Predicting Polycyclic Aromatic Hydrocarbon ... - ACS Publications

Subscriber access provided by UNIV OF NEWCASTLE

Article

PREDICTING POLYCYCLIC AROMATIC HYDROCARBON BIOAVAILABILITY TO MAMMALS FROM INCIDENTALLY INGESTED SOILS USING PARTITIONING AND FUGACITY Kyle Jordan James, Rachel E Peters, Mark Cave, Mark Wickstrom, Eric Lamb, and Steven D Siciliano Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.5b05317 • Publication Date (Web): 07 Jan 2016 Downloaded from http://pubs.acs.org on January 7, 2016

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 free 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 accessible to all readers and 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.

Environmental Science & Technology 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 33

1 2 3 4 5 6 7

Environmental Science & Technology

PREDICTING POLYCYCLIC AROMATIC HYDROCARBON BIOAVAILABILITY TO MAMMALS FROM INCIDENTALLY INGESTED SOILS USING PARTITIONING AND FUGACITY Kyle James1,2, Rachel E Peters1,2, Mark R Cave3, Mark Wickstrom4, Eric G Lamb5 and Steven D Siciliano1*. 1

8 9

2

Toxicology Graduate Program, University of Saskatchewan, Saskatoon, Canada 3

10 11 12

Department of Soil Science, University of Saskatchewan, Saskatoon, Canada

4

British Geological Survey, Nottingham, United Kingdom

Veterinary Biomedical Sciences, University of Saskatchewan, Saskatoon, Canada 5

Department of Plant Sciences, University of Saskatchewan, Saskatoon, Canada

13

Title Running Head

14

Predicting PAH bioavailability using fugacity

15

*Corresponding Author Contact Information

16

Steven Siciliano

17

51 Campus Drive, Department of Soil Science, Agriculture Bldg

18

University of Saskatchewan

19

Saskatoon, SK, Canada, S7N 5A8

20

Phone: 306-966-4035. Fax: 306-966-6881

21

Email: [email protected]

1 ACS Paragon Plus Environment


22 23

Abstract Soil and dust ingestion is one of the major human exposure pathways to contaminated

24

soil; however, pollutant transfer from ingested substances to humans cannot currently be

25

confidently predicted. Soil polycyclic aromatic hydrocarbon (PAH) bioavailability is likely

26

dependent upon properties linked to chemical potential and partitioning such as fugacity,

27

fugacity capacity, soil organic carbon and partitioning to simulated intestinal fluids. We

28

estimated the oral PAH bioavailability of 19 historically contaminated soils fed to juvenile

29

swine. Between soils, PAH blood content, with the exception of benzo(a)pyrene, was not linked

30

to fugacity. In contrast, between individual PAHs, using partitioning explained PAH blood

31

content (Area under the Curve = 0.47 log Fugacity + 0.34, r2 = 0.68, p < 0.005, n = 14). Soil

32

fugacity capacity predicts PAH soil concentration with an average slope of 0.30 (µg PAH g-1

33

soil) Pa-1 and r2’s of 0.61-0.73. Because PAH blood content was independent of soil

34

concentration, soil fugacity correlated to PAH bioavailability via soil fugacity’s link to soil

35

concentration. In conclusion, we can use fugacity to explain PAH uptake from a soil into blood.

36

However, something other than partitioning is critical to explain the differences in PAH uptake

37

into blood between soils.

38

Key words: in vivo bioavailability, polycyclic aromatic hydrocarbons, contaminated soil,

39

fugacity


Page 2 of 33

Page 3 of 33

40 41


Introduction How do pollutants move from soil to our blood? Such a question is what drove us to

42

collect 19 contaminated soils from across the world and feed these soils to swine. This question

43

is essential to answer because the transfer of pollutants from soil to blood following ingestion

44

can be the dominant pathway for human exposure at contaminated sites. For example,

45

incidentally ingesting soil exposes humans to 100 times more carcinogens than inhaling air

46

particles.1 Yet, we have still not developed an acceptable model to quantify movement of

47

carcinogens, such as polyaromatic hydrocarbons (PAHs), from an ingested particle to the human

48

bloodstream.

