Deposition and Source Identification of Nitrogen Heterocyclic

Subscriber access provided by Macquarie University

Characterization of Natural and Affected Environments

Deposition and source identification of nitrogen heterocyclic polycyclic aromatic compounds in snow, sediment, and air samples from the Athabasca Oil Sands Region Leah Chibwe, Carlos A Manzano, Derek Muir, Beau Atkinson, Jane L. Kirk, Christopher H. Marvin, Xiaowa Wang, Camilla Teixeira, Dayue Shang, Tom Harner, and Amila O. De Silva Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.8b06175 • Publication Date (Web): 11 Feb 2019 Downloaded from http://pubs.acs.org on February 17, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 29

Environmental Science & Technology

1

Deposition and source identification of nitrogen heterocyclic polycyclic aromatic

2

compounds in snow, sediment, and air samples from the Athabasca Oil Sands Region

3 4

Leah Chibwe,*1 Carlos A. Manzano,2,3 Derek Muir,1 Beau Atkinson,1 Jane L. Kirk,1 Christopher

5

H. Marvin,1 Xiaowa Wang,1 Camilla Teixeira,1 Dayue Shang,4 Tom Harner,5 Amila O. De Silva1

6 7

1Aquatic

8

ON

9

2Center

for Environmental Science, Faculty of Science, University of Chile, Santiago, Chile

10

3School

of Public Health, San Diego State University, San Diego, CA

11

4Pacific

and Yukon Laboratory for Environmental Testing, Environment & Climate Change

12

Canada, North Vancouver, BC

13 14

5Air

Contaminants Research Division, Environment & Climate Change Canada, Burlington,

Quality Processes Research Division, Environment & Climate Change Canada, Toronto,

ON

15 16

Corresponding author:

17

*Environment and Climate Change Canada, 867 Lakeshore Rd, Room L650A, Burlington ON

18

L7S 1A1, Phone: 905-336-4541

19

Email: [email protected]

20

ABSTRACT

21

Polycyclic aromatic compounds (PACs) can have multiple sources in the Athabasca Oil

22

Sands Region (AOSR). The current study was designed to identify and explore the potential of 1 ACS Paragon Plus Environment


23

nitrogen heterocyclic PACs (NPACs) as source indicators in snowpack, lake sediment and

24

passive air samples from the AOSR during 2014-2015. Source samples including petroleum coke

25

(petcoke), haul road dust and unprocessed oil sands were also analyzed. Samples were analyzed

26

using comprehensive two-dimensional gas chromatography time-of-flight mass spectrometry and

27

liquid chromatography-high resolution Orbitrap mass spectrometry. Over 200 NPACs were

28

identified and classified into at least 24 isomer groups, including alkylated carbazoles,

29

benzocarbazoles and indenoquinolines. Levels of NPACs in environmental samples decreased

30

with distance from the main developments, and with increasing depth in lake sediments, but were

31

detected within 50 km from the major developments. The composition profiles of several NPAC

32

isomer classes, such as dimethylcarbazoles, showed that petcoke had a distinct distribution of

33

NPACs compared to the haul road dust and unprocessed oil sands ores, and was the most similar

34

source material to near-field environmental samples. These results suggest that petcoke is a

35

major contributing source for the identified NPACs, and that these compounds have the potential

36

to be used as source indicators for future research in the AOSR.

37 38

1. INTRODUCTION

39

The Athabasca Oil Sands Region (AOSR) deposits located in Northern Alberta, Canada, are

40

considered one of the world’s largest oil reserves. The oil sands contain a mixture of naturally

41

occurring sands, water and approximately 4-6 % bitumen.1, 2 Bitumen is separated from this

42

mixture using steam and undergoes upgrading to produce light crude oils.3 Petcoke, a

43

carbonaceous by-product of the upgrading process, is often stored in stockpiles within the

44

AOSR.4 According to a report by the Canadian Association of Petroleum Producers, 2.4 million

45

barrels per day of these oil reserves were recovered in 2016 with a projected increase to 4.0

2 ACS Paragon Plus Environment

Page 2 of 29

Page 3 of 29


46

million barrels per day by 2025.3 This presents on-going environmental concerns such as the

47

release of airborne pollutants, including polycyclic aromatic compounds (PACs), associated with

48

the mining and upgrading processes in the AOSR.5, 6

49

PACs are persistent organic contaminants, composed of fused-benzene rings, and many

50

are linked to carcinogenic and/or mutagenic effects.7-9 PACs have been detected in sediments of

51

lakes,6, 10 air particulate matter,11-14 moss,1 lichen,15, 16 and snow,1, 12, 13 both near and far from

52

upgrading facilities within the AOSR. Several sources of PACs have been suggested, including

53

bitumen upgrader emissions, tailing sands, fugitive dust, biomass burning, oil sands ore, and

54

petroleum coke (petcoke).1, 2, 11, 13, 17-20 However, since PACs can have both natural and

55

anthropogenic sources, more knowledge is needed to fully understand and characterize the major

56

industrial sources and environmental burden of PACs in the AOSR.

57

The focus in understanding PAC distributions and sources in the AOSR has mainly been on

58

unsubstituted (unPACs), selected groups of alkylated PACs (aPACs) and dibenzothiophenes

59

(DBTs), even though it is well-known that PACs exist in the environment as complex mixtures.

60

Heterocyclic PACs, containing an oxygen, nitrogen or sulphur atom within a ring, comprise one

61

of the least studied groups of PACs.21 There are few authentic standards available, and the high

62

number of potential isomers present challenges in the separation, identification and quantitation

63

of these more polar PAC analogue compounds in environmental samples. Several studies suggest

64

that these heterocyclic compounds can be just as prevalent as the most routinely monitored PACs

65

in the environment.13, 22-28 N-containing heterocyclic PACs (NPACs) have been detected in air

66

particulate and sediment samples,23 and in soils associated with coal production.28 Specific to the

67

AOSR, Manzano et al. tentatively identified over 250 chromatographic features, mainly

68

associated to S-containing heterocyclic PACs (SPACs) in snowpack, air particulate matter and



69

sediment samples.13 Although there is also paucity in toxicity data for these heterocyclic PACs,

70

several have been shown to be toxic in embryonic zebrafish, marine and land invertebrates, and

71

in plant assays.29-33 Effect threshold concentrations in these assays have been observed to be

72

comparable, if not lower, than the corresponding unsubstituted PAC homologs.29-33 Neglecting to

73

address the contribution of these compounds may greatly underrepresent the overall impact of

74

PACs in monitoring assessments in the AOSR.

75

The objectives of this study were to for the first time (1) identify NPACs in snow, passive air

76

and sediment samples from the AOSR, (2) to characterize the distribution of these compounds in

77

6 source materials (a road dust, oil sands ores, and petcoke samples), and (3) use the profiles of

78

the source materials for source apportionment, using comprehensive two-dimensional gas

79

chromatography mass spectrometry (GC×GC/ToF-MS) and liquid chromatography-high mass

80

resolution (Orbitrap) spectrometry (LC-HRMS). We hypothesized that NPACs are enriched

81

within AOSR source materials, thus providing unique source signatures.

