Subscriber access provided by UNIV OF NEWCASTLE
Article
Bioavailability of carbon nanomaterial-adsorbed polycyclic aromatic hydrocarbons to P. promelas: influence of adsorbate molecular size and configuration Erica N. Linard, Onur G Apul, Tanju Karanfil, Peter van den Hurk, and Stephen J. Klaine Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.7b02164 • Publication Date (Web): 12 Jul 2017 Downloaded from http://pubs.acs.org on July 12, 2017
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
Environmental Science & Technology
Bioavailability of carbon nanomaterial-adsorbed polycyclic aromatic hydrocarbons to P. promelas: influence of adsorbate molecular size and configuration †, ‡
*Erica N Linard, Klaine
, Onur G Apul
§,a
||
, Tanju Karanfil , Peter van den Hurk
†,#
, and Stephen J.
†,^
†
Institute of Environmental Toxicology, Clemson University, Pendleton, South Carolina, USA
‡
Interdisciplinary Graduate Program in Environmental Toxicology, Clemson University
§
School of Sustainable Engineering and the Built Environment, Arizona State University,
Tempe, AZ a
Department of Civil and Environmental Engineering, University of Massachusetts Lowell,
Lowell MA ||
Department of Environmental Engineering and Earth Sciences, Clemson University, Anderson,
South Carolina, USA # ^
Department of Biological Sciences, Clemson University, Clemson, South Carolina, USA Deceased
KEYWORDS. Bioavailability, adsorption, carbon nanomaterials, polycyclic-aromatic hydrocarbons, bioavailability index 1
ACS Paragon Plus Environment
Environmental Science & Technology
1
ABSTRACT. Despite carbon nanomaterials’ (CNMs) potential to alter the bioavailability of
2
adsorbed contaminants, information characterizing the relationship between adsorption behavior
3
and bioavailability of CNM-adsorbed contaminants is still limited. To investigate the influence
4
of CNM morphology and organic contaminant (OC) physicochemical properties on this
5
relationship, adsorption isotherms were generated for a suite of polycyclic aromatic
6
hydrocarbons (PAHs) on multi-walled carbon nanotubes (MWCNTs) and exfoliated graphene
7
(GN) in conjunction with determining the bioavailability of the adsorbed PAHs to Pimphales
8
promelas using bile analysis via fluorescence spectroscopy. Although it appeared that GN
9
adsorbed PAHs indiscriminately compared to MWCNTs, the subsequent bioavailability of GN-
10
adsorbed PAHs was more sensitive to PAH morphology than MWCNTs. GN was effective at
11
reducing bioavailability of linear PAHs by ~ 70%, but had little impact on angular PAHs.
12
MWCNTs were sensitive to molecular size, where bioavailability of two-ringed naphthalene was
13
reduced by ~80%, while bioavailability of the larger PAHs was reduced by less than 50%.
14
Furthermore, the reduction in bioavailability of CNM-adsorbed PAHs was negatively correlated
15
with the amount of CNM surface area covered by the adsorbed-PAHs. This study shows that the
16
variability in bioavailability of CNM-adsorbed PAHs is largely driven by PAH size,
17
configuration and surface area coverage.
18
ABSTRACT ART.
19 2
ACS Paragon Plus Environment
Page 2 of 33
Page 3 of 33
Environmental Science & Technology
20
INTRODUCTION. Among the numerous applications carbon nanomaterials (CNMs) are used
21
for, CNMs’ high adsorption affinity for organic contaminants (OCs) has led to increasing interest
22
in such materials for pollution remediation.1,2 However, due to CNMs’ unique physicochemical
23
characteristics, they interact with OCs in a distinctly different way than natural colloids.
24
Considerable uncertainty exists pertaining to CNMs’ potential to facilitate transport of adsorbed
25
contaminants and likewise alter bioavailability of adsorbed contaminants in the environment.3-6
26
Adsorption capacity and rate of desorption from CNMs are considered the most influential
27
properties driving contaminant transport and likely influence bioavailability of CNM-adsorbed
28
molecules.5,7
29
Several studies have found that CNM type and OC properties significantly impact
30
bioavailability of CNM-adsorbed contaminants, though findings are not conclusive.7-9 Xia et al.