49

In his pivotal 1979 paper titled “Finding Fugacity Feasible”, Professor Don Mackay

50

popularized the use of fugacity to characterize how pollutant movement occurs between

51

environment phases and organisms.2 PAH uptake into human cells (Caco-2)3, 4 and PAH uptake

52

into juvenile swine blood5 can be predicted using fugacity. But using a mouse model, Juhasz et

53

al.6 could not use fugacity to model uptake. Fugacity modelling can be done at both equilibrium

54

and non-equilibrium as well as steady state and non-steady state. Minhas et al.3 and Vasiluk et

55

al.4 calculate fugacity by measuring concentrations over a time course allowing for fugacity

56

calculations at equilibrium and non-equilibrium, whereas Juhasz et al6, like us, did not measure

57

fugacity, instead used published Koc/Kow to calculate fugacity. Minhas et al.3 reports that steady

58

state conditions between soil and aqueous phase (simulated intestinal fluids) is reached after 2

59

hours. Notably, the mean transit time through the stomach (~1.5 hours) and small intestines (~3.5

60

hours) equates to ~5 hours7, suggesting that steady state conditions are reasonable for modelling

61

gastro-intestinal uptake.



62 63

Page 4 of 33

At environmentally relevant concentrations, the concentration (Cphase) of a compound is linearly related to fugacity (f) via the fugacity capacity (Zphase).

64

Cphase = Zphase ×f

65

Mackay2 derived the soil fugacity capacity (Zsoil) of a compound to be:

66

Zsoil = Zwater × soil organic carbon × K oc × soil particle density

67

Where Zwater is calculated as solubility divided by vapour pressure. Assuming a constant

68

soil-organic carbon partitioning coefficient (Koc) between soils allows one to calculate fugacity

69

capacity by determining soil organic carbon content. Routinely, our group and others, would use

70

this assumption because it provided useful results.5, 6, 8, 9 Results from Hawthorne et al.10 suggest

71

that Koc values are not constant between sediments. Across 114 PAH contaminated sediments,

72

Koc values for individual PAHs ranged between 2 and 3 orders of magnitude. Hence, the ability

73

of fugacity models to predict blood uptake may be linked to using a more precise estimate of soil

74

fugacity capacity. To achieve this, one needs to experimentally determine Koc. Koc is the soil-

75

water partitioning coefficient (Kd) normalized to organic carbon content:

(1)

76

K oc = K d ÷ soil organic carbon

77

Substituting equation 3 into equation 2 we obtain equation 4 which eliminates soil

78

(2)

(3)

organic carbon from the fugacity capacity equation.

79

Zsoil = Zwater × K d × soil particle density

80

Therefore, if Zwater and soil particle density remain relatively constant, soil fugacity

81 82

(4)

capacity is primarily dictated by Kd, which can be directly estimated. But our intestines contain intestinal fluid rather than the pure water used to estimate Kd.

83

Intestines contain ingested particles suspended in fluids and these fluids differ in their

84

lipophilicity from pure water. Hence, estimating Ksim, i.e. the partitioning between soil and a


Page 5 of 33


85

synthetic intestinal fluid, should further improve the precision of estimating fugacity for ingested

86

particles11. Improved estimates of fugacity ought to enhance calculations of PAH bioavailability

87

based on partitioning theory.

88

Here, we define bioavailability as the fraction of an administered dose that is present in

89

systemic circulation for a specific time frame12, in our case 48 hours. One can estimate PAH

90

bioavailability by measuring urinary metabolites, PAHs in feces, and PAHs or metabolites in

91

blood (see review by Ramesh et al.13). The use of blood concentrations of parent compounds is

92

criticized because ingested PAHs enter the body via the portal vein, where many PAHs are

93

metabolized by liver enzymes. A series of recent papers14,15 explored how aryl hydrocarbon

94

receptor (AhR) activity, which signals various PAH metabolizing enzymes,16 is linked to

95

toxicity. Hepatic toxicity and detoxification depend on metabolism by cytochrome P450 (CYP)

96

enzymes yet systemic parent PAH compounds cause non-hepatic toxicity. Thus, parent PAHs

97

that reach systemic circulation are what can cause non-hepatic toxicity. This does not suggest

98

that metabolites do not cause toxicity in systemic circulation, only that the majority of systemic

99

toxicity is likely the result of parent compounds metabolizing to reactive species at the site of

100

toxic effect. Thus parent PAHs in the blood provide a meaningful estimate of bioavailability

101

because these PAHs are what pose a risk to the remainder of the body, i.e. not the liver or small

102

intestine which are directly exposed to PAHs and metabolites during ingestion.

