Subscriber access provided by Oregon State University
Article
Influence of Active Layer and Support Layer Surface Structures on Organic Fouling Propensity of Thin-Film Composite Forward Osmosis Membranes Xinglin Lu, Laura H Arias Chavez, Santiago Romero-Vargas Castrillón, Jun Ma, and Menachem Elimelech Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/es5044062 • Publication Date (Web): 07 Jan 2015 Downloaded from http://pubs.acs.org on January 21, 2015
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 26
Environmental Science & Technology
Influence of Active Layer and Support Layer Surface Structures on Organic Fouling Propensity of Thin-Film Composite Forward Osmosis Membranes
Environmental Science & Technology Revised: November 12, 2014
Xinglin Lu1, Laura H. Arias Chavez2, Santiago Romero-Vargas Castrillón2, Jun Ma*,1, and Menachem Elimelech*,2 1
State Key Laboratory of Urban Water Resource and Environment, Harbin Institute of Technology, Harbin 150090, China. 2
Department of Chemical and Environmental Engineering, Yale University, New Haven, Connecticut 06520-8286, USA *Corresponding Authors * E-mail:
[email protected] (J.M.),
[email protected] (M.E.). Tel: +86 451 86283010 (J.M.), +1 203 432 2789 (M.E.). Fax: +86 451 86283010 (J.M.), +1 203 432 4387 (M.E.).
-1ACS Paragon Plus Environment
Environmental Science & Technology
Page 2 of 26
1
Abstract
2
In this study, we investigate the influence of surface structure on the fouling propensity of thin-
3
film composite (TFC) forward osmosis (FO) membranes. Specifically, we compare membranes
4
fabricated
5
(dimethylformamide, DMF and n-methyl-2-pyrrolidinone, NMP) during phase separation. FO
6
fouling experiments were carried out with a feed solution containing a model organic foulant.
7
The TFC membranes fabricated using NMP (NMP-TFC) had significantly less flux decline (7.47
8
± 0.15%) when compared to the membranes fabricated using DMF (DMF-TFC, 12.70 ± 2.62%
9
flux decline). Water flux was also more easily recovered through physical cleaning for the NMP-
10
TFC membrane. To determine the fundamental cause of these differences in fouling propensity,
11
the active and support layers of the membranes were extensively characterized for physical and
12
chemical characteristics relevant to fouling behavior. Polyamide surface roughness was found to
13
dominate all other investigated factors in determining the fouling propensities of our membranes
14
relative to each other. The high roughness polyamide surface of the DMF-TFC membrane was
15
also rich in larger leaf-like structures, while the lower roughness NMP-TFC membrane
16
polyamide layer contained more nodular and smaller features. The support layers of the two
17
membrane types were also characterized for their morphological properties, and the relation
18
between support layer surface structure and polyamide active layer formation was discussed.
19
Taken together, our findings indicate that support layer structure has a significant impact on the
20
fouling propensity of the active layer, and this impact should be considered in the design of
21
support layer structures for TFC membranes.
through
identical
procedures
except
for
the
22
-2ACS Paragon Plus Environment
use
of
different
solvents
Page 3 of 26
Environmental Science & Technology
TOC Art
-3ACS Paragon Plus Environment
Environmental Science & Technology
23
INTRODUCTION
24
Membrane processes hold significant promise for addressing the global challenge of water
25
scarcity and the need for greater sustainability.1-3 Highly permeable and selective polyamide
26
thin-film composite (TFC) membranes have facilitated the rise of reverse osmosis (RO) as the
27
dominant technology for desalination.2, 4 The ability to separately optimize the two membrane
28
layers (i.e., active and support layers) has also allowed the development of TFC membranes for
29
forward osmosis (FO) and pressure retarded osmosis (PRO).5-7 These technologies have wide
30
potential applications in the water, energy, waste reclamation, agricultural, and biomedical
31
industries.8
32
Despite the high water permeability and solute selectivity of TFC polyamide membranes,
33
their performance is significantly hampered by fouling.9-13 Intrinsic hydrophobicity and native
34
carboxyl groups, which interact with foulants, make the aromatic polyamide active layer
35
intrinsically prone to fouling.2 Consequently, extensive efforts in surface modification of the
36
polyamide active layer have sought to mitigate fouling14-17. Other studies have shown that the
37
characteristic nanoscale ridge-and-valley structure of the polyamide layer can also exacerbate
38
membrane fouling.18-20
39
The fabrication of a TFC membrane generally comprises two steps4: (i) fabrication of a
40
support layer through non-solvent induced phase separation and (ii) formation of a thin selective
41
layer on the support layer via interfacial polymerization. Previous publications21-24 suggest that
42
the support layer plays a significant role in determining active layer perm-selectivity through its
43
presence during active layer formation. Singh et al.21 studied the effect of support layer surface
44
pore size on TFC RO membrane salt rejection and water permeability. Kim et al.22 evaluated the
45
influence of support layer hydrophilicity on membrane perm-selectivity. Ghosh et al.23
46
investigated the impacts of support membrane structure and chemistry on RO membrane active
47
layer morphology. In addition, Tiraferri et al.24 fabricated different polysulfone supports by
48
varying casting conditions in the phase separation step and evaluated the perm-selectivity of
49
formed TFC FO membranes. Their results also indicated a significant influence of the
50
polysulfone support on the FO membrane active layer transport properties.
51
The above studies suggest a critical role for the support layer in determining active layer
52
perm-selectivity and morphology, with transport properties having received particular attention. -4ACS Paragon Plus Environment
Page 4 of 26
Page 5 of 26
Environmental Science & Technology
53
It is also known that surface structure, like the intrinsic ridge-and-valley structure of a polyamide
54
thin film, strongly influences membrane fouling propensity25,
55
importance of support layer structure in determining polyamide fouling propensity, through its
56
effect on polyamide structure, has not been evaluated. Only Ramon et al.26, through modeling
57
work, have considered the possibility that the underlying support layer might influence active
58
layer fouling in a significant way. Given the intensity of the current research focus on re-
59
designing support layer structures to maximize transport in FO and PRO, it is of paramount
60
importance to investigate whether these support layer changes might also be contributing to
61
different fouling propensities, which will be highly important for real-world implementation.
62
Knowing the extent to which the support layer surface structure might affect polyamide fouling
63
would also inform efforts to compare polyamide fouling of TFC membranes across RO, FO, and
64
PRO platforms, as these technologies have differently structured support layers.24
26
. However, the relative
65
In this work, we seek to investigate the influence of support layer surface structure on the
66
fouling propensity of TFC FO membranes. Two types of TFC membranes were fabricated with
67
two different support layers, formed using different solvents for the polymer solution used in the
68
phase separation step. Fouling experiments with a model organic foulant were carried out to
69
compare the fouling propensity of the two TFC FO membranes. Our results indicate that support
70
layer structure significantly impacts the polyamide active layer structure and the organic fouling
71
behavior of the TFC FO membranes. The underlying mechanisms for the difference in the
72
fouling behaviors are elucidated and implications for the design of TFC membrane with anti-
73
fouling properties are discussed.
