Science of the Total Environment 777 (2021) 146022 Contents lists available at ScienceDirect Science of the Total Environment j ourna l homepage: www.e lsev ie r .com/ locate /sc i totenv Preparative isolation, fractionation and chemical characterization of dissolved organics from natural and industrially derived bitumen-influenced groundwaters from the Athabasca River watershed☆ Richard A. Frank a,⁎, Anthony E. Bauer b, JamesW. Roy a, Greg Bickerton a, Martina D. Rudy a, Ruth Vanderveen a, Suzanne Batchelor a, Sophie E. Barrett a, Craig B. Milestone a,1, Kerry M. Peru c, John V. Headley c, Pamela Brunswick d, Dayue Shang d, Andrea J. Farwell b, D. George Dixon b, L. Mark Hewitt a a Water Science and Technology Directorate, Environment and Climate Change Canada, 867 Lakeshore Road, Burlington, ON L7S 1A1, Canada b Department of Biology, University of Waterloo, Waterloo, ON N2L 3G1, Canada c Water Science and Technology Directorate, Environment and Climate Change Canada, 11 Innovation Boulevard, Saskatoon, SK S7N 3H5, Canada d Pacific and Yukon Laboratory for Environmental Testing,Water Science and Technology Directorate, Environment& Climate Change Canada,North Vancouver, British Columbia V7H 1B1, Canada H I G H L I G H T S G R A P H I C A L A B S T R A C T • Method for extraction and fractionation of dissolved organics was applied to bitumen-influenced groundwater sam- ples • Dissolved organics are considered prin- cipal toxic components in OSPW • Three large volume organic fractions were generated and characterized by complement of methodologies • Isolated fractions increased in polarity and degree of aromaticity • Chemical composition was similar between groundwaters influenced by OSPW and/or natural bitumen sources ☆ Disclaimer: The views in this paper are held by the au ⁎ Corresponding author at: WSTD, Environment and Cl E-mail address: richard.frank@canada.ca (R.A. Frank). 1 Present address: Faculty of Applied Science and Techn https://doi.org/10.1016/j.scitotenv.2021.146022 0048-9697/Crown Copyright © 2021 Published by Elsevie a b s t r a c t a r t i c l e i n f o Article history: Received 15 November 2020 Received in revised form 27 January 2021 Accepted 16 February 2021 Available online 3 March 2021 Editor: Kevin V. Thomas Keywords: Bitumen-derived organics Mixture fractionation Environmental forensics Groundwater Recent analytical advances have provided evidence that groundwater affected by oil sands process-affected water (OSPW) is reaching the Athabasca River at one location. To understand and discriminate the toxicological risks posed by OSPW-influenced groundwater relative to groundwaters associated with natural oil sands de- posits, these highly complex mixtures of soluble organics were subjected to toxicological characterization through effects directed analysis. A recently-developed preparative fractionation methodology was applied to bitumen-influenced groundwaters and successfully isolated dissolved organics from both industrial and natural sources into three chemically distinct fractions (F1, F2, F3), enabling multiple toxicological assessments. Analyt- ical techniques included electrospray ionization high resolution mass spectrometry (ESI-HRMS), liquid chroma- tography quadrupole time-of-flight mass spectrometry (LC-QToF/MS), gas chromatography mass spectrometry (GC–MS), and synchronous fluorescence spectroscopy (SFS) methods, which did not reveal obvious differences between sources. Comparisons between fractions within each source consistently demonstrated that F3 thors and are not necessarily representative of the official policy of the authors' individual affiliations. imate Change Canada, 867 Lakeshore Road, Burlington, ON L7S 1A1, Canada. ology, Sheridan College, 7899 McLaughlin Road Brampton, ON, Canada L6Y 5H9. r B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). http://crossmark.crossref.org/dialog/?doi=10.1016/j.scitotenv.2021.146022&domain=pdf https://doi.org/10.1016/j.scitotenv.2021.146022 mailto:richard.frank@canada.ca https://doi.org/10.1016/j.scitotenv.2021.146022 http://creativecommons.org/licenses/by-nc-nd/4.0/ http://www.sciencedirect.com/science/journal/ www.elsevier.com/locate/scitotenv R.A. Frank, A.E. Bauer, J.W. Roy et al. Science of the Total Environment 777 (2021) 146022 contained compounds with greater polarity than F2, which in turnwasmore polar than F1. The abundance of O2 species was confined to F1, including naphthenic acids often cited for being the primary toxicants within bitumen-influenced waters. This result is consistent with earlier work on aged OSPW, as well as with other ex- traction methods, suggesting that additional factors other than molecular weight and the presence of acid func- tionalities play a prominent role in defining compound polarities and toxicities within complex bitumen-derived organic mixtures. The similarities in organic abundances, chemical speciation, aromaticity, and double bond equivalents, concomitant with inorganic mixture similarities, demonstrate the resemblances of bitumen- influenced groundwaters regardless of the source, and reinforce the need for more advanced targeted analyses for source differentiation. Crown Copyright © 2021 Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). 1. Introduction Surface mining in Canada's oil sands region of northern Alberta employs an adaptation of the Clark extraction process for the isola- tion of bitumen, which often utilizes a mixture of hot water and NaOH (FTFC, 1995a, 1995b). Water is recycled throughout the ex- traction process, which results in oil sands process-affected water (OSPW) accumulating inorganic and organic constituents from the oil sand material. The OSPW is then stored in large settling basins (Mahaffey and Dube, 2017), often termed tailings ponds. Migration of OSPW-affected groundwater from these tailings ponds to nearby rivers and wetlands is an environmental concern given the toxicity of many of its constituents (Fennell and Arciszewski, 2019), and has been documented for one pond adjacent to the Athabasca River (Frank et al., 2014; Hewitt et al., 2020). Advancement in the chemical characterization of organic com- pounds in OSPW and its associated extracts has greatly improved our understanding of these complex bitumen-derived organic mixtures. A wide range of organic constituents has been revealed, including various diamondoidmono-aromatic acids (Rowland et al., 2011a, 2011b, 2011c, 2012), bicyclic acids (Wilde et al., 2015), and varying abundances of dif- ferent chemical species (Headley et al., 2013; Pereira et al., 2013; Sun et al., 2017). All of these compound classes are derived from the mined ore, and groundwater exposed to natural oil sands deposits can exhibit a strong bitumen influence as well. The exposed groundwater can exhibit a similar chemical composition and concentrations of acid extractable organics (AEOs) quite comparable to OSPW (Frank et al., 2014; Hewitt et al., 2020; Ross et al., 2012; Sun et al., 2017). Select groupings of these organic compounds are under further investigation for their diagnostic potential to differentiate bitumen-influencedwaters from natural and industrial sources (Brunswick et al., 2020; Frank et al., 2014; Hewitt et al., 2020). Laboratory bioassays have demonstrated that natural bitumen, as well as OSPW and extracts of the chemicals within, are toxic to several different classes of aquatic organism (Bartlett et al., 2017; Cardoso et al., 2020; Mahaffey and Dube, 2017; Marentette et al., 2015). While