103

Mice, rats and juvenile swine are common animal models used for bioavailability

104

research. The mouse model has several advantages over the swine model, namely less dilution

105

into body volume of ingested PAHs (thereby lowering the detection limit of bioavailability

106

studies), and that some cancer slope factors are derived from mouse based studies.17 However,

107

there are certain limits to the mouse model compared to juvenile swine models. For example, the



108

mouse AhR is ~10 times more sensitive than the human AhR to AhR ligands such as TCDD,18

109

whereas the swine AhR is comparable to human AhR for benzo(a)pyrene and TCDD.19 Further,

110

the juvenile swine’s digestive tract is very similar to human toddlers20, 21, typically the critical

111

human receptor of concern. In this study, we wished to characterize how toxicokinetics and

112

digestive physiology influenced PAH uptake from soil. Thus, juvenile swine are a good animal

113

model for this purpose.

114

In the core of the manuscript, we focus on the results of five carcinogenic PAHs.

115

Because of their carcinogenicity, these five PAHs are drivers of the majority of contaminated

116

soil risk assessments. Similar results were obtained with another six PAHs and these results are

117

presented in the supplementary material. Other PAHs were not robustly or routinely detected

118

and thus, not included in our discussion.

119

Materials and Methods

120

Soils

121

A total of 19 soils contaminated with hydrocarbons were obtained from sites in the

122

United Kingdom (n = 12), Canada (n = 5) and Sweden (n = 2), as described previously in James

123

et al.5 and Cave et al.22 Soil pH, soil organic carbon content, and particle size were analyzed as

124

previously described by Siciliano et al.23 The physiochemical properties of the 19 soils are

125

displayed in Table 1. Soil locations are provided in supporting information.

126

PAH Soil Extraction

127

The extraction of PAHs from each soil sample was prepared by weighing approximately

128

2 g of soil into a pre-cleaned 50 mL Teflon centrifuge tube. To the tube a 5 mL of a 6:1 (v/v)

129

mix of methanol: toluene was added and the tube was sealed with Teflon lid. The vial was

130

gently shaken to suspend the contents and then sonicated in an ultrasonic bath for 2 hours at


Page 6 of 33

Page 7 of 33


131

50°C. To separate the supernatant solution, the tubes were centrifuged at ~ 3000 rpm (1000 g)

132

for 15 min and filtered through Whatman GMF 0.45 µm filters into a 15 mL amber glass vial. A

133

0.15 mL aliquot of the solvent extract was diluted with 1.35 mL acetonitrile into labeled 2 mL

134

amber HPLC vials and kept at -20°C until analysis. For the quantification of PAHs from soil a

135

sand matrix spike is analyzed every ten samples and the average recovery ranges from 77% to

136

94% with a standard deviation of 12%. Benzo(b)chrysene, present only in very low

137

concentrations in the environmental samples, was added as an internal standard and the recovery

138

ranged between 90-110% with a standard deviation of 11%.

139

Bioavailability of Ingested PAHs to Swine

140

Female Landrace cross swine (8 week-old; ~20kg) were obtained and housed at the

141

Prairie Swine Centre near Saskatoon, Saskatchewan. The animals were housed in individual

142

pens (1.5 x 1.0 x 1.0 m) to prevent cross contamination and for ease of sampling and exposure.

143

The diet consists of standard grower ration, two meals a day, with total daily intake rate limited

144

to 4% body weight. Drinking water is provided ad libitum via self-activated watering nozzles

145

present in each pen. Swine were allowed to acclimate for one week prior to dosing and were

146

trained daily to eat a dough ball treat consisting of molasses, flour, oats and pig feed as described

147

by Casteel et al.24

148

Swine were randomly allocated into treatment groups (n = 6) each week for five weeks.

