Chapter one summarises the development of UK legislation for the protection of the aquatic environment, and highlights current EC legislative requirements for water quality. The need for on-line water quality monitoring and the alternative instrumental approaches to it are discussed, together with the philosophy of "easy care instrumentation" and water industry requirements for online analysers. A simple spectrophotometric FI system is proposed for the on-line determination of a range of water quality parameters. The following chapter details instrumentation used in the FI system, emphasising the solid-state photometric detector. Development of an FI manifold for the determination of aluminium in potable and treated waters is covered in the next chapter. The method, based on complexation of aluminium with pyrocatechol violet is compared with a standard Driscoll procedure. Details of the construction and testing of a fully automated FI instrument are also given. Chapter four describes the development of a modular automated FI monitor with a PC compatible STEbus based computer system. Successful operation of this monitor is illustrated by its application to the determination of residual coagulants (aluminium and iron). Full details of software routines for control, processing and validation are given together with results from a tap water trial for dissolved aluminium. The FI determination of residual iron by its complexation with ferene S, and the application of the optimised method in the STEbus based monitor is detailed in chapter five. In the final chapter the use of on-line FI oxidation procedures for the determination of dissolved organic carbon are examined. The oxidation of a wide range of organic species to carbon dioxide using a silver catalysed persulphate reaction, enhanced with UV irradiation and a stopped-flow procedure is described. The sequential determination of inorganic and organic carbon without separation of the fractions is also investigated.

The hydrocarbons of recent polluted sediments, in-reservoir and laboratory biodegraded crude oils, and certain petroleum products (e.g. lubricating oils) often display 'humps' or Unresolved Complex Mixtures (UCMs) when analysed by gas chromatography (GC). Although widespread and often abundant, to date little is known of their detailed molecular composition. Standard chromatographic methods of isolation of model aliphatic and aromatic hydrocarbon UCMs from lubricating oils followed by conventional methods of analysis provided little compositional detail. Thus GC and GC-electron impact mass spectrometry (GC-EIMS) was limited to an estimate of carbon number ranges and to the identification of certain series of 'biological marker' compounds. However, these were well resolved and were estimated to account for <10% of the total detector response. Further analyses were performed by chemical ionisation-MS (CI-MS), probe distillation EI-MS, field ionisation-MS (FIMS), and elemental analysis; yet the information provided by each was limited to a few 'average' molecular types. In view of the limitations of conventional methods of analysis, alternative methods were adopted. These utilised novel chemical and pyrolytic degradations of the UCM hydrocarbons. Chemical oxidation with Cr03 in glacial acetic acid produced reasonable yields of total recoverable material (40-80%). Furthermore, a high proportion were functionalised (>90%), and many resolved, which allowed their identification by EI and CI GC-MS. Surprisingly, the most abundant products of oxidation of hydrocarbon UCMs were straight chain monocarboxylic acids. This appeared to contradict literature consensus on UCM composition, namely a predominance of highly branched and/or cyclic hydrocarbons. However, from literature reported CrO oxidations of hydrocarbons, potential precursor compounds were proposed. These were monoalkyl substituted 'TO-branched acyclic and monocyclic alkanes for the aliphatic UCM and alkyl 'TO-branched monoaromatic hydrocarbons for the aromatic UCM. Proposed precursor UCM hydrocarbons were confirmed by synthesis and chemical oxidation under the same conditions. Thus each of the synthetic candidate UCM hydrocarbons [7-n-hexylnonadecane, 9-(2-phenylethyl)-heptadecane and 9-(2- cyc 1 ohexyl ethyl j--hep tade cane] produced n-acids on oxidation with Cr03- Further correlations were found for products of other synthetic alkanes and less abundant UCM oxidation products. For example, n-alkan-2-ones. iso alkan-2- ones, and 7-methyl--y-lactones could all be correlated with methyl substituted acyclic alkyl linkages on UCM hydrocarbons. The application of chemical oxidation to aliphatic UCMS of varied origin showed the technique has great potential for fingerprinting such samples. GC-MS analysis of a selected series of resolved product compounds (alkyl ketones, -y-methyl--y-lactones) showed good correlations for samples of the sane origin, yet distinct differences for UCHs from different sources. Biodegradation of the three candidate UCM hydrocarbons alongside acyclic isoprenoid alkanes and normal and monomethyl alkanea showed the UCM hydrocarbons were at least as resistant to microbial degradation as the isoprenoid alkanes. In this context it is therefore concluded that the candidate UCM compounds serve as good molecular models for hydrocarbon UCMs.