the relative toxicities of each of the different chemical classes may not be presently known, extracts of the soluble organic constituents, namely the AEOs which include naphthenic acids, have been long con- sidered to be among the principal toxic components of OSPW (Brown and Ulrich, 2015; MacKinnon and Boerger, 1986; Mahaffey and Dube, 2017). Improved understanding of the toxicity of the bitumen-derived organicmixtures in groundwaters is needed to better assess the relative risks posed from anyOSPWseepage beyond tailings pond containments relative to the discharge of natural groundwaters passing through natu- ral oil sands deposits. To facilitate the required toxicity assessments, an effect directed analysis approach was adopted in the present study. A preparative method that recovers the dissolved organics from bitumen-influenced waters in three fractions of increasing polarity (Bauer et al., 2019) was applied to two sets of groundwaters from previously investigated sites (Frank et al., 2014; Hewitt et al., 2020): one set influenced solely by natural bitumen and another adjacent to a tailings pond with a mixed 2 influence of OSPW and natural bitumen. The objectives of this study in- cluded 1) determining if the method developed for aged OSPW (Bauer et al., 2019) could also be applied to fractionate large volumes of other bitumen-influenced waters for toxicological and chemical evaluations, and 2) to determine if fraction compositions differed between and within sources. 2. Methods and materials 2.1. Sample collection Shallow riparian groundwater was collected in September 2013 from sites previously determined to be influenced by OSPW and/or nat- ural oil sands bitumen (Frank et al., 2014; Hewitt et al., 2020) (Fig. 1). Groundwaters fromDrive-point (DP)-1 and DP-2 sites had been identi- fied as being affected by natural bitumen deposits only, while those from DP-4 and DP-5 sites were identified as being influenced by OSPW from a nearby reclaimed tailings pond mixed with natural groundwater (possibly with natural bitumen influence). Due to the lower water level of the Athabasca River during the 2013 sampling pe- riod, the locations for DP-4 and DP-5 were up to 15m closer to themid- dle of the river than the previous sample locations (Frank et al., 2014). This difference, along with possible temporal changes in groundwater flow patterns, means that the samples from 2013 may not have the same composition as past samples from those same four locations. Groundwater was extracted at depths of 50–90 cm below the riverbed with a stainless steel drive-point system (Roy and Bickerton, 2010). To accommodate the large volume collection, groundwater was pumped slowly over several hours (4–24 h) from a series of drive-points (3–5, spaced over <3m along the bank) at each sample location intomultiple 18-L stainless steel vessels fitted with Viton seals. Sample collection from each drive point commenced following the equilibration of field- measured parameters (electrical conductivity, pH, dissolved oxygen). Once collected, the groundwater samples were maintained in the dark at 4 °C during transport to the Canada Centre for Inland Waters in Bur- lington, ON, as well as in storage, until sub-sampling (within 7 days) and extraction (within 18 months) was completed. 2.2. Centrifugation of groundwater samples prior to extraction Through the drive-point collection process, subsurface fine sedi- ments could bemobilized and introduced to the groundwater collection vessel. As these particulates would not naturally flow with the ground- water, and could also slow filtration through the extraction column, they were removed. A continuous flow centrifuge (Westfalia Model KA 2–06-075) operating at 9470 rpm was used to remove >90% of the suspended sediments, which were collected in a stainless steel bowl (Droppo et al., 2009). 2.3. Isolation and fractionation of dissolved organics Dissolved organicswere isolated and split into three fractions, by dif- ferences in polarity, using a recently developed preparative scale solid http://creativecommons.org/licenses/by-nc-nd/4.0/ DP-2 DP-4 DP-5 DP-1 Drive-point site Tailings pond Reclaimed pond Fig. 1. Map depicting locations of groundwater drive-point sampling sites (red), and approximate (Fennell and Arciszewski, 2019) locations of industrial tailings ponds (blue) and a reclaimed tailings pond site (purple). R.A. Frank, A.E. Bauer, J.W. Roy et al. Science of the Total Environment 777 (2021) 146022 phase extraction (SPE) protocol (Bauer et al., 2019). The base extract- able organics (BEOs; F1, F2) and AEOs (F3) isolated in this method are hereafter referred to as A/BEOs. The fractionation apparatus consisted of a glass column with plunger (10 cm ID x 30 cm height, Spectrum Chromatography, Houston, TX), two 200-L HDPE barrels, and a control- ler and motor (Cole-Palmer, Montreal, QC) with a rotary vane pump head (Procon Pumps, Laval, QC). The column was effectively operated as a SPE cartridge where feedstock flow was directed onto the resin bed using an adjustable plunger. Awater pumpwas used to pull the ini- tial filtered sample from the sample barrel through the resin in the first column (SPE-1, Fig. 2) with negative pressure and transfer of the filtrate to a second barrel. In brief, 100–150 L of centrifuged groundwaterwere filtered through two consecutive 120 g columns of ENV+ (hydroxylated polystyrene 3 divinylbenzene; Biotage®, NC, USA), followed by a total of 3 stages of solvent extraction. Prior to filtering through SPE-1, the centrifuged groundwater sample was adjusted to pH 11.0 ± 0.5 with 10 M sodium hydroxide (NaOH), mixed for approximately 1 h with a hand drill fitted with a PTFE mixing rod, and allowed to equilibrate for 12 h. The pHwas then re-tested, adjusted accordingly, and allowed to equilibrate for 6 h or until stable at 11.0±0.5. Following this, the ENV+ resinwas sequen- tially preconditionedwith 1.5 L each of ethyl acetate (EtOAc) andmeth- anol (MeOH), and 6 L of pH 11 reverse osmosis (RO) water. A barrel containing the 180 L of centrifuged groundwater was then plumbed up- stream into the pre-conditioned ENV+columnand a second empty bar- rel plumbed downstream of the column. Herein and throughout all conditioning and filtering steps, the solvent/water in the column was maintained at a height of 10 cm above the resin bed and the plunger Fig. 2. Fractionationmethod (Bauer et al., 2019) schematic displaying Stage 1 and Stage 2 SPE loading followed by Soxhlet extraction using solvents indicated. The fractionation resulted in the generation of fractions containing dissolved organic constituents of relative lower polarity (F1), intermediate polarity (F2), and higher polarity (F3). R.A. Frank, A.E. Bauer, J.W. Roy et al. Science of the Total Environment 777 (2021) 146022 at a height of 1 cm above the solvent/water to avoid disturbance of the resin, and the filtration rate was maintained at 100 ± 10 mL/min. Following the first extraction of 180 L of groundwater (SPE-1; Fig. 2), the column was disassembled, the resin carefully transferred into a 4-L glass beaker covered with a large Kimwipe® and then allowed to dry in a fume hood for 12–24 h. Once dry, the analytes from SPE-1 were Soxhlet extracted for 12 h using two Soxhlet assemblies, each with 1.5 L of EtOAc and 60 g ENV+ resin packed between 500 g of sodium sulfate (NaSO4). Following the 12 h extraction, the 3 L of EtOAc extract was pooled and dried 4 times through 400 g NaSO4 and 8 μm pore-size filter paper (Whatman grade 