Article pubs.acs.org/est
Characterization and Quantification of Mining-Related “Naphthenic Acids” in Groundwater near a Major Oil Sands Tailings Pond Jason M. E. Ahad,*,† Hooshang Pakdel,‡ Martine M. Savard,† Angus I. Calderhead,‡ Paul R. Gammon,§ Alfonso Rivera,† Kerry M. Peru,∥ and John V. Headley∥ †
Geological Survey of Canada, Natural Resources Canada, Québec City, QC G1K 9A9, Canada INRS Eau Terre Environnement, Québec City, QC G1K 9A9, Canada § Geological Survey of Canada, Natural Resources Canada, Ottawa, ON K1A 0E8, Canada ∥ Water Science and Technology Directorate, Environment Canada, Saskatoon, SK S7N 3H5, Canada ‡
S Supporting Information *
ABSTRACT: The high levels of acid extractable organics (AEOs) containing naphthenic acids (NAs) found in oil sands process-affected waters (OSPW) are a growing concern in monitoring studies of aquatic ecosystems in the Athabasca oil sands region. The complexity of these compounds has substantially hindered their accurate analysis and quantification. Using a recently developed technique which determines the intramolecular carbon isotope signature of AEOs generated by online pyrolysis (δ13Cpyr), natural abundance radiocarbon, and high resolution Orbitrap mass spectrometry analyses, we evaluated the sources of AEOs along a groundwater flow path from a major oil sands tailings pond to the Athabasca River. OSPW was characterized by a δ13Cpyr value of approximately −21‰ and relatively high proportions of O2 and O2S species classes. In contrast, AEO samples located furthest down-gradient from the tailings pond and from the Athabasca River were characterized by a δ13Cpyr value of around −29‰, a greater proportion of highly oxygenated and Ncontaining compound classes, and a significant component of nonfossil and, hence, non-bitumen-derived carbon. The groundwater concentrations of mining-related AEOs determined using a two end-member isotopic mass balance were between 1.6 and 9.3 mg/L lower than total AEO concentrations, implying that a less discriminating approach to quantification would have overestimated subsurface levels of OSPW. This research highlights the need for accurate characterization of “naphthenic acids” in order to quantify potential seepage from tailings ponds.
■
INTRODUCTION Naphthenic acids (NAs), a complex suite of alkyl-substituted acyclic and cycloaliphatic carboxylic acids found naturally in bitumen, are one of the main contaminants of interest associated with the Athabasca oil sands mining operations. Their general chemical formula is CnH2n+ZO2, where n indicates the carbon number and Z is zero or a negative, even integer that specifies the hydrogen deficiency resulting from ring formation.1 NAs negatively affect reproductive physiology, are endocrine disrupting, and are acutely toxic to a range of organisms.2−5 While ambient levels of NAs in northern Alberta rivers in the Athabasca oil sands region are generally below 1 mg/L,6,7 NAs become concentrated in oil sands processaffected waters (OSPW) during bitumen extraction. With concentrations of NAs in OSPW often exceeding 100 mg/L, surface water and shallow groundwater are prone to potential contamination by seepage from tailings ponds.7−9 Owing to their complexity, however, the characterization and quantification of NAs in water samples remains a significant © 2013 American Chemical Society
analytical challenge. This in turn makes it difficult to understand their behavior and fate in the shallow subsurface. Coextractives such as humic and fulvic acids contained within a sample’s acid extractable organic (AEO) fraction interfere with the detection of NAs and contribute to significant discrepancies in concentrations between different quantitative methods.10−14 Additionally, several recent studies utilizing ultrahigh resolution mass spectrometry (MS) and multidimensional comprehensive gas chromatography mass spectrometry (GC×GC-MS) have shown that a large component of the polar organics found in OSPW are not described by the “classical” NA formula given above.11,15−17 For instance, OSPW may contain aromatic carboxylic acids,17 tricyclic diamondoid acids,16 and sulfurcontaining species such as O2S.15 Yet while high and ultrahigh Received: Revised: Accepted: Published: 5023
December 14, 2012 March 6, 2013 April 22, 2013 April 22, 2013 dx.doi.org/10.1021/es3051313 | Environ. Sci. Technol. 2013, 47, 5023−5030
Environmental Science & Technology
Article
Figure 1. (A) Location map of the study area including all surface water and groundwater sampling locations, and (B) zoomed view showing the main groundwater sample locations.