149

At the start of each week the pigs were weighed and fed accordingly. Treatment groups

150

consisted of 1 of 19 soils individually mixed into dough balls, 40 µg of each 16 priority PAHs

151

intra venous (IV) dose in a glycerol tri-octanoate vehicle (each week), and control group

152

receiving no PAH dose. In subsequent weeks, groups were crossed over such that each group

153

was assessed as a control and IV dose at some point during the experiment. Controls were used



154

to quantify background levels and IV dose groups were used to ensure consistency between

155

treatment groups over progressive weeks.

156

Page 8 of 33

Blood samples were collected from the jugular vein at 2, 4, 6, and 8 hours post exposure

157

where n = 4 for each treatment group, while at 0, 12, 24, and 48 hours post exposure, n = 3 for

158

each treatment group. Blood samples were collected from each animal a maximum of five times

159

per week to preserve the integrity of the veins for subsequent weeks. Approximately 10-15 mL

160

of blood was collected using heparinized vacutainer tubes, which prevents the blood from

161

coagulating. Plasma was separated from the blood within one hour after sampling by

162

centrifuging at ~ 3000 rpm (1000 g) for 15 minutes and was stored after decanting at -20˚C until

163

solid phase extraction. PAH tissue concentrations have also been collected and are reported in

164

Peters et al.25

165

PAH bioavailability was calculated according to the following equation

166

Soil Dose 48 H × 100% Bioavailability % = PAH administered from soil μg

167

AUCSoil Dose 48 H is determined using area under the plasma concentration curve (AUC)

168

analysis over a 48 hour period and PAH administered is the amount of PAHs in contaminated

169

soil (µg g-1). Area under the curve calculations used the plasma concentration time course for

170

each compound in individual pigs. AUC was calculated to the 48 hour time point with the MESS

171

package26 in statistical program R27 using the trapezoidal rule and thus, AUC units convert from

172

µg g-1 48h-1 to µg g-1.

173

PAH Plasma Extraction

174

AUC

μg

(5)

The extraction of PAHs from plasma follow the procedures of James et al.5 with minor

175

modifications. To remove impurities from plasma, all samples were extracted using Waters

176

Oasis HLB solid phase extraction columns (waters Corporation, Milford MA). Approximately 5 8 ACS Paragon Plus Environment

Page 9 of 33


177

mL of plasma is diluted with 0.1 M nitric acid (1:2 plasma: nitric acid) and sonicated for 15 min

178

in an ultra sonic bath. The columns are conditioned with 2 mL of Milli-Q water and then with 2

179

mL methanol. Approximately 5 mL of acidified plasma is run through the column, rinsed with 2

180

mL of methanol and is eluted with 9 mL of acetonitrile. The eluate was evaporated to near

181

dryness, re-suspended in 2 mL of acetonitrile, and filtered through Whatman GMF 0.45 µm

182

filters into labeled 2 mL amber HPLC vials and kept at -20°C until analysis. Plasma samples

183

were analyzed in duplicate and the value reported is the average of duplicate samples. In

184

addition, for every 10 samples, we assessed a blank and a matrix spike. For the quantification of

185

PAHs in plasma extracted by SPE columns, a matrix spike is analyzed every ten samples using

186

control plasma and the average recovery ranges from 60% to 75% for individual PAHs with the

187

highest standard deviation of 6.3% for a PAH. The average recovery of PAHs recovered in

188

plasma separated from whole blood ranged from 43% to 59% compared to PAHs recovered from

189

whole blood. Plasma values have been corrected to reflect recovery from whole blood and

190

matrix spike recovery from plasma.

191

Simulated Fluids

192

The simulated fluids used here follows the composition of the Fed ORganic Estimation

193

human Simulation Test (FOREhST) described by Cave et al.22 To summarize, the simulated

194

fluids of the FOREhST model mimic the physiological conditions of the human gastro-intestinal

195

tract during the fed state. There are three stages which include saliva, gastric and intestinal

196

(duodenal and bile) phases where the temperature is held constant at 37°C.

