Hydrocarbons: Review of methods for analysis in sea water, biota, and sediments.

Abstract

The overview presented in the following paragraphs is not a collection of analytical procedures in the sense of a 'cookbook'. It is intended as an introduction to the subject and a collection of references from which a judicious choice has to be made based upon the specific objectives of an investigation. Scientists active in the field of environmental trace analysis are envisaged as potential users, as well as persons on the decision-making level who wish to familiarize themselves with the complexity, range of application, and limitations of contaminant hydrocarbon analyses in the marine environment. Hydrocarbons, a class of chemical substances consisting exclusively of the elements carbon and hydrogen, are trace constituents of all compartments of the marine environment, i.e., water, suspended solids, organisms, and sediments. The sources of hydrocarbons are both natural in the sense that they occur irrespective of man's in- terference, and artificial as their occurrence is linked to a multitude of human activities. Interest, from a water management point of view, in the analysis of hydrocarbons in the marine environment stems from the often-repeated observation that elevated concentrations of non-biosynthetic hydrocarbons have detrimental effects on many marine life forms. A voluminous literature exists on the subject; reviews and collated papers may be found in GESAMP (1977), Connell and Miller (1980a, b), Gundlach and Marchand (1982), and Kuiper and van den Brink (1987). Owing to the ability of carbon atoms to form chemical bonds not only with other elements, but also among themselves, an almost limitless number of different structures can be conceived of, containing chains of practically any length, branched chains, rings, and any combination of these structural elements. It is the characteristics of the sources which, to a certain extent, determine the number of related molecular structures and the range of different structures. Thus, fossil hydrocarbons span a very wide range of molecular weights and structure types, whereas recent biogenic hydrocarbons contain either saturated or olefinic carbon-carbon bonds in straight or branched chains or rings with five or six carbon atoms. In contrast to the multitude of individual compounds in fossil hydrocarbon mixtures, their number in any given source organism is limited owing to specific pathways for their biosynthesis either de DQYQ or by conversion of dietary precursors (Blumer, 1967; Blumer et al., 1971; Connell and Miller, 1980a). Aromatic structures are frequent among fossil hydrocarbons and among those generated by combustion processes. To a certain extent, it is possible to use the degree of alkyl substitution of aromatic hydrocarbons for differentiating between these sources (Blumer and 3 4 Youngblood, 1975; Youngblood and Blumer, 1975; Sporst~l et al., 1983.) In petroleum, alkyl-substituted derivatives usually predomi- nate, whereas combustion-generated hydrocarbon mixtures are richer in the unsubstituted parent compounds. It is not quite clear, however, whether the predominance of unsubstituted aromatic hydrocarbons in environmental samples necessarily indicates the presence of combustion products. Recent analytical results (Davies and Tibbetts, 1987; Ehrhardt and Burns, 1990) suggest that alkyl-substituted aro- matic hydrocarbons may be less refractory under environmental conditions than many unsubstituted nuclei, whose slower rate of decomposition would eventually lead to their preponderance. Aromatic structures, frequent as they are in fossil hydrocarbons, are rare in hydrocarbons biosynthesized by marine organisms. As examples may be cited: the tetralene derivative, calamene, in gor- gonians (Weinheimer et al., 1968); the substituted benzene, laurene, in Laurencia species (Irie et al., 1965); carotenes with benzoid terminal groups in the sponge Reniera ;aponica (Yamaguchi, 1957a,b; 1958a,b); an alkyl-substituted octahydrochrysene in a polychaete (Farrington et al., 1986). These aromatic hydrocarbons occurring in marine organisms have not yet been characterized as components of dissolved organic material in sea water. Since some saturated hydrocarbons (e.g. n-pentadecane, n-heptadecane, n-nonadecane) which are synthesized by marine phytoplankton may also be detected in uncontaminated sea water, the assumption is plausible, however, that these biosynthetic aromatic hydrocarbons eventually find their way into the aqueous phase. In proportion to non-biosynthetic sources of aromatic hydrocarbons the contribution may be insignificant, but their possible presence should be kept in mind when aromatic hydrocarbons indiscriminately are labelled non-biosynthetic