Analysis of Diamonds by FT-IR Spectroscopy

Importance of the topic

Fourier-transform infrared (FT-IR) spectroscopy provides a rapid, nondestructive and highly specific approach to identify diamonds, to classify their point-defect chemistry and to flag evidence of synthetic growth or post-growth treatments. Because diamonds are pure carbon with trace impurities (notably nitrogen, hydrogen and boron) that strongly influence IR absorption signatures, FT-IR is uniquely suited to verify that a faceted stone is genuine diamond and to provide diagnostic information used by gemological laboratories, commercial buyers and quality-control labs in industry.

Objectives and overview

This application note demonstrates practical FT-IR approaches for:

• fast confirmation that a faceted stone is diamond (material ID),
• classification of diamond types based on nitrogen aggregation states (IaA, IaB, Ib, IIa, IIb),
• detection and quantification of diagnostic impurity peaks including platelets (~1360 cm-1), and
• supporting evidence for synthetic (HPHT-grown) or treated diamonds using multivariate and curve‑fitting analyses.

Methods illustrated include automatic similarity matching against a Type IIa reference, Classical Least Squares (CLS) modeling for relative quantitation of nitrogen aggregate components and peak-resolving curve fitting to characterize subtle peak shifts and asymmetry.

Instrumentation

Key instrument configuration and accessories described:

• Thermo Scientific Nicolet 6700 FT-IR spectrometer (benchmarked for gemological use),
• optimized 4X beam condenser or reflection accessory for high-quality spectra from faceted stones,
• TQ Analyst multivariate analysis software for automated similarity-matching and CLS workflows,
• software Peak Resolve™ routine for curve fitting / peak deconvolution.

The described setup enables rapid non-contact measurements of faceted diamonds with acquisition and automated analysis often completed in less than one minute per stone.

Methods and data analysis

Material confirmation: A similarity-match algorithm compares the sample spectrum to a Type IIa diamond reference and computes a match score from 0 to 100. In practice a pass threshold near 80 was found to separate diamonds from non-diamonds reliably while allowing for natural spectral variability. The phonon region (approx. 1500–4000 cm-1) contains the lattice bands characteristic of diamond; Type IIb stones with high boron levels can significantly distort this region and reduce match scores.

Peak deconvolution: Many diamond spectra show a platelet-related absorption near 1360 cm-1. Curve fitting (peak-resolving) using defined line shapes was applied to separate overlapping components, quantify peak positions and widths, and reveal asymmetry. Example results showed a shift from 1359 to 1364 cm-1 with an increase in full width at half maximum (FWHM) from 3.2 to 4.8 cm-1; fits often required two synthetic peaks to account for asymmetric line shapes.

Classical Least Squares (CLS): CLS modeling treated diamond spectra as linear combinations of reference spectra from diamonds with known nitrogen aggregate states (IaA, IaB, Ib). For illustrative reporting, the study assigned an arbitrary 100 ppm value to references so outputs are relative rather than absolute concentrations. CLS also returns standard error terms, allowing a pragmatic decision rule: when the measured peak intensity is roughly three to five times the reported standard error, the presence of that feature is considered confirmed. CLS was used both to quantify aggregate contributions and to detect weak peaks at 3107, 1344 and 1332 cm-1 indicative of trace nitrogen/hydrogen features.

Main results and discussion

Primary practical findings:

• FT-IR readily discriminates diamond from simulants by the presence of diamond lattice phonon bands in the 1500–4000 cm-1 window.
• Similarity matching against a Type IIa reference with an acceptance threshold around 80 provided robust automatic material confirmation with few false positives; exceptions include Type IIb (boron-rich) diamonds which can reduce match quality due to phonon distortion.
• Curve-resolved analysis of the platelet feature near 1360 cm-1 demonstrates that peak position, width and asymmetry vary with sample chemistry; these parameters can serve as additional discriminants.
• CLS modeling permits relative estimation of nitrogen aggregate types (IaA, IaB, Ib) and detection of low-level impurities. The method's standard error output is a useful quality metric for low-concentration features.
• Low-intensity diagnostic peaks at 3107, 1344 and 1332 cm-1 can be detected and confirmed by CLS when signal exceeds several times the method error, aiding identification of treated or synthetic material where nitrogen is scarce.

Together, these approaches provide complementary evidence: material identity (diamond vs. simulant), defect chemistry (nitrogen aggregation and platelets), and anomalies (boron or unusual impurity patterns) that may indicate HPHT synthesis or post-growth treatment.

Benefits and practical applications

FT-IR with optimized optics and multivariate workflows offers several operational advantages:

• rapid, nondestructive screening of large numbers of stones, reducing expert time spent on clear pass/fail samples,
• objective, automatable decision metrics (similarity match scores, CLS coefficients and standard errors) that improve reproducibility and traceability,
• ability to classify diamond types relevant for valuation and treatment assessment (e.g., distinguishing Type Ia from IIa/IIb and detecting HPHT indicators),
• quantitative or semi-quantitative information on defect populations useful for gemological reporting and research.

These capabilities are valuable for gemological laboratories, commercial verification at point-of-sale, and research groups studying growth mechanisms or treatment effects.

Future trends and potential applications

Likely developments and opportunities include:

• integration of larger, standardized spectral reference libraries and community-shared datasets to improve CLS/ similarity-match robustness,
• application of advanced machine learning and chemometric classifiers to handle broader spectral variability and increase sensitivity to subtle synthetic/treatment markers,
• combination of FT-IR with complementary techniques (photoluminescence, Raman, cathodoluminescence) for more definitive synthetic/treatment attribution,
• improvements in accessory design (higher NA condensers, microscope coupling) and detector sensitivity to better resolve weak impurity bands in smaller or low-nitrogen stones,
• automation and high-throughput workflows for routine screening in commercial and auction contexts.

Conclusion

FT-IR spectroscopy, when combined with optimized optics and multivariate analysis (similarity matching, CLS and curve fitting), is a fast, reliable and nondestructive tool for diamond identification, defect-type classification and for providing evidence suggestive of synthetic growth or treatment. The techniques described enable objective screening with confidence metrics, allowing routine samples to be processed quickly while concentrating expert evaluation on ambiguous or suspicious stones.

References

Lowry S. Analysis of Diamonds by FT-IR Spectroscopy, Application Note 51122, Thermo Fisher Scientific, 2006/2008.