Identification of Individual Naphthenic Acids in Oil Sands Process Water

ARTICLE pubs.acs.org/est

Diamonds in the Rough: Identification of Individual Naphthenic Acids in Oil Sands Process Water Steven J. Rowland,*,† Alan G. Scarlett,† David Jones,† Charles E. West,† and Richard A. Frank‡ †

Petroleum and Environmental Geochemistry Group, Biogeochemistry Research Centre, University of Plymouth, Drake Circus, Plymouth PL4 8AA, U.K. ‡ Aquatic Ecosystems Protection Research Division/Water Science & Technology Directorate, Environment Canada, 867 Lakeshore Road, Burlington, Ontario, Canada L7R 4A6

bS Supporting Information ABSTRACT: Expansion of the oil sands industry of Canada has seen a concomitant increase in the amount of process water produced and stored in large lagoons known as tailings ponds. Concerns have been raised, particularly about the toxic complex mixtures of water
soluble naphthenic acids (NA) in the process water. To date, no individual NA have been identified, despite numerous attempts, and while the toxicity of broad classes of acids is of interest, toxicity is often structure-specific, so identification of individual acids may also be very important. Here we describe the chromatographic resolution and mass spectral identification of some individual NA from oil sands process water. We conclude that the presence of tricyclic diamondoid acids, never before even considered as NA, suggests an unprecedented degree of biodegradation of some of the oil in the oil sands. The identifications reported should now be followed by quantitative studies, and these used to direct toxicity assays of relevant NA and the method used to identify further NA to establish which, or whether all NA, are toxic. The two-dimensional comprehensive gas chromatography
mass spectrometry method described may also be important for helping to better focus reclamation/remediation strategies for NA as well as in facilitating the identification of the sources of NA in contaminated surface waters.

’ INTRODUCTION The most extreme examples of bacterial oxidation of petroleum have resulted in the formation of the giant oil sands deposits of Canada, which are an extremely valuable source of energy for North America.1 However, exploitation of the oil sands has also resulted in the accumulation of an estimated trillion liters of process water (oil sands process water, OSPW), and concerns have been raised about the possible impacts, particularly of the somewhat toxic, acid-extractable organic matter, usually known as naphthenic acids (NA2
6 and references therein). Such fractions have proved toxic to fish, trees, birds, and plankton (e.g., refs 4
6), and this has led to the introduction of environmental legislation requiring assessment of the so-called NA. This is driving a resurgence in studies of the acid-extractable organic matter.3 Recently, numerous accurate mass ion cyclotron mass spectrometry and other studies have shown that components other than those adhering to the empirical formula CnH2nþzO2, which is the descriptor used formally to describe petroleumderived carboxylic acids (NA; reviewed in ref 5), are also present in OSPW acid-extractable organic matter.3
11 These include dicarboxylic acids and a variety of nitrogen and sulfur-containing species (e.g., refs 3, , 8, and 11). However, in the meantime even those compounds which do meet the strict formulaic definition of NA, remain identified only in terms of the total carbon number r 2011 American Chemical Society

and the degree of cyclicity. To date most studies have cited the generalized, speculative, structures given in an early review.12 An exception used LC-multistage mass spectrometry of amides of NA to produce fragment ions from which some structural information was gleaned,13 but no acid was conclusively identified. While the focus of former studies has thus, of necessity, been on the toxicity of broad NA classes rather than on the toxicity of individual NA, it is well-known that toxic action can be very structure specific, even to the extent that different isomers have different toxicities. Thus, it may be important to identify the individual NA. The classical mass spectrometry method for structural characterization of unknowns uses electron ionization (EI), but because conventional gas chromatography (GC) produces insufficient resolution of the complex mixtures present in OSPW acid-extractable organic matter, library-searchable mass spectra are not readily obtainable by GC-EI-MS. Therefore, at least two previous attempts have been made to use the increased Received: November 4, 2010 Accepted: February 25, 2011 Revised: February 7, 2011 Published: March 10, 2011 3154

