Rapid Quantitative Determination of Enantiomeric Excess by Vibrational Circular Dichroism Using Noise Based Spectral...

Vibrational circular dichroism (VCD) spectroscopy enables direct determination of enantiomeric excess (ee), but its application as a quantitative analytical tool is limited by low signal-to-noise ratios (SNR), which typically require long measurement times. Using limonene as a model system, predictive models were developed by combining targeted, SNR-based region selection with simple multivariate regression applied to both 30- and 5-min acquisition data. The 30-min data yielded excellent model performance with cross-validated R² of 0.997 and RMSE of ~2% ee. Notably, the 5-min acquisition retained ~97% of predictive performance enabling a sixfold increase in throughput, albeit with increased prediction error. This work therefore demonstrates that data-driven optimisation of spectral selection can enable reliable ee prediction with short acquisition VCD spectra, outperforming equivalent univariate models and providing a practical route toward high-throughput chiral analysis.