Successful PhD Defense of Saer Samanipour

Abstract:
Persistent and bioaccumulative pollutants (PBPs) are continually introduced into the environment as a part of the massive ongoing chemical production that began several decades ago. The PBPs include several chemical families such as: industrial compounds, personal care products, agricultural chemicals, and pharmaceuticals. Most of these PBPs are neutral and more than 60% of them are halogenated. However the concentration distributions in the environment, partitioning properties, and environmental fate and behavior of many PBPs have not been investigated. Recent studies have reported on the lack of information regarding the occurrence, fate, and behavior of these PBPs in the environment (Howard and Muir 2006, 2010 and 2013). The authors emphasized the need for measurements of these PBPs in different environmental compartments in order to better understand their environmental fate and behavior.

In this thesis, I investigated the occurrence of legacy and novel PBPs in a deep aquatic system, Lake Geneva. Measuring these compounds in environmental samples is a challenging task due to their trace level concentrations and due to the complexity of the samples, manifest as matrix effect. I employed comprehensive two-dimensional gas chromatography (GC×GC) to tackle these challenges. Throughout the thesis I refer to “novel PBPs” as PBPs that are neutral, organic, non-legacy, and that have not been measured in the environment. This terminology is similar to that adopted by Howard and Muir, 2010. I report results for several water column and sediment samples that were analyzed for a suite of 69 PBPs, including novel PBPs, PBDEs, PCBs, OCPs and halogenated benzenes. This leads to the first reported detection and quantification for several novel PBPs (i.e. 4-bromobiphenyl (4BBP), tribromobenzene (TBB), and pentachlorothiophenol (PCTP)) in a lake environment. Results for several legacy PBPs, including PBDEs and PCBs, are also reported.

In Chapter 2 of the thesis I developed an analytical protocol for detection, quantification, and identity confirmation of trace level PBPs in environmental samples. This method took advantage of the separation power of GC×GC combined to highly sensitive detectors, including electron capture negative chemical ionization (ENCI)-TOFMS, micro electron capture detector (μECD), and flame ionization detector (FID). Chapter 2 evaluates the effectiveness of the application of GC×GC-μECD for the detection and quantification of trace-level PBPs in the lake environment. In particular, I investigate automated baseline correction and peak delineation algorithms for their ability to remove matrix effect and quantify trace level PBPs in complex environmental samples. By employing a suite of chemometric tests, I systematically assessed different baseline correction and peak delineation algorithms for their confidence and accuracy of target analyte quantification. The results of chemometric tests showed the crucial importance of the baseline correction algorithm for accurate peak integration. An aggressive baseline correction method systematically produced the best results for the chemometric tests, which indicated a better matrix effect removal. The results of the analytical protocol were also validated using a certified reference material. The validated analytical procedure led to the successful detection and quantification of 18 trace level target analytes, including 7 PAHs in a light diesel fuel and 11 chlorinated hydrocarbons in a lake water extract. This chapter resulted in a sensitive and accurate protocol for detection, identity confirmation, and quantification of trace level PBPs in environmental samples. Finally, this chapter also provides guidance for diagnosis of the matrix effect and biased calibration during the quantification of analytes using GC×GC.

Chapters 3 and 4 of the thesis investigate the occurrence of novel and legacy PBPs in the water column and sediments of Lake Geneva, a large and deep lake in the western part of Switzerland. The water column of the lake was sampled by deploying passive samplers at five different depths ranging from 70 m to 166.5 m for three consecutive months during the summer of 2011 (ELEMO research project). Sediment samples were collected in four locations of the lake at depths ranging from 80 m to 310 m. Several novel PBPs (i.e. 4BBP, TBB, HBB and PCTP) and legacy PBPs, such as pentaBDE technical mixture, hexachlorobenzene (HCB) and PCNB, were included in the list of the target analytes. To confidently detect and quantify the target analytes, the analytical procedure developed in chapter 2 was employed. The water column concentrations of two novel brominated PBPs, 4BBP and TBB, were found to be 0.5-1.0 ng/L, whereas the water column levels of PCTP were about 100 ng/L. All four novel PBPs were also detected and quantified in the sediments samples. Suspect screening of the GC×GC-ENCI-TOFMS data additionally revealed the presence of a potential precursor of PCTP, pentachlorothioanisole, in both the water and sediment samples. This is the first report of the levels of these novel PBPs (4BBP, TBB, and PCTP) in a lake environment. These chapters also investigate the potential pathways of introduction and elimination of these novel PBPs in the Lake Geneva environment. The occurrence of these novel PBPs and also their relatively high concentrations warrants future investigations of these compounds in the environment, including the evaluation of their environmental risk.

