Automation, hyphenation and multidimensional chromatography for PAH analysis : Analytical techniques to simplify sample complexity

Sammanfattning: One group of organic environmental pollutants which are potentially hazardous to human health is polycyclic aromatic hydrocarbons (PAHs). These compounds consist of two or more benzene rings in their structure among which benzo[a]pyrene (B[a]P) is classified as a human carcinogen. In order to generate accurate data for PAHs, reliable analytical methods are a prerequisite for PAH monitoring in the environment and cancer risk assessments. However, there are several analytical challenges in PAH analysis because they are a large number of compounds with different physical and chemical properties and occur at varying concentration levels, often in complex matrices. Thus, this thesis aimed at tackling these difficulties in PAH analysis using various analytical techniques such as automation, hyphenation and multidimensional chromatography, particularly to increase the detectability and separation efficiency of PAHs in complex matrices. Furthermore, analytical methods were also presented as a tool for toxicological studies.A fully automated two-dimensional (2D)-liquid chromatography (LC)/2D-gas chromatography system was developed to enable the online sample clean-up, separation and detection, initially with two flame ionization detectors (FIDs) (Paper I). The introduction of mass selective detectors (MSDs) instead of two FIDs further advanced the original 2D system, improving the detectability and selectivity in PAH analysis (Paper II). The PAH levels were determined in two standard reference materials (SRMs) from the National Institute of Standards and Technology (NIST), i.e. urban dust (SRM 1649a) and diesel particulate extract (SRM 1975). The measured PAH levels were in good agreement with those obtained from FID- and MSD-based systems and reported by NIST (Paper I and II). Additionally, the PAH determination was done in particulates from wood burning, which also showed comparable results between FID and MSD systems (Paper II). The FID-based system showed slightly better separation efficiency owing to the use of hydrogen as carrier gas instead of helium and similar detectability to that of MSD (Paper I). However, the MSD-based system was advantageous to detecting late-eluting PAHs and separating interfering peaks in complex matrices (Paper II).Paper III presented an online LC system for B[a]P-selective fractionation as a tool for studying biological interactions. Fractions could be obtained from Stockholm air particulate matter (PM) extract with or without B[a]P. The selective fractionation was achieved using serial LC columns of two pyrenyl phases, resulting in more than 90% of B[a]P in the B[a]P-containing fraction. An analytical method for identification and quantification of benz[j]aceanthrylene (B[j]A) in air PM was developed and described in paper IV. B[j]A is rarely measured due to its low abundance but is known to have high carcinogenic potential. The determination of B[j]A in air PM from Stockholm (Sweden) and Limeira (Brazil), was done together with a series of toxicological studies. The measurement and toxicological data showed an increased number of estimated cancer cases from air PM exposure when levels of B[j]A were taken into account.In conclusion, this thesis presents various analytical approaches to obtain more accurate PAH data as well as the possibilities of using them in toxicological research.