multivariate curve resolution-alternating least squares for comprehensive resolution and quantification of polycyclic aromatic hydrocarbons in oil fractions by means of gc×gc-tofms

Abstract: pahs are a group of organic pollutants with adverse effects on both humans and the environment. they are mainly a fraction of petroleum hydrocarbons and various types of pahs are discharged into the environment by both natural and anthropogenic activities, therefore the determination and quantification of pahs are of prime importance [1]. when comprehensive two-dimensional chromatographic techniques are coupled with a high acquisition rate detector, like a time-of-flight mass spectrometer (tofms), the result is an extremely powerful instrument (e.g., gc × gc−ms and lc × lc−ms) for the analysis of very complex natural samples. however, one of the main challenges in multidimensional chromatography is related to the difficulty of the analysis and interpretation of the enormous amount of data obtained in these cases. additionally, complete separation of all detectable components still cannot be achieved because of the extremely high complexity of natural samples and limitations in experimental and instrumental conditions [2]. chemometric methods can be used to solve this issue and exploit pure analyte signal from a complex mixture (second-order advantage). in this regard, multivariate curve resolution-alternating least squares (mcr-als) is frequently used for gc × gc data [3]. up to now, gc × gc-tofms combined to mcr-als has been proposed for the resolution and quantification of very complex mixtures of compounds such as pahs in heavy fuel oil (hfo) [4]. this approach is based on the analysis of the chromatographic segments. however, segmented analysis is very time consuming and complex, therefore, in this study a new mcr strategy is proposed for full chromatogram analysis for the determination of 13 pahs. gc × gc-tofms data arranged in a column-wise augmented data matrix and then the time direction was reduced using discrete wavelet transform which was allowed reducing the size of the data. then, compressed augmented data matrix was analyzed by mcr-als using spectral normalization and non-negativity constraint during the als optimization procedure. also, orthogonal projection approach (opa) was used to calculate initial spectral estimates to start als optimization. then, calibration curves based on mcr-als built up with standard samples of pahs and analytical figures of merits (afoms) were calculated. finally, determination of target pahs in aromatic fraction of different oil fractions such as north sea oil was performed using proposed strategy and proper quantitative results were also obtained.