In chapters 3 and 4 I also pre-evaluated the environmental fate and behavior of these novel PBPs. I estimated several environmentally relevant partitioning properties of both legacy and novel PBPs. These properties were estimated employing different modeling methods included EPISuite, a GC×GC retention time based method, ACDLab, and a quantum chemistry modeling approach, depending on the analyte. The estimated partitioning properties were used for evaluation of the potential for bioaccumulation, long range transport, and Arctic contamination. We also estimated equilibrium partitioning distributions of these compounds between the water column and sediments of the lake. Based on both the estimated partitioning properties and the limited available occurrence data for these PBPs, we concluded that bioaccumulation, long range transport, and Arctic contamination play an important role in the global fate and behavior of these PBPs. A comparison of PBP levels in lake water and in lake sediments suggested that several target analytes (4BBP, TBB, PCTP and pentaBDE technical mixture) were at or near partitioning equilibrium, whereas for some other target analytes (PCNB, PCBs and OCPs) this was not the case. Finally, avenues are proposed for further investigation of the environmental fate and behavior of these novel PBPs.

In chapter 5, I report on the development of a new fast sampling device for the truly dissolved fraction of hydrophobic compounds in the water column of an aquatic system. Sampling the truly dissolved fraction of PBPs in the water column of an aquatic system is a challenging task, due to the trace level concentrations of these chemicals. Passive sampling approaches are a widely used sampling strategy that can overcome the difficulty of low analyte concentrations in the environment. However, passive sampling techniques necessitate long exposure times, typically 4 to 6 weeks, in order to achieve partitioning equilibrium with the water column. In chapter 5 of this thesis, a fast sampler for the truly dissolved fraction of PBPs in the water column was developed with polydimethylsiloxane (PDMS) as the receiving medium. In this chapter I evaluated the mass transfer between water and the PDMS by measuring the depletion of performance reference compounds (PRCs) with three different flow rates (1, 2 and 4 L/min) at several time intervals for 12 PCBs. In order to explain the mass transfer between the water and PDMS, two modeling approaches were tested. An initial prototype of the sampler was built in-house and fieldtested during the ELEMO field sampling campaign on Lake Geneva. The lab test showed that the rate of mass transfer of hydrophobic compounds was increased by two orders of magnitude compared to conventional passive sampling. The detection of the trace level PCBs in the water column of lake Geneva was achieved only after two hours. The mass transfer of the investigated compounds appeared to be highly sensitive to the water flow rate, and less sensitive to the chemical and physical properties of the compounds. However, further work is needed in order to model the mass transfer process.

Finally, in chapter 6 the water column concentration distributions of trace level PBPs are tentatively evaluated with respect to depth and distance from the shoreline (i.e. vertical and horizontal distribution). The samples collected during the ELEMO sampling campaign were analyzed using the protocol developed in chapter 2. The mass transfer model developed in chapter 5 enabled the estimation of the water column concentrations of PBPs. This enabled the construction of a 3-D map of the concentration distribution for each PBP and also facilitated inferences regarding their potential sources in and around the lake. This is the most comprehensive reported concentration distribution measurement of PBPs in a deep aquatic system such as Lake Geneva.

This thesis lays out a comprehensive protocol for the assessment of novel PBPs in aquatic systems, particularly useful for deep and large lakes. The protocol included the development and optimization of analytical procedures, development of novel methods for the rapid sampling of PBPs in a deep lake, and the assessment of the fate and behavior of novel PBPs based on their environmental partitioning properties.