dx.doi.org/10.1021/es103721b | Environ. Sci. Technol. 2011, 45, 3154–3159

Environmental Science & Technology chromatographic resolving power of multidimensional GC, known as GCxGC, coupled with Time of Flight mass spectrometry(ToF-MS) to study NA mixtures.14,15 However, no individual acids were identified. Recent reviews investigating the potential impacts of the oil sands industry on the surrounding environment have indicated several knowledge gaps that need to be addressed, including the improvement of analytical methods. The Royal Society of Canada published a report16 summarizing the environmental and health impacts of the Canadian oil sands and identified a lack in analytical capability for NA. Consequently, establishing a regulatory standard to monitor the potential leaching of NA in OSPW from tailings ponds into surface waters has not yet proved possible. An oil sands advisory panel established by the Canadian Minister of the Environment reached a similar conclusion17 and noted the need for the development of analytical methods capable of determining the background concentrations of toxic compounds associated with natural bitumen deposits entering the Athabasca River system. Schindler18 has indicated the need for the development of a monitoring program capable of distinguishing contaminants originating from industrial and natural sources. In the present study we have used GCxGC-ToF-MS to study the methyl ester derivatives of OSPW acid-extractable organic matter. We find that hundreds of the components are sufficiently well resolved by the GCxGC method that potentially interpretable, certainly library-searchable, mass spectra could be obtained. As a result, we now report the identification and some tentative identifications, of numerous tricyclic acids in the OSPW, the structures of the acids suggesting that some of the oil from which the oil sands bitumen originates has been highly biodegraded. Several previous reports have indicated that tricyclic acids are major NA in OSPW (e.g., refs 3
11), but none have been identified. In a further report19 we also identify some of the more minor, pentacyclic NA in OSPW.

’ MATERIALS AND METHODS Adamantane-1-carboxylic and 3-ethyl-adamantane-1-carboxylic acid were purchased from Sigma (U.K.). The OSPW NA was obtained from a previous study.20 Acids were derivatized by refluxing with BF3-methanol.15 Two-dimensional comprehensive gas chromatography-timeof-flight-mass spectrometry (GCxGC-ToF-MS) analyses were conducted using an Agilent 7890A gas chromatograph (Agilent Technologies, Wilmington, DE) fitted with a Zoex ZX2 GCxGC cryogenic modulator (Houston, TX, USA) interfaced with an Almsco BenchTOFdx time-of-flight mass spectrometer (Almsco International, Llantrisant, Wales, UK) operated in positive ion electron ionization mode and calibrated with perfluorotributylamine. The scan speed was 50 Hz. The resolution of the mass spectrometer was 1000 at mass 1000. The first-dimension column was a 100% dimethyl polysiloxane 50 m x 0.25 mm x 0.40 μm VF1-MS (Varian, Palo Alto, USA), with an efficiency of 211700 theoretical plates (n-tridecane) and the second-dimension column was a 50% phenyl polysilphenylene siloxane 1.5 m x 0.1 mm x 0.1 μm BPX50 (SGE, Melbourne, Australia) with an efficiency of 5121 theoretical plates per meter (biphenyl). Thus the product efficiency of the GCxGC system was calculated as approximately 1.6 billion theoretical plates. Helium was used as carrier gas, and the flow was kept constant at 0.7 mL min
1.

ARTICLE

Figure 1. Total ion current chromatogram of oil sands process water naphthenic acids (methyl esters) examined by GCxGC-ToF-MS illustrating high chromatographic resolution by GCxGC compared with GCMS (white line on black background). Also shown are structures of some of the acids identified. Note the positions of the substituents are speculative except for I and VI which were verified by comparison with authentic acids.

Samples (1 μL) were injected at 280 °C splitless. The oven was programmed from 40 °C (held for 1 min) and then heated to 300 at 2 °C min
1 and then at 10 °C min
1 to 320 °C (held for 10 min). The modulation period was 5 s. The mass spectrometer transfer line temperature was 280 °C and ion source temperature 300 °C. Data processing was conducted using GC Image v2.1 (Zoex, Houston, TX, USA).

