SURFACTANT EFFECTS ON POLYELECTROLYTE BRUSH SWELLING AND
EMULSION FILTRATION THROUGH BRUSH-COATED MEMBRANES
By
Zhefei Yang

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Chemistry—Doctor of Philosophy
2017

ABSTRACT
SURFACTANT EFFECTS ON POLYELECTROLYTE BRUSH SWELLING AND
EMULSION FILTRATION THROUGH BRUSH-COATED MEMBRANES
By
Zhefei Yang
This dissertation investigates the swelling and fouling resistance of polyanionic and
polyzwitterionic brushes in the presence of surfactants and surfactant-stabilized oil-inwater (O/W) emulsions. Such emulsions are an undesired byproduct of many processes,
the largest of which is the massive oil and gas extraction industry. Among the three existing
forms of oil (free oil, dispersions and emulsions), emulsions are the most difficult to treat
due to the small oil droplet size (< 20 Âľm) and surfactant-enhanced emulsion stability.
Membrane filtration is an effective method for removing oil from emulsions, but fouling
induces a rapid decline in permeate flux that may make such separations impractical.
Polyelectrolyte brushes can potentially resist oil fouling because they are
superoleophobic in aqueous solutions due to their strong water affinity. This research
employs surface-initiated atom transfer radical polymerization for relatively controlled
growth of polyanionic and polyzwitterionic films. Poly(3-sulfopropyl methacrylate
potassium salt) (PSPMK) brushes formed on Au surfaces show underwater hexadecane
contact angles of nearly 180°. Due to electrostatic repulsion, these polyanionic brushes do
not adsorb the anionic surfactant sodium dodecyl sulfate (SDS). Moreover, with SDSstabilized O/W emulsions, microfiltration membranes modified with PSPMK brushes
show ~100% oil rejection and a constant permeate flux, indicating minimal fouling during
12-h, dead-end filtrations. In contrast, electrostatic attraction causes strong adsorption of
the cationic surfactant cetrimonium bromide (CTAB) to PSPMK brushes, and this leads to

brush collapse and no oil rejection during filtration. Thus, polyanionic brushes effectively
resist oil fouling, but only for emulsion stabilized with anionic surfactants.
To investigate whether zwitterionic films can resist adsorption of both cationic and
anionic surfactants, this work employed poly(sulfobetaine methacrylate) (PSBMA)
brushes. Unfortunately, both cationic and anionic surfactants adsorb to these brushes, and
swelling in surfactant solutions depends on surfactant charge and concentration, as well as
brush thickness. Preliminary data show that during the filtration of O/W emulsions, the
permeate flux is much lower than the pure water flux, presumably because of fouling.
Moreover, the permeate flux increases during filtration, implying brush collapse after oil
adsorption. Compared with the strong fouling resistance of anionic PSPMK brushes to
negatively charged oil droplets, the nearly neutral zwitterionic brushes have little charge
repulsion to oil droplets and show low resistance to fouling.
During filtration with PSBMA-modified membranes, oil coalescence occurs. As an
alternative approach to separate O/W emulsions, membrane coalescence traditionally
employs hydrophobic membranes to prompt oil attachment and coalescence on the surface.
Preliminary data show that superoleophobic membranes can also induce oil coalescence
and are less fouled than traditional hydrophobic membranes.

I dedicate this dissertation to my parents, Yi Yang and Hui Ma, for their endless love and
support.

iv

ACKNOWLEDGEMENTS

We gratefully acknowledge financial support from the U.S. Environmental Protection
Agency (Science to Achieve Results grant RC-83518301) and from the U.S. National
Science Foundation (Partnerships for International Education and Research Grant IIA1243433).
This dissertation would not have been completed without the support from so many
people. I would first like to thank my family. They are always believing and supporting me
in my every decision, which is very important to me. I would like to thank my advisor Dr.
Merlin Bruening and my co-advisor Dr. Volodymyr Tarabara. They are always there to
give me suggestions not only in research, also in presenting and writing. I’ve got much
inspiration from them. I would also like to thank my labmates and friends for their help
and company. They made my Ph.D. life much more interesting. Finally, I want to thank
everyone else who gave me help in both research and daily life during these 5 years.

