OPEN Polyelectrolyte interactions enable rapid association ...
Sacrificial polyelectrolyte multilayer coatings as an ...tarabara/docs/pubs...IPC-N 4). Four...
Transcript of Sacrificial polyelectrolyte multilayer coatings as an ...tarabara/docs/pubs...IPC-N 4). Four...
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
1
Sacrificial polyelectrolyte multilayer coatings as an approach to
membrane fouling control: Disassembly and regeneration
mechanisms
Accepted for publication in the Journal of Membrane Science on April 18, 2015.
Published article DOI: 10.1016/j.memsci.2015.04.041
Pejman Ahmadiannamini†, Merlin L. Bruening‡, Volodymyr V. Tarabara†*
† Department of Civil and Environmental Engineering, Michigan State University, East
Lansing, MI 48824, USA
‡ Department of Chemistry, Michigan State University, East Lansing, MI 48824, USA
* Corresponding author:
[email protected]; phone: (517) 432-1755; fax: (517) 355-0250
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
2
Abstract
This study evaluates polyelectrolyte multilayers (PEMs) as sacrificial separation layers
on polysulfone ultrafiltration (UF) membranes. Exposure to surfactants and a swing in
pH can disassemble PEMs by disrupting non-ionic and electrostatic bonds within the
PEM and between the PEM and the support. Trends in frequency and dissipation in
quartz crystal microbalance studies confirm layer-by-layer (LbL) adsorption of PEMs on
the quartz crystal and subsequent PEM removal in response to acid/base treatment.
After disassembly of PEMs on UF membranes, water permeability increases and
methylene blue rejection decreases, in some cases reaching values close to those for
pristine ultrafilters. PEM removal occurs even with fouled membranes. After fouling by
aqueous solutions of bovine serum albumin and alginate, the PEM disassembly via
exposure to acid, base and surfactant results in a 99% decrease in the hydraulic
resistance of the PEM (compared to the unfouled PEM) and, on average, more than
80% recovery of pure water permeability (compared to an uncoated UF membrane).
Repeated cycles of PEM disassembly and readsorption give stable water permeabilities
and methylene blue rejections for the coated membranes.
Keywords:
polyelectrolyte multilayers; fouling; sacrificial films; disassembly; regeneration
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
3
1. Introduction
Membrane-based separations have made remarkable inroads against competing
technologies in applications such as water desalination and reuse, and gas separations
[1, 2]. Significant advantages of membrane processes may include relatively low energy
consumption and a small environmental footprint [3, 4]. Nevertheless, membrane fouling
remains a major challenge that increases the energy costs for separations and limits
membrane lifetime [5]. Managing membrane fouling requires substantial knowledge
and experience, especially because fouling is often feed-specific. In most applications,
periodic hydraulic cleaning limits short-term fouling, whereas more expensive chemical
cleaning partly removes the hydraulically-irreversible fraction of the added resistance
due to foulants [5-7]. With the increasing importance of water reuse and reliance on
low-quality water sources of high fouling propensity, development of new cleaning
approaches and anti-fouling membrane materials becomes even more important [8-12].
Unfortunately, the impact of surface morphology control and chemical modifications that
improve fouling resistance will not endure after formation of the first fouling layer on the
membrane surface. Ultimately, once chemically irreversible fouling exceeds a certain
level, membrane replacement is required.
This work investigates the possible use of PEMs as sacrificial coatings that a membrane
may shed along with foulants that cannot be removed by hydraulic cleaning. LbL
assembly of PEMs is attractive for coating membranes due to its conceptual simplicity
and control over film thickness, composition and surface charge [13, 14]. Thus, a
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
4
number of studies investigated PEMs as membrane skins for nanofiltration [15-21],
reverse osmosis [22, 23], forward osmosis [24] and pervaporation [25]. As with other
specially designed antifouling surfaces, however, PEMs may temporarily resist
adhesion of feed components, but eventually a fouling layer will coat the surface and
negate antifouling properties. Thus, we desire to develop conditions for removing PEMs
from porous substrates prior to adsorption of a new PEM to create a fresh membrane
surface. In this regard, previous studies employed surfactants [26-28] as well as
changes in pH [21, 22], temperature [29] and ionic strength [30, 31] to disassemble
PEMs. Hydrolysis [32], solvent exchange [33], electrochemical methods [34] and
enzymes [35] may also alter PEMs, but such methods may be difficult to apply in
membrane modules.
Previously, we studied the feasibility of regeneration of (PSS/PAH)4 and (PSS/PAH)4.5
PEMs fouled by SiO2 colloids. To regenerate a PEM with a layer of colloids adsorbed
on its surface, we employed a three-step procedure: (i) backflushing with deionized (DI)
water, (ii) soaking in a buffer solution at pH 10, and (iii) PEM re-deposition by the LbL
method [16]. Although the regenerated membranes showed permeabilities similar to
those of the initial PEM-modified membranes, the rejection was 50% lower than that
with the initial PEM. Apparently, residual polyelectrolytes or foulants decreased
rejection, perhaps by creating film defects, but they did not add significant hydraulic
resistance.
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
5
This study aims to develop PEM design and disassembly approaches that afford
membrane regeneration for nearly complete recovery of flux and rejection after
hydraulically-irreversible fouling. We assembled PEMs using poly(vinylsulfonic acid,
sodium salt) (PVS), poly(sodium 4-styrenesulfonate) (PSS), deprotonated poly(acrylic
acid) (PAA) and deprotonated poly(methacrylic acid) (PMAA) as polyanions; and
poly(diallyldimethylammonium chloride) (PDADMAC), poly(allylamine hydrochloride)
(PAH) and protonated chitosan (Chi) as polycations. Quartz crystal microbalance with
dissipation monitoring (QCM-D) was employed to examine PEM assembly and
disassembly kinetics on gold or silicon oxide surfaces, whereas reflectance FTIR
spectroscopy demonstrated film removal from model spin-coated polyethersulfone
substrates. Filtration studies confirmed that PEM-removal approaches restore
membrane flux and rejection to values close to those of uncoated UF support
membranes. Sequential acid/base and surfactant treatments effectively removed PEMs
to give a membrane ready for adsorption of a new polyelectrolyte (PE) coating. Finally,
we investigated whether removal methods are effective in disassembling PEMs fouled
with bovine serum albumin (BSA) and alginate.
