Principle and Overview of Intelligent Peak Deconvolution Analysis (i-PDeA II)

Significance of the Topic

Photodiode array (PDA) detectors provide rich spectral and chromatographic information, but overlapping or co-eluted peaks often complicate component identification and quantitation. The i-PDeA II method integrates chemometric multivariate curve resolution–alternating least squares (MCR-ALS) with a bidirectional exponentially modified Gaussian model to resolve such peaks, improving productivity and data reliability in analytical laboratories.

Objectives and Overview of the Study

This report presents the theory and application of the i-PDeA II algorithm for peak deconvolution of PDA data. It aims to demonstrate how target peaks can be extracted from unresolved chromatograms, even when standard samples for pure components are unavailable, and to evaluate spectral identification and quantitative performance using positional isomers of methylacetophenone.

Methodology and Instrumentation

The core deconvolution approach models the PDA data matrix D as the product of concentration profiles (C) and spectral profiles (S), plus a residual term. By iteratively estimating C and S under least-squares constraints (MCR-ALS), and constraining chromatogram shapes using a bidirectional exponentially modified Gaussian (BEMG) function, i-PDeA II separates co-eluted peaks. Instrumentation used:

• HPLC System: Shimadzu LC-2030C 3D
• Column: Shimadzu Shim-pack XR-ODS III C18 (3.0 × 50 mm, 2.2 μm)
• Mobile Phase: Methanol/water (30/70, v/v)
• Flow Rate: 1.0 mL/min, Oven Temperature: 40 °C
• Detection: PDA at 190–400 nm, slit width 1.2 nm, sampling 240 ms
• Injection Volume: 1.5 μL

Main Results and Discussion

1. Standard isomers (o-, m-, p-methylacetophenone) showed high pairwise spectral similarities (>0.84), yet i-PDeA II distinguished them reliably.
2. In a mixture with relative concentrations 100:1:100 (o:p:m), the p-isomer peak was visually hidden but successfully deconvoluted in the 5.0–7.0 min and 210–320 nm window.
3. Purity analysis revealed a 2.15 % m-isomer impurity in the p-isomer standard, confirmed by remeasurement on a reversed-phase phenyl column.
4. Quantitative evaluation against individual standards yielded errors below ±1 % for major peaks and <±4 % for the minor component, with spectral similarity >0.9996.

Benefits and Practical Applications of the Method

• Enables quantitation without full chromatographic separation, reducing analysis time.
• Separates co-eluted isomers sharing identical molecular weights.
• Detects impurities in commercial standards and complex mixtures.
• Integrates seamlessly with LabSolutions software for full data handling and library searches.

Future Trends and Applications

The integration of advanced chemometric algorithms with high-resolution detectors will continue to expand the scope of PDA applications. Potential developments include real-time deconvolution in complex matrices, coupling i-PDeA II with UHPLC for faster analysis, and extending the approach to other multi-dimensional detectors.

Conclusion

i-PDeA II leverages chemometrics and chromatogram modeling to resolve challenging co-elutions in PDA datasets, achieving accurate spectral identification and quantitation even in trace-level components. Its implementation in LabSolutions enhances laboratory efficiency, making it a valuable tool for analytical chemists tackling complex mixtures.

Reference

1. Hasegawa T. Quantitative Spectral Analysis. Kodansha; 2005.
2. Hasegawa T. Bunseki. 2014;9:460–467.
3. Gemperline P., ed. Practical Guide to Chemometrics. 2nd ed. CRC Press; 2006.
4. Tauler R, Barceló D. Trends Anal Chem. 1993;12:319–327.
5. Tauler R. Chemometr Intell Lab. 1995;30:133–146.
6. Parastar H, Tauler R. Anal Chem. 2014;86:286–297.
7. Arase S, et al. J Chromatogr A. 2016;1469:35–47.
8. Sakuma I, et al. J Chromatogr A. 1990;506:223–243.