74 75
MATERIALS AND METHODS
76
Materials and Chemicals. Polysulfone beads (Mn: 22,000 Da), 1-methyl-2-pyrrolidinone
77
(NMP, anhydrous, 99.5%), N-N-dimethylformamide (DMF, anhydrous, 99.8%), 1,3-
78
phenylenediamine (MPD, >99%), 1,3,5-benzenetricarbonyl trichloride (TMC, 98%), sodium
79
hypochlorite (NaOCl, reagent grade), and sodium bisulfite (NaHSO3, >99%) were used as
80
received (Sigma-Aldrich, St. Louis, MO). The polysulfone support was cast on a commercial
81
poly(ethylene terephthalate) nonwoven fabric (PET, grade 3249, Ahlstrom, Helsinki, Finland)
82
with a thickness of ~40 µm. During interfacial polymerization, TMC was dissolved in Isopar-G,
83
a proprietary non-polar organic solvent (Univar, Redmond, WA). -5ACS Paragon Plus Environment
Environmental Science & Technology
84
Sodium chloride (NaCl, crystals, ACS reagent) and magnesium chloride (MgCl2·6H2O,
85
crystals, ACS reagent) from J.T. Baker (Phillipsburg, NJ) were used for the membrane
86
performance tests and fouling experiments. Unless otherwise specified, all chemicals were
87
dissolved in deionized (DI) water obtained from a Milli-Q ultrapure water purification system
88
(Millipore, Billerica, MA). Sodium alginate (12 to 80 kDa, Sigma-Aldrich, St. Louis, MO), a
89
polysaccharide, was chosen as a model organic foulant. Toluidine blue O (TBO, technical grade,
90
Sigma-Aldrich) was used to characterize carboxyl group density on the TFC membrane surface.
91
Thin-Film Composite Forward Osmosis Membrane Fabrication. Two types
92
of TFC membranes were fabricated. The fabrication protocols were identical except for the
93
choice of solvent (NMP vs. DMF) in the phase separation step, which produces the polysulfone
94
support layer. Fabrication details are presented below.
95
The polysulfone support layer was prepared by non-solvent induced phase separation in
96
accordance with the method described in previous publications.15, 24 First, polysulfone beads (12
97
wt%) were dissolved in a solvent (NMP or DMF), stirred for 8 h, and deaerated in a desiccator
98
for at least 15 h prior to casting. Low-density PET fabric was attached to a clean glass plate (16
99
cm × 24 cm) using waterproof adhesive tape. NMP was applied to pre-wet the PET fabric, and
100
the excess NMP was removed using Kimwipes (Kimberly-Clark, Roswell, GA). A casting knife
101
(Gardco, Pompano Beach, FL), set at a gate height of 10 mils (~250 µm), was used to spread the
102
polysulfone solution over the wetted PET fabric. The whole composite was immediately
103
immersed in a precipitation bath containing 3 wt% solvent in DI water at room temperature
104
(23 °C) to initiate phase separation. The polysulfone support remained in the precipitation bath
105
for 10 min before being transferred to a DI water bath for storage until polyamide fabrication.
106
The polyamide active layer was formed through interfacial polymerization on the
107
polysulfone support. Supports were first immersed in MPD solution (3.4 wt% in DI water) for 2
108
min. After removing the excess MPD from the membrane surface using an air knife, the
109
membrane was immersed in the TMC solution (0.15 wt% in Isopar-G) for 1 min to form the
110
polyamide selective layer on the polysulfone support, followed by vertical draining of excess
111
TMC solution for 2 min. Then, the membrane was cured in a DI water bath at 95 °C for 2 min,
112
immersed in NaOCl solution (0.2 g/L, 2 min) followed by soaking in NaHSO3 solution (1 g/L,
113
30 seconds), and cured again in DI water at 95 °C for 2 min. The fabricated TFC membranes -6ACS Paragon Plus Environment
Page 6 of 26
Page 7 of 26
114
Environmental Science & Technology
were rinsed thoroughly and stored in DI water at 4°C.
115
Characterization of Membrane Transport Properties. The water permeability,
116
A, salt permeability, B, and salt rejection, R, of pristine TFC membranes were determined in a
117
bench-scale crossflow RO unit. After evaluation under RO conditions, each membrane coupon
118
was characterized in a bench-scale crossflow FO unit for the support layer structural parameter, S.
119
The water permeability of the polysulfone support was also characterized using a dead-end
120
filtration system. The details of these characterization methods can be found in the Supporting
121
Information (SI).
122
FO Fouling Experiments. A bench-scale crossflow FO unit, with channel dimensions
123
of 77 mm × 26 mm × 3 mm, was used in the FO fouling experiments. The experiments were
124
conducted without spacers and under co-current crossflow velocities of 8.5 cm/s in both the draw
125
solution and feed solution channels. Solution temperatures were maintained at 25 ± 0.5 °C. All
126
fouling experiments were conducted as follows. First, a clean membrane sample was loaded into
127
the FO cell with the active layer facing the feed solution (FO mode). The coupon was screened
128
for the presence of defects through measurement of water flux and solute flux at a known draw
129
solution concentration with DI water as feed solution. These fluxes were compared with
130
expected values, and the experiment proceeded if membrane performance was as expected,
131
indicating a lack of defects. The system was stabilized with DI water on both feed and draw sides.
132
Next, aliquots of inorganic stock solutions were added to the feed solution to obtain a synthetic
133
wastewater of pH 7.4 and calculated ionic strength of 14.9 mM (Visual MINTEQ 3.0).27 The
134
final concentrations of feed solutions were 0.45 mM KH2PO4, 9.20 mM NaCl, 0.61 mM MgSO4,
135
0.5 mM NaHCO3, 0.5 mM CaCl2, and 0.93 mM NH4Cl. The draw solution was supplemented
136
with concentrated draw solution stock to obtain a bulk draw solution concentration (2–4 M NaCl
137
or 2–3.5 M MgCl2) that produced an initial water flux of ~21 L m-2 h-1.
138
After the water flux became stable, sterile sodium alginate stock solution (10 g/L) was
139
introduced to obtain a feed solution with 250 mg/L of alginate for the start of the accelerated
140
fouling experiment. The experiments were conducted for 15–20 h, until a cumulative permeate
141
volume of 500 mL was collected. Water flux throughout the experiment was monitored by a
142
computer, which recorded the mass of the draw solution at one minute intervals. Water flux data
143
were corrected to eliminate the contribution of draw solution dilution to flux decline;15 data -7ACS Paragon Plus Environment
Environmental Science & Technology
144
below reflect the flux decrease due solely to membrane fouling. Physical cleaning was also
145
performed at the end of fouling experiments to evaluate fouling reversibility. The details of the
146
physical cleaning methodology can be found in the SI.