82 83 84

2. METHODS 2.1. Chemicals and Materials. The following authentic standards for NPACs: carbazole,

85

acridine, phenanthridine, 8,10-dimethylbenzo[a]acridine, benz[a]acridine, 3,6-

86

dimethylcarbazole, benzocarbazole, 7,8-benzo[h]quinolone, 2,4-dimethylbenzo[h]quinolone,

87

benz[c]acridine, 1,5-dimethylcarbazole, 4H-benzo[def]carbazole, and dibenzo[a,j]acridine, were

88

purchased from Cambridge Isotope Labs (Andover, MA) and Sigma Aldrich (Oakville, CA).

89

2.2. Snow, sediment, air, and source materials. Details on the collection, locations and

90

procedures for the environmental samples have been extensively described previously,12, 13 and


Page 4 of 29

Page 5 of 29


91

are provided in further detail in the Supporting Information. Briefly, the sampling information is

92

summarized as follows:

93

Snow. The snowpack samples used in this study were collected as part of the Joint Canada-

94

Alberta Oil Sands Monitoring (JOSM) program in the Regional Municipality of Wood Buffalo,

95

northern Alberta, at varying distances from a widely studied site (AR6) located near oil sands

96

upgrading operations (Figure S1).12, 34, 35 Under JOSM, the unPAC and aPAC concentrations

97

from 2011-2014 have also been reported.12 In this study, 21 samples from 2014 collected

98

between late February and early March, up to 80 km from AR6 were selected for NPAC

99

analysis.

100

Briefly, melted snow was first filtered through a glass fibre filter (GF/F), then passed through

101

a column packed with a XAD-2 resin. Filters were spiked with labeled standards, and the

102

particulate matter on filters were then extracted using pressurized liquid extraction (PLE)

103

(Dionex ASE 350, Thermo Fisher Scientific) with hexane-acetone (1:1), followed by

104

dichloromethane (DCM). The XAD-2 resin was extracted on an elution column with acetone and

105

DCM. The PLE and XAD extracts were both back extracted with 3% sodium chloride solution to

106

remove acetone before drying with sodium sulphate. They were then concentrated, solvent

107

exchanged into cyclohexane, and combined for the total NPACs from particle and the dissolved

108

fractions of the snow samples. The combined extract was then fractionated using silica gel

109

columns that were eluted with hexane followed by benzene to elute the aromatic fraction of

110

interest.

111

Passive air samples. The passive air samples were collected in 2015, also as part of the

112

JOSM monitoring program.13, 14 In brief, polyurethane foam (PUF) disks were deployed at 5

113

sites (AMS 5, 13, 9, 6 and 14) in the AOSR (Figure S1). The PUF disks were spiked with labeled



114

standards and extracted by PLE with two cycles of petroleum ether and acetone (75:25, v/v). The

115

extracts were then purified using silica gel columns (Agilent Technologies) with DCM as the

116

eluent. The extracts were finally concentrated under a steady stream of nitrogen prior to analysis.

117

Sediment cores. A sediment core was collected in 2015 from Lake NE20 located at

118

57.125 oN, 111.56 oW, near upgrading and mining operations (Figure S1). This lake has an area

119

of 0.05 km2 and catchment of 2.1 km2. The sediment core was sampled and dated using isotopes

120

210-Pb and 137-Cs as described by Kurek et al.6

121

Source materials. The source materials investigated in this study were haul road dust,

122

delayed petcoke, and oil sands ores. The representative haul road dust source material was

123

prepared using composite roof and surface dust collected off mining trucks and other vehicles

124

carrying bitumen ore within the mining operations, since dust from haul and transport roads has

125

previously been identified as a major source of air pollution in other mining areas.36, 37

126

Delayed petcoke and oil sands bitumen ore samples were obtained from two companies

127

operating within the AOSR with the cooperation of the Canadian Oil Sands Innovation Alliance.

128

In addition, samples of petcoke collected in the Anzac area (about 30 km south of Fort

129

McMurray, AB), and of delayed and fluidized petcoke, were provided by Yifeng Zhang and

130

Jonathan Martin (University of Alberta/Stockholm University). These samples have been

131

previously characterized for a less extensive suit of unPACs, aPACs and SPACs.1, 13 The

132

fluidized petcoke sample was analyzed and had negligible levels of NPACs, and was excluded

133

from analysis in this study. This result was consistent with previous results for unPACs, aPACs

134

and SPACs in fluidized petcoke.1, 13 The differences in delayed and fluidized petcoke stem from

135

differences in processing methods. While both types are derived from bitumen containing high


Page 6 of 29

Page 7 of 29


136

sulfur content, delayed petcoke is produced at lower temperatures often resulting in higher levels

137

of volatiles than fluidized petcoke.38

138

Sediments and source materials were extracted as described by Shang et al.39 In brief,

139

about 10 g freeze dried subsample was weighed into a polypropylene centrifuge tube. Mass

140

labeled standards of unPACs, about 5 g sodium sulphate, 1 g Florisil, and 15 mL of DCM were

141

also added to the tube. The tube was agitated on a shaker for 10 min, centrifuged, and the solid

142

sample cake was re-extracted with 7 mL of fresh DCM. The combined supernatant was made up

143

to a volume of 20 mL with DCM. Of this extract, 1 mL was passed through a silica gel column,

144

and eluted with 10 mL of DCM. The extract was then concentrated to 1 mL for analysis.

145

2.3. GC×GC-TOF-MS Analysis. All extracts were analyzed using a GC×GC/ToF-MS

146

Pegasus 4D (LECO, St Joseph, MI) instrument. The instrument consisted of an Agilent 7890B

147

gas chromatograph (Palo Alto, CA) with a consumable-free modulator operated at -80 °C. The

148

columns used were the ionic-liquid SLB®-IL60 (20 m × 0.18 mm × 0.14 μm; Sigma Aldrich) and

149

Rxi-17 column (1.2 m × 0.10 mm × 0.10 μm; Restek) in the first and second dimension,

150

respectively. Details on the oven temperature methods are provided in the Supporting

151

Information (Table S2).40 Data processing was performed using the LECO ChromaToF v.4.50.8

152

software.

153

The tentative identification of unknown NPACs features followed a similar approach

154

used previously in the case of SPACs in AOSR samples, using the NIST 11 EI mass spectral

155

library and order in elution profiles of peaks in the 2D chromatogram.13 Isomers were

156

additionally identified using retention indexes (RI) in the first dimension, determined from the

157

compounds: fluorene (RI = 200), phenanthrene (RI = 300), chrysene (RI = 400), benzo[a]pyrene

158

(RI = 450) and benzo[ghi]perylene (RI = 500).



159

The GC×GC/ToF-MS was calibrated using 13 NPAC standards ranging from 0.5 to 2000

160

pg µL-1 (R2 > 0.99). The average response factors (ARF) of the authentic standards were

161

determined using a 10-point calibration curve, relative to the internal standard, chrysene-d12,

162

which was common to all samples. The %RSD for the ARFs for the authentic standards was

163

between 8.1 and 20.6 % (Table S3). The ARFs represent the ratio between the concentrations of

164

a compound, and the response of the detector to that compound, over all concentration levels

165

within a calibration.41 The ARFs for the authentic standards were between 0.38 and 2.61, lowest

166

for the high MW and later eluting compounds (Table S3). Because a large proportion of the

167

identified NPAC features lack authentic standards, it is challenging to determine a response

168

factor for each tentatively identified feature. Therefore, a composite response factor determined

169

from the average ARFs of the 13 NPAC standards was used to estimate concentrations for the

170

243 identified NPAC features.

