The near-infrared (NIR) region of the electromagnetic spectrum spans from approximately 780 nm to 2500 nm (12800 to 4000 cm⁻¹). Unlike mid-infrared spectroscopy, which probes fundamental molecular vibrations, NIR spectroscopy detects overtones and combination bands — absorptions that arise from anharmonic vibrations of bonds involving hydrogen atoms. These overtones (first, second, and third) and combinations of stretching and bending modes produce broad, overlapping bands that require multivariate data analysis for interpretation.
The most prominent absorptions in NIR arise from C-H, O-H, and N-H stretching vibrations, which dominate spectra of organic and biological materials. The molar absorptivity of these overtones is typically 10 to 1000 times weaker than fundamental infrared absorptions, which gives NIR a practical advantage: samples can be measured directly with little or no preparation, even in thick or highly scattering matrices. Measurements are performed in diffuse reflectance mode for solid and powder samples, or in transmittance mode for liquids and clear films.
NIR instrumentation comes in several configurations. Dispersive grating instruments scan wavelengths sequentially using a monochromator. FT-NIR spectrometers use a Michelson interferometer to acquire full spectra rapidly. Diode array detectors enable simultaneous multi-wavelength acquisition, making them suitable for real-time process monitoring. Fiber-optic probes allow remote sampling, enabling in-line and at-line measurements in industrial environments.
Because NIR spectra consist of broad, overlapping peaks, direct assignment of individual bands is rarely possible. Instead, quantitative and qualitative analysis relies on chemometric models constructed using multivariate calibration techniques such as Principal Component Analysis (PCA) for exploratory data analysis and Partial Least Squares (PLS) regression for quantitative prediction. Calibration requires a representative set of reference samples with known values of the property of interest, and model performance is validated through cross-validation and independent test sets.
NIR spectroscopy finds extensive application in agriculture (moisture, protein, and fat content in grains and feed), food (determination of water, sugar, alcohol, and oil content), pharmaceuticals (raw material identification, blend uniformity, and moisture determination in lyophilized products), and polymer analysis (identification of resin types, copolymer composition, and additive content). Its speed, non-destructive nature, and suitability for online measurement make it one of the most widely used process analytical technology (PAT) tools in modern industry.