40 ashless, Sigma-Aldrich®, Oakville, ON). The final extract in EtOAc is hereafter referred to as Fraction 1 (F1), and is ex- pected to contain the least polar soluble organics based on previous OSPW extractions using this approach (Bauer et al., 2019). The resin was then removed from the Soxhlet thimbles, allowed to dry, re-placed in new thimbles with fresh NaSO4, and the extraction and drying process was repeated using a total of 3 L ofMeOH, hereafter referred to as Fraction 2 (F2). This fraction is expected to contain soluble organics with interme- diate polarity relative to the other fractions. The filtrate following SPE-1 was acidified to pH 2 using 12 M HCl as described above for the initial adjustment to pH 11. For prepara- tion of the SPE-2 stationary phase, 120 g fresh ENV+ resin was placed into the cleaned column, conditioned, and equilibrated as de- scribed previously for SPE-1, with the exceptions that only MeOH was used and the final conditioning was with pH 2 ROwater. Follow- ing conditioning, the acidified SPE-1 filtrate was extracted in the SPE-2 column as described above for SPE-1, at 100 ± 10 mL/min. The SPE-2 resin was then collected and dried as for SPE-1, with sub- sequent Soxhlet extraction using a total of 3 L MeOH that was simi- larly dried through NaSO4, and is referred to as Fraction 3 (F3). This 4 fraction is expected to contain the most polar soluble organics rela- tive to the other fractions. 2.4. Synchronous fluorescence spectroscopy (SFS) Synchronous fluorescence spectra were recorded with a Perkin– Elmer Luminescence Spectrometer LS50B, as previously described (Frank et al., 2016; Kavanagh et al., 2009). Samples were filtered through washed disk filters (PES, 25 mm GD/X, O.2 mm pore size; GE Healthcare UK Ltd., Buckinghamshire, UK) to remove particulates and were then scanned in a 1 cm quartz cuvette with PTFE stopper (Hellman, Concord, ON, Canada) at 20 ± 2 °C. All data were collected using FL WinLab 3 software (Perkin–Elmer, Norwalk, CT). The wave- length difference between the excitation and emission monochro- mators (Dk) was optimized by measuring the spectra of dilute soluble organics at various offset values (10–60 nm), with a Dk of 18 nm selected, and synchronous fluorescence spectra were col- lected in the 250–400 nm excitation wavelength range (Kavanagh et al., 2009). Excitation and emission monochromator slit widths were set at 5 nm, scan speed at 50 nm min−1 and resolution at 0.5 nm. The spectra were blank corrected with Milli-Q water and then plotted using Origin software ver. 2019b (OriginLabCorp., Northampton,MA).Detected maxima at 272, 307, and 323 were depictive of bitumen influence (Frank et al., 2016). 2.5. Electrospray ionization high resolution mass spectrometry (ESI-HRMS) An LTQ Orbitrap Elite (Thermo Fisher Scientific) instrument was used for infusion ESI-HRMS analysis with a pre-defined 5-point regres- sion of OSPW-derived organic acids (Frank et al., 2008) at known R.A. Frank, A.E. Bauer, J.W. Roy et al. Science of the Total Environment 777 (2021) 146022 concentrations used to determine resulting dissolved organic concen- trations. Operating in full scan negative-ionmode, themass spectrome- ter ran at a m/z scan range of 100–600. Achieved resolution at m/z 120 = 240,000, m/z 210 = 185,000, m/z 300 = 150,000, and m/z 400= 130,000, and all of the ions in the m/z 100 to 300 range had res- olution between 150,000 to 240,000. The mass accuracy was <2 ppm error for all mass assignments. Operating parameters were as follows; sheath gasflow rate 25 (arbitrary units), spray voltage 2.90 kV, auxiliary gas flow rate 5 (arbitrary units), S lens RF level 67%, heater temperature 50 °C, and capillary temperature 275 °C. Infusion solvent used for loop injection sample introduction was 50:50 acetonitrile:water containing 0.1% ammonium hydroxide at a flow rate of 200 μL/min. Software used for molecular analysis was Xcalibur v 2.1 (Thermo Fisher Scien- tific) and Composer v 1.0.2 (Sierra Analytics, Inc.). 2.6. Liquid chromatography quadrupole time-of-flight mass spectrometry (LC-QToF/MS) quantification Detailed description of this analysis has been described previously (Brunswick et al., 2015, 2016a, 2016b), and is briefly summarized here. Groundwater samples and derived fractions were adjusted to pH 10–11 with ammonium hydroxide and spiked with the internal standard, decanoic-d3 acid. Direct injection reverse phase liquid chromatography was then used to separate the organic compounds in the sample, to- gether with detection by an Agilent 6550 iFunnel quadrupole time-of- flight mass spectrometer (LC/QToF). The BEOs (F1, F2) and AEOs (F3) were ionized in electrospray negative mode and data was acquired by total ion scan (TIC). The instrument qualitative software was able to screen the total ion scan for accurate peak matching using the formula of O2:O3:O4 A/BEO species. It is noted that, due to the presence of isomers, there may be different A/BEO peaks in the reference material compared to the samples. When individual isomer peaks attained acceptable mass accuracy (routinely <5 ppm), reached quantitation limits, and were free of interferences, the resultswere transferred to the quantitative soft- ware program for integration. Final analysis employed aweighted 1/x re- gression standard curve of pooledA/BEO responses in ratio to the internal standard. The calibration range was dependent upon the reference stan- dard employed, with in-house validatedmethods using eitherMerichem Technical mix or a validated extract of OSPW AEOs isolated from fresh OSPW. System suitability standards, blanks, and calibration standards were analyzed at the beginning and end of each analytical sequence with Quality Control samples included within each analytical batch. 2.7. LC-QToF/MS profiling All LC-QToF/MS analyses utilized a methanol mobile phase and re- quired that all samples were dissolved in methanol, therefore EtOAc fraction (F1) aliquots were solvent exchanged into methanol. The analy- sis was carried out in full scan negative ion mode (scan range m/z 100–980, at a sampling rate of 3 spectra/sec and mass resolution of 45,000) using an LC-QToF 6520 (Agilent Technologies, Santa Clara, Cali- fornia, USA) under the following conditions: Gas temp 350 °C, drying gas 10 L/min, nebulizer 35 psi, VCap 3000 V, Fragmentor 130 V, Skimmer 65V, and referencemass recalibration enabled. The LC conditionswere as follows: Column Poroshell 120 EC-C18, 3.0 × 50 mm 2.7 μm, Solvent A Water (0.1% formic acid), Solvent B Methanol (0.1% formic acid), initial conditions 95% A for 2 min, to 100% B at 20 min, hold until 30 min. Sam- pleswere sandwich injectedwith 1 μL of labelled internal standard (9-an- thracene-d9-carboxylic acid, 84.4 pg/μL andDecanoic-d19 acid, 390 pg/μL) to verify against retention time drift. Final extractswere diluted if ion sat- uration of samples and internal standards was encountered. 