Verification of Wavelength Accuracy in an FT-NIR Spectrometer

Significance of the topic

Near-infrared (NIR) spectroscopy increasingly supports material identification, quality control and chemometric modelling across industry and research. Accurate and stable wavelength calibration is critical because even small spectral shifts degrade multivariate models, assignment of spectral features, and inter-instrument comparability. Practical verification procedures that cover both high-resolution laboratory checks and routine on-board validations are therefore essential for reliable FT-NIR performance.

Objectives and overview of the study

This study evaluated wavelength accuracy and stability of the Thermo Scientific Antaris FT-NIR spectrometer family. The goals were to: (1) quantify basic wavelength accuracy and precision using atmospheric water vapor, (2) assess spectral quality under normal operating conditions using the certified NIST SRM 2035 glass standard, and (3) demonstrate a practical, integrated daily verification using an internal polystyrene reference mounted in the instrument validation wheel. The work also compared performance across multiple instruments and across the different optical sampling paths (transmission, reflectance, fiber). Key effects of resolution, apodization and beam path optics on line shapes were examined.

Methodology

• Instrument configuration: Thermo Scientific Antaris FT-NIR with tungsten–halogen source, CaF2 beamsplitter and InGaAs detectors. The instrument supports three selectable optical paths (transmission, reflectance, fiber) with corresponding optimized detector/optics.
• Reference materials and acquisitions:
  • Atmospheric water vapor: high-resolution spectra at 2 cm⁻¹ resolution, two levels of zero filling. Norton–Beer weak apodization used to lower side-lobes while preserving narrow peak shape.
  • NIST SRM 2035: spectra recorded at 8 cm⁻¹ resolution, one level of zero filling, Norton–Beer-Medium apodization; used as a certified reference representing typical NIR operating conditions.
  • Polystyrene reference (0.75 mm thickness) mounted in the instrument validation wheel: measured at 8 cm⁻¹ resolution with the same apodization; intended for integrated daily checks across all optical paths.
• Scanning and processing: 64-scan averages to improve signal-to-noise for SRM and polystyrene measurements. Peak-location algorithms were applied (including aperture-size correction) to determine line centers. Water vapor peak positions were compared to HITRAN96 database values.
• Cross-instrument and temporal tests: SRM spectra acquired on five instruments to estimate inter-instrument variability; repeated measurements over three days to evaluate short-term stability under normal lab environmental conditions.

Instrumentation

• Thermo Scientific Antaris FT-NIR spectrometer (transmission/reflectance/fiber configurations).
• Source: tungsten–halogen lamp.
• Beamsplitter: CaF2.
• Detectors: indium gallium arsenide (InGaAs) optimized for each beam path.
• Internal validation wheel containing a polystyrene reference sample.

Main results and discussion

• Water vapor high-resolution results: well-resolved rotational-vibrational lines were observed in regions near 4000, 5400, 7300 and 8800 cm⁻¹. For ten selected peaks spanning the NIR, the mean difference relative to HITRAN96 line positions was 0.027 cm⁻¹, demonstrating excellent absolute accuracy when compared to a high-resolution line database.
• Inter-instrument comparison using NIST SRM 2035: the sharp SRM feature near 10245 cm⁻¹ was measured on five instruments, giving a standard deviation of 0.08 cm⁻¹, indicating good but measurable inter-instrument variability that may reflect optical-path, detector or alignment differences.
• Short-term wavelength stability: repeated hourly measurements over three days for the 7299 cm⁻¹ water vapor peak produced an average position of 7299.370 cm⁻¹ with a standard deviation of 0.002 cm⁻¹. Observed micro-drift correlated with ambient changes (temperature/humidity cycles), highlighting environmental influences on wavelength stability.
• Internal polystyrene reference across sampling paths: the polystyrene peak near 4570 cm⁻¹ was used to verify wavelength across the instrument’s optical configurations. Results across multiple systems clustered tightly, with one low outlier traced to use of a smaller-diameter test fiber that created a limiting aperture and induced a slight peak shift—demonstrating sensitivity to optical coupling and aperture changes.
• Effects of resolution and apodization: high-resolution water vapor spectra were best for precise line-center checks, while SRM and polystyrene at typical operating resolution (8 cm⁻¹) better represented routine measurement conditions. Norton–Beer apodization variants were chosen to balance side-lobe suppression and peak width.

Benefits and practical applications of the method

• Three-tiered verification strategy: combines the precision of atmospheric water vapor, the traceability of a certified SRM, and the convenience of an integrated polystyrene check for everyday validation—covering research, regulatory validation and routine QA/QC needs.
• Routine, instrument-integrated polystyrene reference provides rapid daily checks without the cost or handling needs of SRMs, improving reproducibility and protecting the reference from contamination.
• High-resolution water vapor scans are a low-cost in-situ probe of absolute wavelength calibration and can detect subtle instrumental shifts beyond typical operating resolution.
• Awareness of optical-path and coupling factors (e.g., fiber diameter, apertures) enables users to reduce inter-instrument variability by standardizing sampling interfaces and alignment procedures.

Future trends and possibilities for application

• Greater standardization across instruments and manufacturers to reduce inter-instrument wavelength offsets, including shared calibration procedures and reference datasets.
• Integration of automated internal references combined with software control charts to provide continuous performance monitoring and automated alerts for drift or aperture-related shifts.
• Improved environmental compensation algorithms and instrument enclosures to decouple minor ambient temperature/humidity variations from wavelength stability.
• Use of updated high-resolution spectral databases and automated line-matching tools (beyond HITRAN96) to refine wavelength calibration and to enable traceable chain-of-custody for regulated applications.
• Further study of fiber-optic coupling effects and development of standardized test fibers or coupling fixtures to minimize aperture-induced spectral shifts in field and at-line configurations.

Conclusion

A complementary approach using atmospheric water vapor, NIST SRM 2035 and an integrated polystyrene reference provides robust verification of wavelength accuracy and stability for FT-NIR spectrometers. Water vapor spectra deliver the highest-precision checks, SRM 2035 provides certified traceability suitable for regulatory validation, and the internal polystyrene standard offers an economical and practical solution for routine instrument monitoring. Attention to optical path, apodization, resolution and sampling apertures is crucial to minimize variability and ensure consistent performance across instruments and operating modes.

References

• Rothman, L. S., et al., HITRAN 1996 Molecular Spectroscopic Database (HITRAN96) - database of high-resolution molecular line positions used for water vapor comparisons.
• NIST Standard Reference Material 2035 — Optical Glass Standard for NIR wavelength validation.
• Thermo Fisher Scientific, Application Note AN50772 (2007): Verification of Wavelength Accuracy in an FT-NIR Spectrometer.