resolution MS and GC×GC-MS characterization will remain invaluable to understanding Athabasca oil sands acids on a molecular level, the data generated by these techniques do not generally lend themselves to an easily interpretable format suitable for quantification and, hence, source apportionment in groundwater systems. One promising method for a more straightforward determination of AEO sources in the subsurface is intramolecular carbon isotope analysis (referred to here as δ13Cpyr), a technique which determines the δ13C signature of the CO2 generated by pyrolysis of carboxyl groups.18−21 In contrast to the total or “bulk” δ13C values measured on the CO2 generated by oxidation of AEOs, which showed no significant variation (−31.0 to −29.0‰), a large range (up to ∼9‰) in δ13Cpyr signatures was observed between process water, groundwater, and surface water samples from the Athabasca oil sands region.18 While the large isotopic discrepancy between different sample types suggested that δ13Cpyr signatures could be used to quantify groundwater concentrations of mining-related AEOs (e.g., in OSPW), the field demonstration of this technique has yet to be reported. This paper reports the first application of intramolecular carbon isotope characterization, combined with high resolution Orbitrap MS and natural abundance radiocarbon analysis, to examine sources of AEO containing “naphthenic acids” along a groundwater flow path from a major oil sands tailings pond to the Athabasca River. The techniques used here were tested at the local scale, in the context of an Athabasca oil sands open pit mining operation, where the host of the bitumen reserves is the Cretaceous McMurray formation (sand and sandstone). The
proportion of mining-related AEOs in the shallow subsurface was quantified using a mass balance based on δ13Cpyr signatures. The results from this study highlight the need for accurate source apportionment in order to evaluate if mining-related AEOs impact groundwater resources in the Athabasca oil sands region.
■
MATERIALS AND METHODS Site Description. The field investigation was carried out along two well alignments near a major oil sands tailings pond (PW1) located north of the city of Fort McMurray, AB, Canada (Figure 1). The exact location and name of the mining company operating at the test site are confidential. The site has been in operation for several decades; thus, sufficient time has passed to allow for contaminants to have potentially moved through the shallow aquifer. The Pleistocene glacio-fluvial aquifer is located above the Cretaceous Clearwater/McMurray formations and Devonian limestone and ranges between ∼5 and 25 m in depth (Figure S1 of the Supporting Information). At the scale of this local study, available data and groundwater flow simulations from the 2D vertical numerical model of the aquifer show that water from the tailings pond, either directly or through the tailings dyke, flows into the Pleistocene units toward the Athabasca River.22 The groundwater flow is generally slow (∼234 years from tailings pond to river) around much of the tailings pond; however, it is much faster along the studied preferential flow path (∼19 years to reach discharge points).22 The underlying Clearwater and McMurray formations act as an aquitard, limiting vertical migration and favoring horizontal flow within the glacio-fluvial formation 5024
dx.doi.org/10.1021/es3051313 | Environ. Sci. Technol. 2013, 47, 5023−5030
Environmental Science & Technology
Article
mation oil sand sample containing ∼0.1% AEOs by weight of bitumen was also extracted and separated into mass fractions. All samples collected using PCGC were then converted back to their original organic acid form following a modified demethylation protocol described by Frank et al.30 Demethylated subsamples from the mass fractions separated by PCGC were analyzed at Environment Canada (Saskatoon, SK) by 5 μL loop injection (flow injection analysis) using a Surveyor MS pump (Thermo Fisher Scientific Inc.) and a mobile phase of 50:50 acetonitrile/water containing 0.1% NH4OH. Mass spectrometry analysis was carried out using a dual pressure linear ion trap−orbitrap mass spectrometer (LTQ Orbitrap Velos, Thermo Fisher Scientific Inc.) equipped with an ESI interface operated in negative mode. Data was acquired in full scan mode from m/z 80 to 600 at 100,000 resolution. The majority of ions were singly charged, and the average mass resolving power (m/Δm50%) was 100,000 at m/z 400. Mass accuracies of less than 1 ppm were obtained using a lock mass compound (n-butyl benzenesulfonamide) for scan-to-scan mass calibration correction. The carbon isotope signature of the CO2 generated by the pyrolytic decarboxylation of AEOs (δ13Cpyr) was determined by thermal conversion (TC)/elemental analysis (EA)-isotope ratio mass spectrometry (TC/EA-IRMS; Thermo-Finnigan Delta+ XL) at the Delta-Lab of the Geological Survey of Canada. The δ13Cpyr values determined by TC/EA-IRMS were analyzed using CO2 calibrated against international carbonate standards (NBS 18 and NBS 19). The uncertainty associated with intramolecular carbon isotope analysis of AEOs was ±0.6‰.18 Radiocarbon Analysis. Natural abundance radiocarbon (14C) analysis can provide valuable insight into the origins of organic compounds in groundwater systems.31−33 Radiocarbon analyses of the “bulk” AEOs (i.e., not solely carboxyl group carbon) from PCGC mass fractions 4 or 5 from ten samples were determined at the University of Georgia’s Center for Applied Isotope Studies. The samples were combusted in quartz tubes containing CuO at 575 °C, and the resulting CO2 was cryogenically purified and catalytically converted to graphite following the method of Vogel et al.34 Graphite 14 C/13C ratios were measured using the CAIS 0.5 MeV AMS (National Electrostatics Corp.) and are compared to the ratio measured from the Oxalic Acid I standard (NBS SRM 4990). Radiocarbon measurements are normalized to δ13C values and reported as Δ14C according to international convention.35 In this context, petroleum has a “14C-free” value of −1000‰ while carbon photosynthesized from the atmosphere over the past couple of decades is closer to the current atmospheric CO2 value of approximately 50‰.36 The uncertainty for Δ14C measurements reported here incorporating both analytical precision and reproducibility was