2.8. Gas chromatography tandem mass spectrometry (GC–MS) All GC–MS analyses were conducted with samples that were meth- ylated using diazomethane and then dissolved in toluene. The analysis 5 was carried out in positive EI full scan mode (scan range m/z 50–500) at nominal mass resolution using a GC 7000A Tandem MS (Agilent Technologies, Santa Clara, California, USA). A 1 μL injection was made into a multimode inlet at 270 °C into a 30 m × 0.25 mm, 0.25 μm DB5 column (Agilent). Oven temperatures were programmed at 90 °C for 0.5 min, ramped to 300 °C at 40 °C/min, with a 5 min hold. 2.9. Inductively coupled plasma-sector field mass spectrometry (ICP-MS) analysis of metals and major ions Total and dissolved metals were analyzed at Environment and Cli- mate Change Canada's National Laboratory for Environmental Testing (NLET) (Burlington, ON) using Inductively Coupled Plasma-Sector Field Mass Spectrometry (SOP 2003 - Standard Operating Procedure for the Analysis of Dissolved, Extractable and Total TraceMetals inWater by “Di- rect Aspiration” or “In Bottle Digestion” Inductively Coupled Plasma- Sector Field Mass Spectrometry (ICP-SFMS; NLET 2008)). The analysis of anions (including chloride and sulphate) was performed by ion ex- change chromatography with conductivity detection (NLET Method 01–1080). The analysis of cations (Na, Ca, K, Mg) was performed by di- rect aspiration using atomic absorption (NLET Method 01–1061). 3. Results 3.1. Indicators of bitumen source in groundwater Although similar geochemical composition of waters collected near Tar Island Dyke have been reported in the same general location in dif- ferent years (Roy et al., 2016), it has also been noted that groundwater conditions can vary spatially in this area. Given the low river level dur- ing sampling for this study, there was some concern whether the two drive-point samples selected for previous OSPW influence (DP-4 and DP-5) in this study would be consistent with our previous work (sites sampled in 2010 (Frank et al., 2014) and in 2012 (Hewitt et al., 2020)). Unfortunately, the GC × GC analysis of Family A/B mono-aromatic acids that has been demonstrated to be highly diagnostic of OSPW influ- ence (Frank et al., 2014; Hewitt et al., 2020) was not available for these samples. However, previous research (Roy et al., 2016) reported on a se- ries of 16 potential indicators for screening purposes (e.g. having high concentrations of fluoride and molybdenum, and high values for sodium-chloride ratio and SFS, but low chloride-fluoride ratios) that co- incided with OSPW-affected groundwater samples from 2010 sampling using Principle Component Analysis (PCA). To test for consistency with the 2013 drive-point samples, a similar PCA, also with log-transformed data sets to improve normal distributions, was performed (Fig. 3). This included the entire data set of background and near-pond samples in the original PCA, including two OSPW samples and drive-point samples (DP 1–6) from 2010, for a total of 111 samples. To this original set was added the drive-point samples collected in 2012 (DP 1,3,4,5; (Hewitt et al., 2020)), and in 2013 (DP 1,2,4,5; this study); eight more samples. Analysis of Family A/B mono-aromatic acids indicated that DP-4 and DP-5 from the two previous campaigns (2010 sampling in Frank et al., 2014; 2012 sampling in Hewitt et al., 2020) contained influence from OSPW. Only 14 parameters were included in this PCA, as selenium and ammonium were not measured in the 2012 and/or 2013 samples. The first three components of this 14-indicator data set PCA account for 33, 19, and 15% of the total variation, respectively, which is nearly identical to the original PCA of Roy et al. (2016). For the combined data set totaling 119 samples assessed here, the plot of the first vs. third components reveals a similar pattern to that reported previously (Roy et al., 2016), both in terms of the parameter vectors and locations of the past samples. This indicates that the additional eight samples and removal of two parameters did not substantially alter the broad relative pattern among the data sets. Notably, DP-5 from 2013 plots in a similar area to DP-5 from 2010 and 2012, and to DP-4 from 2010 and 2012 (Fig. 3b), indicating a similar composition as these known OSPW- Table 1 Concentration of acid/base extractable organics in fractions and filtrate of groundwater sites determined by LC-QToF/MS. Values represent concentrations present in original vol- umes of respective water samples. Acid/base extractable organics (mg/L) DP-1 DP-2 DP-4 DP-5 Whole 1.58 5.93 0.01 9.51 F1 10.46 17.12 0.06 34.02 F2 0.54 1.07 0.00 0.11 F3 1.68 6.21 0.01 1.15 Filtrate O3-containing ions (41.1%). The components pres- ent in site DP-2 contained similar composition of O2 and O4-containing ions, 35.1% and 40.5% respectively, with minor contribution from other species (6.3%). Finally, DP-1 displayed an increase in oxygenated species to a maximum at O4 (39%) with a concurrent decrease in O5-O8. All sites exhibited a shift to a relative increase in oxygenated com- pounds from F1 to F3 (Fig. 5). For example, in DP-5, F1 is composed of predominantly O2-containing ions (75.2%) with lesser contributions from >O2-containing ions (24.4%). Conversely, the distribution of ions in F3 is dominated by O4 (43.3%) with contributions from O3 and ≥O5 (8.6% and 39.8%, respectively), while O2 ions comprised only 3.7%. Sim- ilarly, when comparing oxygenation of components in DP-2 for F1, F2, and F3, contributions by O2-containing ions were 27%, 5.5%, and 2.4%, respectively. This trend is reversed when comparing >O2-containing ion contributions for F1, F2, and F3 with 69.1%, 85.3%, and 89.1%, respectively. Analysis of double bond equivalent (DBE) data, also provided by ESI- HRMS, indicates varying degrees of unsaturation due to hydrogen defi- ciencies which can be in the form of carbon‑carbon double bonds, and rings, whether they are alicyclic or aromatic. Typically, the DBE of O2 or- ganic acid species are representative of cyclicity, where DBE-1 indicates the number of rings present (DBE=1 is present as the carboxyl group). For example, DBE= 2 indicates a compound with a carboxyl group and one saturated ring (DBE = 3 contains 2 rings, etc.) It is quite likely that hydrogen deficiencies observed here result from somedegree of aroma- ticity, as indicated by the SFS analyses. As such, a simple benzene car- boxylic acid (1-ring with 3 double bonds, 1 carboxyl group) presents a DBEof 5. As the degree of aromaticity observedusing SFS (Fig. 3) is qual- itative data, any correlations to DBEmust be cautiously applied. For sim- plicity, DBE data will be interpreted as degree of cyclicity. In the present 8 analyses, only the DBE of O2 dissolved organic species were examined and are presented as percent abundance relative to the total abundance of O2 species (the total percent DBE equals the percent O2 for class dis- tribution) (Fig. 6). Groundwater sites DP-2 (natural groundwater) and DP-5 (natural groundwater + OSPW) contained the greatest relative abundances of O2 species (Fig. 5). The high values for DBE = 3 and DBE = 4 in the whole water and F1 for these two sites are speculated to be due to sub- stances with functionalities other than aromaticity that translate to hy- drogen deficiencies (i.e., substituents such as hydroxyls, double bonds, etc.), while the next highest values at DBE= 6 and DBE= 7 are specu- lated to be due tomono-aromatic acids (Rowland et al., 2011b), a result supported by the SFS maxima at 272 nm (Fig. 4). These results are con- sistent with previous DBE and SFS analyses of acid extractable organics isolated from fresh OSPW (Bauer et al., 2015) and are also con- sistent with compound distributions between fractions observed from spiking experiments (Bauer et al., 2019). For each source type, the ma- jority of O2 species were present in F1 (Figs. 5,6). The DBE in F1 displayed a bimodal distribution, albeit veryminor at sites DP-1 (natural groundwater) and DP-4 (natural groundwater + OSPW). These F1 data indicate a predominance of 2- and 3-ring organic acids, with lesser con- tributions from 4- to 8-ring acids. As noted above, DBE > 5may also in- dicate the presence of low cyclicity of which one ring may be aromatic. Fraction 2 displayed a low overall abundance of O2 species at all sites (Figs. 5,6), with DP-5 being the only sample exhibiting any contribution above 5%, in which 6.3% of compounds are alicyclic. Similar to F2, F3 ex- hibited very low relative abundance of O2 ions, with no contributions greater that 2% at any site. Nonetheless, sites DP-1, DP-2, and DP-5 ex- hibited a distribution maximum at DBE = 3 and 4, suggesting a pre- dominance of saturated 2- and 3-ring organic acids. 