Subscriber access provided by University of Newcastle, Australia
Article
Evaluating activated carbon adsorption of DOM and micropollutants using fluorescence spectroscopy Kyle Koyu Shimabuku, Anthony M. Kennedy, Riley E. Mulhern, and R. Scott Summers Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.6b04911 • Publication Date (Web): 24 Jan 2017 Downloaded from http://pubs.acs.org on January 29, 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 29
1 2 3
4 5
6 7 8
Environmental Science & Technology
Evaluating activated carbon adsorption of DOM and micropollutants using fluorescence spectroscopy , Kyle K. Shimabuku*,†, Anthony M. Kennedy† ‡, Riley E. Mulhern†, and R. Scott Summers†
†
Department of Civil, Environmental and Architectural Engineering, 428 UCB, University of Colorado, Boulder, Boulder, CO 80309, USA
‡
Technical Service Center, US Bureau of Reclamation, Denver Federal Center Building 67, Denver, CO, 80225, USA * Corresponding author:
[email protected] 9
Abstract Dissolved organic matter (DOM) negatively impacts granular activated carbon
10
(GAC) adsorption of micropollutants and is a disinfection by-product precursor. DOM from
11
surface waters, wastewater effluent, and 1 kDa size fractions was adsorbed by GAC and
12
characterized
13
chromatography (SEC). Fluorescing DOM was preferentially adsorbed relative to UV-absorbing
14
DOM. Humic-like fluorescence (peaks A and C) was selectively adsorbed relative to
15
polyphenol-like fluorescence (peaks T and B) potentially due to size exclusion effects. In the
16
surface waters and size fractions, peak C was preferentially removed relative to peak A, whereas
17
the reverse was found in wastewater effluent, indicating that humic-like fluorescence is
18
associated with different compounds depending on DOM source. Based on specific UV-
19
absorption (SUVA), aromatic DOM was preferentially adsorbed. The fluorescence index (FI), if
20
interpreted as an indicator of aromaticity, indicated the opposite, but exhibited a strong
21
relationship with average molecular weight suggesting FI might be a better indicator of DOM
22
size than aromaticity. The influence of DOM intermolecular interactions on adsorption were
23
minimal based on SEC analysis. Fluorescence parameters captured the impact of DOM size on
using
fluorescence
spectroscopy,
UV-absorption,
ACS Paragon Plus Environment
and
size
exclusion
Environmental Science & Technology
24
the fouling of 2-methylisoborneol and warfarin adsorption, and correlated with direct
25
competition and pore blockage indicators.
26
Introduction
27
Dissolved organic matter (DOM) is ubiquitous in natural and anthropogenically-impacted
28
waters such as surface water and wastewater and plays an important role in environmental and
29
engineered aquatic systems. For instance, it can serve as an energy source for bacteria, attenuate
30
light, and influence the fate and transport of contaminants. DOM can also adversely impact water
31
and wastewater treatment processes (e.g., by consuming disinfectants and fouling membranes)
32
and is a disinfection by-product precursor.1-3
33
Granular activated carbon (GAC) is used in water and wastewater treatment plants to
34
remove DOM, as well as specific organic chemicals present at sub part per billion concentrations
35
defined herein as micropollutants. Both DOM concentration, typically measured as dissolved
36
organic carbon (DOC), and character influence its adsorbability as well as the extent to which it
37
decreases micropollutant adsorption through direct competition and pore blockage, collectively
38
termed fouling. The dependence of both DOC and micropollutant breakthrough on initial DOC
39
concentration (DOC0) has been successfully modeled in previous studies.4,5 However, it is more
40
difficult to model how DOM quality influences its adsorption behavior (i.e., adsorbability and
41
micropollutant fouling) because DOM is a heterogeneous mixture of compounds that vary in
42
physical and chemical properties. In addition, DOM indicators that capture the impact of DOM
43
character on its adsorption behavior, particularly in dynamic flow-through systems, are still
44
needed. Several studies have shown that DOM molecular size can determine its adsorption
45
behavior.6-9 However, aromaticity and polarity can also be important,10 and it remains unclear
ACS Paragon Plus Environment
Page 2 of 29
Page 3 of 29
Environmental Science & Technology
46
which characteristic(s) governs DOM adsorption behavior. Also, the extent to which DOM
47
interactions (e.g., supramolecular association) affect DOM adsorption remains unclear.