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
6
2. Materials and Methods
2.1. Reagents
PVS (25 wt.% in H2O); PSS (Mw∼70,000) ; PAA (Mw∼1,800); PMAA (Mw∼9,500,
sodium salt solution, 30 wt.% in H2O); PDADMAC (Mw∼100,000-200,000, 20 wt.% in
H2O), PAH (Mw∼15,000), Chi (medium Mw), BSA, alginate, calcium chloride dihydrate
(CaCl2 · 2H2O) and sodium chloride (NaCl) were purchased from Sigma-Aldrich. HCl
solution (36.5% HCl in water, EMD, Millipore), NaOH (Macron) and Triton X-100
(Sigma-Aldrich) were used to prepare PEM disassembly solutions.
Ethylenediaminetetraacetate (EDTA, Sigma-Aldrich) and sodium dodecyl sulfate (SDS,
Roche) served as components of aqueous cleaning solutions for NF 270 membranes
[36]. DI water used in all experiments was supplied by a commercial ultrapure water
system (Lab Five, USFilter) equipped with a terminal 0.2 µm microfilter (PolyCap,
Whatman Plc); the water resistivity was greater than 16 MΩcm.
2.2. Quartz crystal microbalance with dissipation (QCM-D) measurements
A four-channel quartz crystal microbalance with dissipation (QCM-D, Q-Sense E4) was
employed for real-time studies of PEM adsorption and removal processes. Silicon
oxide- or gold-coated quartz crystals with a resonance frequency of 4.95 MHz ± 50 kHz
served as substrates for PEM deposition. Crystals were cleaned according to standard
protocols provided by the manufacturer (see Supplementary Content (SC) file, section
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
7
S1). PEMs were prepared via the LbL technique by flowing alternating polyanion and
polycation solutions through the QCM-D chamber using a peristaltic pump (Ismatec
IPC-N 4). Four bilayers were deposited for each PE pair by pumping 1 g/L PE solutions
in DI water through the QCM-D chamber and subsequently rinsing with DI water after
each PE deposition step at a flow rate of 0.15 mL/min. PE deposition and rinsing steps
were allowed to proceed until QCM frequency and dissipation reached constant values.
PE solutions contained no added salt, and the pH of the PEM solutions and the pKa
values (when known) of the corresponding PEs were as follows: PAA (pH=3.2, pKa=4.2-
6.5 [37, 38]), PMAA (pH=9.2, pKa=5.5 [39]), PSS (pH=5.4, pKa≈1.0 [40]), PVS (pH=6.5),
Chi (pH=6.0, pKa=6.5 [41]), PAH (pH=4.0, pKa=8.5-9.5 [38, 40]), PDADMAC (pH=4.7).
PEMs were then removed from the substrate by exposing them to HCl (1 M) or NaOH
(1 M) solutions or both. The solutions were pumped through the chambers until the
QCM frequency and dissipation reached constant values. All frequency shifts and
dissipation factors are based on data recorded for the 5th overtone.
2.3 PEM adsorption on membranes and filtration tests
PEMs were adsorbed on PES UF membranes (Pall) with a molecular weight cutoff
(MWCO) of 30 kDa. A small subset of tests was also done using PES membranes with
a MWCO of 10 kDa. A PES membrane was placed in a stirred dead-end filtration cell
(Sterlitech, membrane area of 14.6 cm2), and DI water was filtered through the
membrane at 5 bar until a steady-state flux was achieved. A total of 15 mL of PE
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
8
solution (1 g/L) was then poured into the cell and stirred for 5 min. The solution was
discarded and the membrane was rinsed thoroughly with DI water. The PE-adsorption
and rinsing steps were repeated to form four PE bilayers on the PES substrate surface.
PE solutions contained no added salt and the pH of the solutions was not adjusted.
In filtration experiments, membrane coupons were first compacted by filtering DI water
at a transmembrane pressure of 5 bar until a steady permeate flux was attained. The
permeate flux was calculated based on the time-derivative of the permeate mass, which
was measured by collecting the filtered water on a digital balance (AV8101C, Ohaus)
interfaced with a computer. Methylene blue (MB) served as a rejection probe. To
evaluate rejection, the filtration cell was filled with 300 mL of 6 mg/L aqueous MB, and
the feed solution was filtered under fast-stirring conditions (~1300 rpm) to minimize
concentration polarization. In total, 30 mL of solution were filtered in each experiment
and the first 20 mL was discarded to eliminate the effect of MB adsorption on rejection
data. The MB concentration in the permeate was determined using UV-Vis
spectrophotomery (MultiSpec 1501-Shimadzu). To account for an increase in the feed
concentration during the filtration of the first 30 mL of solution, the rejection was
calculated based on a simple differential equation (see SC, section S4).
The membrane in the cell was then exposed to 1 M HCl or 1 M NaOH or both
sequentially, for 15 min each, to remove the PEM coating. Subsequent exposure to
Triton X-100 solutions (2 wt% in water) for 15 min was also employed in some cases to
disrupt non-ionic interactions. During the backflush treatment, the membrane was
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
9
flipped over in the filtration cell and a surfactant solution was directed from the permeate
side of the membrane to PES-PEM interface at the transmembrane pressure of 5 psi. DI
water flux and MB rejection measurements were repeated for the cleaned membranes.
The efficiency of PEM removal was determined by comparing the DI water permeate
flux and MB rejection values of the support PES membrane, the same membrane
coated with a PEM film, and the same membrane after PEM removal.
To study the effectiveness of the removal methods applied to fouled membranes, dead-
end filtration was performed using a feed solution that contained BSA (0.3 g/L), alginate
(0.3 g/L), NaCl (20 mM) and CaCl2 (10 mM). Disassembly of fouled PEM-coated
membranes was attempted as described above, and DI water flux was measured to
quantify the efficiency of the disassembly methods.
2.4 Reflectance Fourier transform infrared (FTIR) spectroscopy
PES (17 mg) was dissolved in 10 mL of CH2Cl2 and 0.1 mL of this solution was placed
onto Au-coated Si wafers (1.1 × 2.4 cm2) for spin coating at 500 rpm [42]. Using the
same PE solutions and conditions described above for membrane modification, PEMs
were deposited by the LbL technique on the PES-coated wafers. Reflectance FTIR
spectra of PEMs on PES were acquired using a Thermo Nicolet 6700 FTIR
spectrometer equipped with a PIKE grazing angle (80°) attachment. A wafer cleaned
with a combination of UV and ozone served as the IR background, and absorbance
from PES (see SC, section S2) was subtracted from all spectra.
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
10
3. Results and Discussion
This study aims to exploit PEMs as sacrificial coatings and enable detachment of fouling
layers that accumulate during filtration and cannot be fully removed by hydraulic
cleaning. The subsections below describe our strategy for PEM removal, QCM-D
studies that demonstrate PEM loss, MB nanofiltration before and after PEM
disassembly to validate the PEM removal on membranes, and finally detachment of
PEMs after filtration of protein and polysaccharide solutions to show that the strategy
applies to fouled membranes.
