Alberto S. Pereira

Quantitative and Qualitative Analysis of Naphthenic Acids in Natural Waters Surrounding the Canadian Oil Sands Industry

The Canadian oil sands industry stores toxic oil sands process-affected water (OSPW) in large tailings ponds adjacent to the Athabasca River or its tributaries, raising concerns over potential seepage. Naphthenic acids (NAs; C(n)H(2n-Z)O(2)) are toxic components of OSPW, but are also natural components of bitumen and regional groundwaters, and may enter surface waters through anthropogenic or natural sources. This study used a selective high-resolution mass spectrometry method to examine total NA concentrations and NA profiles in OSPW (n = 2), Athabasca River pore water (n = 6, representing groundwater contributions) and surface waters (n = 58) from the Lower Athabasca Region. NA concentrations in surface water (< 2-80.8 μg/L) were 100-fold lower than previously estimated. Principal components analysis (PCA) distinguished sample types based on NA profile, and correlations to water quality variables identified two sources of NAs: natural fatty acids, and bitumen-derived NAs. Analysis of NA data with water quality variables highlighted two tributaries to the Athabasca River-Beaver River and McLean Creek-as possibly receiving OSPW seepage. This study is the first comprehensive analysis of NA profiles in surface waters of the region, and demonstrates the need for highly selective analytical methods for source identification and in monitoring for potential effects of development on ambient water quality.

Characterization of Oil Sands Process-Affected Waters by Liquid Chromatography Orbitrap Mass Spectrometry

Recovery of bitumen from oil sands in northern Alberta, Canada, occurs by surface mining or in situ thermal recovery, and both methods produce toxic oil sands process-affected water (OSPW). A new characterization strategy for surface mining OSPW (sm-OSPW) and in situ OSPW (is-OSPW) was achieved by combining liquid chromatography with orbitrap mass spectrometry (MS). In electrospray positive and negative ionization modes (ESI(+)/ESI(-)), mass spectral data were acquired with high resolving power (RP > 100,000-190,000) and mass accuracy (<2 ppm). The additional chromatographic resolution allowed for separation of various isomers and interference-free MS(n) experiments. Overall, ∼3000 elemental compositions were revealed in each OSPW sample, corresponding to a range of heteroatom-containing homologue classes: Ox (where x = 1-6), NOx (where x = 1-4), SOx (where x = 1-4), NO₂S, N, and S. Despite similarities between the OSPW samples at the level of heteroatom class, the two samples were very different when considering isomer patterns and double-bond equivalent profiles. The chromatographic separations also allowed for confirmation that, in both OSPW samples, the O₂ species detected in ESI(-) (i.e., naphthenic acids) were chemically distinct from the corresponding O₂ species detected in ESI(+). In comparison to model compounds, tandem MS spectra of these new O₂ species suggested a group of non-acidic compounds with dihydroxy, diketo, or ketohydroxy functionality. In light of the known endocrine-disrupting potential of sm-OSPW, the toxicity of these O₂ species deserves attention and the method should be further applied to environmental forensic analysis of water in the region.

Effects-Directed Analysis of Dissolved Organic Compounds in Oil Sands Process-Affected Water

Acute toxicity of oil sands process-affected water (OSPW) is caused by its complex mixture of bitumen-derived organics, but the specific chemical classes that are most toxic have not been demonstrated. Here, effects-directed analysis was used to determine the most acutely toxic chemical classes in OSPW collected from the worlds first oil sands end-pit lake. Three sequential rounds of fractionation, chemical analysis (ultrahigh resolution mass spectrometry), and acute toxicity testing (96 h fathead minnow embryo lethality and 15 min Microtox bioassay) were conducted. Following primary fractionation, toxicity was primarily attributable to the neutral extractable fraction (F1-NE), containing 27% of original organics mass. In secondary fractionation, F1-NE was subfractionated by alkaline water washing, and toxicity was primarily isolated to the ionizable fraction (F2-NE2), containing 18.5% of the original organic mass. In the final round, chromatographic subfractionation of F2-NE2 resulted in two toxic fractions, with the most potent (F3-NE2a, 11% of original organic mass) containing predominantly naphthenic acids (O2(-)). The less-toxic fraction (F3-NE2b, 8% of original organic mass) contained predominantly nonacid species (O(+), O2(+), SO(+), NO(+)). Evidence supports naphthenic acids as among the most acutely toxic chemical classes in OSPW, but nonacidic species also contribute to acute toxicity of OSPW.