197

Soil-Water and Soil-Simulated Fluid Partitioning Coefficient (Kd and Ksim)

198 199

The Kd was determined by weighing approximately 3 g of soil into pre-cleaned 50 mL Teflon centrifuge tubes. To the tube, 8 mL of de-ionized water is added and the tubes are sealed



Page 10 of 33

200

and placed into a shaker at ~ 30 rpm (0.1 g) for a period of fourteen days. To separate the

201

supernatant solution, tubes are centrifuged at ~3000 rpm (1000 g) for 15 minutes and the

202

overlaying water is decanted and filtered through Whatman GMF 0.45 µm filters into 15 mL

203

amber vials. Approximately 1 mL of toluene is added to the water and the mixture is evaporated

204

to near dryness and re-suspended in 2 mL of acetonitrile and transferred into labeled 2 mL amber

205

HPLC vials. The leftover soil pellet is air-dried overnight and the PAHs remaining in soil are

206

extracted using a 6:1 mix of methanol: toluene in an ultrasonic bath for 2 hours at 50°C. To

207

separate the supernatant solution, the tubes are centrifuged at ~ 2000 rpm (600 g) for 15 min and

208

are filtered through Whatman GMF 0.45 µm filters into a 15 mL amber glass vial. A 0.15 mL

209

aliquot of the methanol:toluene mixture is diluted with 1.35 mL acetonitrile into labeled 2 mL

210

amber HPLC vials and kept at -20°C until analysis. The Ksim was determined similarly to the Kd,

211

except only 0.3 g of soil was used with 30 mL of simulated fluids to mimic the procedures of

212

Cave et al.22

213

The Kd and Ksim was calculated by dividing the PAH concentration remaining in the soil

214

by the concentration of PAHs in the water and simulated fluids, respectively.

215

Fugacity Analysis The soil fugacity capacity (Zsoil) was calculated as described by Mackay.28 Solubility and

216 217

vapour pressure (i.e. Zwater) for each PAH are the median values obtained from Mackay.29

218

Fugacity capacity was calculated using equation 4 and measured Kd values. Soil particle density

219

was assumed to be 2.65. Zsoil = Zwater × &' × soil particle density

220 221

HPLC


(4)

Page 11 of 33

222


Prepared PAH samples were analyzed using an Agilent 1260 Infinity High Pressure

223

Liquid Chromatography coupled with Fluorescence Detection (HPLC-FD).30 A 10 µL aliquot

224

was injected onto an Agilent PAH Pursuit column (3 µm particle size, 100mm length, and 4.6

225

mm internal diameter) guarded with an Agilent Pursuit C18 MetaGuard (3 µm particle size, 2mm

226

internal diameter). The column temperature is maintained at 25°C for the duration of the 25 min

227

run. The mobile phase consists of acetonitrile and water with a flow rate of 1.5 mL min-1. At the

228

start of the run the solvent gradient is 60:40 acetonitrile: water, gradually increases to 95:5

229

acetonitrile: water at 20 m, and is held constant until the end of the run. The fluorescence

230

detector employs a constant excitation wavelength of 260 nm and four emission wavelengths of

231

350, 420, 440, and 500 nm. Detection limits for anthraceneis 0.70 pg µL, fluoranthene is 1.71 pg

232

µL, pyrene 0.43 pg µL, benzo(a)anthracene is 2.45 pg µL, chrysene is 5.27 pg µL,

233

benzo(b)fluoranthene is 5.58, benzo(k)fluoranthene is 2.77 pg µL, benzo(a)pyrene is 13.02 pg

234

µL, dibenzo(a,h)anthracene is 7.79 pg µL, benzo(g,h,i)perylene is 1.78 pg µL, and indeno(1,2,3-

235

cd)pyrene is 1.80 pg µL.

236

Statistical Analysis

237

We selected regression analysis for our estimates of the nature of the relationship

238

between blood content, bioavailability and soil fugacity because (i) we wished to imply causality

239

and (ii) the variation in soil concentration, measured soil fugacity, or soil fugacity capacity was

240

much less than that associated with bioavailability or blood content. Where the variability of the

241

x-axis exceeds the variability on the y-axis 95% confidence intervals are provided. Regression

242

analyses were performed in SigmaPlot (version 10.0).