’ RESULTS AND DISCUSSION We examined, by GCxGC-ToF-MS, a sample of the methyl ester derivatives of OSPW NA described previously.20 The OSPW was collected en route to storage in an in-pit settling basin.20 By conventional GC-MS the OSPW NA were almost completely unresolved, as has been reported frequently and as is shown by the black and white background in Figure 1. In contrast, GCxGC resolution under the optimized conditions was very high (Figure 1, colored foreground), allowing EI mass spectra containing molecular and fragment ions of many individual acids to be obtained (Figure 2). The normal background of ions which is caused by the thermal desorption of the GC stationary phases (so-called ‘bleed’ ions) was also well separated by GCxGC from ions produced by ionization of the acid methyl esters, which further improved the quality of the mass spectra of 3155

dx.doi.org/10.1021/es103721b |Environ. Sci. Technol. 2011, 45, 3154–3159

Environmental Science & Technology

ARTICLE

Figure 2. Mass spectra of examples of methyl esters of tricyclic acids identified in oil sands process water NA. Two acids (I and VI) were identified by comparison of the spectra and GC x GC retention times with those of purchased reference acids (methyl esters).

the unknown esters and the well-known GCxGC chromatographic ‘tiling’21 aided the assignment of structurally related groups of acids. Thus, due to these factors and the over 1.6

billion theoretical plates calculated for the combined GCxGC columns, the mass spectra obtained for many of the OSPW acids (methyl esters) were essentially those of individual compounds, 3156

dx.doi.org/10.1021/es103721b |Environ. Sci. Technol. 2011, 45, 3154–3159

Environmental Science & Technology as was also shown by the close similarities with the spectra of some relevant authentic acids (methyl esters). The mass spectra contained clear molecular ions which showed that the OSPW NA comprised mainly C11
19 bi- to pentacyclic acids, fitting the formula CnH2nþzO2. Although numerous other compounds have been suggested to be present in OSPW NA (e.g., refs 3 and references therein), we detected overwhelmingly NA (methyl esters) fitting the above formula, with only a few minor hydrocarbons and other constituents. Unrefined oil sands bitumen has been reported previously to contain 90% tricyclic acids, and electrospray mass spectra of OSPW NA have routinely shown that tricyclic and bicyclic acids are the major components (e.g., refs 4 and 7); therefore, we concentrated on identifying the tricyclic components in this study. (The tetra- and pentacyclic acids are described elsewhere.19) The mass spectra of the tricyclic OSPW acids did not match those in mass spectral libraries or in any published literature available to us. However, we were able to identify several of them by interpreting the spectra from first principles, then obtaining as many reference acids as were available, esterifying these, and obtaining the spectra of the methyl esters by GCxGC-ToF-MS for comparison with the spectra and GCxGC retention times of the unknowns. Many more NA remain to be identified. The results thus far show that the OSPW comprises an extensive series of diamondoid tricyclic acids (structures, Figure1). Thus, adamantane-1-carboxylic acid (structure I) was identified by comparison of the mass spectrum (Figure 2) and GCxGC retention times with that of a reference sample. The spectrum contained a minor molecular ion (m/z 194) and was dominated by a base peak ion m/z 135 due, we suggest, to fragmentation and loss of the methylated carboxy group. The corresponding adamantane-2-carboxylic acid (II) was also identified. This isomer is known to have a later retention time on the first apolar GC column than the 1-isomer, and the mass spectrum (Figure 2) was also characterized by a molecular ion (Mþ. 194) and major fragment ions due to loss of methanol (m/z 162), typical of methyl esters and again, loss of the methylated carboxy group (m/z 135). However, a noteworthy difference to the spectrum of the 1-isomer, in which the carboxy group is substituted at a quaternary center (C-1), was the base peak ion at m/z 134 in the putative 2-isomer. We interpret formation of this ion as due to loss of the methylated carboxy group followed by H-transfer at the tertiary center to form an even mass alkenyl ion (m/z 134). This dominance of an even mass base peak ion might prove to be a useful feature for distinguishing isomers of diamondoid acids substituted at tertiary centers (e.g., C-1) compared with those substituted at quaternary centers (e.g., C-2). Indeed, the same phenomenon was observed when the mass spectra of the methyl esters of authentic samples of diamantane-1-carboxylic acid (carboxy group substituted at a quaternary center, C-1) and diamantane-3-carboxylic acid (carboxy group substituted at a tertiary center, C-3) were obtained in our related studies of pentacyclic OSPW NA.19 We were also able to identify the methyl ester of 3-ethyl adamantane-1-carboxylic acid (VI) by comparison of the spectrum with that of a reference sample (Figure 2). The mass spectrum (Figure 2) like that of a reference sample contained a molecular ion m/z 222 and ions due to loss of the ethyl group (m/z 193) and the methylated carboxy group (m/z 163). In addition, but more tentatively, we were able to identify numerous methyl, dimethyl and ethyladamantane carboxylic