3.1 Strategy for removal of PEMs
Layer-by-layer adsorption of polyelectrolytes typically occurs through multiple
electrostatic interactions, although non-ionic interactions sometimes help stabilize PEMs
[43-45]. However, when the first PE layer and the substrate possess charges of the
same sign, the initial adsorption occurs only due to non-ionic bonding. Strategies for
removal of PEMs from substrates must take into account the forces that stabilize films
and bind them to the substrate (Fig. 1).
For PEMs containing a weak-base polycation or a weak-acid polyanion, exposure to
high or low pH solutions promotes disassembly by decreasing the degree of PE
ionization and limiting the extent of electrostatic cross-linking in the film [19, 46, 47].
Nevertheless, pH changes will not remove the initial layer from the substrate when
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
11
Figure 1: Mechanisms operative in the assembly of PEMs and their attachment to a
supporting substrate in two cases: a) The first PE layer has the same
charge as the substrate; and b) The first PE layer has a charge opposite
to that of the substrate. For clarity, the diagram does not show the
overlap that occurs over several layers or the release of charge-
compensating counterions that helps to drive adsorption [48-50].
adsorption occurs primarily through non-ionic interactions. In this case desorption will
require disruption of non-ionic bonds, for example through exposure of the PE-substrate
interface to a surfactant. With membranes, this can occur during crossflow filtration of a
surfactant solution through the PEM-coated membrane or, when the PEM is
impermeable to the surfactant, by a backflush with the same solution. Figure 2
summarizes the disassembly strategy employed in this study when the first PE layer
and support have charges of the same sign. Note that surfactant alone may be
insufficient to remove a highly ionically cross-linked film from the surface without
acid/base treatment.
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
12
Figure 2: Proposed PEM disassembly algorithm for the case when the first PE layer
and support have charges of the same sign.
3.2. Examination of film formation and disassembly on a quartz crystal microbalance
Figure 3 shows QCM-D data for the assembly and disassembly of (Chi/PAA)4 on a
QCM-D crystal coated with silicon oxide. After introduction of PE solutions into the
QCM-D chamber, the crystal oscillation frequency decreased and then plateaued,
indicating successful PE adsorption. Frequency drops after both Chi and PAA
adsorption increased exponentially (𝑟2 = 0.9973) with the number of adsorbed layers
(-28, -97, -284 and -765 Hz for 1, 2, 3 and 4 bilayers, respectively), consistent with prior
studies that showed exponential growth for low charge-density PEs such as Chi [51].
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
13
Figure 3: QCM-D data for adsorption and disassembly of a (Chi/PAA)4 multilayer on a SiO2-coated crystal.
Chi, PAA, HCl, and NaOH labels denote times when solutions of the corresponding compounds
were introduced into the QCM-D chamber. The rinse label points to times when the QCM-D
chamber was rinsed with DI water. Note: The data set presented in Fig. 3 shows a PEM removal of ~97%, which is lower than the 100% we observed in most
other experiments. This specific data set was selected to facilitate graphical illustration of ∆𝜈2.
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
14
After depositing each PE layer, the QCM-D crystal was rinsed with DI water to remove
loosely bound PEs from the surface. As expected, the oscillation frequency increased
during rinsing, especially in the case of PEMs that included PAA.
We attempted to disassemble PEMs containing weak polyanions or weak polycations
through treatments with dilute HCl and NaOH solutions, respectively. We did not
employ surfactants in QCM-D tests because the interaction with the silica surface of the
QCM-D crystal should be primarily electrostatic. Because Chi is a weak polybase and
PAA is a weak polyacid, electrostatic interactions in (Chi/PAA)4 should decline when the
PEM is exposed to either HCl or NaOH. Figure 3 shows a dramatic increase in
frequency for the (Chi/PAA)4-modified QCM-D crystal after exposure to HCl, clearly
demonstrating removal of most of the film from the substrate. Subsequent exposure to
NaOH followed by rinsing decreases the frequency to nearly the same level as for a film
with only a single PAA layer. We note that the SiO2 coating on the crystal might be
unstable in 1 M NaOH. However, the oscillation frequency returned to nearly its original
value, suggesting minimal SiO2 dissolution.
We carried out the same experiments with PEMs built from several different PE pairs
(see the SC file for QCM-D frequency shifts for these PEMs (Figures S-1 – S-12) as
well as dissipation data (section S2)). Table 1 summarizes QCM-D data on the
efficiency of PEM removal, :
=∆𝜈1−∆𝜈2
∆𝜈1100% ,
(1)
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
15
where ∆𝜈1 and ∆𝜈2 are QCM frequency shifts (relative to DI water prior to PEM
deposition) before and after PEM disassembly, respectively (see Figure 3 for an
example).
Table 1: Efficiency of PEM disassembly, (wt% removed) in response to sequential
treatment by acid/base solutions as measured in QCM-D experiments with
4-bilayer PEMs.
Notes: * weak PE
** strong PE
# PEMs assembled on Au-coated QCM-D crystals; all other PEMs were
assembled on SiO2-coated crystals
↕ Based on one QCM-D test
Polyanion
Polycation
Chi * PAH * PDADMAC **
PAA * 99.4 ± 1.3 92.3 ± 3.0 94.3 ± 4.4
PMAA * 95.4 # ↕ 86.0 ± 4.9 82.4 ± 9.0
PSS ** 84.3 ± 13.9 # 86.5 ± 5.9 7.5 ± 3.2
PVS ** 94.5 ± 7.8 72.1 ± 6.1 17.8 ± 11.3
Among the PEMs, removal was highest for (PAA/Chi)4, which contains two weak PEs.
The high water solubilities of Chi and PAA and the very low MW of PAA may also lead to
the essentially 100% disassembly of films containing these polymers. Moreover, all the
films that contained at least one weak PE had removal efficiencies higher than 72%. As
expected, disassembly was least effective with PEMs composed of two strong PEs:
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
16
(PSS/PDADMAC)4 and (PVS/PDADMAC)4. In these cases, a change in pH had little
effect on ionization and removed on average 20% or less of the mass of the PEMs.
We note that, ideally, the surface for QCM-D tests should be the same as the surface of
the UF supports for PEM adsorption. Unfortunately, such an ideal model is not feasible.
