Correlation constrained multivariate curve resolution for determination of polycyclic aromatic hydrocarbons in oil contaminated sea waters extracted by dispersive liquid-liquid micro-extraction and analyzed by gas chromatography

Polycyclic Aromatic Hydrocarbons (PAHs) which are coming from incomplete combustion processes are very toxic organic pollutants and they can cause serious effects on the human health [1]. Due to the complexity of sample matrices including PAHs (e.g., oil contaminated waters), development of sensitive and selective analytical methods including a sample preparation technique followed by a separation technique (e.g., gas chromatography (GC)) is required for determination of PAHs. In this study, ultrasonic assisted extraction-dispersive liquid-liquid micro-extraction (UAE-DLLME) followed by gas chromatography-flame ionization detector (GC-FID) was proposed for simultaneous determination of thirteen polycyclic aromatic hydrocarbons (PAHs) in sea waters exposed to oil spills. The optimum extraction conditions obtained by response surface methodology (RSM) were 10 min sonication time, 14μL of tetrachloroethylene as extraction solvent and 1.2% (w/v) salt concentration. In order to have a comprehensive analysis, correlation constrained multivariate curve resolution-alternating least squares (CC-MCR-ALS) as a first-order calibration algorithm was used for resolution and quantification of the 13 target PAHs in the presence of uncalibrated interferences [2]. In this way, orthogonal projection approach (OPA) was used for estimation of the initial spectral values to start ALS optimization. In this regard, calibration set for external calibration was randomly designed in concentration range 0.1-100 ngmL-1. Then, multivariate analytical figures of merit (AFOM) including sensitivity (SEN), selectivity (SEL) and limit of detection (LOD) were calculated [3]. In all cases, the AFOMs were in acceptable ranges which confirm the validity of the proposed method. Finally, the proposed chemometrics-based strategy was successfully applied for determination of 13 PAHs in oil contaminated sea water after removal of oil spill using standard addition calibration. On this matter, all 13 PAHs were found in sea water in concentration range 8.0-300.0 ngmL-1 with relative standard deviation (RSD) below 5.6 %. It is concluded that combination of multivariate calibration methods with DLLME-GC-FID can be a straightforward and rapid technique to improve the identification and quantification ability of these methods for the simultaneous determination of target compounds in different sample matrices.