ARTICLE

Figure 3. Two-dimensional mass chromatograms illustrating GCxGC chromatographic resolution of selected adamantane carboxylic acids (A. m/z 134 þ135. B. m/z 163 þ193 C. m/z 177 methyl esters) in oil sands produced waters, as examined by GCxGC-ToF-MS.

acids and adamantane ethanoic acid isomers by interpretation of the mass spectra from first principles (Figure 2). The spectra of the methyl esters of the methyl adamantane carboxylic acids were characterized by molecular ions (m/z 208) and were dominated by a base peak ion m/z 149 due to fragmentation and loss of the methylated carboxy group; those of the esters of the dimethyladamantane carboxylic acids (numerous isomers were present, separated by GCxGC; Figure 3) by a molecular ion (m/z 222), dominated by a base peak ion m/z 163. (Common components such as phthalate esters, which also have dominant m/z 149 ions in their mass spectra, were well separated from such NA by GCxGC as they are esters of aromatic acids, the aromaticity resulting in good separation from the methyl esters of the NA on the second, more polar, GC phase. Thus the NA were easily differentiated from common laboratory contaminants, such as phthalates). Other isomers of ethyl adamantane carboxylic acids were identified by the presence of the latter ions in different relative abundances (Figure 3S). Adamantane ethanoic acids were also present; for example, the spectrum of a methyladamantane ethanoic acid contained a molecular ion (m/z 222) consistent with the methyl ester of a C13 tricyclic acid, but the base peak ion was m/z 149, indicative of a methyl substituted (C11) tricyclic core, rather than the m/z 163 characteristic of the dimethyl C12 adamantane core. This formation of ethanoic acids is consistent with an origin from biodegradation of adamantane hydrocarbons (ref 13, Figure 1S). 3157

dx.doi.org/10.1021/es103721b |Environ. Sci. Technol. 2011, 45, 3154–3159

Environmental Science & Technology The distributions of multiple series of adamantane carboxylic acids (methyl esters) could be easily displayed by selected ion mass chromatography of key ions in the spectra (Figure 3). Thus GCxGC-ToF-MS allows the distributions of multiple individual acids in different OSPW or OSPW and environmental samples to be compared routinely. We propose that most, if not all, of the diamondoid acids are biotransformation products of methyl, dimethyl, ethyl, and ethylmethyladamantane hydrocarbons (Figure 1S). Although an origin from oxidation during oil sands processing may also be feasible since the hydrocarbons are widely reported in crude oils including in Western Canada oils, laboratory and field studies have shown that both adamantane and the methyl adamantane hydrocarbons are indeed slightly biotransformed by some bacteria.23,24 However, the acidic products have never been identified previously, and discussions of NA in oil sands have therefore not considered the involvement of such acids to our knowledge. Despite reports of unidentified diacids in OSPW based on data from nuclear magnetic resonance studies,8 we could not detect adamantane dicarboxylic acids using the spectra of reference compounds as a guide, suggesting that further degradation of the alkyladamantane monoacids does not proceed by this route. However, it is known that biotransformation of adamantane proceeds with formation of a lactone25 which when hydrolyzed would ring open to form a bicyclic (hydroxy) acid. If a similar process also occurred during biodegradation of the alkyladamantanes in addition to, or in place of, alpha oxidation of the alkyl substituents (Figure 1S), the products might be bicyclic acids, such as those we detected in the OSPW NA. Synthetic analogues of such acids are known, and the facile ring-opening and closing has been studied;26 confirmation of these identifications will require such syntheses to be repeated. Toxicity assays can then also be conducted. The discovery of adamantane diamondoid acids in OSPW NA allows identification, we believe, not only of the first structural class of NA in oil sands but also of some the first individual isomers of oil sands NA to be made, including members of the abundant tricyclic constituents. It is clear from our data that many more NA, including other tricyclic compounds, are present in OSPW and since two-dimensional comprehensive GCxGC-ToF-MS has been shown to be a powerful tool in elucidating the complex distributions of at least some NA in OSPW, the method can now be used to identify further acids and to quantify the concentrations of the individual acids. The approach will also allow directed syntheses of further relevant acids to be conducted and the toxicity of acids responsible for any adverse biological effects to be much better constrained. If adamantane carboxylic acids are not toxic, it will be important to exclude them from environmental measurements of NA concentrations. Furthermore, the applicability of this method for profiling NA distributions will be helpful for more accurately determining the sources of NA in surface and groundwater samples and thus may help to resolve controversies regarding the origins of NA in regions where both natural and pollutant NA sources exist, which is a critical development for monitoring programs responsible for the detection of potential leaching of oil sands process materials into the Athabasca River system.16
18 The same methods can be deployed for studies of NA from other sources, including offshore produced water.