Spin-coating a QCM-D crystal with polysulfone would produce a non-porous polymer
layer that would interact with the 1st PE layer differently than a porous layer of the UF
support’s skin. In principle, one can attempt to replicate the phase inversion process to
prepare a UF polysulfone membrane on the surface of a QCM-D crystal. While the
desired surface morphology might form, the membrane would be too thick (300 μm)
and, therefore, too heavy to serve as a coating for a QCM-D crystal. In addition, the
small size of the crystal would lead to boundary effects, and insufficient post-processing
(i.e. no possibility to pre-compact the membrane or to flow water through the membrane
to flush out the porogen) would make true replication impossible.
3.3. Filtration before and after PEM disassembly
To initially assess disassembly of a range of PEMs on membranes, we measured DI
water flux and MB rejection before and after PEM disassembly by the acid/base
treatment. Figure 4 shows results of these tests with the data plotted in the order of
decreasing DI water permeate flux values for membranes with as-assembled PEMs.
Adsorption of PEMs on the UF substrates decreased permeability to DI water and
increased MB rejection.
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
17
Figure 4: DI water flux (top) and MB rejection (bottom) of PEM-coated 30 kDa PES
membranes before (dark gray) and after (light gray) PEM removal. For
comparison the figure shows data for pristine PES membranes at the far
left (hashed gray). All PEMs were disassembled by sequential treatment
with acid and base solutions.
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
18
Upon PEM disassembly, flux and rejection did not return to the levels typical of the
unmodified UF support suggesting incomplete removal of the PEM. In the few cases
where disassembly conditions decreased flux, e.g. (PAH/PMAA)4, or increased
rejection, e.g. (PDADMAC/PMAA)4, the acid and base likely induce conformational
changes in the film without removing sufficient PE to enhance permeability [52].
Flux recovery after PEM disassembly (flux after disassembly divided by the flux through
the pristine membrane) does not correlate directly with the extent of polyelectrolyte
removal because both the PEM and the underlying PES membrane provide resistance
to flow. Thus, we employed a resistance-in-series model to determine the resistances
of the UF substrate, the PEM and residual material left on the membrane after PEM
removal. These data allow definition of the resistance removal efficiency (𝜀𝑅), defined
as follows:
𝜀𝑅 = 1 −𝑅𝑟𝑒𝑠
𝑅𝑃𝐸𝑀 ,
(2)
where 𝑅𝑃𝐸𝑀 and 𝑅𝑟𝑒𝑠 are hydraulic resistances of the PEM and residual film left after
PEM removal, respectively. The resistance removal efficiency is indicative of the extent
of PEM removal, so we also term it the PEM removal efficiency.
Table 2 provides values for both flux recovery after PEM removal and 𝜀𝑅. Note that for
(Chi/PAA)4 (top entry in the table), even though flux recovery is only 54%, the PEM
removal efficiency (i.e. resistance removal efficiency) is 98%.
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
19
Table 2: Recovery and regeneration of PEM properties (DI water permeate
flux and rejection) in response to four methods of PEM disassembly.
PEM
MWCO of PES support,
kDa
PEM removal method Flux recovery,
%
PEM removal
efficiency 𝜀𝑅, % M1 M2 M3 M4
1
PE
Ms w
ith a
poly
catio
n a
s t
he 1
st l
ayer
(Chi/PAA)4
30
x x 54 98
2 (Chi/PMAA)4 x x 32 0
3 (Chi/PSS)4 x x 54 78
4 (Chi/PVS)4 x x 22 25
5 (PAH/PAA)4 x x 82 100
6 (PAH/PMAA)4 x x 39 14
7 (PAH/PSS)4 x x 52 -8
8 (PAH/PVS)4 x x 90 86
9 (PDADMAC/PAA)4 x x 18 90
10 (PDADMAC/PMAA)4 x x 17 -28
11
PE
Ms w
ith a
poly
anio
n a
s t
he 1
st l
ayer
(PAA/Chi)4
30 x x 42 ± 7* 96.5 ± 0.3*
12
10
x x 59 ± 9* 98.7 ± 0.5*
13 x x x 94 99.8
14 (PSS/Chi)4 x x 86 96.5
15 (PSS/PDADMAC)4
x 56 75.5
16 x 95.9 98.2
17 (PAA/Chi)4/fouled† x x x 110 ± 16* 100.3 ± 0.5*,‡
18 (PSS/Chi)4/fouled† x x 80 ± 15* 97.8 ±2.2*,‡
19 (PSS/PDADMAC)4/fouled† x 93 ± 12* 99.6 ± 0.5*,‡
Notes: M1 = exposure to high pH; M2 = exposure to low pH; M3 = exposure to Triton X-100;
M4 = backflush with Triton X-100.
* Uncertainties represents standard deviations based on sets of 3 replicate measurements
performed with different membranes. Other data in the table were obtained in experiments
with only one membrane for each PE pair.
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
20
† Prior to PEM removal, these membranes were fouled by filtering solutions containing BSA,
alginate, and Ca2+. ‡ Note that in the calculations of εR
(i) for fouled membranes, RPEM represents the hydraulic
resistance of fouled PEM membranes.
This occurs because 𝑅𝑃𝐸𝑀 is very high compared to the resistance of the untreated
membrane (see the SC, section S3, for a discussion of the series resistance model).
Similarly, for (PAH/PAA)4, the flux recovery is only 82%, but the resistance recovery
rounds to 100%. All other PE pairs show lower removal efficiencies. Thus, of all the PE
pairs, (Chi/PAA)4 and (PAH/PAA)4 are the best choices for disassembly by acid/base
treatment and regeneration. However, the DI water flux through membranes coated
with (PAH/PAA)4 is very low, so the work that follows focuses on membranes coated
with (Chi/PAA)4. The (Chi/PAA)4 system also showed the highest removal efficiency in
QCM-D studies (Table 1), and these films on membranes show relatively high MB
rejections. We note that zero or negative 𝜀𝑅 values in Table 1 represent membranes
whose flux does not increase after a disassembly treatment. A pH swing or exposure to
surfactant may induce PEM conformational changes that are not accompanied by
sufficient removal of PEs to decrease PEM resistance, so 𝜀𝑅 is not positive.
3.4. Effects of substrate MWCO on PEM disassembly and regeneration
To understand how the pore size of the substrate affects PEM removal, we examined
assembly and disassembly of (Chi/PAA)4 films on PES substrates with MWCOs of 10
kDa and 30 kDa. Further, to study the durability of membranes under chemical
treatments and whether the PEM slowly builds up on the membrane, we repeated the
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
21
assembly-disassembly cycle three times (Fig. 5). Flux and rejection data show similar
trends for filtration through PEMs prepared on both 10 kDa and 30 kDa PES supports,
although flux is lower and rejection higher with the 10 kDa support. However,
disassembly does not lead to complete flux recovery with either membrane, suggesting
partial blocking of pores by residual PEs. The incomplete PEM removal even on the 10
kDa membrane suggests that a mechanism other than physical entrapment retains
some PE on the membrane. In particular, non-ionic interactions may prevent
detachment of the first PE layer from the PES substrate. Nevertheless, the 𝜀𝑅 values in
Table 2 (see entries 11 and 12) suggest that a higher membrane MWCO may result in
greater polyelectrolyte entrapment and lower removal efficiency. Most importantly,
fluxes and rejections with the PEM-coated membranes are similar for the initial film and
films formed after one or more regeneration cycles (i.e. disassembly-buildup), although
both flux and rejection decline after the second disassembly and third film-formation
step.
