Subscriber access provided by UNIVERSITY OF ADELAIDE LIBRARIES
Environmental Processes
New Insights and Model for Understanding NOM Adsorption onto Mixed Adsorbents Siamak Modarresi, and Mark Benjamin Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.8b00849 • Publication Date (Web): 08 May 2018 Downloaded from http://pubs.acs.org on May 8, 2018
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
1
Environmental Science & Technology
New Insights and Model for Understanding NOM Adsorption onto Mixed Adsorbents
2 3 4
By
5
Siamak Modarresi and Mark M. Benjamin*
6 7
Dept. of Civil and Environmental Engineering, Box 352700, University of Washington, Seattle, WA 98195
ACS Paragon Plus Environment
1
Environmental Science & Technology
8
Page 2 of 29
Abstract
9
Adsorption-based processes are commonly used to remove natural organic matter (NOM)
10
from drinking water sources and thereby mitigate its impacts on other water treatment
11
processes and the quality of the finished water. These processes are complicated by the fact
12
that NOM comprises multiple fractions that can exhibit disparate adsorption behaviors. Prior
13
modeling of NOM adsorption has invariably focused on systems with a single adsorbent, but
14
results presented here demonstrate surprising and counter-intuitive behavior in systems
15
containing two or more adsorbents. Specifically, if the sequence of affinities of the different
16
NOM fractions are the same for two adsorbents, then overall adsorption changes
17
monotonically as one adsorbent is gradually replaced by the other. However, if the sequence
18
differs for the two adsorbents, overall adsorption can increase even when the nominally
19
stronger adsorbent is gradually replaced by the weaker one. This work demonstrates and
20
explains such behavior for a particular mixture of adsorbents and introduces a mathematical
21
model that illustrates how other mixtures might behave.
22
ACS Paragon Plus Environment
2
Page 3 of 29
23
Environmental Science & Technology
Abstract/TOC Art DOC adsorbed (mg /l)
2.5
PAC dose = Total dose − HAOPs dose
2.0
50 mg/l 1.5
20 mg/l
1.0
Model 1 0.5
Total adsorbent dose: 10 mg/l
0.0 0
24
10
20
Model 2 30
40
50
HAOPs dose (mg/l)
25
ACS Paragon Plus Environment
3
Environmental Science & Technology
26
1.
Page 4 of 29
Introduction
27
Natural organic matter (NOM) comprises a diverse group of organic compounds arising
28
from plant degradation and microbial metabolism1,2. It can impart taste, odor and color to water;
29
react with disinfectants to produce toxic and/or carcinogenic disinfection by-products (DBPs)3–5;
30
increase the required doses of coagulant, adsorbent and disinfectant for effective water treatment;
31
and foul microfiltration and ultrafiltration membranes6–10. For all these reasons, NOM is a major
32
target of water treatment operations.
33
The most commonly used process for NOM removal is coagulation with Al- or Fe-based
34
salts, in which the primary removal mechanism is adsorption onto the hydrous metal oxide
35
precipitates that the coagulants generate. Coagulation works well for removing high-molecular
36
weight NOM with high specific UV absorbance at 254 nm (SUVA254), but it is less effective at
37
removing low-molecular weight NOM with low SUVA25411. Adsorption by activated carbon has
38
also been extensively studied for NOM treatment and has been found to depend strongly on the
39
volume of pores in the secondary micropore and mesopore regions (0.8–5nm)12,13. In most
40
studies, activated carbon adsorption and metal oxide coagulation have been found to target
41
different NOM fractions, with coagulation preferentially collecting larger molecules (>30 kDa)
42
and PAC more efficient at collecting smaller ones
43
combined with PAC adsorption, DOC removal over a broad range of molecular weights can be
44
attained. When this approach has been applied to enhance removal of DBP precursors or mitigate
45
membrane fouling, the results have often been encouraging in laboratory studies, but those in
46
pilot- or full-scale applications have been mixed17–24.
13–16
. As a result, when coagulation is
47
Heated aluminum oxide particles (HAOPs) are a novel adsorbent first synthesized by Kim et
48
al.25 At low doses, HAOPs remove more NOM than conventional coagulants (alum and ferric
ACS Paragon Plus Environment
4
Page 5 of 29
Environmental Science & Technology
49
chloride) do, but at high doses, the removal efficiency reaches a plateau, indicating that a fraction
50
of the NOM cannot adsorb onto HAOPs26. Much of the NOM that is not adsorbable on HAOPs
51
can adsorb onto PAC26, so applying a mixture of HAOPs and PAC might be an effective
52
treatment strategy.
53
NOM is frequently characterized by fractionation based on the molecules’ hydrophobicity,
54
acid/base characteristics, and/or molecular weight27. By comparing the concentrations of
55
different fractions before and after treatment of the whole water, or by subjecting the isolated
56
fractions to various treatment steps, researchers have attempted to infer which components of
57
NOM are selectively targeted by the treatments, and which ones are responsible for adverse
58
impacts such as DBP formation and membrane fouling5,14,28–30. While these studies have greatly
59
enhanced our understanding of the chemical nature and reactivity of NOM, they have not led to
60
many practical improvements in water treatment applications. This limitation arises both because
61
the molecular properties of the NOM are distributed more or less continuously across a wide
62
spectrum, rather than falling neatly into the well-defined ranges implied by the fractionation
63
procedure, and because treatment processes are similarly imprecise in the types of molecules that
64
they remove. As a result, designing and operating a practical treatment process that selectively
65
targets a particular subset of the NOM molecules has proven extremely challenging.