Ozonation degrades all detectable organic compound classes in oil sands process‐affected water; an application of high‐performance liquid chromatography/obitrap mass spectrometry

RATIONALE Surface mining of bitumen in Northern Alberta, Canada, results in large volumes of toxic oil sands process-affected water (OSPW) that must be contained in tailings ponds. Ozonation has shown great promise as an OSPW treatment process, by decreasing its toxicity and increasing its biodegradability, but the effect of ozonation on the thousands of dissolved organic chemical groups has not yet been examined. METHODS Reversed-phase liquid chromatography with ultrahigh-resolution linear ion trap-orbitrap mass spectrometry was applied to the characterization of treated (utilized ozone doses of 20 and 50 mg O3/L) and untreated OSPW. The analysis was performed in positive and negative electrospray ionization modes for each sample (ESI(+)/ESI(-)). RESULTS Semi-quantitative analysis of ozonated and unozonated samples allowed degradation to be monitored for naphthenic acids (i.e. O2 species in ESI(-) mode) and >2000 other organic species belonging to various heteroatom-containing classes: Ox (where x = 1 to 6), NOx (where x = 1 to 4), SOx (where x = 1 to 4), NO2 S, N, and S. No chlorinated byproducts were detected in any treated sample, but at the low dose (20 mg O3/L) some compound classes increased in abundance (e.g. the O5 class), indicating that they were formed as byproducts at faster rates than they were degraded. Nevertheless, all organic compound classes subsequently diminished at the higher dose (50 mg O3/L). For several Ox and SOx classes, species observed in ESI(+) mode (e.g. O2(+) species) were often more recalcitrant to ozonation than the corresponding species detected in ESI(-) mode (e.g. O2(-) species; naphthenic acids). CONCLUSIONS Ozonation appears to be a very suitable treatment option for OSPW, but the more recalcitrant groups of compounds may help to explain the residual toxicity of ozonated OSPW. Analysis of OSPW constituents in both ionization modes is warranted in all future OSPW fate studies.

Discovery of C5–C17 Poly- and Perfluoroalkyl Substances in Water by In-Line SPE-HPLC-Orbitrap with In-Source Fragmentation Flagging

The presence of unknown organofluorine compounds in environmental samples has prompted the development of nontargeted analytical methods capable of detecting new perfluoroalkyl and polyfluoroalkyl substances (PFASs). By combining high volume injection with high performance liquid chromatography (HPLC) and ultrahigh resolution Orbitrap mass spectrometry, a sensitive (0.003-0.2 ng F/mL for model mass-labeled PFASs) untargeted workflow was developed for discovery and characterization of novel PFASs in water. In the first step, up to 5 mL of water is injected to in-line solid phase extraction, chromatographed by HPLC, and detected by electrospray ionization with mass spectral acquisition in parallel modes cycling back and forth: (i) full scan with ultrahigh resolving power (RP = 120,000, mass accuracy ≤3 ppm), and (ii) in-source fragmentation flagging scans designed to yield marker fragment ions including [C2F5](-) (m/z 118.992), [C3F7](-) (m/z 168.988), [SO4H](-) (m/z 96.959), and [Cl](-) (m/z 34.9). For flagged PFASs, plausible empirical formulas were generated from accurate masses, isotopic patterns, and fragment ions. In the second step, another injection is made to collect high resolution MS/MS spectra of suspect PFAS ions, allowing further confirmation of empirical formulas while also enabling preliminary structural characterization. The method was validated by applying it to an industrial wastewater, and 36 new PFASs were discovered. Of these, 26 were confidently assigned to 3 new PFAS classes that have not previously been reported in the environment: polyfluorinated sulfates (CnFn+3Hn-2SO4(-); n = 5, 7, 9, 11, 13, and 15), chlorine substituted perfluorocarboxylates (ClCnF2nCO2(-); n = 4-11), and hydro substituted perfluorocarboxylates (HCnF2nCO2(-); n = 5-16). Application of the technique to environmental water samples is now warranted.