The occurrence of unresolved complex mixtures of hydrocarbons (UCMs) in the aliphatic fraction of marine sediments and organisms from areas impacted by petroleum hydrocarbons is well documented and widely accepted as an indication of fossil fuel contamination. In contrast, the presence of an aromatic UCM is often ignored and environmental concentrations of aromatic UCM hydrocarbons in marine biota and sediments are rarely reported. The aims of this study were to establish the quantitative significance of aromatic UCMs in environmental samples and to assess the toxicological significance of both aliphatic and aromatic UCMs. A reproducible method was developed and validated for the analysis and quantification of petroleum hydrocarbons in mussel (Mytilus edulis) tissue. Emphasis was placed upon development of a method which minimized losses of more volatile, lower molecular weight, toxicologically significant hydrocarbons, without compromising recovery of higher molecular weight compounds which are useful for source identification in environmental monitoring schemes. Analysis of mussels from a small number of U. K. coastal locations indicated that aromatic hydrocarbon UCMs may form a significant proportion (ca 20 %) of the total hydrocarbon body burden of mussels from areas contaminated with petroleum hydrocarbons. Aromatic UCM hydrocarbons were not observed in mussels from relatively uncontaminated areas but concentrations of 430 μg g'' (dry wt tissue) were measured in mussels from heavily impacted areas. Aliphatic UCM concentrations ranged from 7- 3445 μg gg' (dry wt tissue). For the purposes of toxicological studies, a low molecular weight model aliphatic UCM hydrocarbon, 4-propyloctane (4-PO) was synthesised. Two low molecular weight model aromatic hydrocarbons 7- cyclohexyltetralin and 7-cyclohexyl-l-propyltetralin were also synthesised using a modification of the Haworth synthesis. All three target compounds and synthetic intermediates were characterised by NMR, MSandIR. Exposure of M. edulis to 4-PO caused a significant reduction in mussel ciliary feeding activity indicating that 4-PO was indeed toxic as measured by this bioassay. The demonstrable narcotic activity of 4-PO is presumably related to the greater aqueous solubility of branched hydrocarbons compared with similar straight chain hydrocarbons. Further experiments investigating the effect of 4- PO over exposure periods up to 120h provided a unique and detailed insight into the relationship between concentration of toxicant in the gills of M. edulis and observed feeding rate. The established method of mussel feeding rate determination was modified in light of this relationship to produce an " improved rapid and reproducible screening technique. Both of the model aromatic UCM hydrocarbons were also found to be toxic to mussel ciliary feeding activity. This appears to be the first report of investigations into the toxicity of the aromatic UCM and suggests that previous studies have ignored an environmental burden of toxicological significance. Estimates of the tissue effective concentration (TEC50) for the model UCM hydrocarbons gave comparable values with those reported for the effect of other narcotic hydrocarbons upon mussel feeding rate, providing support for the theory that non-specific narcosis occurs at a relatively constant tissue concentration of toxicant. The demonstrated narcotic activity of each of the three model UCM hydrocarbons has extended the molecular weight range of narcotic hydrocarbons studied to date. The results presented herein suggest that a small proportion of low molecular weight aliphatic UCMs and perhaps a greater proportion of aromatic UCMs are of toxicological significance.