147
AFM Adhesion Force Measurements. Adhesion force measurements were
148
performed in a multimode AFM (Bruker, Santa Barbara, CA) operating in contact mode using
149
SiN cantilevers (Bruker NP-O10, Santa Barbara, CA; spring constant 0.06 N/m). Cantilevers
150
were inspected under an optical microscope for breaks and cracks before use, and cleaned in a
151
UV/Ozone cleaner for 20 minutes (BioForce Nanosciences, Ames, IA). A 4.0-µm carboxyl-
152
modified latex particle (CML, carboxyl content 19.5 µeq/g, Life Technologies, Eugene, OR) was
153
glued to the tip of the cantilever using UV-curable adhesive (Norland Optical Adhesive 68,
154
Norland Products, Cranbury, NJ), and subsequently cured for 20 minutes in the UV/Ozone
155
cleaner. In order to investigate foulant−membrane interactions, the functionalized cantilever was
156
immersed in a 4.0 g/L alginate solution for at least 16 hours prior to the measurements. During
157
this period of time, alginate molecules adsorbed on the surface of the colloidal particle. Force
158
measurements were collected at a rate of 0.5 Hz and a resolution of 512 samples per line. The
159
deflection sensitivity, in nm/V, was determined from the slope of the compliance region. All
160
force curves were collected in relative trigger mode at a deflection of 100 nm. To avoid large
161
variations between cantilever spring constants, all measurements were performed with the same
162
cantilever, immersing the tip in fresh 4.0 g/L alginate for at least 16 hours between
163
measurements. Adhesion forces were determined by converting curves of cantilever deflection vs.
164
piezoelectric stage retraction to force vs. particle−membrane separation.28 All measurements
165
were performed in a liquid cell filled with ~2 mL of synthetic wastewater with the same solution
166
composition as indicated above. Adequate sampling was ensured by collecting measurements in
167
5 randomly chosen locations on each surface. The total number of measurements for each
168
membrane type was ~200.
169
Characterization of Membrane Structural and Material Properties. The
170
surface hydrophilicity of the TFC membranes was evaluated by measuring the contact angle for
171
DI water using the sessile drop method. A 1-µL droplet was placed on the air-dried membrane
172
surface, and contact angles were measured after 10 s (VCA, Optima XE, AST Products, Billerica,
173
MA). For each type of membrane to be characterized, we performed twelve droplet -8ACS Paragon Plus Environment
Page 8 of 26
Page 9 of 26
174
Environmental Science & Technology
measurements on each of three independently cast membrane coupons.
175
The membrane surface roughness was characterized by a multimode atomic force
176
microscope (AFM, Bruker, Santa Barbara, CA). Imaging of the air-dried samples was performed
177
in tapping mode with silicon probes coated with 30-nm-thick back side aluminum (Tap300A,
178
Bruker Nano, Inc., Camarillo, CA). The probe had a spring constant of 40 N/m, resonance
179
frequency of 300 kHz, tip radius of 8 ± 4 nm, and cantilever length of 125 ± 10 µm.
180
Membrane surface morphology was observed through a field emission scanning electron
181
microscope (FE-SEM, SU-70, Hitachi, Japan). Membrane samples were air-dried overnight prior
182
to the measurements and sputter-coated (DESK V, Denton Vacuum, LLC, Moorestown, NJ) with
183
a 10-nm-thick layer of chromium. Micrographs of support layer surfaces were quantitatively
184
analyzed for porosity and selected pore metrics using ImageJ 1.46r software (National Institutes
185
of Health, Bethesda, Maryland, USA).
186
The surface carboxyl group density of TFC membrane polyamide surfaces was quantified
187
via the TBO technique developed by Tiraferri et al.29 Briefly, the support surface of the TFC
188
membrane was sealed with waterproof tape to leave only the active layer exposed. Then, the
189
active layer of the TFC membrane was contacted with a freshly-prepared solution of TBO (2
190
mM) and NaOH (pH 11) to bind positively-charged TBO molecules to deprotonated carboxylic
191
acid groups on the polyamide surface. After thorough rinsing with a dye-free NaOH solution (pH
192
11) to remove unbound dye molecules, the membrane coupon was immersed into a NaCl
193
solution at pH 2 to elute the bonded TBO dye from the polyamide surface. The absorbance of the
194
eluent was measured at a 630 nm wavelength to determine the surface carboxyl group density.
195 196
RESULTS AND DISCUSSION
197
Membrane Intrinsic Transport Properties. The hand-cast TFC FO membranes were
198
first characterized for intrinsic transport properties (Table 1). The water permeability and salt
199
permeability are 1.73 ± 0.33 L m-2 h-1 bar-1 and 0.50 ± 0.11 L m-2 h-1, respectively, for the TFC
200
membrane cast on the polysulfone support for which NMP was the solvent used during phase
201
separation (NMP-TFC membrane). The TFC membrane cast on polysulfone support formed
202
using DMF as the solvent (DMF-TFC membrane) has higher values of water permeability and
203
salt permeability (3.14 ± 0.02 L m-2 h-1 bar-1 and 0.60 ± 0.31 L m-2 h-1, respectively), indicating a -9ACS Paragon Plus Environment
Environmental Science & Technology
204
more permeable active layer for the DMF-TFC membrane. The structural parameter, S, which is
205
an intrinsic property of the polysulfone support, was comparable for the NMP-TFC membrane
206
and the DMF-TFC membrane (817 ± 146 µm vs. 953 ± 2 µm), indicating the severity of internal
207
concentration polarization is similar.
Page 10 of 26
TABLE 1
208 209
Fouling Behavior of FO-TFC Membranes. Alginate, a model polysaccharide,
210
was chosen to evaluate the fouling propensity of the membranes in order to mimic extracellular
211
polymeric substances commonly present in wastewater effluents. Representative fouling curves
212
for the NMP-TFC and DMF-TFC membranes are presented in Figure 1A, and the summarized
213
flux decline data with duplicate experiments can be found in the SI (Figure S1 and Table S1).
214
FIGURE 1
215
FIGURE S1
216
TABLE S1
217
Generally, the loss of water flux due to fouling was relatively small for both membranes
218
(Figure 1A), considering the extremely high foulant concentrations (250 mg/L) used. This
219
observation is consistent with previous FO fouling studies, which demonstrated low fouling
220
propensity of FO membranes.30, 31 The NMP-TFC membranes exhibited significantly smaller
221
water flux decline (7.47 ± 0.15%) than the DMF-TFC membranes (12.7 ± 2.62%). Physical
222
cleaning also led to greater recovery of water flux for the NMP-TFC membrane (Figure S1 and
223
Table S1). Both of these results indicate a lower fouling propensity for the NMP-TFC membrane
224
than for the DMF-TFC membrane.