Estimates of Octanol-Water Partitioning for Thousands of Dissolved Organic Species in Oil Sands Process-Affected Water.

In this study, the octanol-water distribution ratios (DOW, that is, apparent KOW at pH 8.4) of 2114 organic species in oil sands process-affected water were estimated by partitioning to polydimethylsiloxane (PDMS) coated stir bars and analysis by ultrahigh resolution orbitrap mass spectrometry in electrospray positive ((+)) and negative ((-)) ionization modes. At equilibrium, the majority of species in OSPW showed negligible partitioning to PDMS (i.e., DOW <1), however estimated DOWs for some species ranged up to 100,000. Most organic acids detected in ESI- had negligible partitioning, although some naphthenic acids (O2(-) species) had estimated DOW ranging up to 100. Polar neutral and basic compounds detected in ESI+ generally partitioned to PDMS to a greater extent than organic acids. Among these species, DOW was greatest among 3 groups: up to 1000 for mono-oxygenated species (O(+) species), up to 127,000 for NO(+) species, and up to 203,000 for SO(+) species. A positive relationship was observed between DOW and carbon number, and a negative relationship was observed with the number of double bonds (or rings). The results highlight that nonacidic compounds in OSPW are generally more hydrophobic than naphthenic acids and that some may be highly bioaccumulative and contribute to toxicity.

Mass spectral characterisation of a polar, esterified fraction of an organic extract of an oil sands process water

RATIONALE Characterising complex mixtures of organic compounds in polar fractions of heavy petroleum is challenging, but is important for pollution studies and for exploration and production geochemistry. Oil sands process-affected water (OSPW) stored in large tailings ponds by Canadian oil sands industries contains such mixtures. METHODS A polar OSPW fraction was obtained by silver ion solid-phase extraction with methanol elution. This was examined by numerous methods, including electrospray ionisation (ESI) Fourier transform ion cyclotron resonance mass spectrometry (FTICRMS) and ultra-high-pressure liquid chromatography (uHPLC)/Orbitrap MS, in multiple ionisation and MS/MS modes. Compounds were also synthesised for comparison. RESULTS The major ESI ionisable compounds detected (+ion mode) were C15-28 SO3 species with 3-7 double bond equivalents (DBE) and C27-28 SO5 species with 5 DBE. ESI-MS/MS collision-induced losses were due to water, methanol, water plus methanol and water plus methyl formate, typical of methyl esters of hydroxy acids. Once the fraction was re-saponified, species originally detected by positive ion MS, could be detected only by negative ion MS, consistent with their assignment as sulphur-containing hydroxy carboxylic acids. The free acid of a keto dibenzothiophene alkanoic acid was added to an unesterified acid extract of OSPW in known concentrations as a putative internal standard, but attempted quantification in this way proved unreliable. CONCLUSIONS The results suggest the more polar acidic organic SO3 constituents of OSPW include C15-28  S-containing, alicyclic and aromatic hydroxy carboxylic acids. SO5 species are possibly sulphone analogues of these. The origin of such compounds is probably via further biotransformation (hydroxylation) of the related S-containing carboxylic acids identified previously in a less polar OSPW fraction. The environmental risks, corrosivity and oil flow assurance effects should be easier to assess, given that partial structures are now known, although further identification is still needed.

Temporal trends of perfluorooctanesulfonate isomer and enantiomer patterns in archived Swedish and American serum samples

Human perfluorooctanesulfonate (PFOS) body burdens are attributable to both direct PFOS and indirect PFOS precursor (PreFOS) exposure. The relative importance of these two pathways has been estimated, but the relative temporal trajectory of exposure to PFOS and PreFOS has not been examined. Here, two hypothesized biomarkers of PreFOS exposure, PFOS isomer profiles (quantified as percent branched PFOS, %br-PFOS) and chiral 1m-PFOS enantiomer fractions (1m-PFOS EF) were analyzed in archived human serum samples of individual American adults (1974-2010) and pooled samples of Swedish primiparous women (1996-2010). After correcting for potential confounders, significant correlations between %br-PFOS and 1m-PFOS EFs were observed in American samples and in Swedish samples for the 1996-2000 period, supporting the hypothesis that both %br-PFOS and 1m-PFOS EF are biomarkers of PreFOS exposure. Significant trends of increasing %br-PFOS, from 2000 to 2010, and increasingly non-racemic 1m-PFOS EFs, from 1996 to 2000, were detected in Swedish samples. No statistically significant trend for %br-PFOS or 1m-PFOS EF was observed in American samples, but American males had significantly higher %br-PFOS and significantly lower 1m-PFOS EF (i.e. more non-racemic) than females, and a similar significant difference was shown in the older age group, relative to the younger age group. These temporal trends in %br-PFOS and 1m-PFOS EF are not easily explained and the results highlight uncertainties about how humans are exposed to PFOS.