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
22
(a)
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
23
(b)
Figure 5: DI water permeate flux and MB rejection data during filtration of a 17.5 μM
MB solution through (Chi/PAA)4 films on the surface of (a) 30 kDa PES
membranes and (b) 10 kDa PES membranes. (Each error bar represents
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
24
a standard deviation for a set of 3 replicate measurements with different
membranes.)
3.5. Disassembly with surfactants
Surfactants may disrupt non-ionic interactions between the PES substrate and PEs to
enable more complete PEM removal [26-28]. However, if both electrostatic and non-
ionic interactions immobilize the initial PE, removal of this layer may require
simultaneous treatment with base/acid and surfactant rather than sequential treatments.
Thus, to prevent electrostatic interactions between the PES membrane, which carries a
negative surface charge, and the first PE layer, we employed a polyanion as the initial
layer in PEM-coated membranes on a PES substrate with a 10 kDa MWCO. After
treating the (PAA/Chi)4-modified membrane with low and high pH solutions, we filtered a
surfactant solution through the membrane for 15 min and reexamined its filtration
properties. Although the pH treatment results in only 59 ± 9% recovery of the DI water
permeate flux through the bare membrane (Table 2, entry 12), subsequent treatment
with surfactant leads to 94% flux recovery (Table 2, entry 13). The MB rejection
decreased from 99% to 55% after PEM removal by pH swing and to 36% after PEM
removal by pH swing and surfactant. Note that MB rejection by the PEM-free UF
membrane (10 kDa MWCO) is 31%, or nearly the same as by the membrane after PEM
disassembly.
We also examined the effectiveness of surfactant treatment for the disassembly of
(PSS/Chi)4, a PEM that contains one strong PE. For these films, pH affects only the
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
25
ionization of the weakly basic Chi, so the disassembly procedure included sequential
treatment with NaOH (no HCl) and surfactant to break electrostatic and non-ionic
bonds, respectively. DI water filtration through (PSS/Chi)4-coated PES membranes
showed flux recoveries (relative to the bare PES membrane) of 28% after treatment with
NaOH and 86% after subsequent treatment with surfactant (Table 2, entry 14). MB
rejection decreased from 73% for (PSS/Chi)4-coated membranes, to 54% after
treatment with NaOH and 34% after subsequent surfactant filtration. Clearly, surfactant
treatment enhanced the PEM removal.
Finally, we studied surfactant-induced disassembly of membranes composed of two
strong PEs (PSS and PDADMAC). Because pH does not affect the charge density of
strong PEs, the disassembly procedure only included exposure to the surfactant.
Remarkably, after surfactant treatment, flux recovery was 96% (Table 2, entry 16) while
MB rejection decreased from 48% to 18%, a rejection slightly lower than that of even
the pristine membrane. We speculate that in this case, the PEM film remains integral
and is “peeled off” the substrate.
3.4. Investigation of PEM disassembly on PES-coated wafers using FTIR spectroscopy
(PAA/Chi)4, (PSS/Chi)4 and (PSS/PDADMAC)4 films were also prepared on PES-
coated wafers to enable corroborating FTIR spectroscopy studies of PEM removal from
a surface similar to that of PES membranes (although without pores). Figures 6 (a-c)
present reflectance FTIR spectra for the three PEMs before and after different
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
26
disassembly treatment steps. In the spectrum of (PAA/Chi)4 (Fig. 6a), peaks at ~1550
and ~1400 cm−1 most likely stem from the PAA asymmetrical and symmetrical –COO-
stretches, respectively.
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
27
Figure 6: FTIR spectra of PEMs prepared on Au-coated silicon wafers
modified with spin-coated PES and treated with various steps in the
PEM buildup/disassembly sequence: after PEM adsorption, after pH
treatment, and after surfactant treatment. The PEMs were a)
(Chi/PAA)4; b) (PSS/PAA)4; c) (PSS/PDADMAC)4.
The peak at 1720 cm−1 results from the acid carbonyl stretch, and the absorbance at
1457 cm−1 is a -CH2 band. Most importantly, these peaks disappear after either HCl or
NaOH treatment, and the spectrum is featureless after the NaOH treatment, implying
complete film removal.
The spectra of PEMs containing PSS show absorbance peaks at 1127, 1036 and 1009
cm-1, which correspond to sulfonic acid group (SO3-) stretches (Fig. 6b and Fig. 6c).
With PSS/Chi films, these peaks disappear after treatment with NaOH. (We did not
treat PSS/Chi films with HCl because they do not contain a weak acid.) However, the
surfactant treatment alone did not appear to completely remove the films from PES-
coated wafers. In contrast to filtration results, the spectra of (PSS/PDADMAC)4 coatings
suggest no removal after surfactant treatment. This could be due to the high density of
the (PSS/PDADMAC)4 films, which does not allow the surfactant molecules to diffuse
through the PEM and reach the substrate-PEM interface to break the non-ionic bonds
between substrate and first PE layer. Note that surfactant treatment for the
(PSS/PDADMAC)4 membrane was performed via back flushing or filtration to bring the
surfactant to the PEM-substrate interface.
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
28
Figure 7: DI water flux through (1) 10 kDa PES membranes, (2) NF 270 or PEM-
modified PES membranes, (3) NF270 or PEM-coated membranes fouled by
a solution of BSA and alginate, (4) hydraulically cleaned NF270 or PEM-
coated membranes, (5) PES membranes after PEM removal, (6) PES
membranes with re-deposited PEMs or chemically cleaned NF 270. (a)
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
29
(Chi/PAA)4, (b) (PSS/PAA)4, (c) (PSS/PDADMAC)4 and (d) NF 270. PEM
removal protocols are described in Table 2, entries 17-19. Notes:
1) Each error bar represents a standard deviation for a set of 3 replicate measurements with
different membranes.
2) DI water fluxes for PAA/Chi, PSS/Chi, PSS/PDADMAC and NF 270 membranes are 1.67·10-
11, 4.44·10-11, 4.67·10-11, 3.17·10-11 m/s/Pa (6.0, 16.0, 26.8 and 11.4 LMH/bar), respectively.