225
Foulant–Membrane Interaction Forces. The observed fouling propensities of the
226
membranes were in agreement with the corresponding foulant–membrane interaction forces.15, 32
227
The distribution of probe–membrane adhesion forces for the NMP-TFC membrane (Figure 2A)
228
was shifted toward the right (i.e., lower adhesion forces) relative to that of the DMF-TFC
229
membrane (Figure 2B). This resulted in a relative lower mean adhesion force of −0.38 ± 0.31
230
mN m-1 for the NMP-TFC membrane compared to that of −0.60 ± 0.34 mN m-1 for the DMF-
231
TFC membrane. In addition, a significantly higher fraction of the measurements performed on
232
the NMP-TFC membrane (22.6% vs. 3.3% for the DMF-TFC membrane) exhibited no detectable - 10 ACS Paragon Plus Environment
Page 11 of 26
Environmental Science & Technology
233
adhesion force (counted in the “NO” column), implying the absence of adhesion probe–
234
membrane interactions. As shown in Figure 2C, D, the DMF-TFC membrane had more “sticky”
235
sites, indicated by more frequent measurement (50%) of rupture distances larger than 100 nm
236
(compared to 20% for the NMP-TFC membrane). Accordingly, the average rupture distance for
237
the DMF-TFC membrane (90.8 nm) was nearly twice the value observed for the NMP-TFC
238
membrane (54.9 nm). These AFM characterizations all indicate that the NMP-TFC membrane
239
exhibits a lower fouling propensity than the DMF-TFC membrane, in agreement with our
240
observed fouling trend.
241
FIGURE 2
242
Determining the Causes for Different Fouling Propensities. In the above
243
sections, we observed a significant difference in the fouling propensity of the two TFC
244
membranes. Possible causes for the observed difference are systematically analyzed in this
245
section. Our analysis is organized according to general mechanisms through which the properties
246
of our membranes and experiments might have influenced the overall fouling result. For each of
247
these mechanisms, we considered how specific characteristics of the membranes would be
248
involved and selected methods to capture those possibilities. The particular circumstances of our
249
fouling experiments were also used to rule out certain factors. In this way, we were able to
250
identify a set of techniques that would achieve a comprehensive evaluation of fouling propensity
251
in an efficient manner. An example of our selection process appears in Supporting Information
252
(S6).
253
(a) Is it reverse solute diffusion? In FO, reverse diffusion of draw solute from the
254
draw solution to the feed solution causes an accumulation of salts at the active layer–feed
255
solution interface. This phenomenon occurs in addition to the concentration of feed solutes due
256
to external concentration polarization. The increase in salt concentration reduces the effective
257
osmotic pressure driving force, resulting in water flux decline. Foulants that accumulate on the
258
active layer surface exacerbate these effects through cake-enhanced osmotic pressure (CEOP).30
259
If the NMP-TFC membrane permitted less reverse diffusion of draw solutes than the DMF-TFC
260
membrane, this could explain why its decline in water flux was less than that of the DMF-TFC
261
membrane.
262
Because the two TFC membranes have different transport properties, different initial draw - 11 ACS Paragon Plus Environment
Environmental Science & Technology
263
solution concentrations were required to reach the same initial water flux (~21 L m-2 h-1). To
264
analyze the effect of reverse solute diffusion on fouling, we determined the reverse solute flux
265
selectivity33:
266
Jw A = nRgT Js B
Page 12 of 26
(1)
267
where Jw is water flux, Js is the reverse solute flux, A and B are water and salt permeability
268
coefficients, respectively, n is the number of dissolved species created by the draw solute (2 for
269
NaCl), Rg is the ideal gas constant, and T is the absolute temperature. The reverse solute flux
270
selectivity is independent of draw solution concentration, and can be obtained simply from the
271
ratio A/B. The calculated A/B values (Table 1) are 3.47 ± 0.43 bar-1 for the NMP-TFC and 6.16 ±
272
2.18 bar-1 for the DMF-TFC membrane. Given the same initial water flux (~21 L m-2 h-1) for
273
both types of membranes in the fouling experiments, a lower A/B value implies greater reverse
274
solute diffusion, Js, for the NMP-TFC membrane, which would cause greater water flux decline.
275
However, the decline in water flux was less severe for the NMP-TFC membrane in our fouling
276
experiments, which rules out this mechanism as the cause for the different fouling propensities.
277
(b) Is it disruption of calcium–alginate complexes by sodium? The calcium ions
278
in our simulated wastewater feed solution facilitate bridging and complexation between carboxyl
279
groups on the polyamide surface and alginate molecules.34-36 The gel network that consequently
280
forms on the active layer could impede water flux. At the same time, sodium ions could disrupt
281
calcium bridges between foulant molecules through cation exchange.37-39 Given the higher
282
reverse solute diffusion for the NMP-TFC membrane, and our use of NaCl as the draw solute,
283
the lower fouling propensity of this membrane could be attributed to greater disruption of
284
calcium bridging and complexation by the sodium ions. To further evaluate this possibility, we
285
repeated our fouling experiments with MgCl2 as the draw solute, which does not have the cation
286
exchange capability of NaCl. The fouling behaviors of the TFC membranes evaluated with
287
MgCl2 draw solution are presented in the Figure 1B. Even without the influx of sodium ions
288
from the draw solution to the foulant layer, the NMP-TFC membrane still exhibited a lower flux
289
decline (78.8 ± 5.7%) than the DMF-TFC membrane (66.5 ± 4.5%). Therefore, the lower fouling
290
propensity of the NMP-TFC membrane cannot be attributed to differences in the degree of cation
291
exchange occurring between sodium and calcium ions. - 12 ACS Paragon Plus Environment
Page 13 of 26
Environmental Science & Technology
292
(c) Is it membrane surface hydrophilicity or chemistry? Surface energy and
293
surface chemistry strongly influence the interaction of organic foulants with the membrane
294
surface. The layer of strongly-bound water molecules that forms on a hydrophilic surface
295
through hydrogen bonding provides an enthalpic penalty for foulant adhesion that lowers fouling
296
propensity.40,
297
through the availability of sites for calcium–alginate complexation with the surface.42,
298
determine whether differences in surface hydrophilicity or chemistry might explain the observed
299
difference in fouling propensity, we measured water contact angle and the density of native
300
carboxyl groups on the polyamide surfaces (Table 1). Although identical interfacial
301
polymerization procedures were performed on both membrane types, the DMF-TFC membrane
302
had lower value (P-value < 0.001) in water contact angle (55.7 ± 6.6° vs. 66.4 ± 2.3°,
303
respectively) and comparable carboxyl group density (18.96 ± 3.66 nm-2 vs. 20.24 ± 6.17 nm-2)
304
compared to the NMP-TFC membrane. This would suggest less propensity for fouling due to
305
greater hydrophilicity and fewer sites for carboxyl-group-facilitated foulant attachment. However,
306
the observed overall fouling propensity is higher for the DMF-TFC membrane, thereby ruling
307
out both surface hydrophilicity and surface chemistry explanations for the observed fouling
308
behavior.
