Application of Principal Component Analysis for the Data Mining of High Resolution Mass Spectrometry Datasets

Abstract

The release of oil sands process-affected water (OSPW) from tailings ponds is a major environmental issue that oil sands companies must consider over the next decade. Advanced oxidation processes (AOPs) and biological treatment processes have been shown to be able to degrade contaminant compounds and reduce toxicity of this OSPW. However, these processes are also associated with by-products generation which may be of environmental concern. This study successfully combined High Resolution Mass Spectrometry (HRMS) as an analytical tool to detect organic compounds (markers) in OSPW samples and Principal Component Analysis (PCA) as a statistical tool to manage the extensive HRMS datasets (over 1000 markers per sample). The HRMS and PCA were used to determine the markers most significantly changed during ozonation in different conditions and biological treatment processes and to determine their potential by-products. Based on m/z values, all the significant markers selected by PCA were designated into groups including naphthenic acids (NAs), oxidized NAs and unknown compounds. Of these markers, the main focus in this study was the unknown compounds given their trends in OSPW treatment processes have not been assessed previously. The significant unknown markers which decreased over treatment time were degraded during treatments; while those which increased over time were by-products of organic compounds found in raw OSPW treated by ozonation in different conditions and biological treatment processes. There were negligible or very low concentrations of compounds which were identified as by-products in ozonation in different conditions and biological treatment processes found in different raw OSPWs (Syncrude West in Pit, Suncor Pond 7 and CNRL OSPW). These compounds in raw OSPWs showed negative correlations to NAs concentrations and positive correlations to oxidized NAs concentrations, which indicate their close association with NAs degradation via oxidation. This study demonstrates an advanced approach to determining by-products that can be further used for any chemical or biological treatment process. Further research aimed at the identification of by-products structures and determination of potential degradation mechanisms will be useful in assessing treatment efficiency of OSPW compounds.