3.5. Removal of fouled PEMs
We also examined removal of fouled PEMs and whether a second PEM deposition
gives the same separation properties as the originally deposited coating. These
experiments included filtering mixtures of BSA and alginate through (PAA/Chi)4-,
(PSS/Chi)4- and (PSS/PDADMAC)4-coated PES membranes, application of PEM
removal protocols, coating the cleaned substrate with a new PEM, and reexamining
filtration performance. Figure 7 shows that in all cases, removal of the fouled PEM
(chemical treatment column) led to DI water fluxes within ±10% of those through the
bare membrane. Adsorption of a fresh PEM gave fluxes within ± 20% of those through
the initial PEM-coated membrane. However, we should note that fouling is partly
hydraulically reversible as indicated by the significant (73 ± 19 %) flux recovery
observed for the NF 270 membrane. Treatment of fouled NF 270 with a cleaning
solution containing 1 mM EDTA and 10 mM SDS at pH 10 resulted in complete flux
recovery (115 ± 15%).
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
30
3.6. Limitations and possible applications of the proposed approach
The proposed PEM removal procedures may not be compatible with some membrane
materials. For example, cellulose acetate and, to a lesser extent, its di- and tri-acetate
derivatives hydrolyze at high and low pH. However, PEMs can adsorb to almost any
material and serve as sacrificial coatings on polymeric or ceramic membranes that are
stable over a wide range of pH. This work focused on surface fouling, but for
membranes with very large pores PEMs may serve as sacrificial coating to mitigate
intrapore fouling. When pores are too large to be bridged by PEs, their internal pore
surface can be coated. Perhaps the biggest limitation of PEMs is the multistep removal
and film formation process. Simplification of the deposition procedure will likely be
necessary to make the regeneration process economical. Although polyelectrolyte
redeposition may require more than one hour, replacement of only the top layers of a
film [27, 28] could shorten the process.
The removal strategy has several potential advantages. First, regenerating a
membrane’s selective layer in situ can allow for on-demand assembly of a membrane
with desired surface and separation properties. Second, regeneration of the selective
layer is an alternative method of managing irreversible fouling; this method may remove
hydraulically-irreversible fouling as a substitute for or in addition to chemical cleaning.
Third, the membrane can be “replaced” without changing the support layer (UF
membrane) or the membrane module, potentially leading to cost savings.
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
31
4. Conclusions This work examined controlled disassembly of PEMs on several substrates and
explored the use of sacrificial PEMs as an approach to managing membrane fouling.
On the surface of QCM crystals, exposure to 1 M HCl or 1 M NaOH or both effectively
removed most (70-100%) of the mass of PEMs composed of weak PEs. In contrast,
PEMs made of only strong PEs were stable under these conditions. For the same
PEMs assembled on the surface of PES UF membranes (MWCO 10 kDa and 30 kDa),
exposure to 1 M NaOH and 1 M HCl did not completely restore permeate flux to the
value of a PEM-free support. A subsequent treatment with a surfactant successfully
removed PEMs consisting of weak/weak, weak/strong and strong/strong PE pair
combinations and restored >86% of the flux, presumably by disrupting non-ionic
interactions between the PES support and PEs that remain adsorbed after exposure to
low or high pH or both. If a PEM contains strong PEs only, the PEM can be removed
from the substrate by disrupting non-ionic bonding between the substrate and the first
PE layer by a surfactant treatment. The proposed method is also efficient in removing
PEMs that have been fouled by a mixture of BSA and alginate.
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
32
Acknowledgements
This work was supported in part by the NSF Partnerships for International Education
and Research Grant IIA-1243433.
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
33
List of Figures
Figure 1: Mechanisms operative in the assembly of PEMs and their attachment to a
supporting substrate in two cases: a) The first PE layer has the same
charge as the substrate; and b) The first PE layer has a charge opposite
to that of the substrate. For clarity, the diagram does not show the
overlap that occurs over several layers or the release of charge-
compensating counterions that helps to drive adsorption [48-50].
Figure 2: Proposed PEM disassembly algorithm for the case when the first PE layer
and support have charges of the same sign.
Figure 3: QCM-D data for adsorption and disassembly of a (Chi/PAA)4 multilayer on
SiO2-coated crystal. Chi, PAA, HCl, and NaOH labels denote times where
solutions of the corresponding compounds were introduced into the QCM-
D chamber. The rinse label points to times when the QCM-D chamber
was rinsed with DI water.
Note:
The data set presented in Fig. 3 shows a PEM removal of ~97%, which is lower than the
100% we observed in most other experiments. This specific data set was selected to
facilitate graphical illustration of ∆𝜈2.
Figure 4: DI water flux (top) and MB rejection (bottom) of PEM-coated 30 kDa PES
membranes before (dark gray) and after (light gray) PEM removal. For
comparison the Figure shows data for pristine PES membranes at the far
left (hashed gray). All PEMs were disassembled by the sequential
treatment with acid/base solutions.
Figure 5: DI water permeability and MB rejection data during filtration of a 17.5 μM
MB solution through (Chi/PAA)4 films on the surface of (a) 30 kDa PES
membranes and (b) 10 kDa PES membranes. (Each error bar represents
a standard deviation for a set of 3 replicate measurements with different
membranes.)
Figure 6: FTIR spectra of PEMs prepared on Au-coated silicon wafers modified with
spin-coated PES and treated with various steps in the PEM
buildup/disassembly sequence: after PEM adsorption, after pH treatment,
and after surfactant treatment. The PEMs were a) (Chi/PAA)4; b)
(PSS/PAA)4; c) (PSS/PDADMAC)4.
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
34
Figure 7: DI water permeate flux through (1) 10 kDa PES membrane, (2) NF 270 or
PEM-modified PES membrane, (3) NF270 or PEM-coated membranes
fouled by a solution of BSA and alginate, (4) hydraulically cleaned NF270
or PEM-coated membranes, (5) PES membranes after PEM removal, (6)
PES membranes with re-deposited PEMs or chemically cleaned NF 270.
a) (Chi/PAA)4, b) (PSS/PAA)4, c) (PSS/PDADMAC)4 and NF 270. PEM
removal protocols are described in Table 2.
Notes:
1) Each error bar represents a standard deviation for a set of 3 replicate
measurements with different membranes.
2) DI water fluxes for PAA/Chi, PSS/Chi, PSS/PDADMAC and NF 270
membranes are 1.67·10-11, 4.44·10-11, 4.67·10-11, 3.17·10-11 m/s/Pa (6.0, 16.0,
26.8 and 11.4 LMH/bar), respectively.