3.5. Fraction profiles by GC–MS Each fraction was profiled for all sites using GC–MS. The relative abundances for each fraction are only comparable between sites, as the fractionation method did not generate equivalent fraction Fig. 6. Double bond equivalents for whole water and dissolved organic fractions of the O2 ion class (classical naphthenic acids) in sites DP-1, DP-2, DP-4, and DP-5 as determined by ESI- HRMS. Graphs present double bond equivalents as a function of hydrogen deficiencies (x-axis) versus percent relative abundance of the total O2 ion class (y-axis). R.A. Frank, A.E. Bauer, J.W. Roy et al. Science of the Total Environment 777 (2021) 146022 concentrations. Comparison of profiles for each fraction across sites showed greater abundances of unresolved organics in F1 and F3, except at site DP-4, which displayed a relatively low abundance of organics in all fractions (Fig. 7). Sites DP-1, DP-2, and DP-5 exhibited a similar broad distribution of organics with only slightly different abundances, Fig. 7. GC–MS ion chromatograms of relative percent abundance vs. retention time for F1 ( 9 as displayed by the peak maxima. For the most polar fraction (F3), DP- 1 exhibited the highest peak maxima, while DP-2 and DP-5 showed lower maxima indicating lower relative abundance of organics (Fig. 7). The more polar organics in F3 displayed a broad distribution in DP-1, while DP-2 and DP-5 had reduced distributions in comparison. blue), F2 (green), and F3 (red) for each groundwater site DP-1, DP-2, DP-4, and DP-5. R.A. Frank, A.E. Bauer, J.W. Roy et al. Science of the Total Environment 777 (2021) 146022 3.6. Fraction profiles by LC-QToF/MS In order to elucidate relative abundances for the polar organic com- ponents present in each fraction, LC-QToF/MSwas utilized. As with GC– MS, the relative abundance of each fraction is only comparable between sites and not between fractions, as the fractionation method did not generate equivalent fraction concentrations thatwould allow such com- parisons. Since reverse phase chromatography was employed, compo- nents which elute earlier are more polar than those retained longer (Fig. 8). For all sites, F3 contained the greatest relative abundance of organics profiled using this method (ESI-), which is expected based on the pH conditions used and the polarities of the compounds within (Fig. 8). The differences in the F3 profiles between sites are indicative of the source differences and consistent with differences noted above for ion class distributions (Fig. 5) and DBE (Fig. 6). In general, when observing the maxima for each fraction, there is a shift to earlier retention times from the least polar (F1) to the most polar (F3) fractions. The peaks at 14 min and 17 min are internal labelled standards (9-anthracene car- boxylic acid and decanoic acid). For the least polar fraction (F1), sites DP-1 (natural groundwater) and DP-5 (natural groundwater + OSPW) displayed the broadest distribution of organics compared to DP-2 (natural groundwater) and DP-4 (natural groundwater + OSPW) (Fig. 8). Site DP-2 displayed a reduced distribution of organics with a similar peak retention time asDP-5. Fraction 2 showed low abun- dance of organics at all sites except DP-1, which displayed a broad dis- tribution of compound polarities. Similarly in F3, DP-1 and DP-5 exhibited broad distributions of compound polarity, while DP-2 and DP-4 displayed a relatively reduced range in polarity. Worthy of note, DP-2 displayed a unique, bimodal distribution for F3 under these ESI- conditions. Similar to GC–MS (EI+) analysis, LC-QToF/MS (ESI-) analysis of DP- 1 and DP-5 exhibited a distribution of organics in F1 and F3 which Fig. 8. LC-QToF/MS ion chromatograms of relative percent abundance vs. retention time fo 10 encompassed a broad range in retention times. DP-2 contained organics which exhibited a more polar distribution of organics at all three frac- tions, and DP-4 displayed a relatively low abundance of organic compo- nents in all fractions, with a minor peak in F3. 4. Discussion This investigation applied a recently developed extraction method (Bauer et al., 2019) to isolate and fractionate the soluble organic com- pounds within groundwater samples previously identified as having significant bitumen influence (Frank et al., 2014; Hewitt et al., 2020). Two samples (DP-1 and DP-2) were selected due to previous determi- nation of their bitumen influence being solely natural, and two samples (DP-4 and DP-5) were selected due to previous determinations of being influenced by both OSPW and with natural bitumen. The first objective of this investigation was to determine if the method developed for aged OSPW could be applied to fractionate, in large volume (~150 L), bitumen-influenced groundwater. The method created 3 distinct fractions with no detectable organic components in the final Stage 2 filtrate by SFS analysis (Fig. 4) and measurements of total A/BEOs (Table 1), thus indicating success at recovering the soluble organics from all investigated bitumen-influenced groundwater sam- ples at this scale. Note that it was not possible to add surrogate spikes to each groundwater sample to quantify total recovery as all generated fractions were to be toxicologically assessed. Consequently, matrix ef- fects (e.g. ion suppression or enhancement) could not be accounted for when making comparisons of LC-QToF/MS concentrations of the acid/base extractable organics between the whole groundwater sam- ples and the isolated fractions (Table 1).While the DP-4 location had previously been identified as having influence from OSPW (Frank et al., 2014; Hewitt et al., 2020), there was little to no signal detected with several of themethodologies previously demonstrated as being di- agnostic of bitumen influence (Figs. 2, 3, 6, and 7; Table 1).We therefore r F1 (blue), F2 (green), F3 (red) for each groundwater site DP-1, DP-2, DP-4, and DP-5. R.A. Frank, A.E. Bauer, J.W. Roy et al. Science of the Total Environment 777 (2021) 146022 conclude that the water collected from DP-4 for this study did not con- tain appreciable amounts of bitumen, and so it likely does not represent OSPW-influenced groundwater. This conclusion is not overly surprising given the large spatial variability in groundwater composition adjacent to this tailings pond (Roy et al., 2016), with few locations showing evidence of OSPW influence, and considering the likely altered groundwater discharge patterns present at the lower river stage for this 2013 sample. Therefore, this sample was removed from further comparisons. The secondobjective of this studywas to determine if fraction chem- ical compositions differed within and between sources. Analyses by GC and LC for fractions derived fromDP-1, DP-2, and DP-5 indicated that F1 contained the greatest abundance of organic compounds, followed by F3 and then F2. This result is somewhat different from