Decoding PFAS contamination via Raman spectroscopy: A combined DFT and machine learning investigation

Area of Science:

• Environmental Chemistry
• Analytical Chemistry
• Computational Chemistry

Background:

• Perfluoroalkyl substances (PFASs) are prevalent environmental contaminants requiring robust detection methods.
• Existing analytical techniques for PFAS identification can be complex and time-consuming.
• Understanding the unique spectral properties of different PFAS is crucial for accurate environmental monitoring.

Purpose of the Study:

• To compute and analyze Raman spectra for 40 key Perfluoroalkyl substances (PFASs) using Density Functional Theory (DFT).
• To identify characteristic Raman peaks, vibrational modes, and spectral regions associated with specific PFAS chemical bonds and functional groups.
• To investigate the influence of molecular structure (isomer branching, chain length, functional groups) on PFAS Raman spectra.

Main Methods:

• Density Functional Theory (DFT) calculations were performed to simulate Raman spectra.
• Systematic comparison of spectral features, including peak locations and vibrational modes, was conducted.
• Chemometric techniques, Principal Component Analysis (PCA) and t-distributed Stochastic Neighbor Embedding (t-SNE), were applied to spectral data.
• A spectral database was created with controlled noise for enhanced differentiation.

Main Results:

• Specific Raman spectral regions were identified for key chemical bonds (C-C, CF2, CF3) and functional groups (-COOH, -SO3H, etc.).
• Molecular structure variations (isomer branching, chain length, functional groups) were shown to significantly impact spectral features and peak locations.
• PCA and t-SNE effectively distinguished between the 40 PFAS compounds and their isomers based on their computed Raman spectra.

Conclusions:

• Raman spectroscopy, combined with DFT computations and chemometric analysis, offers a powerful approach for discriminating between various PFAS compounds.
• The study provides a foundation for developing advanced, rapid, and accurate methods for PFAS detection and characterization in environmental samples.
• The generated spectral database and analytical methodologies hold significant promise for improving environmental monitoring and regulatory compliance.