243

Structure equation modeling (SEM) was performed using R software27 and the 'lavaan'31

244

package. SEM is a statistical approach similar to path analysis and is used for testing hypotheses



245

where the relationships between many variables are confounded by intercorrelations.32 Soil

246

organic carbon content is removed from the fugacity capacity equation (see equation 4) however

247

soil organic carbon is known to influence both partitioning33 and soil PAH concentration.34

248

Furthermore, Kd and Ksim are correlated due to similar experimental conditions. Using SEM, the

249

influence of soil organic carbon on partitioning and soil concentration can be accounted for,

250

while simultaneously accounting for the correlation between Kd and Ksim. The initial SEM was

251

set up with soil organic carbon as an endogenous variable. Direct paths from soil organic carbon

252

to soil PAH and simulated fluid partitioning were included because soil organic carbon is known

253

to influence partitioning33 and soil concentration.34 Soil simulated fluid partitioning and soil

254

fugacity capacity are inherently related due to partitioning.11, 33 Soil fugacity was then included

255

as a direct cause of soil PAH concentration based on results generated within. Finally, all four

256

variables, soil organic carbon, soil fugacity capacity, soil simulated fluid partitioning and soil

257

PAH concentration were included as direct causes of PAH exposure. PAH exposure (AUC48

258

plasma content) was calculated with R software with the 'MESS'26 package using the trapezoidal

259

rule.

260

The assumptions of normality were tested using the Shapiro-Wilk test. All data that were

261

not normally distributed were typically log-normally distributed therefore they were log-

262

transformed prior to analysis. Data which were not log-normally distributed were transformed to

263

normality using box-cox transformations.

264

Results

265

Bioavailability

266

The in vivo bioavailability estimates of the 5 PAH compounds analyzed from the 19 soils

267

were largely below 30% with an average bioavailability of 12% for benzo(a)anthracene, 11% for


Page 12 of 33

Page 13 of 33


268

chrysene, 29% for benzo(b)fluoranthene, 27% for benzo(k)fluoranthene, and 21% for

269

benzo(a)pyrene. The bioavailability of PAHs without a benzo(a)pyrene equivalency factor are

270

reported in Supplemental material. Five of the soils, COT1-COT5, had very low PAH soil

271

concentrations, ranging from 0.008 to 0.02 µg/g (Table 1). Because of low soil PAH

272

concentrations in these soils, a non-detect in plasma led to bioavailability of 0% whereas a

273

detection in plasma led to bioavailability of 100%. BGS2 was the only other soil that had a 0%

274

bioavailability for benzo(k)fluoranthene. Trimming these samples from the data set, results in

275

lower average bioavailability of 12% for benzo(a)anthracene, 11% for chrysene, 8.8% for

276

benzo(b)fluoranthene, 6.5% for benzo(k)fluoranthene, 4.4% for benzo(a)pyrene.

277

Partitioning Coefficients (Kd and Ksim)

278

As a pre-requisite to calculating fugacity capacity, we determined the partitioning

279

coefficients. As expected, between soils, the PAH partitioning coefficient between soils and

280

water (Kd) varied over five orders of magnitude. For example, for benzo(a)pyrene, the log Kd

281

varied between 1.34 and 5.05 (Supporting Information Table S2). In contrast, between PAHs

282

within a soil, log Kd values varied at most one order of magnitude.

283

Partitioning between simulated intestinal fluid (Ksim) and soils remained relatively similar

284

between soils. For example, for benzo(a)pyrene, the log Ksim only varied between 1.59 and 2.97.

285

The Ksim between PAHs was also relatively similar with Ksim varying slightly less than an order

286

of magnitude within a soil (Supporting Information Table S3). The two measures of partitioning

287

were not linked to one another across soils. For example, Kd and Ksim did not correlate (r < 0.1)

288

for all PAHs, except benzo(a)pyrene in which Kd and Ksim were negatively correlated (r = -0.43,

289

p < 0.05).

290

Soil Fugacity Capacity



291

Soil fugacity capacity predicts, with an average slope of 0.30 with standard error of the

292

slopes between 0.054-0.073, soil PAH concentration (r2 between 0.61-0.73, p

Predicting Polycyclic Aromatic Hydrocarbon ... - ACS Publications

Recommend Documents