31
found the bioaccumulation of aliphatic perfluorochemicals (PFCs) in C. plumosus was mainly
32
driven by PFC structure rather than type of CNM.9 Similarly, the bioavailability and toxicity of
33
phenanthrene adsorbed to five different types of carbon nanotubes (CNTs) in D. magna was not
34
significantly different despite differences in CNTs’ surface area, functionalization and
35
diameter.10 Conversely, studies have observed phenanthrene to become more bioaccessible as
36
CNMs surface curvature decreases and with increased dispersion. 7,9,11,12 The bioaccumulation of
37
several classes of aromatic OCs, including polycyclic aromatic hydrocarbons (PAHs),
38
polychlorinated biphenyls (PCBs), and polybrominated diphenyl ethers (PBDEs) in benthic
39
invertebrates was more effectively moderated when sediment was amended with CNTs than
40
black soot; this was attributed to higher adsorption affinity for OCs by CNTs.13 Although
41
without inclusion of adsorption data it was not possible to discern how carbonaceous material
42
type contributed to differences in OC bioaccumulation. While the bioavailable fraction of CNM-
43
adsorbed contaminants is often assumed to be directly related to the un-adsorbed OC 3
ACS Paragon Plus Environment
Environmental Science & Technology
44
concentration in the system, bioaccumulation enhancement of diuron and ionizable
45
pentachlorophenol indicate this may not necessarily be the case.14,15 By examining how OC-
46
CNM interactions and adsorption behavior is linked with resulting bioavailability of the adsorbed
47
contaminant, a more holistic understanding of how CNMs may act as a “contaminant carrier”
48
can be established.
49
Although CNTs are essentially graphene sheets rolled into tubes and graphene has
50
become increasingly popular, few studies exist investigating the influence of graphene on
51
bioavailability of adsorbed contaminants. Adsorption of OCs by graphene and CNTs is largely
52
driven by hydrophobic forces and π-π bonding16,17, but due to different interactions with
53
adsorbates, graphene and CNTs can have different adsorption mechanisms.18,19 To the best of our
54
knowledge, no study exists examining how such differences in adsorption interactions of
55
graphene and CNTs may differentially affect bioavailability of adsorbed contaminants.
56
Therefore, the main objectives of this study were to (i) compare adsorption of four OCs by
57
exfoliated graphene (GN) and multi-walled carbon nanotubes (MWCNTs), (ii) investigate the
58
role of OC physicochemical characteristics on adsorption interaction differences between the two
59
materials, and (iii) compare the impact of GN and MWCNT on OC bioavailability to Pimphales
60
promelas as a function of adsorption behavior, in environmentally relevant conditions. PAHs
61
were selected as a model class of OCs for this study because they are ubiquitous in the
62
environment, on the EPAs priority pollutant list, and have a wide range of physicochemical
63
properties.
64
MATERIALS AND METHODS
65
Chemicals and particles. Multi-walled carbon nanotubes (MWCNTs) were purchased
66
from Nanostructured & Amorphous Materials (Houston, TX) and mechanically exfoliated
67
graphene (GN) was prepared at Clemson University. Details on the characterization techniques 4
ACS Paragon Plus Environment
Page 4 of 33
Page 5 of 33
Environmental Science & Technology
68
are described in the Supplemental Information (SI) and Table S1 summarizes CNMs’
69
morphological characteristics. Four polycyclic aromatic hydrocarbons (PAHs) were used in this
70
study: naphthalene (NAP), 99+% purity, from Alfa Aesar; phenanthrene (PHEN), 97% purity,
71
and anthracene (ANT), 99% purity from ACROS organics; fluoranthene (FLU), 98% purity,
72
from ULTRA scientific. Physicochemical properties of PAHs are presented in Table S2. Stock
73
solutions were prepared in HPLC grade methanol. Suwannee River natural organic matter
74
(NOM) collected from Suwannee River Visitor’s Center (Fargo, GA) had a dissolved organic
75
carbon concentration of 64 mg C/L, determined using a Shimadzu total organic carbon analyzer
76
and had a specific ultraviolet absorbance (SUVA254) value of 4.08 L mg-1m-1. All solutions were
77
prepared in moderately hard water following the U.S. Environmental Protection Agency standard
78
recipe; 96 mg/L NaHCO3, 60 mg/L CaSO4, 60 mg/L MgSO4, 4 mg/L KCl in 18 mega-ohm
79
Millipore Milli-Q water.