41
The density of carboxyl groups on a surface also affects fouling propensity 43
To
309
(d) Is it membrane surface structure? During interfacial polymerization, MPD
310
diffuses into the organic phase, where it reacts with TMC to form the polyamide layer with its
311
characteristic ridge-and-valley structure.4, 44 The inherent roughness of this polyamide structure
312
can enhance membrane fouling through (i) greater surface area for foulant attachment and (ii)
313
accumulation of foulants in valley features that can hinder their removal during physical
314
cleaning18. We used AFM to quantify membrane surface roughness (Figure 3) and SEM to
315
inspect surface morphology (Figure 4).
316
FIGURE 3
317
FIGURE 4
318
The NMP-TFC membrane had a root-mean-square roughness (Rrms) of 94.41 ± 3.59 nm, an
319
average roughness (Ra) of 74.64 ± 3.22 nm, and a maximum roughness (Rmax) of 696.42 ± 34.84
320
nm. These values are comparable to those of membranes fabricated under similar conditions.31, 45
321
The surface roughness of the DMF-TFC membrane was higher (P-value < 0.001) than that of the - 13 ACS Paragon Plus Environment
Environmental Science & Technology
322
NMP-TFC membrane (Figure 3A), with Rrms, of 124.39 ± 5.68 nm, Ra of 98.19 ± 3.79 nm, and
323
Rmax of 969.64 ± 144.53 nm. This is the first comparison between the membranes for which the
324
observed fouling trend agrees with what we would predict from a specific surface characteristic.
Page 14 of 26
325
SEM micrographs of the NMP-TFC and DMF-TFC membrane polyamide surfaces show
326
the expected ridge-and-valley structure (Figure 4A, B, additional replicate in SI, Figure S3).
327
These micrographs were obtained with a 45° stage tilt in order to supplement the top-down
328
perspective provided by SEM. The surface of the NMP-TFC membrane contained many small,
329
nodular features, while the DMF-TFC membrane surface included some larger leaf-like
330
structures (these unique features are indicated by yellow arrows in Figure 4A, B). These leaf-like
331
structures overhang the membrane surface, potentially providing regions where foulants can
332
accumulate and be sheltered from the shear forces produced during physical cleaning.18 The
333
difficulty of quantifying the morphology of the polyamide layer prevents us from conclusively
334
tying this factor to our observed fouling propensity, as we did for roughness. However, we can
335
recognize the possibility that the larger overhanging features of the DMF-TFC polyamide layer
336
might make the build-up of foulants on the DMF-TFC membranes more rapid and less reversible,
337
as observed in our fouling experiments. If this mechanism is at work as we propose, its effects on
338
fouling are combined with those associated with roughness differences to determine the overall
339
fouling propensity.
340
Relating Polyamide Surface Properties to Support Layer Structure. Our
341
finding that subtle changes to the support layer affect polyamide structure in a way that
342
significantly alters fouling propensity warrants a closer look at the support layer itself. Below we
343
employ further characterization of the support layers to examine the current understanding of the
344
role of support layer characteristics on polyamide formation and foulant accumulation on the
345
membrane surface.
346
(a) Origin of polyamide layer structure. Figure 3B depicts surface roughness of the
347
polysulfone supports, which is much lower than the roughness of the polyamide surfaces (Figure
348
3A). Slightly lower values in Rrms, Ra, and Rmax of the DMF-polysulfone indicate that this support
349
is smoother than the NMP-polysulfone support. We note that the converse is observed for
350
polyamide roughness, i.e., the DMF-TFC polyamide surface is rougher than the NMP-TFC
351
polyamide surface. These results indicate that the high roughness features of the polyamide - 14 ACS Paragon Plus Environment
Page 15 of 26
352 353
Environmental Science & Technology
active layers are neither a translation nor simple amplification of the support layer roughness. High magnification SEM micrographs of the polyamide and polysulfone surfaces were also
354
compared (Figure 4, middle and bottom rows). Previous studies23,
46
355
individual ridge (or nodule) in the polyamide layer arises from one underlying pore in the
356
support layer; we referred to it herein as the “one pore – one ridge” model. This proposed
357
mechanism for polyamide roughness formation was based on an observed trend between pore
358
size and polyamide feature size. While we note the same trend — that the NMP-TFC membrane
359
has both smaller polysulfone surface pores and smaller polyamide features — when micrographs
360
of the polyamide and polysulfone surfaces are compared side-by-side at equal magnification, it
361
becomes very clear that the one pore – one ridge proposed model is not consistent with our SEM
362
micrographs. The polyamide features (Figure 4C, D) are at least an order of magnitude larger
363
than the pores on the polysulfone support (Figure 4E, F), and the base of each polyamide ridge
364
covers an area occupied by many polysulfone pores ( >10 ).
speculated that each
365
(b) Importance for active layer fouling behavior. A recent modeling effort26
366
proposed that a more permeable support might contribute to a more uniform spatial distribution
367
of water permeation across the TFC membrane surface, resulting in a surface with lower fouling
368
propensity. Conversely, a less permeable support would constrain water transport to occur over a
369
reduced effective area, through scattered points with relatively high local water fluxes, referred
370
to as hot spots. Such a membrane would have a higher fouling propensity.
371
In our experiments, we found that the support layer of the lower fouling propensity
372
membrane (NMP-TFC) did have a higher water permeability (Table 1, 1107 ± 193 vs. 290 ± 40
373
L m-2 h-1 bar-1 for the DMF-TFC support), in basic agreement with the hot-spots theory. However,
374
further analysis (SI, section S4) showed that overall water permeability of the support layers was
375
primarily determined by the bulk structure rather than by the dense skin layer. This indicated that
376
the overall support permeability might not accurately reflect water transport through the dense
377
skin layer and adjacent polyamide layer. We therefore turned to more direct characterizations of
378
the support layer surface to inform the comparison of our fouling observations with the proposed
379
hot-spots mechanism.
380
The surfaces of the support layers had comparable pore density (Table 1), but the DMF-
381
TFC membrane support surface had a larger pore size (as shown in the Figure 4E, F and Figure - 15 ACS Paragon Plus Environment
Environmental Science & Technology
Page 16 of 26
382
S2) and higher surface porosity (1.99 ± 0.60% vs. 1.01 ± 0.60%, Table 1). Therefore, the DMF-
383
TFC membrane would have a greater amount of area available for water permeation at the
384
polysulfone–polyamide interface. Fouling propensity would consequently be expected to be
385
lower for the DMF-TFC membrane, while we observed the opposite. This finding does not
386
negate the plausibility of the ‘hot-spots' theory. Rather, it shows that, if the theory is correct, the
387
effect of polysulfone surface pore size and porosity on polyamide layer fouling in our
388
membranes was overcome by more influential differences in the structure of the polyamide
389
layers.