171

Field, method and solvent (toluene) blanks, as well as calibration check standards for

172

every set of 8 samples were used for quality assurance and control, and also to test for carry over

173

during analysis.

174

2.4. LC-HRMS analysis. The LC-HRMS analysis was performed on a Vanquish

175

UHPLC coupled to a Q Exactive Focus Orbitrap MS (Thermo Fisher Scientific).

176

Chromatographic separation was performed using a Poroshell 120, EC-C18 column (3.0 x 100

177

mm x 2.7 µm particle size, Agilent Technologies). The mobile phase consisted of a gradient

178

elution using water (A) and methanol (B). The heated electrospray ionization (HESI-II) source

179

was operated in both positive and negative ionization modes, and source conditions were

180

optimized using the NPAC authentic standards and were: sheath gas flow rate, 20 arbitrary units

181

(a.u.); auxiliary gas, 5 a.u.; auxiliary temperature, 300 °C; and capillary temperature, 300 °C.


Page 8 of 29

Page 9 of 29


182

The ESI-QTOF-MS was operated in full scan data-dependent MS2 (MS/dd-ms2) discovery mode

183

and stepwise dd-ms2 fragmentation was conducted at collision energies: 20, 30 and 50 eV. Mass

184

calibration was performed using Pierce® LTQ Velos ESI positive and negative ion calibration

185

solutions (Thermo Fisher Scientific).

186

For the identification of unknowns, the XCalibur and Compound Discoverer 2.0 (Thermo

187

Fisher Scientific) software were used. Filters were applied to the Compound Discoverer to

188

process the data, including signal to noise (S/N) ratio greater than 5, mass tolerance of 5 ppm,

189

and a minimum ratio between samples and blanks of 5.

190

2.5 Data analysis. Principal component analysis (PCA) was used as an exploratory tool to

191

classify the environmental and source samples based on their NPACs profiles. PCA was

192

performed using the composition fraction (concentration of compound relative to the total

193

NPACs). The data was log transformed and standardized prior to PCA analysis. A one-way

194

analysis of variance (ANOVA) on Ranks non-parametric test with Tukey’s honestly significant

195

differences (HSD) post hoc test was used to determine significant differences in estimated NPAC

196

concentrations between the different source materials. All analyses were performed using

197

SigmaPlot version 14 (Systat Software Inc).

198 199 200

3. RESULTS AND DISCUSSION 3.1 Identification and characterization of NPACs in Environmental Samples. The snow

201

sample, closest to the main upgrading areas in the AOSR (site AR6) was used as reference to

202

identify NPACs.5, 12, 13 The column combination with an ionic-liquid stationary phase in

203

GC×GC/ToF-MS was effective in separating the NPACs from other components in the samples,

204

including SPACs (Figure S2).42 In total, 243 individual features, characterized into 24 isomer 9 ACS Paragon Plus Environment


205

classes, were identified based on the NIST 11 mass spectral similarity matching and elution

206

profiles in the 2D chromatogram for a given isomer class (Supporting Information S1, Figure

207

S3). Classes were assigned names if the isomers had a >80 % average mass spectral similarity

208

match, but otherwise were designated as “unknown” (U). RIs in addition to the elution profiles

209

were used to identify isomers across samples and were useful in identifying compounds in

210

samples that did not contain all isomers for a given class. The average RSD for RIs across all

211

samples was 0.1 % (range between 0 and 0.4 %) (Supporting Information S1).

212

The high mass resolution capability of LC-HRMS was used as a supplementary technique

213

to confirm features. Low NIST mass spectral matches in GC×GC/ToF-MS can occur for a

214

number of reasons; the mass spectra for the given compound could be absent in the library or the

215

unknown peak could have low S/N values insufficient to obtain a good library match. For

216

example, in GC×GC/ToF-MS, a group of ordered isomers with molecular weight (MW) 237

217

were matched by the NIST library as features corresponding to phenylthio-isoquinolines,

218

C15H11NS (66% NIST score), and isopropyl acridones, C16H15NO (75% NIST score). However,

219

given these low mass spectral similarity scores, other features corresponding to formulas

220

C14H11N3O, C15H15N3, C15H11NO2 and C16H18N2 were also suggested. In LC-HRMS, peaks

221

corresponding to MW 237, and with ring double bond equivalents (RDBEs) characteristic of

222

PACs, were identified as C17H19N (3.4 ppm, RDBE 9), C15H11NS (-1.11 ppm, RDBE 11), and

223

C16H15NO (2.8 ppm, RDBE 10) (Supporting Information S1). While these LC-HRMS results

224

also suggested the presence of additional PAC-related compounds, the technique allowed for the

225

elimination of the other suggested formulas by the NIST library in GC×GC/ToF-MS, and an

226

improved characterization of probable structural features. Additionally, dd-ms2 experiments,

227

particularly in ESI positive mode, provided further structural information on NPACs. Methylated


Page 10 of 29

Page 11 of 29


228

NPACs were distinguished by the initial loss of m/z 15 (-CH3), followed by a loss of m/z 28/27

229

(-H2CN) (Figure S4). However, the loss of the m/z 28/27 fragment was dependent on the number

230

of methyl-groups, in which case, multiple –CH3 losses were observed. For non-alkylated

231

NPACs, generally, the initial main fragment loss corresponded to the m/z 28 ion (-H2CN). These

232

fragmentation pathways have been previously observed for NPACs.28, 43

233

The Kendrick Mass Defect plot (KMD) from LC-HRMS analysis was used to distinguish

234

the unknown features of the same homologous series in the reference snow sample.44 The

235

abundances were normalized to the total areas of detected peaks in the sample. The LC-HRMS

236

results were in general agreement with GC×GC/ToF-MS in that the most abundant isomer

237

classes in the snow reference sample were isomer classes: c2-carbazoles (c2-Car), c1-carbazoles

238

(c1-Car), benzocarbazoles (BCar), indenoquinolines (InQ), and unknown 231 (U231)

239

(Supporting Information S1, Table S4). These are labelled A-E respectively, in the mass defect

240

plots (Figure 1). Additionally, the presence of other homologous series, separated by -CH2,

241

observed in the KMD plots from the LC-HRMS data suggested other hetero-containing PACs,

242

which could include additional NPACs as well as nitro- and oxy-PACs (Figure 1). It is likely that

243

these compounds were not detected or chromatographically separated with the current

244

GC×GC/ToF-MS conditions, but were able to be detected with the high mass resolution

245

capabilities of the LC-HRMS. Aside from a study by Manzano et al. which identified 259

246

SPACs in environmental samples in the AOSR,13 monitoring efforts in the AOSR are focused on

247

unPACs, aPACs and several DBTs. Combined with the results of this study, there is evidence

248

that at least 500 heterocyclic PACs are present in environmental samples in the AOSR along

249

with the routinely studied PACs.