analyses of frac- tions isolated from aged OSPW in which F3 had the greatest measured concentrations of dissolved organics followed by F1 (Bauer et al., 2019). Although not yet understood, these observed differences could be due to multiple factors, including the caustic extraction processes used in the mining of bitumen, as well as the ageing, degradation, and sorption processes that can happen within tailings ponds and as groundwater permeates through geological formations. Assessment of chemical speciation in the present study revealed increased oxygena- tion from F1 to F2 to F3, consistent with increases in polarity that were expected given the extraction protocol and consistent with earlier results with aged OSPW (Bauer et al., 2019). However, analyses of F1 in this previous OSPW fractionation, as well as in the current groundwater investigation, revealed that themajority of O2 species,which include or- ganic acids such as naphthenic acids, are collected in F1. Therefore, the resulting distribution of bitumen-derived soluble organic compounds between the three fractions appears to be a function of factors other than pH driven protonation and deprotonation of carboxylic acid moie- ties. These additional factors could include functional groups, water sol- ubility, surfactant properties, and molecular size and structure. Spiking experiments with the fractionation method utilized here also suggest that additional factors, including water solubility, supersede other properties such as molecular weight and pKa in differentiating the polarities of the compounds present within complex bitumen- derived organic mixtures (Bauer et al., 2019). This observation is also consistent with liquid-liquid extractions of fresh OSPW (Morandi et al., 2015). While these studies employed two different methods (SPE, liquid-liquid) and used different sources of bitumen-influenced waters (fresh OSPW, groundwater, aged OSPW), they both resulted in the abundance of dissolved organics being isolated in the initial alkaline extraction, and not in the fraction expected to contain the most polar components. Furthermore, this result of larger, more complex, com- pounds being more polar than smaller, simpler acids is also consistent with HRMS analyses of a previous fractionation of fresh OSPW by distil- lation (Bauer et al., 2015; Frank et al., 2008, 2009). These same function- alities that are impacting compound polarity may very well play a role in their relative bioavailability and toxicological properties. Using advanced separation and high resolution analytical methodol- ogies, previous investigations had identified anOSPW influence at DP-5, and solely natural bitumen influence at the sites of DP-1 and DP-2 (Frank et al., 2014; Hewitt et al., 2020). In many respects, the individual analytical profiles for DP-2 and DP-5 closely resemble each other, in- cluding SFS maxima, DBE plots of O2 species, LC-QToF profiles, and spe- ciation plots. Furthermore, the lack of contribution of inorganic constituents in the OSPW-influenced sample, and the lack of noticeable differences in organic and compound abundance, chemical speciation, aromaticity, and double bond equivalents, demonstrate the chemical similarity of bitumen-influenced groundwaters regardless of the source being natural or industrial. This result reinforces previous determina- tion that a complementary approach using analyses with advanced sep- aration and high mass resolution is necessary for source differentiation between natural bitumen and OSPW influences (Frank et al., 2014; Hewitt et al., 2020; Milestone et al., 2021). 11 Further characterization of the generated fractions from these groundwater sources using advanced separation and high resolution analyticalmethodologiesmay allow for the identification of compounds unique to OSPWand/or natural sources (Milestone et al., 2021). In addi- tion, toxicological assessment of the isolated fractions is needed to iden- tify principal drivers of toxicity in bitumen-influenced groundwaters and may also help to identify sensitive species and endpoints; informa- tion that is supportive of monitoring and remediation research initiatives. CRediT authorship contribution statement Richard A. Frank: Conceptualization, Methodology, Software, Vali- dation, Formal analysis, Investigation, Resources, Data curation,Writing – original draft, Writing – review & editing, Visualization, Supervision, Project administration, Funding acquisition.AnthonyE. Bauer: Concep- tualization,Methodology, Software, Validation, Formal analysis, Investi- gation, Data curation, Writing – original draft, Writing – review & editing, Visualization. James W. Roy: Conceptualization, Methodology, Validation, Formal analysis, Investigation, Data curation, Writing – re- view & editing, Project administration. Greg Bickerton: Conceptualiza- tion, Methodology, Validation, Formal analysis, Investigation, Data curation, Project administration. Martina D. Rudy: Methodology, Vali- dation, Formal analysis, Investigation, Data curation. Ruth Vanderveen: Methodology, Validation, Formal analysis, Investigation, Data curation. Suzanne Batchelor: Methodology, Validation, Formal analysis, Investi- gation, Data curation. Sophie E. Barrett:Methodology, Validation, For- mal analysis, Investigation, Data curation. Craig B. Milestone: Methodology, Validation, Formal analysis, Investigation, Data curation. KerryM. Peru:Methodology, Validation, Formal analysis, Investigation, Data curation. John V. Headley: Methodology, Validation, Formal anal- ysis, Investigation, Data curation. Pamela Brunswick: Methodology, Validation, Formal analysis, Investigation, Data curation. Dayue Shang: Methodology, Validation, Formal analysis, Investigation, Data curation. Andrea J. Farwell: Supervision. D. George Dixon: Supervision, Project administration, Funding acquisition. L. Mark Hewitt: Conceptualiza- tion, Methodology, Software, Validation, Formal analysis, Investigation, Resources, Data curation, Writing – review & editing, Visualization, Su- pervision, Project administration, Funding acquisition. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influ- ence the work reported in this paper. Acknowledgements This work was funded under the Oil SandsMonitoring Program, and is a contribution to the Program, but does not necessarily reflect the po- sition of the Program. Internal resources from Environment and Climate Change Canada were also used to fund this research. Appendix A. Supplementary information Supplementary data to this article can be found online at https://doi. org/10.1016/j.scitotenv.2021.146022. References Bartlett, A.J., Frank, R., Gillis, P.L., Parrott, J., Marentette, J.R., Brown, L.R., Hooey, T., Vanderveen, R., McInnis, R., Brunswick, P., Shang, D., Headley, J.V., Peru, K.M., Hewitt, L.M., 2017. Toxicity of naphthenic acids to invertebrates: extracts from oil sands process-affected water versus commercial mixtures. Environ. Pollut. 