390
Implications for TFC Membrane Fabrication. Much research effort over the last
391
few years has focused on obtaining a support layer with a structure that minimizes internal
392
concentration polarization in forward osmosis. Our work demonstrates that changes made to the
393
support layer may have unintended and unexamined effects on fouling propensity for TFC
394
membranes. Our comparison of just two slightly different membranes does not support
395
development of a comprehensive model to quantitatively relate support layer structure or
396
fabrication choices to the fouling propensity ultimately observed on an overlying active layer.
397
However, it does highlight the importance of evaluating supposed improvements to support
398
layers for their impact on fouling propensity or on factors known to influence fouling propensity,
399
like those we have presented here.
400 401
ACKNOWLEDGMENTS
402
Financial support from the Department of Defense through the Strategic Environmental
403
Research and Development Program (SERDP, Project No. 12 ER01-054/ER-2217) and the
404
National Science and Technology Pillar Program of China (Project No. 2012BAC05B02) is
405
gratefully acknowledged. We also acknowledge the use of SEM and AFM facilities supported by
406
the Yale Institute for Nanoscience and Quantum Engineering (YINQE) under NSF MRSEC
407
DMR 1119826. This publication was developed under a graduate fellowship awarded by the
408
China Scholarship Council (CSC) to Xinglin Lu and STAR Fellowship (Agreement No. FP-
409
91733801-0) awarded by the U.S. Environmental Protection Agency (EPA) to L.H. Arias
410
Chavez. It has not been reviewed or endorsed by EPA.
- 16 ACS Paragon Plus Environment
Page 17 of 26
Environmental Science & Technology
411 412
SUPPORTING INFORMATION
413
Details on the evaluation of TFC membrane intrinsic properties (S1); evaluation of polysulfone
414
support properties (S2); physical cleaning methodology (S3); polysulfone support thickness
415
measurements (S4); analysis on support layer structure and resistance to water transport (S5);
416
Selection or Exclusion of Potential Techniques for Evaluating Fouling Propensity (S6); FO
417
fouling and cleaning experiment results (Figure S1 and Table S1); pore diameter distribution of
418
support surfaces (Figure S2); and replicate SEM micrographs of polyamide layers (Figure S3).
419
This material is available free of charge via the Internet at http://pubs.acs.org.
420 421
- 17 ACS Paragon Plus Environment
Environmental Science & Technology
Page 18 of 26
422
REFERENCES
423 424 425
(1) Shannon, M. A.; Bohn, P. W.; Elimelech, M.; Georgiadis, J. G.; Marinas, B. J.; Mayes, A. M., Science and technology for water purification in the coming decades. Nature 2008, 452, (7185), 301-310.
426 427
(2) Elimelech, M.; Phillip, W. A., The Future of Seawater Desalination: Energy, Technology, and the Environment. Science 2011, 333, (6043), 712-717.
428 429 430
(3) Hoover, L. A.; Phillip, W. A.; Tiraferri, A.; Yip, N. Y.; Elimelech, M., Forward with osmosis: emerging applications for greater sustainability. Environmental Science & Technology 2011, 45, (23), 9824-30.
431 432
(4) Petersen, R. J., Composite Reverse-Osmosis and Nanofiltration Membranes. Journal of Membrane Science 1993, 83, (1), 81-150.
433 434 435
(5) Yip, N. Y.; Tiraferri, A.; Phillip, W. A.; Schiffman, J. D.; Elimelech, M., High Performance Thin-Film Composite Forward Osmosis Membrane. Environmental Science & Technology 2010, 44, (10), 3812-3818.
436 437 438
(6) Wei, J.; Qiu, C.; Tang, C. Y.; Wang, R.; Fane, A. G., Synthesis and characterization of flat-sheet thin film composite forward osmosis membranes. Journal of Membrane Science 2011, 372, (1–2), 292-302.
439 440 441
(7) Wang, R.; Shi, L.; Tang, C. Y.; Chou, S.; Qiu, C.; Fane, A. G., Characterization of novel forward osmosis hollow fiber membranes. Journal of Membrane Science 2010, 355, (1-2), 158167.
442 443
(8) Cath, T. Y.; Childress, A. E.; Elimelech, M., Forward osmosis: Principles, applications, and recent developments. Journal of Membrane Science 2006, 281, (1-2), 70-87.
444 445 446
(9) Kumar, M.; Adham, S. S.; Pearce, W. R., Investigation of Seawater Reverse Osmosis Fouling and Its Relationship To Pretreatment Type. Environmental Science & Technology 2006, 40, (6), 2037-2044.
447 448 449
(10) Tang, C. Y.; Kwon, Y.-N.; Leckie, J. O., Fouling of reverse osmosis and nanofiltration membranes by humic acid—Effects of solution composition and hydrodynamic conditions. Journal of Membrane Science 2007, 290, (1–2), 86-94.
450 451 452
(11) Xu, P.; Bellona, C.; Drewes, J. E., Fouling of nanofiltration and reverse osmosis membranes during municipal wastewater reclamation: Membrane autopsy results from pilotscale investigations. Journal of Membrane Science 2010, 353, (1–2), 111-121.
453 454
(12) Ang, W. S.; Elimelech, M., Protein (BSA) fouling of reverse osmosis membranes: Implications for wastewater reclamation. Journal of Membrane Science 2007, 296, (1-2), 83-92.
455 456 457
(13) Li, Q.; Xu, Z.; Pinnau, I., Fouling of reverse osmosis membranes by biopolymers in wastewater secondary effluent: Role of membrane surface properties and initial permeate flux. Journal of Membrane Science 2007, 290, (1–2), 173-181.
458 459
(14) Rana, D.; Matsuura, T., Surface Modifications for Antifouling Membranes. Chemical Reviews 2010, 110, (4), 2448-2471.
460
(15)
Lu, X.; Romero-Vargas Castrillón, S.; Shaffer, D. L.; Ma, J.; Elimelech, M., In Situ - 18 ACS Paragon Plus Environment
Page 19 of 26
Environmental Science & Technology
461 462 463
Surface Chemical Modification of Thin-Film Composite Forward Osmosis Membranes for Enhanced Organic Fouling Resistance. Environmental Science & Technology 2013, 47, (21), 12219-12228.
464 465 466
(16) Yu, H.-Y.; Kang, Y.; Liu, Y.; Mi, B., Grafting polyzwitterions onto polyamide by click chemistry and nucleophilic substitution on nitrogen: A novel approach to enhance membrane fouling resistance. Journal of Membrane Science 2014, 449, (0), 50-57.
467 468 469
(17) Romero-Vargas Castrillón, S.; Lu, X.; Shaffer, D. L.; Elimelech, M., Amine enrichment and poly(ethylene glycol) (PEG) surface modification of thin-film composite forward osmosis membranes for organic fouling control. Journal of Membrane Science 2014, 450, 331-339.