80
Suspension of CNM. Pristine MWCNTs and GN were suspended in 2 mg C/L NOM at a
81
concentration of 10 mg/L following previously published methods.20 CNMs were sonicated for 1
82
hour at 60 watts using a 0.5 microtip affixed Branson model 450 digital sonifier. Mass based
83
calibration curves of MWCNTs and GN in solution are supplied in Supplemental Information
84
(Figure S2); concentration was measured via visible light absorbance at 800 and 660 nm for
85
MWCNT and GN solutions respectively, using a Molecular Devices Spectramax 190 microplate
86
spectrophotometer.21,22 CNM solutions were diluted to ~1.5 mg/L in 2 mg C/L NOM background
87
solution for all adsorption and bioavailability assays. Absorbance measurements of CNM
88
solutions at beginning and end of each experiment showed little change in concentration,
89
indicating a stable solution.
90
Adsorption Isotherms. Adsorption experiments were run in triplicate following the
91
batch approach we have previously published.20 Briefly, 10 mL glass centrifuge tubes containing 5
ACS Paragon Plus Environment
Environmental Science & Technology
92
either 1.5 mg/L MWCNT or GN suspension was spiked with each PAH individually, sealed with
93
aluminum-foil-lined Teflon screw caps and allowed to come to equilibrium on a rotary tumbler
94
at room temperature (25ºC± 1°C) for 96 hours (preliminary tests showed that equilibrium was
95
reached within 48 hours). PAH stock solutions used to spike experimental vials were prepared so
96
that the volume percentage of methanol was 0.1% to avoid co-solvent effects. Initial PAH
97
concentrations of NAP, PHEN, ANT and FLU ranged from 5-1000, 5-400, 0.25-40, and 1-200
98
µg/L, respectively. Vials without CNMs were identically prepared to serve as positive controls;
99
controls for the entire system were spiked with methanol only. A combination of centrifugation
100
and fluorescence quenching techniques were used to quantify the un-adsorbed PAH
101
concentration left in the system. After equilibrium was reached vials with and without CNM
102
were centrifuged (Eppendorf 5804 R) at ~1500 x g for 1 hour to sediment any CNM
103
agglomerates that may have formed during tumbling. The remaining supernatant was analyzed in
104
black polystyrene 96-well plates with Molecular Devices Gemini fluorescence microplate reader
105
at excitation/emission wavelengths 276/330, 252/399, 250/363, and 280/440 for NAP, ANT,
106
PHEN and FLU respectively. PAHs act as π-donors when interacting with graphitic materials
107
and when adsorbed no longer are in an excited state. Therefore, fluorescence of CNM-adsorbed
108
PAHs is statically quenched and the remaining unbound PAH concentration in the aqueous
109
solution could be determined via fluorescence intensity.23-25 Analysis of a methanol rinse of
110
positive control vials showed minimal loss of PAHs due to sorption to vial walls. CNM sorption
111
of all PAHs was calculated as differences in aqueous PAH concentration between positive
112
controls and CNM samples using fluorescent standard curves developed for each PAH in the
113
background solution containing 2 mg/L NOM. Temperature, pH, and ionic strength were
114
controlled across each experiment and reflected minimal change. The adsorption data was fit
115
with the nonlinear Freundlich model (model details presented in the SI). 6
ACS Paragon Plus Environment
Page 6 of 33
Page 7 of 33
116
Environmental Science & Technology
Bioavailability Assays. All exposures were conducted for 16 h in static, non-aerated
117
conditions with full water renewals every 4 h (refer to SI for details on background conditions).
118
Assays assessing the influence of MWCNT and GN on PAH bioavailability to P. promelas were
119
replicated four times where each aquarium was stocked with ~4 fish per treatment. Response of
120
fish from each aquarium were averaged to represent a single experimental unit per replicate per
121
treatment. The prepared treatments for each PAH were 1) control, consisting of background
122
solution only; 2) PAH only, a positive control consisting of PAH-spiked background solution
123
(i.e. 10 µg/L of NAP, PHEN and FLU; 5 µg/L of ANT); and 3) PAH and CNM, consisting of the
124
same PAH-spiked background solution with either ~1.5 mg/L of suspended MWCNT or GN.
