Verifying the Performance of the Fiber Optic Reflectance Probe on the Thermo Scientific Antaris FT-NIR Analyzer

Significance of the topic

The use of fiber optic reflectance probes with FT-NIR instruments enables remote, non-destructive sampling of solids, powders and packaged products. This capability is important for rapid quality checks in production, at-line process monitoring and manual bench checks where sample handling must be minimized. Verifying probe performance (noise, spectral fidelity, wavelength accuracy and repeatability) is essential to ensure analytical robustness when measurements are taken at the end of several meters of optical fiber rather than inside an integrating sphere or other dedicated sampling accessory.

Objectives and study overview

This technical note evaluates the performance of the SabIR fiber optic diffuse reflectance probe used with the Thermo Scientific Antaris FT-NIR analyzer. The goals were to compare spectral noise and response versus an integrating sphere, verify wavelength accuracy using a reference containing talc (KTA-1920x, related to NIST SRM 1920a), and assess sampling precision and stability during extended repeated measurements on a lactose/talc mixture. Tests were performed at 2 cm-1 resolution and under typical, non-controlled environmental conditions to represent realistic use.

Methodology

Key experimental approaches included:

• Background and sample spectra collected for both the SabIR probe and the instrument-integrating sphere using a diffuse gold reference over the sapphire window.
• Spectral noise assessment by taking repeat background measurements and calculating RMS noise in the 6000 cm-1 region.
• Comparison of polystyrene reference spectra measured with the SabIR and integrating sphere and analysis of spectral subtraction residuals to separate sampling effects from spectrometer linearity.
• Wavelength accuracy evaluation against a talc-containing standard (KTA-1920x) and high-resolution spectra of pure talc.
• Precision testing via 100 consecutive measurements of a lactose/talc mixture over 12 hours without active temperature control and with the optical fiber free to move, using a classical least squares quantitative model to track percent talc.

Instrumentation used

The Antaris FT-NIR analyzer configuration and probe details used in the study were:

• Antaris FT-NIR analyzer with CaF2 beamsplitter, Tungsten-Halogen source and high-sensitivity InGaAs detector.
• SabIR diffuse reflectance fiber optic probe: bifurcated, randomly mixed low-OH fiber bundle mounted in stainless steel probe head with an angled sapphire window; half the fibers deliver excitation light and half collect reflected light.
• Optical interface providing remote start-of-scan and indicator lights on the probe head to report analysis outcomes.

Major results and discussion

Noise and spectral response

Both the SabIR probe and the integrating sphere achieved low spectral noise, with RMS noise in the 6000 cm-1 region below 20 micro-absorbance units. This indicates that the SabIR probe, when coupled to the Antaris system, does not impose a noise penalty relative to the integrating sphere for the tested conditions.

Spectral fidelity and polystyrene comparison

Comparison of polystyrene spectra obtained with the SabIR and integrating sphere showed close agreement. Spectral subtraction of the two measurements produced residual artifacts under ±2 milli-absorbance units, implying that the spectrometer linearity and detector response are consistent across sampling interfaces and that remaining differences are dominated by sampling geometry rather than instrument nonlinearity.

Sampling geometry effects

Differences between SabIR and integrating sphere spectra were attributed to the probe’s expanding beam and acceptance angle, which change the effective pathlength distribution in the sample and introduce wavelength-dependent differences. Consequently, methods developed on an integrating sphere may not be directly transferrable to a fiber optic probe without adaptation or recalibration.

Wavelength accuracy

Wavelength accuracy evaluated using the KTA-1920x talc reference and high-resolution talc spectra showed good agreement, supporting use of the probe for qualitative and quantitative assignments that depend on correct peak positions.

Precision and stability

During 100 repeat measurements of a lactose/talc mixture over 12 hours (single background prior to the run, no temperature control, fiber allowed to move naturally), the quantitative model produced stable results with no observable drift. The reported standard deviation for the repeated measurements was 0.037 (in the units of the quantitative output used in the test), demonstrating good short-term precision under realistic handling conditions.

Benefits and practical applications of the method

The combination of the Antaris FT-NIR analyzer and the SabIR fiber optic probe provides:

• Remote sampling capability that supports in-container, at-line and in-process checks without transferring or destructively sampling material.
• Low-noise spectral acquisition comparable to an integrating sphere, enabling sensitive analyses in reflectance mode.
• Robust mechanical design and an electronic interface for remote triggering and pass/fail indication, facilitating rapid screening workflows.

Practical considerations include the need to account for probe-specific sampling geometry when transferring methods from integrating spheres or different probe designs; calibration updates or transfer models are recommended to maintain quantitative accuracy.

Future trends and applications

Expected directions and potential uses include:

• Wider adoption of fiber-coupled FT-NIR probes for inline and at-line process analytics in pharmaceutical, food and chemical production.
• Improved method transfer protocols and calibration transfer algorithms to compensate for sampling geometry differences between spheres and probe heads.
• Integration with process control systems using the probe’s remote trigger and indicator signals for automated decision-making and faster lot release.
• Advances in fiber technology and detector sensitivity that could extend usable wavelength range and further reduce noise for more challenging low-concentration analyses.

Conclusion

The SabIR fiber optic reflectance probe coupled to the Antaris FT-NIR analyzer demonstrated performance comparable to an integrating sphere in terms of spectral noise, wavelength accuracy and short-term repeatability. Minor spectral differences arise from probe sampling geometry and must be considered when transferring methods. Overall, the probe is suitable for reliable, remote reflectance measurements on containers, packages and process streams, enabling rapid screening and at-line monitoring with minimal sample handling.

Reference

Lowry S., McCarthy B., Verifying the Performance of the Fiber Optic Reflectance Probe on the Thermo Scientific Antaris FT-NIR Analyzer, Technical Note 51670, Thermo Fisher Scientific, TN51670_E 05/08M, 2008.