIR provided an alternative technology for soil analysis. For more details, see Chapter 6.
The method relies on knowledge that there are a number of spectral peaks in the MIR spectrum that are characteristic of and sensitive to gypsum. Beech et al. (2003) noted these can be assigned to the fundamental gypsum (-O-S-O) vibrations of SO42– at 1260 cm-1, water of hydration (-OH) stretch of 3500 cm-1, and deformation vibrations (1690–1640 cm-1). Weaker peaks associated with overtones and combinations of fundamental vibrations (2234 and 5145 cm-1) may also contribute. The soil matrix adds considerable spectral complexity, which could adversely affect the usefulness of relationships between actual and predicted soil gypsum concentrations.
Laboratories using this secondary method must initially confirm the existence of strong, robust and reliable relationships between outputs from their MIR analyser and conventional measurements of soil gypsum content. Alternatively, regional or national calibration datasets may become available, derived by ‘fingerprinting’ modal examples from ‘legacy’ and new soil collections (Viscarra Rossel et al. 2008). If the conventional measurements used were originally expressed on an oven-dry basis, then the MIR results for gypsum will also allow the direct reporting of results on an oven-dry basis. Sequential scans typically take around 1–2 min/sample. This analytical option should only be used, however, when instrument calibrations, commonly but not exclusively based on partial least square regressions (Stevens et al. 2008), are known to be repeatable and of reasonably high precision (e.g. R2 values ≥0.85). Soil samples should be ≈40°C air-dry and finely ground (<0.5 mm).
Stabilise and verify the set-up and operating performance of the MIR spectrometer, usually incorporating an auto-focussing diffuse reflectance accessory or capability. Also confirm the ‘standard calibration’ for gypsum, likely based on Method 11A1.
Load the instrument’s auto-sampler with ‘unknown soils’, previously dried to ≈40°C and finely ground (≥0.5 mm) to assist with sample uniformity. Typically, the instrument or its associated computer will integrate the spectral signals with the calibration equations to provide the result without further calculation, if the conventional method’s results were expressed on an oven-dry basis. If the calibration equations were based on conventional measurements expressed on an air-dry basis (≈40°C), then the MIR results will need to be adjusted for residual moisture using relevant air-dry moisture to oven-dry moisture ratios, guided by Method 2A1.
Report as gypsum (%) in oven-dry (105°C) soil.