66
Rather than focusing on the chemical distinctions among different NOM fractions, some
67
researchers have used adsorption directly to characterize the fractions31–34. In this approach, the
68
NOM is represented as a mixture of several “fictive” components that compete for the available
69
surface sites. The components are assigned adsorptive properties that, when considered in
70
conjunction with a specific model for competitive adsorption, cause the overall NOM adsorption
71
computed using the model to mimic the experimental findings, without ascribing any particular
ACS Paragon Plus Environment
5
Environmental Science & Technology
Page 6 of 29
72
molecular properties to the different components. The research presented here evaluated the
73
removal of NOM from water by mixtures of HAOPs and PAC and modeled the results in terms
74
of fictive NOM components. The work yielded unexpected results and provided new insights
75
into competitive adsorption in systems containing not only multiple adsorbates, but also multiple
76
adsorbents.
77
2.
Materials and methods
78
Freshwater for the tests was collected from Lake Union (LU) in Seattle, WA. The water
79
contained 2.4±0.3 mg/l DOC and had UV254 of 0.057±0.03 cm–1; these parameters were closely
80
correlated with one another in the subsequent adsorption tests, and both were used as indicators
81
of NOM concentration. The pH of the samples was adjusted to 7.0±0.05, and 0.5 mM each of
82
NaCl and NaHCO3 was added to increase the ionic strength and buffer capacity. HAOPs were
83
synthesized following the approach of Kim et al.25, and Norit SA SUPER PAC was purchased
84
commercially (Cabot Co. Boston, MA). Some key properties of the adsorbents are provided in
85
Table SI-1 of the Supplementary Information (SI).
86
Batch adsorption was investigated by adding either one or both adsorbents to flasks
87
containing the test water and mixing the suspensions on a rotary shaker for two hours. This
88
mixing time was chosen both because it is at the upper end of exposure times for adsorption at
89
water treatment plants and because, as shown in Figure SI-1, the rate of adsorption onto both
90
adsorbents was very low at the end of that period. The solids were then removed by filtration
91
through 0.45-µm filters and analyzed for DOC and UV254. UV254 was measured with a dual-
92
beam Lambda-18 spectrophotometer (Perkin-Elmer Gmbh., Überlingen Germany) using a 1-cm
93
quartz cell, and DOC was determined with a Sievers 900 TOC analyzer (GE Analytical
94
Instruments, Boulder, CO).
ACS Paragon Plus Environment
6
Page 7 of 29
Environmental Science & Technology
95
3.
96
Batch adsorption of NOM
97 98
Results and discussion
The removal of DOC by PAC and HAOPs in single-adsorbent systems is characterized in Figure 1. PAC removed more DOC than HAOPs did at all the adsorbent doses investigated.
99
Because DOC removal by HAOPs approached a plateau at high adsorbent doses, adsorption
100
in that system was modeled by treating the NOM as comprising two fictive components – one
101
component (A) that is adsorbable on HAOPs and accounts for 45% of the DOC, and a second
102
component (B) that cannot adsorb on HAOPs and accounts for the remaining 55%. With that
103
assumption, the conceptual model corresponds to the model that leads to the classical Langmuir
104
isotherm for a system containing a single adsorbate that binds to a single type of adsorbent site
105
(Eqn.1):
106
qH,A =
qmax,H K H,A cA
(1)
1 + K H,A cA
107
where qH,A is the equilibrium adsorption density (mg DOC/g) of NOM (fictive) component A on
108
HAOPs (‘H’), qmax,H is the maximum adsorption density on HAOPs, KH,A is the adsorptive
109
equilibrium constant for binding of A to HAOPs, and cA is the dissolved concentration of A at
110
equilibrium. The data could be fit very well by this isotherm equation, as shown in Figure 1.
111
Adsorption of the NOM onto PAC was then modeled using the same two fictive
112
components and, again, only one type of surface site. In this case, however, both the A and B
113
components of the NOM were assumed to adsorb, albeit with different affinities for the surface;
114
that is, the two fictive components competed with one another for the surface sites, so adsorption
115
was modeled using the two-component competitive Langmuir isotherm (Eq.2a,b):
116
qP,A =
qmax,P K P,A cA
(2a)
1 + K P,A cA + K P,BcB
ACS Paragon Plus Environment
7
Environmental Science & Technology
117
qP,B =
qmax,P K P,BcB
Page 8 of 29
(2b)
1 + K P,A cA + K P,BcB
118
This modeling approach again yielded a very good fit to the experimental data. However, two
119
sets of Langmuir parameters fit the data equally well – one with NOM component A adsorbing
120
more strongly than component B (KP,A>KP,B), and one with the opposite order of affinities
121
(KP,AKH,B (=0), if KP,A>KP,B, the NOM
236
component that binds more strongly to HAOPs also binds more strongly to PAC (maximum
237
overlap), and if KP,AKP,B) 0.876 0.065 0.221
HAOPs 1.140 0.000 0.185
PAC (KP,A