48
Fluorescence spectroscopy is one technique that is growing in popularity to characterize
49
DOM as it is simple and provides a large amount of information. Molecular size, aromaticity,
50
and the presence of specific classes of DOM components have been related to different
51
fluorescence parameters such as the peak C intensity to UV-absorption ratio (see Table S1),
52
fluorescence index (FI), and fluorescence peaks, respectively.11-13 However, the physicochemical
53
properties associated with many fluorescence indices are still unknown. GAC is mainly
54
composed of graphite-like sheets that form a heterogeneous, microporous network, and it serves
55
as a largely hydrophobic, highly conjugated sieve that simultaneously separates DOM
56
components by molecular size and chemical characteristics such as aromaticity.14 However,
57
GAC can exhibit variable pore size distributions and surface functional groups and could, e.g.,
58
separate DOM as it passes through a column by selectively adsorbing DOM that does not
59
experience electrostatic repulsion. Thus, fractionating DOM with GAC provides a unique
60
method to characterize the physicochemical properties of fluorescence parameters compared to
61
studies that have employed size exclusion chromatography (SEC) and/or high performance
62
liquid chromatography (HPLC) where DOM is separated solely by molecular size or
63
hydrophobicity, respectively.
64
There is growing interest in using fluorescence to characterize DOM adsorption behavior.
65
Recent studies have found that fluorescence metrics can be used to predict DOM fouling,15-17
66
fluorescing DOM can be more adsorbable than non-fluorescing DOM,15,18 and PAC can
67
selectively remove DOM that fluoresces at longer wavelengths.19 Previously, we showed that
68
fluorescence parameters correlated with the competitive nature of DOM in batch-powdered
ACS Paragon Plus Environment
Environmental Science & Technology
69
activated carbon (PAC) systems.17 However, whether or not fluorescence indicators can also be
70
used in flow-through systems to capture the influence of DOM character on micropollutant
71
fouling is unknown considering that in batch-PAC systems micropollutants and DOM are
72
simultaneously exposed to PAC, while in GAC adsorbers DOM can preload adsorption sites.
73
The objectives of this study were to: (i) assess the preferential adsorption of fluorescing
74
DOM components by GAC, (ii) relate physicochemical and fluorescence properties of DOM,
75
(iii) examine if DOM intermolecular interactions influence DOM adsorption by GAC, (iv)
76
develop relationships between fluorescence indices with DOC and micropollutant breakthrough,
77
and (v) compare the significance of DOM molecular size and aromaticity on DOM adsorption
78
and micropollutant fouling. The findings presented here provide new interpretations of
79
fluorescence metrics and DOM adsorption phenomena. They also demonstrate that fluorescence
80
metrics, which can be determined with on-line sensors or conventional analyzers that
81
characterize discrete samples, could be used in the future to predict and monitor the control of
82
micropollutants and DOM by GAC. This study focused on the adsorption behavior of surface
83
water DOM fractioned with a 1 kDa ultrafilter and the unfractionated DOM, which were also
84
examined in a companion study by Kennedy and Summers.20 These waters were chosen because
85
they vary distinctly in molecular size and aromaticity. Supplementary experiments were
86
conducted with additional surface waters and a wastewater effluent, and in batch-mode with
87
three different PACs.
88
Materials and Methods
89
Waters. Water collected from a non-wastewater impacted mountain reservoir at Big Elk
90
Meadows (BEM), CO, which has high DOC (~15 mg/L), low alkalinity (~21 mg/L as CaCO3),
91
and low conductivity (~119 µS/cm), was concentrated using reverse-osmosis. If a surface water
ACS Paragon Plus Environment
Page 4 of 29
Page 5 of 29
Environmental Science & Technology
92
is wastewater impacted is distinguished here because wastewater effluent has a significant
93
influence on fluorescence signatures.21,22 The reverse-osmosis concentrate was fractionated using
94
a tangential flow ultrafilter (Amicon regenerated cellulose 1 kDa cartridge) into 1
95
kDa nominal molecular size fractions. Nominal should be emphasized because DOM that is 1 kDa can be detected both in the 1 kDa ultrafilter retentate and permeate (Figure S1).
97
The reverse-osmosis concentrate was also coagulated and used to create two matrices, one only
98
spiked with the micropollutants 2-methylisoborneol (MIB) and warfarin (described below) and
99
the other was spiked with 27 different micropollutants described in Table S2. The total mass
100
concentration of the micropollutants was 3.0 ± 0.9 µg/L. The 27 micropollutants were not
101
directly monitored in this study, but were added to make a representative anthropogenic-
102
impacted water.23,24 Details describing the isolation, coagulation, and ultrafiltration conditions
103
are provided in the Supporting Information (SI).
104
The unfractionated concentrate, the 1 kDa fraction, the coagulated
105
concentrate, and the micropollutant-spiked coagulated concentrate were each diluted into
106
reverse-osmosis treated tap water to target a DOC of 4 mg/L and are designated herein as BEM,
107
1K BEM, CB, and CBMP, respectively. Tertiary wastewater effluent collected prior
108
to UV disinfection is designated herein as WWef. Water was also collected after alum
109
coagulation from the Betasso Water Treatment Plant in Boulder, Colorado, which treats a non-
110
wastewater impacted source water and is designated herein as Betasso. Influent water quality
111
characteristics are provided in Table 1.