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
35
List of Tables
Table 1: Efficiency of PEM disassembly, (wt% removed) in response to sequential
treatment by acid/base solutions as measured in, based on QCM-D
experiments with measurements for films 4-bilayer PEMs composed of 4
bilayers.
Notes: * weak PE
** strong PE
# PEMs assembled on Au-coated QCM-D crystals; all other PEMs were
assembled on SiO2-coated crystals
↕ Based on one QCM-D test
Table 2: Recovery and regeneration of PEM properties (DI water permeate flux and
rejection) in response to the four methods of PEM disassembly.
Notes: M1 = exposure to high pH; M2 = exposure to low pH; M3 = exposure to Triton
X-100;
M4 = backflush with Triton X-100.
* Uncertainties represents standard deviations based on sets of 3 replicate
measurements performed with different membranes. Other data in the table
were obtained in experiments with only one membrane for each PE pair. † Prior to PEM removal, these membranes were fouled by filtering solutions
containing BSA, alginate, and Ca2+. ‡ Note that in the calculations of εR
(i) for fouled membranes, RPEM represents
the hydraulic resistance of fouled PEM membranes.
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
36
References
[1] V. Abetz, T. Brinkmann, M. Dijkstra, K. Ebert, D. Fritsch, K. Ohlrogge, D. Paul, K.-V. Peinemann, S. Pereira-Nunes, N. Scharnagl, Developments in membrane research: from material via process design to industrial application, Adv. Eng. Mater., 8 (2006) 328-358.
[2] E.M.V. Hoek, V.V. Tarabara, Encyclopedia of Membrane Science and Technology, Volume 3: Applications, John Wiley & Sons, 2013.
[3] M. Elimelech, W.A. Phillip, The future of seawater desalination: energy, technology, and the environment, Science, 333 (2011) 712-717.
[4] S. Judd, The MBR book: principles and applications of membrane bioreactors for water and wastewater treatment, 2nd ed., Elsevier, 2010.
[5] AWWA, Committee Report: Recent Advances and Research Needs in Membrane Fouling, J. Am. Water Works Ass., 97 (2005) 79-89.
[6] P. Le-Clech, V. Chen, T.A.G. Fane, Fouling in membrane bioreactors used in wastewater treatment, J. Membr. Sci., 284 (2006) 17-53.
[7] G.L. Amy, M.M. Clark, J. Pellegrino. NOM rejection by, and fouling of, NF and UF membranes, Project report, American Water Works Association Research Foundation, 2001.
[8] M. Elimelech, Z. Xiaohua, A.E. Childress, S. Hong, Role of membrane surface morphology in colloidal fouling of cellulose acetate and composite aromatic polyamide reverse osmosis membranes, J. Membr. Sci., 127 (1997) 101-109.
[9] I. Pinnau, B.D. Freeman, Formation and modification of polymeric membranes: overview, in: Membrane Formation and Modification, American Chemical Society, 1999, pp. 1-22.
[10] V. Kochkodan, D.J. Johnson, N. Hilal, Polymeric membranes: surface modification for minimizing (bio)colloidal fouling, in: Adv. Colloid Interface Sci., 2014, pp. 116-140.
[11] C.-C. Ho, A.L. Zydney, Overview of fouling phenomena and modeling approaches for membrane bioreactors, Sep. Purif. Technol., 41 (2006) 1231-1251.
[12] Y. Ye, P.L. Clech, V. Chen, A.G. Fane, Evolution of fouling during crossflow filtration of model EPS solutions, J. Membr. Sci., 264 (2005) 190-199.
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
37
[13] N. Joseph, P. Ahmadiannamini, R. Hoogenboom, I.F.J. Vankelecom, Layer-by-layer preparation of polyelectrolyte multilayer membranes for separation, Polym. Chem., 5 (2014) 1817-1831.
[14] Q. Zhao, Q.F. An, Y. Ji, J. Qian, C. Gao, Polyelectrolyte complex membranes for pervaporation, nanofiltration and fuel cell applications, J. Membr. Sci., 379 (2011) 19-45.
[15] P. Ahmadiannamini, X. Li, W. Goyens, B. Meesschaert, I.F.J. Vankelecom, Multilayered PEC nanofiltration membranes based on SPEEK/PDDA for anion separation, J. Membr. Sci., 360 (2010) 250-258.
[16] W. Shan, P. Bacchin, P. Aimar, M.L. Bruening, V.V. Tarabara, Polyelectrolyte multilayer films as backflushable nanofiltration membranes with tunable hydrophilicity and surface charge, J. Membr. Sci., 349 (2010) 268-278.
[17] R. Malaisamy, A. Talla-Nwafo, K.L. Jones, Polyelectrolyte modification of nanofiltration membrane for selective removal of monovalent anions, Sep. Purif. Technol., 77 (2011) 367-374.
[18] B.W. Stanton, J.J. Harris, M.D. Miller, M.L. Bruening, Ultrathin, multilayered polyelectrolyte films as nanofiltration membranes, Langmuir, 19 (2003) 7038-7042.
[19] P. Ahmadiannamini, X. Li, W. Goyens, N. Joseph, B. Meesschaert, I.F.J. Vankelecom, Multilayered polyelectrolyte complex based solvent resistant nanofiltration membranes prepared from weak polyacids, J. Membr. Sci., 394-395 (2012) 98-106.
[20] P. Ahmadiannamini, X. Li, W. Goyens, B. Meesschaert, W. Vanderlinden, S. De Feyter, I.F.J. Vankelecom, Influence of polyanion type and cationic counter ion on the SRNF performance of polyelectrolyte membranes, J. Membr. Sci., 403-404 (2012) 216-226.
[21] L. Ouyang, R. Malaisamy, M.L. Bruening, Multilayer polyelectrolyte films as nanofiltration membranes for separating monovalent and divalent cations, J. Membr. Sci., 310 (2008) 76-84.
[22] W. Jin, A. Toutianoush, B. Tieke, Use of polyelectrolyte layer-by-layer assemblies as nanofiltration and reverse osmosis membranes, Langmuir, 19 (2003) 2550-2553.
[23] J. Xu, X. Feng, C. Gao, Surface modification of thin-film-composite polyamide membranes for improved reverse osmosis performance, J. Membr. Sci., 370 (2011) 116-123.
[24] P. Pardeshi, A.A. Mungray, Synthesis, characterization and application of novel high flux FO membrane by layer-by-layer self-assembled polyelectrolyte, J. Membr. Sci., 453 (2014) 202-211.
