Asbestos identification in bulk samples using FTIR and multivariate data analysis

Area of Science:

• Analytical Chemistry
• Materials Science

Background:

• Asbestos identification in asbestos-containing materials (ACMs) is crucial for health and safety.
• Traditional methods like polarized light microscopy (PLM) can be time-consuming and require significant expertise.

Purpose of the Study:

• To develop and validate a rapid, cost-effective laboratory procedure for identifying asbestos types in ACMs.
• To combine Fourier Transform Infrared (FTIR) spectroscopy with Partial Least Squares-Discriminant Analysis (PLS-DA) for automated asbestos identification.

Main Methods:

• FTIR spectroscopy utilizing the diffuse reflectance infrared Fourier transform (DRIFT) technique was employed.
• A PLS-DA model was trained using six regulated asbestos reference materials.
• The model's predictive performance was assessed using laboratory-generated and industry-sourced ACM samples, comparing results with standard PLM analysis.

Main Results:

• The PLS-DA model achieved 100% correct classification for single asbestos-type samples and 80% for mixed-asbestos samples.
• High accuracy (96%) was observed for chrysotile-containing samples after specific pre-treatment steps.
• Accuracy decreased for samples with multiple asbestos types, indicating a need for further model optimization.

Conclusions:

• The proposed FTIR-PLS-DA method provides a rapid, cost-effective, and potentially less experience-dependent approach for asbestos identification.
• Further refinement with larger datasets is necessary to improve accuracy for complex mixed-asbestos samples.
• This technique shows promise for enhancing asbestos detection efficiency in industrial and remediation settings.