250

3.2 Snow, Sediment and Passive Air Samples. The ΣNPACs in snowpack samples were

251

between 0.311 and 10,100 pg L-1, generally decreasing with distance from AR6 (Table S5). The

252

concentrations of NPACs were substantially lower than ΣunPACs and ΣaPACs concentrations,

253

which have been previously determined to be between 28,800 and 36.8x106 pg L-1, and 111,000

254

and 252x106 pg L-1, respectively.12 However, for many of these compounds, in the absence of

255

toxicity data and their interactions in mixtures, further work is needed to discern the

256

environmental relevance of these levels.

257

We evaluated NPACs in Lake NE20 sediments which is 12 km from AR6, 7 and 9 km

258

from a Syncrude coke storage and upgrader stack, respectively (Figure S1, Table S1).6, 45

259

Maximum estimated ΣNPACs were found in the most recent horizons dated post ~2000 (2,910-

260

49,600 pg g-1 dry wt and lowest in the deep sediments, dated pre-1970 (i.e., pre-development)

261

(817-1,910 pg g-1 dry wt) (Figure 2, Table S6). The high levels of NPACs in the most recent

262

horizons confirm results from studies that have shown an increasing trend in PAC concentrations

263

from the mid-20th century to modern times, in sediment cores from lakes near upgrading and

264

mining facilities in the AOSR.6 High molecular weight unPACs and aPACs were shown to

265

dominate the top sections of a sediment core from lake NE20 collected in 2011.3,45 The heavily

266

contaminated sections of the sediment core were enriched with isomer classes: U231, c2-Car,

267

BCar, c1-Car, InQ and carbazole (Car) (Table S7).

268

The estimated ΣNPACs in the passive air samples ranged between 0.01 and 0.27 pg m-3

269

(Table S8). For comparison, Jariyasopit et al. determined the ΣunPAC, ΣaPACs and

270

ΣNitro+Oxy-PACs in the same passive air samples to be between 2,110 and 6,750 pg m-3,

271

11,000 and 51,500 pg m-3, and 230 and 770 pg m-3, respectively (Table S8).14 The PUF passive

272

air samples were enriched in the lower molecular weight isomer classes, particularly carbazole,


Page 12 of 29

Page 13 of 29


273

and isomer classes c1-Car, c2-Car and c3-Car (Table S7). This observation is consistent with

274

results that show that these samplers collect both ambient gas- and particle associated

275

compounds but not larger particles (i.e. > 10 µm) that are subject to near-source deposition from

276

air.14 The relationship between ΣNPAC concentrations and distance from AR6, a widely

277 278

studied site near main developments in the AOSR, was explored. In addition, the geographical

279

coordinates for each snowpack site were used to determine the distance to the nearest source

280

(e.g., coke storage, mine, stack) (Table S1), as multiple oil sands developments spatially exist

281

within the AOSR and likely contribute differently to NPAC emissions.46 For the snowpack

282

samples, an exponential decay function relationship of the form, f = ae (-bx), where f is the natural

283

logarithm transformed concentration and x is the distance (km) from AR6 was assumed.

284

ΣNPACs declined exponentially with increasing distance from AR6, with a decay constant of

285

0.027 (R2 = 0.70, p < 0.0001) (Figure 3A). This decay constant is lower than what has been

286

previously reported for unPACs and aPACs (0.069-0.766 km-1),5, 12, 17 and for SPACs (1.14 km-

287

1),13

288

AR6 as observed for unPACs and aPACs. The exponential relationship with distance from the

289

nearest source similarly showed a decline in NPACs with increasing distance, but a higher decay

290

constant of 0.049 suggesting more rapid concentration gradient concentrated around other point

291

sources beyond AR6 (R2 = 0.68, p < 0.0001) (Figure 3A).

suggesting a more gradual concentration gradient as opposed to highly concentrated around

292

In the passive air samples, ΣNPACs were highest at site AMS 5, which is near, AR6 and

293

at site AMS 9, which is 20 km from AR6 but within 10 km of two mines (Figure 3B, Table S8).

294

The lowest concentrations of ΣNPACs were at residential/recreational sites AMS 6 and AMS 14.

295

This trend has been similarly observed for unPACs, aPACs, DBTs, and SPACs in passive air 13 ACS Paragon Plus Environment


296

samples, with increasing distance from the main developments in the AOSR.11-13 The elevated

297

levels confirm that NPACs, like other PACs, largely associate with sources near bitumen mining

298

and processing sites.

299

3.3. Distinguishing between source materials. The source materials investigated in this

300

study were delayed petcoke, oil sands ore, and road haul dust. The differences in estimated

301

concentrations of NPAC isomers between sources samples were statistically significant

302

(P 0.5) (Figure S5).

304

Generally, the levels of NPACs between petcoke and the oil sands ores were comparable within

305

a company (i.e., petcoke 1 and oil sands ore 1 are from company 1). The estimated

306

concentrations in the road dust source material were comparable to those in the petcoke 2 and oil

307

sands ore 2 samples.

308

To study the distribution of the 24 NPAC isomer classes between the source materials,

309

the composition fraction (ratio of concentration of class to total NPACs) was determined (Figure

310

S6, Table S9). The most abundant isomer classes in the petcoke samples by percent composition

311

were: U231 (16-21%), InQ (5.2-17%), BCar (12-23%), c2-Car (12-23%), c1-Car (9.7-19%) and

312

Car (5.4-19%). Among the oil sands ore samples, the most abundant classes were: U223 (27-

313

28%), c3-Car (19-32%), U237A (8.5-12%), c2-Car (8.2-13%), InQ (6.5-9.7%) and U259 (7.6-

314

9.4%). The most abundant classes in the road dust sample were: c3-Car (31%), U223 (20%), and

315

c2-Car (18%). An important distinction between the source materials was that carbazole

316

(confirmed with an authentic standard) represented 5.4 – 19.3 % of total NPACs in the petcokes,

317

whereas in the oil sands ores and haul road dust it had negligible levels between 0 – 0.63%

318

composition (Table S9). It is likely that carbazole is one of the volatiles produced during bitumen


Page 14 of 29

Page 15 of 29


319

upgrading. Since petcoke is a by-product of bitumen upgrading, carbazole potentially associates

320

more with it than with the oil sands ores. Although carbazole can have both natural and

321

anthropogenic origins, it is most often linked to the latter when found in the environment.47-52

322

Carbazole has been associated with thermal processes, including as a product from coking, oil

323

refining and the incineration of industrial wastes.47-52 The results of the current study suggest that

324

carbazole has the potential to be used as a chemical marker to distinguish between petcoke and

325

oil sands sources, just as DBTs, phenanthrenes and fluorenes are established chemical markers of

326

oil for bitumen, and coke combustion sources, respectively.5, 53, 54

327

Generally, the distribution profiles of the NPAC isomer classes revealed differences

328

between the source materials, suggesting they could be used to provide signatures required for

329

source apportionment of PACs in environmental samples. The availability and chemical profiling

330

of source materials related to industrial activities in the AOSR is beneficial in elucidating the

331

contributions of contaminants resulting from emissions, especially when multiple natural and

332

anthropogenic sources exist, as is the case for PACs.1, 5, 13, 16, 46

333

3.4. Comparison of Source Materials to Environmental Samples. To compare NPAC

334

distributions between the source materials and environmental samples, the composition fractions

335

of each isomer class were evaluated. Visually, there was similarity in profiles between the

336

petcoke source materials and, especially, the near-field snowpack and sediment environmental

337

samples (Figure S6). To further explore these distributions, we considered the patterns of

338

individual carbazole-related features, which included Car, the alkylated-carbazoles, and the

339

benzocarbazoles (Figure 4).