112
Target adsorbates. Influent concentrations for micropollutants MIB and warfarin were
113
107±3 and 100±6 ng/L, respectively. MIB and warfarin (pKa = 4.5) were predominately neutral
114
and anionic, respectively. 14C MIB and 3H warfarin were obtained from American Radiolabeled
ACS Paragon Plus Environment
Environmental Science & Technology
115
Chemicals, Inc. (St. Louis, MO). They were analyzed with liquid scintillation counting, which
116
yielded method detection limits of 19 ng/L for MIB and 8 ng/L for warfarin as detailed in the SI.
117
GAC column tests. Rapid small-scale column tests were used to evaluate GAC
118
adsorption, and details are provided in the SI. >1K BEM, 1K BEM, 1K BEM fraction, 1K BEM, 1K BEM, 1K BEM, 380 nm in an HPLC.37 Korak et al. showed that Suwanee River NOM,
188
Suwanee River Fulvic Acid, and Suwanee River Humic Acid fluoresced in the peak A and C
189
regions, though the relative intensities of peak A and C varied in each water.38 Therefore, it is
190
more likely that there are different fluorescing substances that exhibit unique peak A/C ratios and
191
adsorbabilities, and compounds with smaller A/C ratios are more adsorbable in the surface
192
waters and size fractions, while the opposite is true for the WWef.
193
The preferential removal of the compounds with different peak A/C ratios in GAC
194
columns can be attributed to differences in their molecular size or chemical properties (e.g.,
195
aromaticity and/or hydrophobicity). Li et al. showed that peaks A and C simultaneously eluted in
196
an HPLC for a surface water and a wastewater effluent,36 which suggests that compounds with
197
different A/C ratios exhibit similar polarities. Thus, the compounds in the surface waters that had
198
higher peak A/C ratios might be larger and experience more size exclusion than compounds with
199
smaller A/C ratios, and the reverse could be true for the WWef. However, Li et al. also found
200
that fluorescence intensity in regions near peaks A and C simultaneously eluted using SEC. The
201
reason that GAC could fractionate DOM components with different A/C ratios and Li et al. was
202
unable to do so with SEC and HPLC could be that GAC can separate DOM through both size
203
exclusion and adsorption. For instance, it is conceivable that when GAC adsorbs DOM,
204
intermolecular bonds in DOM supramolecular structures are disrupted, which SEC is unable to
205
provoke.36 These separated DOM molecules that have different A/C ratios might exhibit
ACS Paragon Plus Environment
Environmental Science & Technology
206
different MWs and experience varying degrees of size exclusion, which could cause these
207
compounds to breakthrough at different rates. Such separations would not happen in an HPLC if
208
these compounds had similar polarities.
Page 10 of 29
209
DOM breakthrough for the seven waters as assessed by the ratio of the sum of peaks T+B
210
to the sum of peaks A+C is shown in Figure 2 and Figure S6. For each water, the peaks
211
T+B/peaks A+C ratio values initially broke through above the influent peaks T+B/peaks A+C
212
ratio values indicating that peaks T+B are less adsorbable by GAC than peaks A+C in waters
213
that vary in MW and source. Several studies have found that the peaks in the T+B regions exhibit
214
greater hydrophobicities than peaks in the A+C regions in both surface waters and
215
wastewater.36,37,39 According to Li et al., peaks A+C could be more hydrophilic because they are
216
dominated by humic substances that are anionic and, therefore, hydrophilic at circumneutral pH
217
whereas compounds associated with peaks B+T (e.g., tannins) can be neutral and
218
hydrophobic.36,40 However, because peaks A, B, C, and T can be associated with different classes
219
of compounds (see below), their hydrophobicities could vary depending on the water source.
220
Thus, further research is needed to confirm if peaks T+B are more hydrophobic than A+C
221
regardless of the surface water or wastewater effluent. If peaks B+T are more hydrophobic than
222
A+C here, the faster peak T+B breakthrough suggests that they are associated with larger DOM
223
compounds that experience more size exclusion because GAC preferentially adsorbs small,
224
hydrophobic compounds. Studies using SEC with fluorescence detection have found that peaks
225
in the T+B regions exhibit greater MWs than peaks in the A+C regions.39,41
226
While peaks A+C and peaks T+B are often referred to as humic-like and protein-like,
227
respectively, DOM components that do not fit into these classifications can fluoresce in these
228
regions.42 Although the hypothesis that peaks T+B are greater in MW than peaks A+C is in
ACS Paragon Plus Environment
Page 11 of 29
Environmental Science & Technology
229
agreement with peaks T+B being associated with proteins, as proteins typically exhibit greater
230
MWs (>10 kDa) than humic substances (average MWs < 5~ kDa), high MW polyphenolic
231
compounds other than proteins (e.g., tannins MWs 0.5-20 kDa) can also exhibit “protein-like”
232
fluorescence in surface waters and WWef.41-44 In addition, peaks T+B were concentrated in