’ ASSOCIATED CONTENT

bS

Supporting Information. The figures show a scheme for the bacterial oxidation of alkyladamantanes and additional

ARTICLE

electron ionization mass spectra of methyl esters of adamantane carboxylic acids. This material is available free of charge via the Internet at http://pubs.acs.org.

’ AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected].

’ ACKNOWLEDGMENT Funding of this research was provided by an Advanced Investigators Grant (no. 228149) awarded to S.J.R. for project OUTREACH, by the European Research Council, to whom we are extremely grateful. We thank the University of Plymouth for a Ph.D. scholarship (D.J.) and B. White, D. Rosser, A. Cole, and N. Bukowski of Almsco International for valuable discussions on aspects of multidimensional GC-MS. ’ REFERENCES (1) Kean, S. Eco-alchemy in Alberta. Science 2009, 326, 1052–1055. (2) Whitby, C. Microbial naphthenic acid degradation. Adv. Appl. Microbiol. 2010, 70, 93–125. (3) Grewer, D. M.; Young, R. F.; Whittal, R. M.; Fedorak, P. M. Naphthenic acids and other acid-extractables in water samples from Alberta: What is being measured? Sci. Total Environ. 2010, 408, 59976010. (4) Frank, R. A.; Kavanagh, R.; Burnison, B. K.; Arsenault, G.; Headley, J. V.; Peru, K. M. P.; Van Der Kraak, G.; Solomon, K. R. Toxicity assessment of collected fractions from an extracted naphthenic acid mixture. Chemosphere 2008, 72, 1309–1314. (5) Clemente, J. S.; Fedorak, P. M. A review of the occurrence, analyses, toxicity, and biodegradation of naphthenic acids. Chemosphere 2005, 60, 585–600. (6) Headley, J. V.; McMartin, D. W. A review of the occurrence and fate of naphthenic acids in aquatic environments. J. Environ. Sci. Health, Part A: Toxic/Hazard. Subst. Environ. Eng. 2004, 39, 1989–2010. (7) Headley, J. V.; Peru, K. M.; Barrow, M. P. Mass spectrometric characterization of naphthenic acids in environmental samples: a review. Mass Spec. Rev. 2009, 28, 121–134. (8) Frank, R. A.; Fischer, K.; Kavanagh, R.; Burnison, B. K.; Arsenault, G.; Headley, J. V.; Peru, K. M.; Van Der Kraak, G.; Solomon, K. R. Effect of carboxylic acid content on the acute toxicity of oil sands naphthenic acids. Environ. Sci. Technol. 2009, 43, 266–271. (9) Bataineh, M.; Scott, A. C.; Fedorak, P. M.; Martin, J. W. Capillary HPLC/QTOF-MS for characterizing complex naphthenic acid mixtures and their microbial transformation. Anal. Chem. 2006, 78, 8354–8361. (10) Han., X.; Mackinnon, M. D.; Martin, J. W. Estimating the in situ biodegradation of naphthenic acids in oil sands process waters by HPLC/HRMS. Chemosphere 2009, 76, 63–70. (11) Headley, J. V.; Peru, K. M.; Armstrong, S.; Han, X.; Martin, J. W.; Mapolelo, M. M.; Smith, D. F.; Rodgers, R. P.; Marshall, A. G. Comparison of Aquatic Plant Derived Changes in Aqueous Naphthenic Acid Profiles Determined by ESI/MS, HPLC/QTOF MS and FT-ICR MS. Rapid Commun. Mass Spectrom. 2009, 23, 515–522. (12) Brient, J. A.; Wessner, P. J.; Doyle, M. N. In Encyclopaedia of Chemical Technology 1017; John Wiley: 1995. (13) Smith, B. E.; Rowland, S. J. A derivatisation and liquid chromatography/electrospray ionisation multistage mass spectrometry method for the characterisation of naphthenic acids. Rapid Commun. Mass Spectrom. 2008, 22, 3909–3927. (14) Hao, C.; Headley, J. V.; Peru, K. M.; Frank, R.; Yang, P.; Solomon, K. R. Characterization and pattern recognition of oil sands naphthenic acids using comprehensive two dimensional gas chromatography/time of flight mass spectrometry. J. Chromatogr., A 2005, 1067, 277–284. 3158