227, 271–279. Bauer, A.E., Frank, R.A., Headley, J.V., Peru, K.M., Hewitt, L.M., Dixon, D.G., 2015. Enhanced characterization of oil sands acid-extractable organics fractions using electrospray ionization-high-resolution mass spectrometry and synchronous fluorescence spec- troscopy. Environ. Toxicol. Chem. 34, 1001–1008. https://doi.org/10.1016/j.scitotenv.2021.146022 https://doi.org/10.1016/j.scitotenv.2021.146022 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0005 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0005 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0010 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0010 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0010 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0010 R.A. Frank, A.E. Bauer, J.W. Roy et al. Science of the Total Environment 777 (2021) 146022 Bauer, A.E., Frank, R.A., Headley, J.V., Milestone, C.M., Batchelor, S., Peru, K.M., Rudy, M.D., Barrett, S.E., Vanderveen, R., Dixon, D.G., Hewitt, L.M., 2019. A preparative method for the isolation and fractionation of dissolved organic acids from bitumen-influenced waters. Sci. Total Environ. 671, 587–597. Brown, L.D., Ulrich, A.C., 2015. Oil sands naphthenic acids: a review of properties, mea- surement, and treatment. Chemosphere 127, 276–290. Brunswick, P., Shang, D., van Aggelen, G., Hindle, R., Hewitt, L.M., Frank, R.A., Haberl, M., Kim, M., 2015. Trace analysis of total naphthenic acids in aqueous environmental ma- trices by LC/MS-QToF direct injection. J. Chromatogr. A 1405, 49–71. Brunswick, P., Hewitt, L.M., Frank, R.A., Kim, M., Van Aggelen, G., Shang, D., 2016a. A trace- able reference for direct comparative assessment of total naphthenic acid concentra- tions in commercial and acid extractable organic mixtures derived from oil sands process water. Journal of Environmental Science and Health. Part A - Toxic/Hazardous Substances and Environmental Engineering 52, 1–7. Brunswick, P., Hewitt, L.M., Frank, R.A., Van Aggelen, G., Kim, M., Shang, D., 2016b. Spec- ificity of high resolution analysis of naphthenic acids in aqueous environmental ma- trices. Anal. Methods 8, 6764–6773. Brunswick, P., Shang, D., Frank, R.A., Van Aggelen, G., Kim, M., Hewitt, L.M., 2020. A new concept for the tracking of oil sands process-affected water naphthenic acids and other water-soluble organics in surface waters. Environ. Sci. Technol. 54, 2228–2243. Cardoso, D.N., Soares, A.M.V.M., Wrona, F.J., Loureiro, S., 2020. Assessing the acute and chronic toxicity of exposure to naturally occurring oil sands deposits to aquatic or- ganisms using Daphnia magna. Sci. Total Environ. 729. Droppo, I.G., Liss, S.N., Williams, D., Nelson, T., Jaskot, C., Trapp, B., 2009. Dynamic exis- tence of waterborne pathogens within river sediment compartments. Implications for water quality regulatory affairs. Environ. Sci. Technol. 43, 1737–1743. Fennell, J., Arciszewski, T., 2019. Current knowledge of seepage from oil sands tailings ponds and itsenvironmental influence in northeastern Alberta. Sci. Total Environ. 686, 968–985. Frank, R.A., Kavanagh, R., Burnison, B.K., Arsenault, G., Headley, J.V., Peru, K.M., Solomon, K.R., 2008. Toxicity assessment of generated fractions from an extracted naphthenic acid mixture. Chemosphere 72, 1309–1314. 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., 2009. Effect of carboxylic acid content on the acute toxicity of oil sands naphthenic acids. Environ. Sci. Technol. 43, 266–271. Frank, R.A., Roy, J.W., Bickerton, G., Rowland, S.J., Headley, J.V., Scarlett, A.G., West, C.E., Peru, K.M., Conly, F.M., Hewitt, L.M., 2014. Profiling oil sandsmixtures from industrial developments and natural groundwaters for source identification. Environ. Sci. Technol. 48, 2660–2670. Frank, R.A., Milestone, C.B., Rowland, S.J., Headley, J.V., Kavanagh, R.J., Lengger, S.K., Scarlett, A.G., West, C.E., Peru, K.M., Hewitt, L.M., 2016. Assessing spatial and temporal variability of acid-extractable organics in oil sands process-affected waters. Chemosphere 160, 303–313. FTFC. Volume I: clark hot water extraction fine tailings. In: Advances in Oil Sands Tailings Research. Alberta Department of Energy, Oil sands and research division, Edmonton, AB, Canada: FTFC (Fine Tailings Fundamentals Consortium), 1995a. FTFC. Volume II: fine tails and process water reclamation. In: Advances in Oil Sands Tail- ings Research. Alberta Department of Energy, Oil sands and research division, Ed- monton, AB, Canada: FTFC (Fine Tailings Fundamentals Consortium), 1995b. Headley, J.V., Peru, K.M., Fahlman, B., Colodey, A., McMartin, D.W., 2013. Selective solvent extraction and characterization of the acid extractable fraction of Athabasca oil sands process waters by Orbitrap mass spectrometry. Int. J. Mass Spectrom. 345-347, 104–108. Hewitt, L.M., Roy, J.W., Rowland, S.J., Bickerton, G., De Silva, A., Headley, J.V., Milestone, C.B., Scarlett, A.G., Brown, S., Spencer, C., West, C.E., Peru, K.M., Grapentine, L., Ahad, J.M.E., Pakdel, H., Frank, R.A., 2020. Advances in distinguishing groundwater influ- enced by Oil Sands Process-affected Water (OSPW) from natural bitumen- influenced groundwaters. Environ. Sci. Technol. 54, 1522–1532. 12 Kavanagh, R., Burnison, B.K., Frank, R.A., Solomon, K.R., Van Der Kraak, G., 2009. Detecting oil sands process-affected waters in the Alberta oil sands region using synchronous fluorescence spectroscopy. Chemosphere 76, 120–126. MacKinnon, M.D., Boerger, H., 1986. Description of two treatment methods for detoxify- ing oil sands tailings pond water. Water Pollution Research Journal of Canada 21, 496–512. Mahaffey, A., Dube, M., 2017. Review of the composition and toxicity of oil sands process- affected water. Environ. Rev. 25, 97–114. Marentette, J.R., Frank, R.A., Bartlett, A.J., Gillis, P.L., Hewitt, L.M., Peru, K.M., Headley, J.V., Brunswick, P., Shang, D., Parrott, J.L., 2015. Toxicity of naphthenic acid fraction com- ponents extracted from fresh and aged oil sands process-affected waters, and com- mercial naphthenic acid mixtures, to fathead minnow (Pimephales promelas) embryos. Aquat. Toxicol. 164, 108–117. Milestone, C.B., Sun, C., Martin, J.W., Bickerton, G., Roy, J.W., Frank, R.A., Hewitt, L.M., 2021. Untargeted profiling of bitumen influenced waters for the identification of tracers of oil sands processed water (OSPW) migrations in the Athabasca watershed of Alberta Canada. Rapid Commun. Mass Spectrom. 35, e8984. Morandi, G.D., Wiseman, S.B., Pereira, A., Mankidy, R., Gault, I.G.M., Martin, J.W., Giesy, J.P., 2015. Effects-directed analysis of dissolved organic compounds in oil sands process- affected water. Environ. Sci. Technol. 49, 12395–12404. Pereira, A.S., Bhattacharjee, S., Martin, J.W., 2013. Characterization of oil sands process- affected waters by liquid chromatography orbitrap mass spectrometry. Environ. Sci. Technol. 47, 5504–5513. Ross, M.S., dos Santos, Pereira A., Fennell, J., Davies, M., Johnson, J., Sliva, L., Martin, J.W., 2012. Quantitative and qualitative analysis of naphthenic acids in natural waters sur- rounding the Canadian oil sands industry. Environ. Sci. Technol. 46, 12796–12805. Rowland, S.J., Scarlett, A.G., Jones, D., West, C.E., Frank, R.A., 2011a. Diamonds in the rough: identification of individual naphthenic acids in oil sands process water. Envi- ron. Sci. Technol. 45, 3154–3159. Rowland, S.J., West, C.E., Jones, D., Scarlett, A.G., Frank, R.A., Hewitt, L.M., 2011b. Steroidal aromatic ‘naphthenic acids’ in oil sands process-affected water: structural compari- sons with environmental estrogens. Environ. Sci. Technol. 45, 9806–9815. Rowland, S.J., West, C.E., Scarlett, A.G., Jones, D., Frank, R.A., 2011c. Identification of indi- vidual tetra- and pentacyclic naphthenic acids in oil sands process water by compre- hensive two-dimensional gas chromatography-mass spectrometry. Rapid Commun. Mass Spectrom. 25, 1198–1204. Rowland, S.J., West, C.E., Scarlett, A.G., Ho, C., Jones, D., 2012. Differentiation of two indus- trial oil sands process-affectedwaters by two-dimensional gas chromatography/mass spectrometry of diamondoid acid profiles. Rapid Commun. Mass Spectrom. 