125
Solutions for each treatment were tumbled in 4L amber glass containers on a rotary tumbler to
126
reach equilibrium prior to fish exposure. Water samples from each exposure were collected at the
127
beginning and end of each water renewal; analysis of water samples verified the treatment PAH
128
concentrations throughout duration of the exposure. At the end of each exposure, fish were
129
euthanized, the gallbladder harvested, and the collected bile analyzed for PAH metabolites to
130
determine the bioavailable fraction of PAH in the system.
131
Bioavailability is defined as the total amount of PAH that is taken up by the organism
132
from the system and biotransformed26,27; including aqueous and CNM-adsorbed PAH exposure.
133
Due to highly efficient metabolism of PAHs by fish, the bioaccumulation potential of these
134
compounds in traditionally monitored tissues such as muscles is typically low and not always
135
well correlated with exposure.28 Instead, PAHs are quickly biotransformed into water-soluble
136
metabolites in the liver and concentrated in the gallbladder for later excretion. Most PAHs have
137
distinct fluorometric properties allowing for easy detection and identification in complex
138
matrices like bile.28 Dose-response relationships showed a linear correlation between PAH
139
exposure and bile fluorescence intensity (Figure S3), therefore change in bile fluorescence was 7
ACS Paragon Plus Environment
Environmental Science & Technology
140
directly related to change in PAH bioavailability. Extracted bile was diluted, vortexed, and
141
centrifuged. The supernatant was analyzed via fixed wavelength fluorescence (FF) analysis at
142
discreet excitation and emission wavelength pairs for each PAH.28,29 PHEN, ANT, and FLU
143
metabolites were detected using a Molecular Devices Gemini fluorescence microplate reader at
144
excitation/emission wavelengths of 250/363, 252/399, and 280/440 nm; NAP metabolites were
145
detected via HPLC with a UV-fluorescence detector 276/330 nm. See SI for more details on bile
146
preparation and analysis.
147
Statistical Analysis. JMP Pro 12 software, SAS Institute Inc. was used to perform all
148
statistical tests. Adsorption isotherms were fit with the non-linear Freundlich model. Multiple
149
regression was used to assess and compare the influence of PAH physicochemical characteristics
150
on adsorption of PAHs by CNMs. Linear regression and correlation coefficients (R2) were used
151
to assess goodness of fit of predicted fish responses to actual response in fish exposures and
152
CNM bioavailability assays. Analysis of variance (ANOVA) followed by Student’s T test were
153
used to compare response of fish across all treatments for a single PAH; linear squared contrasts
154
were used to compare differences between treatments across the different PAHs. Two-tailed
155
mean hypothesis testing was used to determine whether bioavailability indices differed from 1
156
for each PAH_CNM combination. Student’s T test was used to compare bioavailability indices
157
to one another across all PAHs. Shapiro-Wilk’s test was used to test normality of bile
158
fluorescence residuals; Brown and Forsythe test was used to test homogeneity of variance within
159
treatments. Unless otherwise denoted, the significance level of all tests was set at 0.05.
160
RESULTS AND DISCUSSION
161
Influence of CNM type on adsorption. Adsorption isotherms of PAHs by MWCNT and
162
GN in the presence of 2 mg C/L NOM are represented in the form of a distribution curve
163
following methods by Zhu et al., where adsorption (qe) of the PAHs by GN was plotted against 8
ACS Paragon Plus Environment
Page 8 of 33
Page 9 of 33
Environmental Science & Technology
164
the adsorption of the PAHs by MWCNT at the same experimental concentration (Figure 1).30
165
The corresponding Freundlich parameters of individual adsorption isotherms are on Table 1. To
166
help conceptualize differences in GN/MWCNT distribution among the four PAHs, lines at 0.33,
167
1 and 3 are provided on Figure 1 to represent GN/MWCNT distribution coefficients (KGN/MWCNT)
168
where qGN = 0.33 qMWCNT, qGN = qMWCNT, or qGN = 3qMWCNT, respectively. On a mass basis,
169
distribution coefficients of PHEN and FLU fall along 0.33 distribution line (Figure 1 [A]) though
170
NAP and ANT fall closer to the distribution line equal to one. Further comparison of the
171
Freundlich KF values confirms that MWCNT adsorption capacity for FLU and PHEN was
172
significantly greater (p-value