340 341

The first subset of classes (I), comprising the lower MW carbazole and the c1-Cars, were minimally present in the oil sands samples but enriched in the petcoke, road dust, and



342

environmental samples. This is in contrast to a previous study, that observed an enrichment of

343

lower MW (and more volatile) alkylated PACs in oil sands ore as opposed to petcoke samples.14

344

However, similarly to the aforementioned study,14 the lower MW aPACs like the lower MW

345

NPACs were predominant in the passive air samples, as well as in the snowpack and sediment

346

samples. The predominance of lower MW and usually more volatile PACs, in snow particulate

347

matter has been shown to be typical for compounds stemming from combustion sources.5

348

Furthermore, the profiles of subset class I was similar between the petcoke and the

349

environmental samples, suggesting petcoke as a likely source for these features.

350

The second and third subsets of classes (II, III), comprising isomer classes of c2-Car and

351

c3-Car, showed some distinction between the petcoke (specifically petcokes #1 and #2), oil

352

sands ores and road dust source materials. The isomers in class III were minimally present in

353

petcoke #3, not allowing for comparison with the other source materials. However, the

354

distribution of classes II and III in the snowpack and sediment samples was more similar to the

355

petcoke samples (Figure 4). The last subset of classes (IV), comprising the higher MW

356

compounds, BCar and phenylcarbazoles (PhCar), would be expected to be less volatile than the

357

other carbazoles and primarily exist in the particle phase. These later eluting compounds (IV)

358

were minimally detected in the oil sands ores and road dust source materials, but were more

359

abundant in the petcoke, snowpack and top sediment cores. The distribution profiles of this class

360

were similar between petcoke and the aforementioned environmental samples, suggesting a

361

petcoke influence.

362

The NPAC compositions between the source materials and environmental samples were

363

evaluated by PCA, using features from the most abundant isomer classes in each source material

364

(Car, c1-Car, c2-Car, c3-Car, BCar, U223, U237A and InQ) (133 of the 243 identified features).


Page 16 of 29

Page 17 of 29


365

The first 2 principal components (PC1, PC2) accounted for 48.5% of the total variance (Figure 5,

366

S7), while up to 10 PCs accounted for 80.6% (Table S10). The PCA component plot for PC1 vs

367

PC2 showed that the snowpack samples within 50 km of AR6, the top sections of the sediment

368

cores (dated post ~2000), and the petcoke samples tended to group together, suggesting influence

369

from petcokes (Figure 5A, S7). The snowpack samples further away from AR6 (> 50 km), and

370

the sediment cores at deeper depths (dated pre ~1960), and the passive air samples, including

371

AMS 5 which is near main developments, mostly clustered separately together. This could

372

indicate the influence of other sources, especially at distal snowpack sites, the influence of a

373

mixture of sources, and/or be representative of a background NPAC signal.

374

The distinction between source materials was also evident as the petcoke samples

375

grouped together and the oil sands samples clustered together, suggesting similarities in physical

376

characteristics as well as with interactions with NPACs (Figure 5A). However, the road dust

377

sample was isolated, indicating it was compositionally different from the other source materials.

378

Comparison of PC1 and the third principal component (PC3) also showed isolation of the road

379

dust sample from the other source materials (Figure 5B). Since the road dust source material was

380

representative of fugitive dust collected off moving vehicles and surfaces in the AOSR, it is

381

likely that there is combined influence from vehicle emissions, combustion and other industrial

382

processes. Additionally, PC1 vs PC3 showed a clustering of the petcokes and oil sands ore

383

source materials and near-field snowpack and sediment samples, suggesting contribution from

384

both petcoke and oil sands ores in the near-field samples (Figure 5B). Using chemical profiling

385

of unPACs, aPACs, oxy- and nitro-PACs, and energy dispersive X-ray spectroscopy of source

386

materials and passive air samples, a recent study observed combined contribution from both

387

petcoke and oil sands ore in passive air samples near mining sites.14



388

Attributing the sources of PACs in the AOSR is an on-going challenge for a number of

389

reasons; multiple anthropogenic sources, many usually in close proximity to each other, have to

390

be considered. Moreover, PACs interactions with different materials are influenced by their

391

physicochemical properties, and may be greatly impacted by degradation and transformation

392

processes in the environment. The general consensus from multiple studies is that there is a

393

substantial contribution of PACs as a result of mining and upgrading activities, 1, 2, 10, 13-16, 46, 54, 55

394

and consequently that the accurate understanding of sources is necessary to determine the

395

environmental burden of PACs in the AOSR.

396

Petcoke has been suggested as a source for PACs in environmental sources in the AOSR.

397

Manzano et al. 13 identified 259 SPACs and observed that relative distributions of these

398

compounds were similar between delayed petcoke and snow samples less than 30 km from main

399

development areas, as well as near-field sediment and air samples. Priority unPAC and aPAC

400

distributions combined with chemical mass balance (CMB) modelling were evaluated in several

401

source materials (exposed bitumen sample, oil sands ore, fine tailings, delayed and fluid petcoke

402

samples), moss and peat ore samples from the AOSR by Zhang et. al.1 Results from the study

403

showed similarities in PAC profiles between delayed petcoke and moss samples, confirming

404

petcoke as a major source of PACs at near-field sites.1

405

Implications and Limitations. In the current study, 243 non-routinely analyzed NPACs

406

were identified in environmental samples in the AOSR. Together with the study by Manzano et

407

al.,13 there are over 500 N- and S-PACs present in the AOSR and not monitored. This could

408

underestimate the PAC burden to the surrounding environment from industrial and mining

409

activities associated with the AOSR. Moreover, the chemical distribution of these compounds in

410

several source materials showed that they provided unique signatures, from which we were able


Page 18 of 29

Page 19 of 29


411

to identify petcoke as a major source for these compounds in the AOSR. One of the limitations is

412

that many of the identified heterocyclic PACs lack authentic standards, thus it can be difficult to

413

accurately quantify their contribution, or evaluate their toxicity and bioavailability in the

414

environment. Further work is needed to fully characterize hetero-PACs in the environmental

415

samples from the AOSR. Additionally, LC-HRMS is a promising approach for identifying these

416

classes of compounds in environmental media.

417 418

Supporting Information

419

Sampling and source sites; GC×GC/ToF-MS chromatogram of c3-Carbazoles; First (1tR) and

420

second (2tR) dimension time plots of isomer classes; MS/MS fragmentation patterns for NPACs;

421

Boxplot of log ∑NPAC concentrations between source materials; Log composition fraction of

422

isomer classes; Vector loadings plot NPAC isomers classes in source materials; Table of distance

423

of sites to nearest source in the AOSR; GC×GC/ToF-MS oven temperature program; Correlation

424

co-efficients, response factors and retention indexes of 13 NPAC authentic standards; Percent

425

composition of NPAC isomer classes in snowpack samples; Estimated ΣNPACs concentrations

426

in snowpack; Percent composition of NPAC isomer classes in sediment cores and passive air

427

samples; Estimated ΣNPACs concentrations in passive air samples; Percent composition of

428

NPAC isomer classes in source materials; Principal components of NPACs.