dx.doi.org/10.1021/es103721b |Environ. Sci. Technol. 2011, 45, 3154–3159

Environmental Science & Technology

ARTICLE

(15) Shepherd, A. G.; van Mispelaar, V.; Nowlin, J.; Genuit, W.; Grutter, M. Analysis of naphthenic acids and derivatization agents using two dimensional gas chromatography and mass spectrometry: impact on flow assurance predictions. Energy Fuels 2010, 24, 2300–2311. (16) Royal Society of Canada. Environmental and health impacts of Canada’s oil sands industry. 2010. Available at http://www.rsc.ca/ expertpanels_reports.php (accessed 25th Jan 2011). (17) Oilsands Advisory Panel. A foundation for the future: Building an environmental monitoring system for the oil sands. A report submitted to the Minister of Environment. 2010. (18) Schindler, D. Tar sands needs solid science. Nature 2010, 468, 499–501. (19) Rowland, S. J.; West, C. E.; Scarlett, A. G.; Jones, D.; Frank, R. A. Identification of individual tetra- and pentacyclic naphthenic acids in oil sands process water by comprehensive two-dimensional gas chromatography-mass spectrometry. Rapid Commun. Mass Spectrom. 2011 (doi: 10.1002/rcm.4977). (20) Frank, R. A.; Kavanagh, R.; Burnison, B. K.; Headley, J. V.; Peru, K. M.; Van Der Kraak, G.; Solomon, K. R. Diethylaminoethyl-cellulose clean-up of a large volume naphthenic acid extract. Chemosphere 2006, 64, 1346–1352. (21) Adachour, M.; Beens, J.; Brinkman, U. A. Th. Recent developments in the application of comprehensive two-dimensional gas chromatography. J. Chromatogr., A 2008, 1186, 67–108. (22) Mishra, S.; Meda, V.; Dalai, A. K.; McMartin, D. W.; Headley, J. V.; Peru, K. M. Photocatalysis of naphthenic acids in water. J. Water Resour. Prot. 2010, 2, 644–650. (23) Grice, K.; Alexander, R.; Kagi, R. I. Diamondoid hydrocarbon ratios as indicators of biodegradation in Australian crude oils. Org. Geochem. 2000, 31, 67–73. (24) Wang, Z.; Yang, C.; Hollebone, B.; Fingas, M. Forensic fingerprinting of diamondoids for correlation and differentiation of spilled oil and petroleum products. Environ. Sci. Technol. 2006, 40, 5636–5646. (25) Selifonov, S. Microbial oxidation of adamantanone by Pseudomonas putida carrying the camphor catabolic plasmid. Biochem. Biophys. Res. Commun. 1992, 186, 1429–1436. (26) Sasaki, T.; Eguchi, S.; Toru, T. Novel substitution reaction of adamantanone. A simple synthesis of bicyclo[3.3.0.]non-2-ene-7-carboxylic acid. J. Am. Chem. Soc. 1969, 91, 3390–3391.

3159

dx.doi.org/10.1021/es103721b |Environ. Sci. Technol. 2011, 45, 3154–3159