Application of a Fourier Transform Infrared (FTIR) Principal Component Regression (PCR) Chemometric Method for the Quantification of Respirable Crystalline Silica (Quartz), Kaolinite, and Coal in Coal Mine Dusts from Australia, UK, and South Africa

Abstract

This article describes the approach used to assess the performance of a Fourier transform infrared (FTIR) and principal component regression (PCR) chemometric method when measuring respirable quartz, kaolinite, and coal in samples from a variety of mines from different countries; relative to target assigned values determined using X-ray diffraction (XRD). For comparison, FTIR results using the partial least squares regression (PLSR) method are also available. Bulk dusts from 10 Australian mines were scanned using XRD and grouped into three sets based on the levels of quartz, kaolinite, and feldspar within their crystalline mineral composition. Prediction samples were generated from 5 of these Australian mine dusts, Durrans coal dust, 2 mine dusts from the UK, and a single South African mine dust (71 samples in total) by collecting the aerosolized respirable dust onto 25-mm diameter polyvinylchloride filters using the Safety in Mines Personal Dust Sampler (SIMPEDS) operating at 2.2 l min−1. The predicted values from the FTIR chemometric methods were compared with assigned target values determined using a direct on-aerosol filter XRD analysis method described in Method for the Determination of Hazardous Substances (MDHS) 101. Limits of detection (LOD) and uncertainty values for each analyte were calculated from a linear regression between target and predicted values. The uncertainty was determined using the calibration uncertainty equation for an unweighted regression. FTIR results from PCR and PLSR are very similar. For the PCR method, the LOD for quartz, kaolinite, and coal were 5, 25, and 71 µg, respectively. For quartz, an LOD of 5 µg corresponds to an airborne quartz concentration of 10 µg m−3, assuming a 4-h sampling time and collection flow rate of 2.2 l min−1. The FTIR measurement met the expected performance criteria outlined in ISO 20581 when sampling quartz for more than 4 h using a flow rate of 2.2 l min−1 at a concentration of 0.1 mg m−3 (100 µg m−3), the current workplace exposure limit in Great Britain. This method met the same performance criteria when measuring exposures at the Australian Workplace Exposure Standard (WES) concentration of 0.05 mg m−3, although in this case a sampling period greater than 8 h was needed.

Atomic spectrometry update – a review of advances in environmental analysis

“…The authors recommended that, to obtain better FTIR data, spectra should be examined for potential matrix interferences, a peak height ratio method should be used for quantification and filters should not be overloaded. In order to minimise such interferences, a PCR chemometric model was developed 54 using coal dust mixtures on filter samples and verified by comparison with the results from XRD analyses. This model allowed quartz to be measured in several coal dust types with a LOD of 5 μg per filter and met the method performance requirements set out in ISO 20581 if airborne silica concentrations of 100 μg m −3 were to be sampled using a nominal 500 L sample volume.…”

Chemometric Techniques in the Assessment of Ambient Air Quality and Development of Air Quality Index of Coal Mining Complex: A Statistical Approach

“…In some reported studies, industrial chemometric analysis is used to understand the movement of pollutants (Grabowski et al, 2021;Vakarelska et al, 2021). Various research has been reported in quantifying the respirable silica emerging from coal mines (Stacey et al, 2022). Still, there needs to be more research on spatial analysis of pollutants emerging from the various activities of the mining operations using chemometric techniques.…”