Simple Method Transfer from FOSS NIR to Antaris FT-NIR Analyzers

Significance of the topic

Near-infrared (NIR) spectroscopy is widely used in industry for raw material identification and multicomponent quantitative analysis because it is rapid, non-destructive, and suitable for in-line or at-line monitoring. Large investments in calibration libraries and primary laboratory reference data create a strong need to transfer validated methods between instrument models and vendors without re-collecting thousands of spectra. Demonstrating reliable, low-effort transfers between dispersive NIR systems (FOSS) and Fourier-transform (FT) NIR analyzers (Thermo Scientific Antaris) reduces vendor lock-in, lowers cost, and accelerates adoption of newer, higher-performance instruments.

Objectives and overview of the study

This technical note demonstrates a practical workflow to transfer two common types of NIR applications from FOSS dispersive instruments (WinISI and Vision software) to Thermo Scientific Antaris FT-NIR analyzers. The goals were to show that (1) spectral and library data can be converted reliably between instrument technologies, (2) primary laboratory values can be preserved and imported automatically, and (3) the transferred methods perform equivalently when used on Antaris instruments. The note emphasizes automation and reproducibility by using a Standards Converter utility together with TQ Analyst chemometric software and RESULT instrument control/validation software.

Methodology

The transfer workflow is composed of three principal functions performed by the Standards Converter and subsequent steps in TQ Analyst:

• File format conversion: export from FOSS WinISI or Vision into a single JCAMP multifile (.jcp) or ASCII (.txt), then convert to JCAMP-DX multifiles (.jdx) suitable for import.
• Spectral unit and baseline conversion: automatically convert wavelength units (nanometers or microns) to wavenumbers (cm-1) and convert spectra reported as transmittance or reflectance into absorbance units expected by Antaris FT-NIR/TQ Analyst workflows.
• Standards and metadata import: automatically launch TQ Analyst and populate the standards table with spectra plus corresponding primary laboratory (reference) values, preserving the original method’s calibration targets.

For multicomponent PLS models (WinISI): the WinISI calibration was exported to JCAMP multifile format, converted via Standards Converter to JCAMP-DX, and imported into TQ Analyst. The transferred method was re-created in TQ Analyst with the original preprocessing and model parameters (first/second derivatives, SNV, MSC, smoothing). The method was then inoculated (augmented) with spectra collected on the Antaris to improve model performance when required.

For raw material identification libraries (Vision): Vision libraries were exported as ASCII (.txt) and converted to JCAMP-DX via the Standards Converter. The FOSS concept of a product (a set of standards for a material) was mapped to a TQ Analyst class. Spectra were converted to wavenumber and absorbance units and automatically imported; the library was re-created in TQ Analyst using the original identification settings and then validated on Antaris-collected spectra.

Used instrumentation

• Thermo Scientific Antaris FT-NIR analyzers (Antaris II shown)
• Thermo Scientific Standards Converter software (file and unit conversion, automated import)
• TQ Analyst chemometric package (spectral preprocessing, PLS modeling, library management)
• Thermo Scientific RESULT software for instrument control and validation reporting
• Source systems and formats: FOSS NIR instruments using WinISI and Vision software; JCAMP (.jcp/.jdx) and ASCII (.txt) file formats

Main results and discussion

The note reports successful, straightforward transfers for both application classes with minimal user effort. Key outcomes and observations include:

• Automated conversion preserved primary laboratory values and inserted them directly into TQ Analyst standards tables, eliminating manual re-entry and reducing transcription risk.
• Unit conversions (nm/µm to cm-1) and transformation to absorbance were handled automatically and consistently, overcoming perceived incompatibilities between dispersive and FT spectra.
• TQ Analyst supports the common preprocessing options used in WinISI (1st/2nd derivative, SNV, MSC, smoothing), enabling faithful reproduction of original model parameters.
• Inoculation of the transferred calibration with a subset of spectra collected on the Antaris is recommended and commonly improved accuracy and robustness.
• Validation using RESULT software showed successful positive identification for a caffeine challenge sample, demonstrating that transfer did not degrade identification performance.

The authors conclude that differences in instrument technology (dispersive vs FT) are not a fundamental barrier to method transfer when proper file and unit conversions are applied and when spectra are represented in compatible units and baseline conventions.

Benefits and practical applications

Practical advantages of the described transfer approach include:

• Time and cost savings by avoiding recollection of extensive libraries and re-analysis of primary reference data.
• Improved vendor independence — users can upgrade hardware or change platforms without losing validated methods.
• Reduced human error due to automatic import of standards and metadata.
• Rapid deployment: a simple three-step process (convert files, convert units/absorbance, import into TQ Analyst) enables scalable migration even for very large libraries.
• Applicability to both identification libraries (hundreds to thousands of spectra) and quantitative PLS calibrations.

Future trends and possibilities

Potential developments that will further simplify and strengthen method transfer and cross-platform NIR analytics include:

• Standardized, vendor-neutral spectral and metadata formats to streamline interoperability across instrument families.
• Cloud-based centralized calibration repositories that support versioning, traceability, and distributed instrument networks.
• Automated validation workflows and continuous monitoring (transfer-validation pipelines) for routine assurance of method performance after transfer.
• Machine-learning approaches to adjust or re-calibrate models automatically when systematic spectral differences remain after conversion.
• Tighter international standards and certification for inter-instrument consistency to facilitate regulated-industry adoption.

Conclusion

This technical note demonstrates that transferring NIR methods from FOSS dispersive instruments to Thermo Scientific Antaris FT-NIR analyzers is feasible, efficient, and reliable using a Standards Converter plus TQ Analyst workflow. Automated file and unit conversions together with direct import of primary laboratory values remove the primary logistical obstacles to transfer. With appropriate preprocessing reproduction and optional inoculation using Antaris spectra, transferred methods maintain or improve performance for both identification libraries and multicomponent PLS calibrations. The approach reduces cost, time, and vendor lock-in while enabling laboratories to adopt FT-NIR technology without sacrificing existing validated methods.

Reference

Technical Note 51869: Simple Method Transfer from FOSS NIR to Antaris FT-NIR Analyzers, Chris Heil, Thermo Fisher Scientific, Madison, WI, USA (2010).