Inhibition of ABC transport proteins by oil sands process affected water

The ATP-binding cassette (ABC) superfamily of transporter proteins is important for detoxification of xenobiotics. For example, ABC transporters from the multidrug-resistance protein (MRP) subfamily are important for excretion of polycyclic aromatic hydrocarbons (PAHs) and their metabolites. Effects of chemicals in the water soluble organic fraction of relatively fresh oil sands process affected water (OSPW) from Base Mine Lake (BML-OSPW) and aged OSPW from Pond 9 (P9-OSPW) on the activity of MRP transporters were investigated in vivo by use of Japanese medaka at the fry stage of development. Activities of MRPs were monitored by use of the lipophilic dye calcein, which is transported from cells by ABC proteins, including MRPs. To begin to identify chemicals that might inhibit activity of MRPs, BML-OSPW and P9-OSPW were fractionated into acidic, basic, and neutral fractions by use of mixed-mode sorbents. Chemical compositions of fractions were determined by use of ultrahigh resolution orbitrap mass spectrometry in ESI(+) and ESI(-) mode. Greater amounts of calcein were retained in fry exposed to BML-OSPW at concentration equivalents greater than 1× (i.e., full strength). The neutral and basic fractions of BML-OSPW, but not the acidic fraction, caused greater retention of calcein. Exposure to P9-OSPW did not affect the amount of calcein in fry. Neutral and basic fractions of BML-OSPW contained relatively greater amounts of several oxygen-, sulfur, and nitrogen-containing chemical species that might inhibit MRPs, such as O(+), SO(+), and NO(+) chemical species, although secondary fractionation will be required to conclusively identify the most potent inhibitors. Naphthenic acids (O2(-)), which were dominant in the acidic fraction, did not appear to be the cause of the inhibition. This is the first study to demonstrate that chemicals in the water soluble organic fraction of OSPW inhibit activity of this important class of proteins. However, aging of OSPW attenuates this effect and inhibition of the activity of MRPs by OSPW from Base Mine Lake does not occur at environmentally relevantconcentrations.

Effect of Lipid Partitioning on Predictions of Acute Toxicity of Oil Sands Process Affected Water to Embryos of Fathead Minnow (Pimephales promelas)

Dissolved organic compounds in oil sands process affected water (OSPW) are known to be responsible for most of its toxicity to aquatic organisms, but the complexity of this mixture prevents use of traditional bottom-up approaches for predicting toxicities of mixtures. Therefore, a top-down approach to predict toxicity of the dissolved organic fraction of OSPW was developed and tested. Accurate masses (i.e., m/z) determined by ultrahigh resolution mass spectrometry in negative and positive ionization modes were used to assign empirical chemical formulas to each chemical species in the mixture. For each chemical species, a predictive measure of lipid accumulation was estimated by stir-bar sorptive extraction (SBSE) to poly(dimethyl)siloxane, or by partitioning to solid-supported lipid membranes (SSLM). A narcosis mode of action was assumed and the target-lipid model was used to estimate potencies of mixtures by assuming strict additivity. A model developed using a combination of the SBSE and SSLM lipid partitioning estimates, whereby the accumulation of chemicals to neutral and polar lipids was explicitly considered, was best for predicting empirical values of LC50 in 96-h acute toxicity tests with embryos of fathead minnow (Pimephales promelas). Model predictions were within 4-fold of observed toxicity for 75% of OSPW samples, and within 8.5-fold for all samples tested, which is comparable to the range of interlaboratory variability for in vivo toxicity testing.