429

Acknowledgements

430

We would like to thank the ECCC Air Quality and Research Division laboratory (Ottawa,

431

ON) for the silica gel cleanup and GC-MS analysis of PACs in the 2014 snow samples, the

432

ECCC Pacific and Yukon Laboratory for Environmental Testing (Vancouver, BC) for the

433

extraction of the source materials and sediment samples, and Fan Yang (ECCC, Burlington ON) 19 ACS Paragon Plus Environment


434

for sediment core dating. We thank Neal Tanna (Canadian Oil Sands Innovation Alliance

435

(COSIA), Calgary AB) and Kelly Munkittrick (formerly with COSIA, now Wilfrid Laurier

436

University, Waterloo ON) and staff of SUNCOR and CNRL, for provision of AOSR mining

437

ores, pet coke and haul road dust, Jonathan Martin (Stockholm University) and Yifeng Zhang

438

(University of Alberta) for provision of the Anzac and fluidized petcoke source samples, and

439

Narumol Jariyasopit (ECCC, Toronto ON) for concentration calculations on the passive air

440

samples.

441 442

References

443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469

1. Zhang, Y.; Shotyk, W.; Zaccone, C.; Noernberg, T.; Pelletier, R.; Bicalho, B.; Froese, D. G.; Davies, L.; Martin, J. W., Airborne petcoke dust is a major source of polycyclic aromatic hydrocarbons in the Athabasca Oil Sands Region. Environ. Sci. Technol. 2016, 50, (4), 1711-1720. 2. Jautzy, J. J.; Ahad, J. M.; Gobeil, C.; Smirnoff, A.; Barst, B. D.; Savard, M. M., Isotopic evidence for oil sands petroleum coke in the Peace–Athabasca Delta. Environ. Sci. Technol. 2015, 49, (20), 1206212070. 3. Canadian Association of Petroleum Producers (CAPP) 2017. https://www.capp.ca/canadian-oiland-natural-gas/oil-sands 4. Oil Change International. Petroleum coke:The coal hidingin the tar sands. http://priceofoil.org/content/uploads/2013/01/OCI.Petcoke.FINALSCREEN.pdf 5. Kelly, E. N.; Short, J. W.; Schindler, D. W.; Hodson, P. V.; Ma, M.; Kwan, A. K.; Fortin, B. L., Oil sands development contributes polycyclic aromatic compounds to the Athabasca River and its tributaries. Proc. Natl. Acad. Sci. 2009, 106, (52), 22346-22351. 6. Kurek, J.; Kirk, J. L.; Muir, D. C.; Wang, X.; Evans, M. S.; Smol, J. P., Legacy of a half century of Athabasca oil sands development recorded by lake ecosystems. Proc. Natl. Acad. Sci. 2013, 110, (5), 1761-1766. 7. Grimmer, G., Environmental carcinogens: polycyclic aromatic hydrocarbons. Crc Press: 2018. 8. Jennings, A. A., Worldwide regulatory guidance values for surface soil exposure to carcinogenic or mutagenic polycyclic aromatic hydrocarbons. J. Environ. Manage. 2012, 110, 82-102. 9. Masiol, M.; Hofer, A.; Squizzato, S.; Piazza, R.; Rampazzo, G.; Pavoni, B., Carcinogenic and mutagenic risk associated to airborne particle-phase polycyclic aromatic hydrocarbons: a source apportionment. Atmos. Environ. 2012, 60, 375-382. 10. Ahad, J. M.; Jautzy, J. J.; Cumming, B. F.; Das, B.; Laird, K. R.; Sanei, H., Sources of polycyclic aromatic hydrocarbons (PAHs) to northwestern Saskatchewan lakes east of the Athabasca oil sands. Org. Geochem. 2015, 80, 35-45. 11. Schuster, J. K.; Harner, T.; Su, K.; Mihele, C.; Eng, A., First results from the oil sands passive air monitoring network for polycyclic aromatic compounds. Environ. Sci. Technol. 2015, 49, (5), 2991-2998.


Page 20 of 29

Page 21 of 29

470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517


12. Manzano, C. A.; Muir, D.; Kirk, J.; Teixeira, C.; Siu, M.; Wang, X.; Charland, J.-P.; Schindler, D.; Kelly, E., Temporal variation in the deposition of polycyclic aromatic compounds in snow in the Athabasca Oil Sands area of Alberta. Environ. Monit. Assess. 2016, 188, (9), 542. 13. Manzano, C. A.; Marvin, C.; Muir, D.; Harner, T.; Martin, J.; Zhang, Y., Heterocyclic aromatics in petroleum coke, snow, lake sediments, and air samples from the Athabasca oil sands region. Environ. Sci. Technol. 2017, 51, (10), 5445-5453. 14. Jariyasopit, N.; Zhang, Y.; Martin, J. W.; Harner, T., Comparison of polycyclic aromatic compounds in air measured by conventional passive air samplers and passive dry deposition samplers and contributions from petcoke and oil sands ore. Atmospheric Chem. Phys. 2018, 18, (12), 9161-9171. 15. Landis, M.; Pancras, J.; Graney, J.; Stevens, R.; Percy, K.; Krupa, S., Receptor modeling of epiphytic lichens to elucidate the sources and spatial distribution of inorganic air pollution in the Athabasca Oil Sands Region. In Developments in Environmental Science, Elsevier: 2012; Vol. 11, pp 427467. 16. Studabaker, W.; Krupa, S.; Jayanty, R.; Raymer, J., Measurement of polynuclear aromatic hydrocarbons (PAHs) in epiphytic lichens for receptor modeling in the Athabasca Oil Sands Region (AOSR): A pilot study. In Developments in Environmental Science, Elsevier: 2012; Vol. 11, pp 391-425. 17. Cho, S.; Sharma, K.; Brassard, B.; Hazewinkel, R., Polycyclic aromatic hydrocarbon deposition in the snowpack of the Athabasca oil sands region of Alberta, Canada. Water, Air, Soil Pollut. 2014, 225, (5), 1910. 18. Galarneau, E., Source specificity and atmospheric processing of airborne PAHs: implications for source apportionment. Atmos. Environ. 2008, 42, (35), 8139-8149. 19. Hsu, Y.-M.; Harner, T.; Li, H.; Fellin, P., PAH measurements in air in the Athabasca oil sands region. Environ. Sci. Technol. 2015, 49, (9), 5584-5592. 20. Parajulee, A.; Wania, F., Evaluating officially reported polycyclic aromatic hydrocarbon emissions in the Athabasca oil sands region with a multimedia fate model. Proc. Natl. Acad. Sci. 2014, 111, (9), 3344-3349. 21. Andersson, J. T.; Achten, C., Time to say goodbye to the 16 EPA PAHs? Toward an up-to-date use of PACs for environmental purposes. Polycyclic Aromat. Compd. 2015, 35, (2-4), 330-354. 22. Achten, C.; Andersson, J. T., Overview of polycyclic aromatic compounds (PAC). Polycyclic Aromat. Compd. 2015, 35, (2-4), 177-186. 23. Delhomme, O.; Millet, M., Azaarenes in atmospheric particulate matter samples of three different urban sites in east of France. Atmos. Environ. 2012, 47, 541-545. 24. Lundstedt, S.; Bandowe, B.; Wilcke, W.; Boll, E.; Christensen, J. H.; Vila, J.; Grifoll, M.; Faure, P.; Biache, C.; Lorgeoux, C., First intercomparison study on the analysis of oxygenated polycyclic aromatic hydrocarbons (oxy-PAHs) and nitrogen heterocyclic polycyclic aromatic compounds (N-PACs) in contaminated soil. TrAC, Trends Anal. Chem. 2014, 57, 83-92. 25. Siemers, A.-K.; Mänz, J. S.; Palm, W.-U.; Ruck, W. K., Development and application of a simultaneous SPE-method for polycyclic aromatic hydrocarbons (PAHs), alkylated PAHs, heterocyclic PAHs (NSO-HET) and phenols in aqueous samples from German Rivers and the North Sea. Chemosphere 2015, 122, 105-114. 26. Siemers, A.-K.; Palm, W.-U.; Faubel, C.; Mänz, J. S.; Steffen, D.; Ruck, W., Sources of nitrogen heterocyclic PAHs (N-HETs) along a riverine course. Sci. Total Environ. 2017, 590, 69-79. 27. Stout, S. A.; Emsbo-Mattingly, S. D.; Douglas, G. S.; Uhler, A. D.; McCarthy, K. J., Beyond 16 priority pollutant PAHs: A review of PACs used in environmental forensic chemistry. Polycyclic Aromat. Compd. 2015, 35, (2-4), 285-315. 28. Tian, Z.; Vila, J.; Wang, H.; Bodnar, W.; Aitken, M. D., Diversity and abundance of high-molecularweight azaarenes in PAH-contaminated environmental samples. Environ. Sci. Technol. 2017, 51, (24), 14047-14054. 21 ACS Paragon Plus Environment