470 471 472
(18) Vrijenhoek, E. M.; Hong, S.; Elimelech, M., Influence of membrane surface properties on initial rate of colloidal fouling of reverse osmosis and nanofiltration membranes. Journal of Membrane Science 2001, 188, (1), 115-128.
473 474 475
(19) Hobbs, C.; Hong, S. K.; Taylor, J., Effect of surface roughness on fouling of RO and NF membranes during filtration of a high organic surficial groundwater. Journal of Water Supply Research and Technology-Aqua 2006, 55, (7-8), 559-570.
476 477 478
(20) Gu, Y.; Wang, Y.-N.; Wei, J.; Tang, C. Y., Organic fouling of thin-film composite polyamide and cellulose triacetate forward osmosis membranes by oppositely charged macromolecules. Water Research 2013, 47, (5), 1867-1874.
479 480 481 482
(21) Singh, P. S.; Joshi, S. V.; Trivedi, J. J.; Devmurari, C. V.; Rao, A. P.; Ghosh, P. K., Probing the structural variations of thin film composite RO membranes obtained by coating polyamide over polysulfone membranes of different pore dimensions. Journal of Membrane Science 2006, 278, (1-2), 19-25.
483 484 485
(22) Il Kim, H.; Kim, S. S., Plasma treatment of polypropylene and polysulfone supports for thin film composite reverse osmosis membrane. Journal of Membrane Science 2006, 286, (1-2), 193-201.
486 487 488
(23) Ghosh, A. K.; Hoek, E. M. V., Impacts of support membrane structure and chemistry on polyamide-polysulfone interfacial composite membranes. Journal of Membrane Science 2009, 336, (1-2), 140-148.
489 490 491
(24) Tiraferri, A.; Yip, N. Y.; Phillip, W. A.; Schiffman, J. D.; Elimelech, M., Relating performance of thin-film composite forward osmosis membranes to support layer formation and structure. Journal of Membrane Science 2011, 367, (1-2), 340-352.
492 493
(25) Hoek, E. M. V.; Bhattacharjee, S.; Elimelech, M., Effect of membrane surface roughness on colloid-membrane DLVO interactions. Langmuir 2003, 19, (11), 4836-4847.
494 495 496
(26) Ramon, G. Z.; Hoek, E. M. V., Transport through composite membranes, part 2: Impacts of roughness on permeability and fouling. Journal of Membrane Science 2013, 425–426, 141148.
497
(27)
498 499
(28) Senden, T. J., Force microscopy and surface interactions. Current Opinion in Colloid & Interface Science 2001, 6, (2), 95-101.
500
(29)
Ben-Asher, J., Irrigation with saline water. GeoJournal 1987, 15, (3), 267-272.
Tiraferri, A.; Elimelech, M., Direct quantification of negatively charged functional - 19 ACS Paragon Plus Environment
Environmental Science & Technology
Page 20 of 26
501
groups on membrane surfaces. Journal of Membrane Science 2012, 389, 499-508.
502 503
(30) Lee, S.; Boo, C.; Elimelech, M.; Hong, S., Comparison of fouling behavior in forward osmosis (FO) and reverse osmosis (RO). Journal of Membrane Science 2010, 365, (1-2), 34-39.
504 505 506
(31) Tiraferri, A.; Kang, Y.; Giannelis, E. P.; Elimelech, M., Highly Hydrophilic Thin-Film Composite Forward Osmosis Membranes Functionalized with Surface-Tailored Nanoparticles. Acs Applied Materials & Interfaces 2012, 4, (9), 5044-5053.
507 508 509
(32) Li, Q.; Elimelech, M., Organic Fouling and Chemical Cleaning of Nanofiltration Membranes: Measurements and Mechanisms. Environmental Science & Technology 2004, 38, (17), 4683-4693.
510 511 512
(33) Phillip, W. A.; Yong, J. S.; Elimelech, M., Reverse Draw Solute Permeation in Forward Osmosis: Modeling and Experiments. Environmental Science & Technology 2010, 44, (13), 5170-5176.
513 514
(34) Lee, S.; Elimelech, M., Relating organic fouling of reverse osmosis membranes to intermolecular adhesion forces. Environmental Science & Technology 2006, 40, (3), 980-987.
515 516 517
(35) Liu, Y.; Mi, B., Combined fouling of forward osmosis membranes: Synergistic foulant interaction and direct observation of fouling layer formation. Journal of Membrane Science 2012, 407–408, 136-144.
518 519 520
(36) Xiang, Y.; Liu, Y.; Mi, B.; Leng, Y., Molecular Dynamics Simulations of Polyamide Membrane, Calcium Alginate Gel, and Their Interactions in Aqueous Solution. Langmuir 2014, 30, (30), 9098-9106.
521 522
(37) Lee, S.; Elimelech, M., Salt cleaning of organic-fouled reverse osmosis membranes. Water Research 2007, 41, (5), 1134-1142.
523 524
(38) Skjåk-Bræk, G.; Grasdalen, H.; Smidsrød, O., Inhomogeneous polysaccharide ionic gels. Carbohydrate Polymers 1989, 10, (1), 31-54.
525 526
(39) Matsumoto, T.; Mashiko, K., Viscoelastic properties of alginate aqueous solutions in the presence of salts. Biopolymers 1990, 29, (14), 1707-1713.
527 528
(40) van Oss, C. J.; Chaudhury, M. K.; Good, R. J., Monopolar Surfaces. Advances in Colloid and Interface Science 1987, 28, (1), 35-64.
529 530 531
(41) van Oss, C. J.; Wu, W.; Docoslis, A.; Giese, R. F., The interfacial tensions with water and the Lewis acid-base surface tension parameters of polar organic liquids derived from their aqueous solubilities. Colloids and Surfaces B-Biointerfaces 2001, 20, (1), 87-91.
532 533 534
(42) Jin, X.; Huang, X. F.; Hoek, E. M. V., Role of Specific Ion Interactions in Seawater RO Membrane Fouling by Alginic Acid. Environmental Science & Technology 2009, 43, (10), 35803587.
535 536 537
(43) Wu, J.; Contreras, A. E.; Li, Q., Studying the impact of RO membrane surface functional groups on alginate fouling in seawater desalination. Journal of Membrane Science 2014, 458, 120-127.
538 539
(44) Freger, V., Kinetics of film formation by interfacial polycondensation. Langmuir 2005, 21, (5), 1884-1894.
540
(45)
Tang, C. Y. Y.; Kwon, Y. N.; Leckie, J. O., Effect of membrane chemistry and coating - 20 ACS Paragon Plus Environment
Page 21 of 26
Environmental Science & Technology
541 542 543
layer on physiochemical properties of thin film composite polyamide RO and NF membranes II. Membrane physiochemical properties and their dependence on polyamide and coating layers. Desalination 2009, 242, (1-3), 168-182.
