Electrospray Ionization Efficiency Predictions and Analytical Standard Free Quantification for SFC/ESI/HRMS.

Area of Science:

• Analytical Chemistry
• Chromatography
• Mass Spectrometry

Background:

• Supercritical fluid chromatography (SFC) offers a sustainable alternative to liquid chromatography (LC).
• Coupling SFC with high resolution mass spectrometry (HRMS) is effective for nontarget screening (NTS).
• Predictive models for LC/ESI/HRMS enable quantification without analytical standards.

Purpose of the Study:

• To evaluate analytical standard-free quantification in SFC/ESI/HRMS.
• To assess the transferability of LC/ESI/HRMS ionization efficiency models to SFC/ESI/HRMS.
• To develop and validate a predictive model using SFC/ESI/HRMS data.

Main Methods:

• Trained random forest regression models using PaDEL descriptors for ionization efficiency prediction.
• Collected data for 127 chemicals using SFC/ESI/HRMS and LC/ESI/HRMS.
• Compared predictive model performance using SFC-specific and LC-derived data.

Main Results:

• Ionization efficiency predictions showed statistically significant correlation with measured response factors (Spearman's rho = 0.584 for SFC, 0.669 for LC).
• Models trained on SFC data yielded high prediction accuracy (median error 2.20×).
• Models pre-trained on LC data showed utility for SFC/ESI/HRMS, despite higher median error (5.11×).

Conclusions:

• Analytical standard-free quantification is feasible for SFC/ESI/HRMS.
• SFC-specific models offer superior accuracy for predicting ionization efficiency.
• LC/ESI/HRMS data can inform and improve SFC/ESI/HRMS ionization behavior predictions.