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
38
[25] G. Zhang, W. Gu, S. Ji, Z. Liu, Y. Peng, Z. Wang, Preparation of polyelectrolyte multilayer membranes by dynamic layer-by-layer process for pervaporation separation of alcohol/water mixtures, J. Membr. Sci., 280 (2006) 727-733.
[26] X. Huang, A.B. Schubert, J.D. Chrisman, N.S. Zacharia, Formation and tunable disassembly of polyelectrolyte-Cu2+ layer-by-layer complex film, Langmuir, 29 (2013) 12959-12968.
[27] J. Irigoyen, N. Politakos, R. Murray, S.E. Moya, Design and fabrication of regenerable polyelectrolyte multilayers for applications in foulant removal, Macromol. Chem. Phys., 215 (2014) 1543–1550.
[28] J.I. Ramos, I. Llarena, S. Moya, Selective removal of hybrid di-block polyelectrolyte multilayers by means of quaternary ammonium surfactants, J. Membr. Sci., 45 (2010) 4970-4976.
[29] R. Steitz, V. Leiner, K. Tauer, V. Khrenov, R.v. Klitzing, Temperature-induced changes in polyelectrolyte films at the solid-liquid interface, Appl. Phys. A Mater. Sci. Process, 74 (2002) s519-s521.
[30] N. Wang, G. Zhang, S. Ji, Z. Qin, Z. Liu, The salt-, pH- and oxidant-responsive pervaporation behaviors of weak polyelectrolyte multilayer membranes, J. Membr. Sci., 354 (2010) 14-22.
[31] A.A. Antipov, G.B. Sukhorukov, H. Möhwald, Influence of the ionic strength on the polyelectrolyte multilayers' permeability, Langmuir, 19 (2003) 2444-2448.
[32] Y. Zhu, J. Shi, W. Shen, X. Dong, J. Feng, M. Ruan, Y. Li, Stimuli-responsive controlled drug release from a hollow mesoporous silica sphere/polyelectrolyte multilayer core-shell structure, Angew. Chem., 117 (2005) 5213-5217.
[33] Y. Lvov, A.A. Antipov, A. Mamedov, H. Möhwald, G.B. Sukhorukov, Urease encapsulation in nanoorganized microshells, Nano Lett., 1 (2001) 125-128.
[34] R. Zahn, J. Vörös, T. Zambelli, Swelling of electrochemically active polyelectrolyte multilayers, Curr. Opin. Colloid Interface Sci., 15 (2010) 427-434.
[35] V.H. Orozco, V. Kozlovskaya, E. Kharlampieva, B.L. López, V.V. Tsukruk, Biodegradable self-reporting nanocomposite films of poly(lactic acid) nanoparticles engineered by layer-by-layer assembly, Polymer, 51 (2010) 4127-4139.
[36] Q. Li, M. Elimelech, Organic fouling and chemical cleaning of nanofiltration membranes: measurements and mechanisms, Environ. Sci. Technol., 38 (2004) 4683-4693.
[37] C. Echeverria, N.A. Peppas, C. Mijangos, Novel strategy for the determination of UCST-like microgels network structure: effect on swelling behavior and rheology, Soft Matter, 8 (2012) 337-346.
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
39
[38] S.W. Cranford, C. Ortiz, M.J. Buehler, Mechanomutable properties of a PAA/PAH polyelectrolyte complex: rate dependence and ionization effects on tunable adhesion strength, Soft Matter, 6 (2010) 4175-4188.
[39] J. Zhang, N.A. Peppas, Synthesis and characterization of pH- and temperature-sensitive poly(methacrylic acid)/poly(N-isopropylacrylamide) interpenetrating polymeric networks, Macromolecules, 33 (1999) 102-107.
[40] S.R. Lewis, S. Datta, M. Gui, E.L. Coker, F.E. Huggins, S. Daunert, L. Bachas, D. Bhattacharyya, Reactive nanostructured membranes for water purification, PNAS, 108 (2011) 8577-8582.
[41] J. Mao, S. Kondu, H.-F. Ji, M.J. McShane, Study of the near-neutral pH-sensitivity of chitosan/gelatin hydrogels by turbidimetry and microcantilever deflection, Biotechnol. Bioeng., 95 (2006) 333-341.
[42] P. Jain, J. Dai, S. Grajales, S. Saha, G.L. Baker, M.L. Bruening, Completely Aqueous Procedure for the Growth of Polymer Brushes on Polymeric Substrates, Langmuir, 23 (2007) 11360-11365.
[43] F. Caruso, H. Lichtenfeld, E. Donath, H. Möhwald, Investigation of electrostatic interactions in polyelectrolyte multilayer films: binding of anionic fluorescent probes to layers assembled onto colloids, Macromolecules, 32 (1999) 2317-2328.
[44] J.B. Schlenoff, S.T. Dubas, Mechanism of polyelectrolyte multilayer growth: charge overcompensation and distribution, Macromolecules, 34 (2001) 592-598.
[45] P.T. Hammond, Recent explorations in electrostatic multilayer thin film assembly, Curr. Opin. Colloid Interface Sci., 4 (1999) 430-442.
[46] J.B. Schlenoff, H. Ly, M. Li, Charge and mass balance in polyelectrolyte multilayers, J. Am. Chem. Soc., 120 (1998) 7626-7634.
[47] S.S. Shiratori, M.F. Rubner, pH-dependent thickness behavior of sequentially adsorbed layers of weak polyelectrolytes, Macromolecules, 33 (2000) 4213-4219.
[48] G. Ladam, P. Schaad, J. Voegel, P. Schaaf, G. Decher, F. Cuisinier, In situ determination of the structural properties of initially deposited polyelectrolyte multilayers, Langmuir, 16 (2000) 1249-1255.
[49] R.v. Klitzing, Internal structure of polyelectrolyte multilayer assemblies, Phys. Chem. Chem. Phys., 8 (2006) 5012-5033.
[50] P.A. Neff, A. Naji, C. Ecker, B. Nickel, R. v. Klitzing, A.R. Bausch, Electrical detection of self-assembled polyelectrolyte multilayers by a thin film resistor, Macromolecules, 39 (2006) 463-466.
© 2015. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0
40
[51] C. Picart, J. Mutterer, L. Richert, Y. Luo, G. Prestwich, P. Schaaf, J.-C. Voegel, P. Lavalle, Molecular basis for the explanation of the exponential growth of polyelectrolyte multilayers, PNAS, 99 (2002) 12531-12535.
[52] R.A. Ghostine, R.M. Jisr, A. Lehaf, J.B. Schlenoff, Roughness and salt annealing in a polyelectrolyte multilayer, Langmuir, 29 (2013) 11742-11750.