26, 572–576. Roy, J.W., Bickerton, G., 2010. A proactive screening approach for detecting groundwater contaminants along urban streams at the reach-scale. Environ. Sci. Technol. 44, 6088–6094. Roy, J.W., Bickerton, G., Frank, R.A., Grapentine, L., Hewitt, L.M., 2016. Assessing risks of shallow riparian groundwater quality near an oil sands tailings pond. Groundwater 54, 545–558. Sun, C., Shotyk, W., Cuss, C.W., Donner, M.W., Fennell, J., Javed, M., Noernberg, T., Poesch, M., Pelletier, R., Sinnatamby, N., Siddique, T., Martin, J.W., 2017. Characterization of naphthenic acids and other dissolved organics in natural water from the Athabasca oil sands region, Canada. Environ. Sci. Technol. 51, 9524–9532. Wilde, M.J., West, C.E., Scarlett, A.G., Jones, D., Frank, R.A., Hewitt, L.M., Rowland, S.J., 2015. Bicyclic naphthenic acids in oil sands process water: Identification by comprehensive multidimensional gas chromatography-mass spectrometry. J. Chromatogr. A 1378, 74–87. http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0015 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0015 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0015 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0020 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0020 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0025 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0025 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0030 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0030 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0030 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0030 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0030 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0035 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0035 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0035 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0040 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0040 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0040 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0045 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0045 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0045 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0055 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0055 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0055 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0060 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0060 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0060 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0065 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0065 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0070 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0070 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0075 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0075 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0075 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0080 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0080 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0080 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0085 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0085 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0085 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0085 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0090 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0090 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0090 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0095 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0095 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0095 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0100 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0100 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0100 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0105 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0105 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0110 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0110 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0110 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0110 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0115 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0115 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0115 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0120 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0120 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0125 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0125 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0125 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0130 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0130 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0135 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0135 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0135 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0140 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0140 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0140 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0145 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0145 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0145 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0145 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0150 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0150 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0150 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0150 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0155 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0155 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0155 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0160 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0160 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0160 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0165 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0165 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf0165 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf9550 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf9550 http://refhub.elsevier.com/S0048-9697(21)01089-5/rf9550 Preparative isolation, fractionation and chemical characterization of dissolved organics from natural and industrially deri... 1. Introduction 2. Methods and materials 2.1. Sample collection 2.2. Centrifugation of groundwater samples prior to extraction 2.3. Isolation and fractionation of dissolved organics 2.4. Synchronous fluorescence spectroscopy (SFS) 2.5. Electrospray ionization high resolution mass spectrometry (ESI-HRMS) 2.6. Liquid chromatography quadrupole time-of-flight mass spectrometry (LC-QToF/MS) quantification 2.7. LC-QToF/MS profiling 2.8. Gas chromatography tandem mass spectrometry (GC–MS) 2.9. Inductively coupled plasma-sector field mass spectrometry (ICP-MS) analysis of metals and major ions 3. Results 3.1. Indicators of bitumen source in groundwater 3.2. Acid/base extractable organics (LC-QToF/MS) 3.3. Aromaticity of fractions (SFS) 3.4. Organic ion class distributions (ESI-HRMS) 3.5. Fraction profiles by GC–MS 3.6. Fraction profiles by LC-QToF/MS 4. Discussion CRediT authorship contribution statement Declaration of competing interest Acknowledgements Appendix A. Supplementary information References