518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565

29. Bleeker, E.; Pieters, B.; Wiegman, S.; Kraak, M., Comparative (photoenhanced) toxicity of homocyclic and heterocyclic PACs. Polycyclic Aromat. Compd. 2002, 22, (3-4), 601-610. 30. Bleeker, E. A.; Wiegman, S.; de Voogt, P.; Kraak, M.; Leslie, H. A.; de Haas, E.; Admiraal, W., Toxicity of azaarenes. Rev. Environ. Contam. Toxicol. 2001, 173, 39-83. 31. Chlebowski, A. C.; Garcia, G. R.; La Du, J. K.; Bisson, W. H.; Truong, L.; Massey Simonich, S. L.; Tanguay, R. L., Mechanistic investigations into the developmental toxicity of nitrated and heterocyclic PAHs. Toxicol. Sci. 2017, 157, (1), 246-259. 32. Pašková, V.; Hilscherová, K.; Feldmannová, M.; Bláha, L., Toxic effects and oxidative stress in higher plants exposed to polycyclic aromatic hydrocarbons and their N-heterocyclic derivatives. Environ. Toxicol. Chem. 2006, 25, (12), 3238-3245. 33. Wassenberg, D. M.; Nerlinger, A. L.; Battle, L. P.; Di Giulio, R. T., Effects of the polycyclic aromatic hydrocarbon heterocycles, carbazole and dibenzothiophene, on in vivo and in vitro cypia activity and polycyclic aromatic hydrocarbon-derived embryonic deformities. Environ. Toxicol. Chem. 2005, 24, (10), 2526-2532. 34. Kirk, J. L.; Muir, D. C.; Gleason, A.; Wang, X.; Lawson, G.; Frank, R. A.; Lehnherr, I.; Wrona, F., Atmospheric deposition of mercury and methylmercury to landscapes and waterbodies of the Athabasca oil sands region. Environ. Sci. Technol. 2014, 48, (13), 7374-7383. 35. Canada-Alberta Oil Sands Environmental Monitoring Program. https://www.canada.ca/en/environment-climate-change/services/oil-sands-monitoring.html 36. Cong, X.; Yang, G.; Qu, J.; Dai, M., Evaluating the dynamical characteristics of particle matter emissions in an open ore yard with industrial operation activities. Environ. Sci. Pollut. Res. Int. 2016, 23, (21), 21336-21349. 37. Mandal, K.; Kumar, A.; Tripathi, N.; Singh, R.; Chaulya, S.; Mishra, P.; Bandyopadhyay, L., Characterization of different road dusts in opencast coal mining areas of India. Environ. Monit. Assess. 2012, 184, (6), 3427-3441. 38. Anthony, E., Fluidized bed combustion of alternative solid fuels; status, successes and problems of the technology. Prog. Energy Combust. Sci. 1995, 21, (3), 239-268. 39. Shang, D.; Kim, M.; Haberl, M., Rapid and sensitive method for the determination of polycyclic aromatic hydrocarbons in soils using pseudo multiple reaction monitoring gas chromatography/tandem mass spectrometry. J. Chromatogr. A 2014, 1334, 118-125. 40. Manzano, C. A.; Muir, D.; Marvin, C., Separation of thia-arenes and aza-arenes from polycyclic aromatics in snowpack samples from the Athabasca oil sands region by GC× GC/ToF-MS. Int. J. Environ. Anal. Chem. 2016, 96, (10), 905-920. 41. Samanipour, S.; Dimitriou-Christidis, P.; Gros, J.; Grange, A.; Arey, J. S., Analyte quantification with comprehensive two-dimensional gas chromatography: Assessment of methods for baseline correction, peak delineation, and matrix effect elimination for real samples. J. Chromatogr. A 2015, 1375, 123-139. 42. Manzano, C.; Hoh, E.; Simonich, S. L. M., Improved separation of complex polycyclic aromatic hydrocarbon mixtures using novel column combinations in GC× GC/ToF-MS. Environ. Sci. Technol. 2012, 46, (14), 7677-7684. 43. Lintelmann, J.; França, M. H.; Hübner, E.; Matuschek, G., A liquid chromatography–atmospheric pressure photoionization tandem mass spectrometric method for the determination of azaarenes in atmospheric particulate matter. J. Chromatogr. A 2010, 1217, (10), 1636-1646. 44. Hughey, C. A.; Hendrickson, C. L.; Rodgers, R. P.; Marshall, A. G.; Qian, K., Kendrick mass defect spectrum: a compact visual analysis for ultrahigh-resolution broadband mass spectra. Anal. Chem. 2001, 73, (19), 4676-4681. 45. Thienpont, J. R.; Desjardins, C. M.; Kimpe, L. E.; Korosi, J. B.; Kokelj, S. V.; Palmer, M. J.; Muir, D. C.; Kirk, J. L.; Smol, J. P.; Blais, J. M., Comparative histories of polycyclic aromatic compound 22 ACS Paragon Plus Environment