and, hence, contaminants. For the various objectives of surveillance and monitoring programmes as well as activities related to basic research, it has been found expedient to analyse and quantitate hydrocarbons separately in different matrices. Thus, water as the principal agent for transport and dispersal as well as the medium in which marine organisms live is analysed to gather information on the sources, inputs, distri- bution, and concentrations of hydrocarbons to which its inhabitants are exposed. Marine organisms are analysed to investigate the chemical nature of biosynthetic hydrocarbons, the accumulation of contaminant hydrocarbons from the surrounding water owing to the higher lipophilicity of living tissue as compared with sea water, and the associated stress to organisms. A global ocean monitoring programme consisting of many regional components, but relying on common strategies, is based on the use of the sedentary filter-feeding blue mussels and oysters as biological concentrators. Assessment of contamination by hydrocarbons is one component of this programme (Farrington et al., 1982; Goldberg, 1986; Murray and Law, 1980; Reynolds et al., 1981; Risebrough et al., 1983). The analysis of sediments for hydrocarbons is a component of many investigations and monitoring programmes because concentrations are generally higher, and thus easier to measure, than in water and also less variable (less patchy) in the short term. Although similar concentrations may be found in organisms, the biological lipid matrix from which they have to be separated is far more complex. Nevertheless, analyses of hydrocarbons in sediments have often been found to be just as challenging as in any other matrix. Around point sources, such as offshore oil production platforms or refinery outfalls, gradients of concentration may be established which help to determine the maximum area of effect. Hydrocarbons de- posited in sediments may persist for a long period of time, particularly under anoxic conditions. The hydrocarbon composition within the sediments can be altered both by degradation, which leads to the loss of some components, and by early diagenetic reactions in shallow sediments which lead to the in situ production of particular compounds, such as perylene and retene. Biogenic precursor molecules may be altered by chemical and microbiological processes to yield a variety of compounds, such as steranes and pentacyclic triterpanes (Aizenshtat, 1973; Hites et al., 1980; NRC, 1985; Venkatesan, 1988). The range of hydrocarbon concentrations found in sediments is very wide, total hydrocarbon concentrations varying from approximately 1~g/g dry weight in clean offshore sand deposits to >10\in areas impacted by oil spills or close to platforms discharging cuttings resulting from the use of oil-based drilling muds. In addition, different particle sizes and types within a given sediment may have different hydrocarbon compositions (Thompson and Eglinton, 1978; Prahl and Carpenter, 1983). The wide range of boiling points and polarities of hydrocarbon compounds found in sediments also complicate the analysis, as no one method can efficiently extract and concentrate all hydrocarbons present. To some extent, therefore, the analytical method chosen will determine the types of hydrocarbons found. 5 The compilation of methodologies which follows does not include analysis of the sea-surface microlayer, nor does it specifically address analysis of suspended particles. This may appear as a serious shortcoming, because hydrocarbon concentrations in the microlayer tend to surpass those in bulk water by at least an order of magnitude (Burns, 1986; Marty and Saliot, 1976). Particles as principal carriers for vertical transport also claim the attention of environmental analysts, but in both cases methodological differences with respect to the procedures described rest in the proper collection of samples. This is straightforward for particles which are collected by filtration on suitable filters, usually made of glass fibre. The extraction of hydrocarbons then parallels sediment extraction. Procedures for their analysis may be selected from the methods given for other matrices. Sampling of the sea-surface microlayer is more difficult. A useful procedure is delineated in IOC Manuals and Guides No.15 (UNESCO, 1985); Knap et al. (1986) describe its application. A detailed study on the composition of petroleum hydrocarbons in the microlayeris given by Butler and Sibbald (1987) who use a Teflon disk for collecting samples practically free of a separate aqueous phase. Carlson et al. (1988) present a new micro-layer sampling device based on the rotating drum principle. The collected material, of course, is appropriate for analysis by any method selected for a specific purpose.