IonicX Portable XRF Analyzer Device and Algorithm

Importance of the topic

The ability to rapidly and reliably verify ionic salt identity at point-of-use is important for quality control, supply chain verification and regulatory compliance in pharmaceutical, chemical and industrial settings. Portable X-ray fluorescence (XRF) devices that combine material-specific elemental and scattering signatures with automated algorithms can replace lengthy wet-chemistry assays for many inbound chemical checks, reducing cost, turnaround time and handling risk while providing audit-ready pass/fail outcomes.

Objectives and overview of the study

This white paper describes the Thermo Scientific IonicX Portable XRF Analyzer and the algorithmic approach used to authenticate ionic salts. Goals are to explain the underlying XRF physics exploited by the instrument, to show how spectral features (fluorescence and scatter) are used to distinguish chemically similar salts, and to demonstrate the device’s selectivity and clear pass/fail decision logic for a set of common ionic compounds.

Methodology and algorithm

• Spectral acquisition strategy: A fast low-voltage pre-scan is used first to elicit fluorescence peaks from medium-Z elements commonly present in salts (K, Cl, Mg, Ca, S). Detected fluorescence peaks are used to pre-filter the calibration library, eliminating candidate compounds lacking the observed elements.
• Limitations of fluorescence: Handheld ED-XRF typically cannot detect elements lighter than magnesium by their fluorescence lines. This limitation means elemental detection alone is insufficient when different compounds contain only the same detectable elements.
• Use of scattered continuum: The algorithm leverages differences in the Rayleigh and Compton scatter continuum (Bremsstrahlung scatter from the X-ray tube) to discriminate compounds that have identical detectable fluorescence peaks but different chemical composition (for example, NaClO3 vs NaClO4). Differences in mass attenuation coefficients caused by additional atoms (oxygen content differences) produce subtle but reproducible changes in the scatter region (e.g., 25–40 keV) after normalization.
• Library matching and linear fit metric: For candidate matches remaining after pre-filtering, IonicX performs a linear least-squares comparison of the unknown spectrum to each library spectrum using a linear fitting model (slope and intercept). A correlation-like score (C-Val) is computed as C-Val = 1.0 – sqrt(slope^2 + offset^2) (where offset is the intercept). A perfect match would approach slope = 1 and intercept = 0, giving C-Val near 1.0.
• Decision threshold: A simple binary decision is applied: C-Val ≥ 0.7 is treated as a pass (match) and C-Val < 0.7 as a fail. This yields clear, unambiguous outcomes for the tested methods/compounds.
• Multi-voltage interrogation: If elemental pre-scan indicates a mixed-element signature (e.g., Mg and Cl), the system performs a high-voltage scan and uses the primary algorithm including scatter-region analysis to distinguish true salts (e.g., MgCl2) from physical mixtures (e.g., MgO + NaCl) which produce different scatter profiles.

Instrumentation used

• Thermo Scientific IonicX Portable XRF Analyzer (handheld ED-XRF device).
• Silver (Ag) anode X-ray tube used in the examples; characteristic Ag lines (Ag La1 at 22.16 keV and K at higher energies) contribute both characteristic peaks and Rayleigh/Compton scattered features in measured spectra.
• Energy-dispersive semiconductor X-ray detector for capturing characteristic fluorescence and scattered X-rays.
• Spectral normalization is commonly performed to a primary Compton or Ag K feature to compare continuum shapes between samples.

Main results and discussion

• Spectral examples: A KCl spectrum measured with an Ag tube shows distinct K and Cl fluorescence peaks superimposed on a continuum of scattered Bremsstrahlung with Rayleigh and Compton components; Ag tube characteristic lines also appear as Rayleigh-scattered peaks. The Ag La1 line (22.16 keV) can produce a Compton-scatter feature at lower energy (~20.6 keV) in the incoherent region.
• Discrimination using scatter: NaClO3 and NaClO4—both only show detectable Cl fluorescence on IonicX—exhibit small but reproducible differences in the normalized scatter region between ~25 and 40 keV due to the extra oxygen atom in NaClO4, enabling correct differentiation when full-spectrum matching is applied.
• Algorithm performance (representative C-values): When each method in the device’s library was challenged by samples of the five tested chemicals, the correct method returned high C-values while other method-sample pairs returned near-zero values. Representative diagonal (correct match) C-values reported were: NaCl -> 0.94, KCl -> 0.96, MgCl2 -> 0.91, CaCl2 -> 0.83, NaOH -> 0.88. Using the 0.7 threshold produced clear pass/fail outcomes with minimal ambiguity.
• Robustness to mixtures: The combined strategy of elemental pre-filtering plus scatter-region analysis reduces false positives from simple physical mixtures (e.g., MgO + NaCl) by exploiting differences in mass absorption and continuum shape that mixtures produce versus single-compound spectra.

Benefits and practical applications

• Speed and efficiency: Portable XRF verification dramatically shortens incoming material checks compared with wet-chemistry methods.
• Non-destructive and safe sample handling: Minimal sample preparation and no consumptive reagents reduce handling risk and waste.
• Clear decision logic: The C-Val thresholding approach provides straightforward pass/fail outputs for operational use and audit trails suitable for QA/QC workflows.
• Portability and field use: Small form factor enables verification at receiving docks, warehouses and in-field environments.
• Applicability: Especially useful for pharmaceutical raw-material verification, manufacturing QA, customs/supply chain checks and any context where rapid identity confirmation of ionic salts is needed.

Future trends and potential applications

• Algorithmic enhancements: Incorporation of multivariate or machine learning classifiers trained on full-spectrum features could further improve discrimination among complex mixtures and lower C-Val ambiguity for edge cases.
• Expanded libraries and metadata: Larger, curated spectral libraries including known mixtures, polymorphs and typical contaminants would broaden applicability and robustness.
• Detector and source improvements: Advances that lower the detectable-energy floor or increase energy resolution would permit detection of lighter elements and improve sensitivity where oxygen or sodium differences are critical.
• Multi-modal verification: Combining XRF with complementary portable techniques (e.g., Raman, NIR) can provide orthogonal confirmation for compounds with weak fluorescence signatures.
• Integration with digital QA systems: Automated reporting, cloud-managed libraries, and traceable audit logs will improve compliance and operational efficiency in regulated environments.

Conclusion

The IonicX Portable XRF Analyzer leverages energy-dispersive XRF physics combined with a pragmatic algorithm that uses elemental pre-filtering, scatter-region analysis and linear-spectrum matching to deliver rapid, unambiguous verification of common ionic salts. While ED-XRF cannot directly detect light elements below Mg by fluorescence, the exploitation of Rayleigh/Compton scatter differences and normalized full-spectrum matching provides effective discrimination between chemically distinct salts and common mixtures. The device offers a practical on-site alternative to slower wet-chemistry assays for many industrial verification tasks.

Reference

Thermo Fisher Scientific. IonicX Portable XRF Analyzer Device and Algorithm. White paper MCS-AN1649-EN, November 2025.