Page 22 of 29

Page 23 of 29

566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597


accumulation in lake sediments near petroleum operations in western Canada. Environ. Pollut. 2017, 231, 13-21. 46. Landis, M. S.; Studabaker, W. B.; Pancras, J. P.; Graney, J. R.; Puckett, K.; White, E. M.; Edgerton, E. S., Source apportionment of an epiphytic lichen biomonitor to elucidate the sources and spatial distribution of polycyclic aromatic hydrocarbons in the Athabasca Oil Sands Region, Alberta, Canada. Sci. Total Environ. 2019, 654, 1241-1257. 47. Altarawneh, M.; Dlugogorski, B. Z., Formation and chlorination of carbazole, phenoxazine, and phenazine. Environ. Sci. Technol. 2015, 49, (4), 2215-2221. 48. Glarborg, P.; Jensen, A.; Johnsson, J. E., Fuel nitrogen conversion in solid fuel fired systems. Prog. Energy Combust. Sci. 2003, 29, (2), 89-113. 49. Guo, J.; Li, Z.; Ranasinghe, P.; Bonina, S.; Hosseini, S.; Corcoran, M. B.; Smalley, C.; Rockne, K. J.; Sturchio, N. C.; Giesy, J. P., Spatial and temporal trends of polyhalogenated carbazoles in sediments of upper Great Lakes: Insights into their origin. Environ. Sci. Technol. 2016, 51, (1), 89-97. 50. Ishikawa, S.; Sakazaki, Y.; Eguchi, Y.; Suetomi, R.; Nakamura, E., Identification of chemical substances in industrial wastes and their pyrolytic decomposition products. Chemosphere 2005, 59, (9), 1343-1353. 51. OECD The 2007 OECD List of High Production Volume Chemicals. http://www.oecd.org/officialdocuments/publicdisplaydocumentpdf/?cote=ENV/JM/MONO(2009)40&d oclanguage=en 52. Sumpter, W. C.; Miller, F. M., Heterocyclic compounds with indole and carbazole systems. John Wiley & Sons: 2009; Vol. 16. 53. Duval, M. M.; Friedlander, S. K. Source resolution of polycyclic aromatic hydrocarbons in the Los Angeles atmosphere: application of a chemical species balance method with first order chemical decay. Final report Jan-Dec 80; California Univ., Los Angeles (USA). Dept. of Chemical, Nuclear, and Thermal Engineering: 1981. 54. Jautzy, J.; Ahad, J. M.; Gobeil, C.; Savard, M. M., Century-long source apportionment of PAHs in Athabasca oil sands region lakes using diagnostic ratios and compound-specific carbon isotope signatures. Environ. Sci. Technol. 2013, 47, (12), 6155-6163. 55. Studabaker, W. B.; Puckett, K. J.; Percy, K. E.; Landis, M. S., Determination of polycyclic aromatic hydrocarbons, dibenzothiophene, and alkylated homologs in the lichen Hypogymnia physodes by gas chromatography using single quadrupole mass spectrometry and time-of-flight mass spectrometry. J. Chromatogr. A 2017, 1492, 106-116.



Page 24 of 29

Tables and Figures TOC/Abstract Art

Athabasca Oil Sands

Snow

1

0.1

Oil sand

Petcoke

0.01

0.001

0.0001

NH

N


Page 25 of 29


1

A.

Kendrick Mass Defect (-CH2)

0.95

E

0.9

C A .

D

0.85

B .

0.8 0.75 0.7 0.65 0.6

160

180

200

220

240

m/z

260

280

1

B. Kendrick Mass Defect (-CH2)

0.95

E

0.9

C A

D

B

0.85

0.8

0.75

0.7 160

180

200

220

240

260

280

m/z

Figure 1. Kendrick mass defect plot (KMD) of snowpack sample nearest main upgrading and mining activities in the Athabasca Oil Sands with data from in LC-HRMS analysis in (A) negative and (B) positive mode, showing the NPAC isomer classes (red). The 5 most abundant classes by composition detected by GC×GC/ToF-MS are labelled as follows: A (C17H13N, 19%), B (C18H15N, 15%), C (C14H13N, 13%), D (C16H11N, 12%), E (C13H11N, 11%).



Page 26 of 29

60

Total NPACs, ng/g

50 40 30 20 10 0

1905

1920

1935

1950

1965

1980

1995

2010

2025

Year

Figure 2. ΣNPAC concentrations (ng/g dry wt) in a dated sediment core from Lake NE20 vs year. Horizontal error bars represent 1 SD for the estimated median age of the sediment sample.


Page 27 of 29


0.30

A.

Distance from AR6, R2 = 0.70 Distance to nearest source, R2 = 0.68

0.25

6

0.20

4

0.15

2

0.10

0

0.05

40

60

80

100

Distance, km

60.1 (AMS 14)

20

24.5 (AMS 6)

0.00 0

8 (AMS 13)

-2

6.6 (AMS 9)

Ln (Total NPACs, pg/uL)

8

B.

3.7 (AMS 5)

10

Distance from nearest source, km (site)

Figure 3. ΣNPAC concentrations in (A) snowpack samples as a function of distance from AR6 (red dashed line), and from nearest source (i.e. mine, stack) (black dashed line) based on exponential decay function, f = ae (-bx), and (B) passive air samples vs distance to nearest source.



1

0.1

A. Petcoke 1

I.

G. Roaddust

D. Oil Sands Ore 1

II.

III.

IV. I.

Page 28 of 29

II.

III.

IV. I.

J. Sediment (0-8 cm), n = 3

II.

III.

IV. I.

II.

III.

IV.

0.01

0.001

log (Composition fraction)

0.0001 1

B. Petcoke 2

E. Oil Sands Ore 2

C. Petcoke 3

F. Oil Sands Ore 3

H. Snowpack (0-10 km), n = 8

H. K. Passive air (AMS 5, 6.6 km)

I. Snowpack (50 - 80 km), n = 7

L. Passive air (AMS 13, 16.1 km)

0.1

0.01

0.001

0.0001 1

C.

X Data

0.1

0.01

0.001

0.0001

Figure 4. Log composition fraction (individual concentration divided by total NPACs) of carbazoles in (A-G) delayed petcoke, oil sands mineral ores and road dust source materials, average composition fraction of snowpack sites within (H) 0-10 km and (I) 50-80 km of AR6, (J) average composition fraction of top 3 sections of sediment core and (K, L) passive air samples. The x-axis in each plot represents the individual isomers sub-grouped as I- carbazole and methylcarbazoles, II – dimethylcarbazoles and ethylcarbazoles, III – trimethylcarbazoles, and IV – benzocarbazoles and phenylcarbazoles (increasing molecular weight from left to right). Source materials were provided by the Canadian Oil Sands Innovation Alliance and obtained from a site near an upgrading facility near Anzac, Alberta.


Page 29 of 29


A. 5

10

Snow (AR6)

Sediment (top sections)

B.

5 Passive air (AMS-5)

PC 3 (7.76%)

PC 2 (22.99%)

0

-5

Snow (AR6)

Sediment (top sections)

0

-5 -10 Passive air (AMS-5)

-10

-15

-15

-20 -10

-5

0

5 PC 1 (25.51%)

10

15

-10

-5

0 5 PC 1 (25.51%)

10

Figure 5. (A) PC1 vs PC2 and (B) PC1 vs PC3 score plots for 133 compounds of the most abundant NPAC isomer classes in each source material. (Loadings plot is shown in Figure S7).


15

Deposition and Source Identification of Nitrogen Heterocyclic

Recommend Documents