544 545 546 547
(46) Klaysom, C.; Hermans, S.; Gahlaut, A.; Van Craenenbroeck, S.; Vankelecom, I. F. J., Polyamide/Polyacrylonitrile (PA/PAN) thin film composite osmosis membranes: Film optimization, characterization and performance evaluation. Journal of Membrane Science 2013, 445, 25-33.
548 549
- 21 ACS Paragon Plus Environment
Environmental Science & Technology
Page 22 of 26
550
Table 1. Intrinsic properties, salt rejection, and surface characteristics of the fabricated TFC FO
551
membranes.
NMP
DMF
1.73 ± 0.33 0.50 ± 0.11 3.47 ± 0.43 817 ± 146 95.94 ± 0.60 98.07 ± 1.56
3.14 ± 0.02 0.60 ± 0.31 6.16 ± 2.68 953 ± 2 98.18 ± 0.53 98.84 ± 0.89
TFC Membrane A a (L m-2 h-1 bar-1) B b (L m-2 h-1) A/B (bar-1) S c (µm) R (NaCl) d (%) R (MgCl2) e (%)
Active Layer Contact Angle f (°) Carboxyl group density g (nm-2)
66.4 ± 2.3 20.24 ± 6.17
55.7 ± 6.6 18.96 ± 3.66
Support Layer Water permeability h (L m-2 h-1 bar-1) Surface pore area i (nm2) Nominal surface pore diameter i (nm) Surface porosity i (%) Surface pore number density i (103 µm-2) t j (µm) ε/τ k (-)
1107 ± 193 37.22 ± 53.39 6.0 1.01 ± 0.40 26 ± 8 130.0 ± 7.4 0.159 ± 0.009
290 ± 40 70.55 ± 108.71 8.0 1.99 ± 0.60 28 ± 6 90.9 ± 4.5 0.095 ± 0.005
Values are presented as mean ± one standard deviation. a Water permeability, determined by RO experiment with DI water at 25 °C and 27.6 bar with three coupons cut from three independently cast membranes. b Salt permeability, measured by RO experiment with NaCl (50 mM) at 25 °C and 27.6 bar. c Structural parameter, determined in the FO experiment with 1 M NaCl as draw solution and DI water as feed solution. d Salt rejection, measured by RO experiment with 50 mM NaCl at 25 °C and 27.6 bar. e Salt rejection, measured by RO experiment with 50 mM MgCl2 at 25 °C and 27.6 bar. f Contact angle measurements performed with DI water at room temperature (23 °C). g Measured by the TBO method. h Measured by ultrafiltration with DI water at 25 °C and 0.35 bar. i Determined through SEM images analysis by ImageJ software. j Micrometermeasured thickness. k Porosity/tortuosity, obtained by dividing S by t.
- 22 ACS Paragon Plus Environment
Page 23 of 26
Environmental Science & Technology
Normarlized Water Flux Jw/Jw,0(%)
NaCl Draw Solution NMP DMF
100 95 90 85
A 0
100
200
300
400
500
Cumulative Permeate Volume (mL) MgCl2 Draw Solution Normarlized Water Flux Jw/Jw,0(%)
100
NMP DMF
95 90 85 80 75 70
B
65 0
100
200
300
400
500
Cumulative Permeate Volume (mL)
552
Figure 1. FO alginate fouling results with (A) NaCl and (B) MgCl2 draw solutions. Fouling
553
conditions were as follows: feed solution chemistry simulating domestic wastewater (pH 7.4) as
554
described in Materials and Methods, supplemented with 250 mg/L alginate as model organic
555
foulant; draw solutions of (A) 2–4 M NaCl and (B) 2–3.5 M MgCl2, resulting in an initial
556
permeate water flux of ~21 L m-2 h-1. The system temperature was maintained at 25 °C, and the
557
crossflow velocity of the feed and draw streams was set to 8.5 cm/s. To reduce experimental
558
noise, data were smoothed using a 5-point window moving average. Note that y-axis scales differ.
- 23 ACS Paragon Plus Environment
Environmental Science & Technology
A
-1 40 Mean Force = -0.38 ± 0.31 mN m
B
Frequency Count
Frequency Count
NMP
30 20 10 0
-2.5 -2.0 -1.5 -1.0 -0.5
0.0
NMP 25
Mean Rupture Distance = 54.9 ± 54.9 nm
20 15 10
0
NO
D
DMF 40 Mean Force = -0.60 ± 0.34 mN m
-1
Frequency Count
Frequency Count
30
5
50
30 20 10 0
559
C
50
Page 24 of 26
0
50
100
150
200
250
300
250
300
30
DMF 25
Mean Rupture Distance = 90.8 ± 54.5 nm
20 15 10 5
-2.5 -2.0 -1.5 -1.0 -0.5
0.0
0
NO
0
50
100
150
200
Rupture Distance (nm)
Adhesion Force (mN/m)
560 561
Figure 2. Distribution of foulant–membrane adhesion forces and rupture distance for active
562
layers formed on polysulfone support using different solvents: NMP (upper row) and DMF
563
(lower row). Measurements were performed by contact mode AFM at room temperature in a
564
liquid cell filled with synthetic wastewater (pH 7.4). At least 125 retraction force measurements
565
distributed over five randomly chosen locations were performed for each sample. The columns
566
labeled ‘NO’ indicate the fraction of the population of force measurements for which no
567
adhesion was observed.
568
- 24 ACS Paragon Plus Environment
Page 25 of 26
Environmental Science & Technology
140 130 120 110 100 90 80 70 10 0
A
NMP-TFC DMF-TFC
Rrms
Ra
30
Roughness (nm)
Roughness (nm)
569
25
NMP-PSf DMF-PSf
20 15 10 5 0
Rmax/10
B
Rrms
Ra
Rmax/10
570 571 572
Figure 3. Roughness parameters measured by AFM tapping mode analysis. (A) Polyamide
573
surface and (B) polysulfone surface. Rrms is the root mean square of roughness, Ra is the average
574
roughness, and Rmax/10 is the maximum roughness divided by a factor of 10. Roughness values
575
shown averaged together from a total of 6 random spots on three separately cast membrane
576
samples. Error bars indicate one standard deviation. Note that in y-axis scales differ.
577
- 25 ACS Paragon Plus Environment
Environmental Science & Technology
578
579 580 581
Figure 4. SEM micrographs displaying the structure of the TFC FO membranes at the (top row)
582
surface of the polyamide layer imaged at 45 degree angle of incidence and at lower
583
magnification, with yellow arrows indicating the nodular (NMP-TFC) and leaf-like (DMF-TFC)
584
structures of the polyamide surfaces; (middle row) surface of the polyamide layers at equal
585
magnification; and (bottom row) the top surface of the bare polysulfone support layers (yellow
586
arrows indicate pores). Membranes formed using NMP are shown on the left and those formed
587
using DMF are shown on the right. - 26 ACS Paragon Plus Environment
Page 26 of 26