Synthesis, Fractionation, Characterisation and Toxicity of Naphthenic Acids from Complex Mixtures

Abstract

Amongst the polar organic compounds occurring in unrefined and refined crude oils and the associated polluted production waters, complex mixtures of acids, known historically as naphthenic acids (NAs), have achieved prominence. This is particularly because NAs have been designated a toxicant class of concern in the oil sands process-affected water (OSPW) that has accumulated in vast quantities following exploitation of the oil sands of Northern Alberta, Canada in recent years. However, though there have been calls for NAs to be added to pollutant inventories, at the initiation of the current study, little knowledge existed of the exact composition of refined or unrefined NAs. The overall aim of the current study was therefore to identify individual NAs in refined (commercial) and unrefined (e.g. oil sands process-derived) complex mixtures of acids and then to assess the toxicity of any identified NAs. Individual NAs were tentatively identified by interpretation of the electron ionisation mass spectra of methyl ester derivatives, following comprehensive multidimensional gas chromatography-mass spectrometry (GCxGC-MS). Reference acids were then either purchased, or more commonly, where they were not commercially-available, synthesised, mainly by micro-hydrogenation methods, for co-chromatography and comparison of mass spectra of methyl esters with those of unknowns. The synthetic NAs, purified to >97% were then subjected to toxicological assessments using the Microtox™ assay. In all, 34 compounds were obtained pure enough for testing. Microtox results revealed that the toxicity endpoint (50% Inhibition Concentration, IC50) was between 0.004 and 0.7 mM. Exponential and other correlations were noted between carbon number and toxicity in several of the structural groups of acids assayed, which may be beneficial for predictions of toxicity of non-synthesised acids. Although n-hexanoic acid (IC50 0.7 mM) had the lowest toxicity, adamantane-type acids were the least toxic as a group overall. Conversely, the decahydronaphthalene (decalin)-type acids had the largest range of toxicities (IC50 0.004 to 0.3 mM) and the most toxic acid assayed was 3-decalin-1-yl-propanoic acid. According to USEPA guidelines many individual acids can be said to show low to medium toxicity. Since the acids in commercial and unrefined NAs occur in complex mixtures, an attempt was also made to assess mixture toxicity. Mixtures of individual structural groups of acids (e.g. acyclic isoprenoid acids, n-acids) and a mixture of all 34 acids were assessed. Apart from the adamantane sub-group of acids, all of the mixtures showed toxicities lower than the sum of the parts when calculated using equations for Concentration Addition and Model Deviation Ratios (simply the predicted IC50/Observed IC50). A hypothesis that achievement of a critical micelle concentration is required to produce toxicity was proposed to explain the lower than expected results. Some of the mass spectra of NA present in the commercial and unrefined mixtures were inconsistent with those of any of the alicyclic acids synthesised or purchased. These were hypothesised to be aromatic acids. Fractionation experiments of the NA mixtures using silver ion thin layer chromatography and solid phase extraction (Ag+TLC and Ag+SPE) were carried out in order to provide further evidence for aromatic acids. Ag+TLC allowed separation of a methylated NA mixture from OSPW into three distinct fractions; Ag+SPE resulted in eleven fractions, through the use of a wider range of solvents and differential solvent ratios. Analysis of the fractions by GC-MS revealed that each fraction was largely still made up of unresolved acids (as esters), although one or two fractions revealed some resolved acids. Use of averaged mass spectra and mass chromatography on each fraction revealed further resolved chromatographic peaks and associated interpretable mass spectra. Each of eight of the eleven sub-fractions were examined by GC-MS, in some cases by GCxGC-MS, and all by infrared spectroscopy, ultraviolet visible spectrophotometry and elemental analysis. A number of structures were proposed for the aromatic acids, including those with sulphur-containing moieties. It was noted that far from being minor components, aromatic acids comprised ca.25-40% of the OSPW acid extracts.