v

TABLE OF CONTENTS

LIST OF TABLES ........................................................................................................... ix
LIST OF FIGURES .......................................................................................................... x
KEY TO ABBREVIATIONS...................................................................................... xviii
Chapter 1. Introduction.................................................................................................... 1
1.1. O/W Emulsions in Produced Water ..................................................................... 2
1.1.1. Oil and Grease in Produced Water................................................................ 2
1.1.2. Surfactant-Stabilized Oil-in-Water (O/W) Emulsions .................................. 2
1.2. Separation of O/W Emulsions .............................................................................. 4
1.2.1. Common Oil-Water Separation Methods ..................................................... 4
1.2.2. Conventional Methods for Demulsification.................................................. 5
1.2.3. Membrane Filtration ..................................................................................... 6
1.3. Polyelectrolyte Brushes ........................................................................................ 8
1.3.1. Synthesis via Surface-Initiated Atom Transfer Radical Polymerization (SIATRP).... ...................................................................................................................... 8
1.3.2. Wettability of Polyelectrolyte Brushes ....................................................... 10
1.4. Membrane Coalescence for Separation of O/W Emulsions ............................... 12
1.5. Outline of This Dissertation ............................................................................... 14
REFERENCES .............................................................................................................. 16
Chapter 2. Adsorption of Anionic or Cationic Surfactants in Polyanionic Brushes and
Its Effect on Brush Swelling and Fouling Resistance during Emulsion Filtration ... 22
2.1. Introduction ........................................................................................................ 22
2.2. Experimental ...................................................................................................... 25
2.2.1. Materials ..................................................................................................... 25
2.2.2. Preparation of PSPMK Brushes on Au-Coated Wafers and Nylon
Membranes ................................................................................................................ 25
2.2.2.1.
Initiator Immobilization....................................................................... 25
2.2.2.2.
Polymerization of SPMK..................................................................... 26
2.2.3. Characterization of Modified Au-Coated Wafers and Nylon Membranes . 27
2.2.4. Emulsion Preparation and Characterization................................................ 28
2.2.5. Dead-End Filtration .................................................................................... 29
2.3. Results & Discussion ......................................................................................... 32
2.3.1. Brush Synthesis on Au-Coated Wafers and Nylon Membranes ................. 32
2.3.2. Film Swelling and Contact Angles ............................................................. 36
2.3.3. Surfactant Interactions with PSPMK Brushes ............................................ 39
2.3.3.1.
Anionic Surfactant ............................................................................... 39
2.3.3.2.
Cationic Surfactant .............................................................................. 41
2.3.3.2.1. Effect of CTAB Concentration ......................................................... 41

vi

2.3.3.2.2. Effect of the Density of PSPMK Brushes on CTAB Adsorption and
Changes in Swelling ........................................................................................... 44
2.3.4. Emulsion Filtration with PSPMK-Modified Membranes ........................... 45
2.3.4.1.
Filtration of SDS-Stabilized Emulsions .............................................. 45
2.3.4.2.
Filtration of CTAB-Stabilized Emulsions ........................................... 49
2.3.5. Comparison of Filtration with PSPMK-Modified Membranes and NF270
Nanofiltration Membranes......................................................................................... 51
2.4. Conclusions ........................................................................................................ 53
REFERENCES .............................................................................................................. 54
Chapter 3. Aqueous Swelling of Zwitterionic Poly(Sulfobetaine Methacrylate)
Brushes in the Presence of Ionic Surfactants ............................................................... 60
3.1. Introduction ........................................................................................................ 60
3.2. Experimental ...................................................................................................... 62
3.2.1. Materials ..................................................................................................... 62
3.2.2. Preparation of PMEDSAH Brushes on Au-Coated Wafers and Nylon
Membranes ................................................................................................................ 63
3.2.2.1.
Initiator Immobilization....................................................................... 63
3.2.2.2.
Polymerization of MEDSAH............................................................... 64
3.2.3. Characterization of Modified Au-Coated Wafers and Nylon Membranes . 65
3.2.3.1.
Fourier Transform Infrared Spectroscopy (FTIR) ............................... 65
3.2.3.2.
Ellipsometry......................................................................................... 65
3.2.3.3.
X-Ray Photoelectron Spectroscopy (XPS) .......................................... 66
3.2.3.4.
Scanning Electron Microscopy (SEM) ................................................ 67
3.2.3.5.
Contact Angle Measurements .............................................................. 67
3.2.4. Determination of Membrane Permeability and Brush Swelling on
Membranes ................................................................................................................ 68
3.3. Results and Discussion ....................................................................................... 69
3.3.1. Brush Synthesis on Au-Coated Wafers and Nylon Membranes ................. 69
3.3.2. Brush Swelling and Wettability .................................................................. 73
3.3.3. Interactions between Ionic Surfactants and PMEDSAH Brushes ................... 75
3.3.3.1.
Anionic Surfactant (SDS) .................................................................... 76
3.3.3.1.1. Effect of PMEDSAH Brush Thickness on Swelling ........................ 76
3.3.3.1.2. Swelling as a Function of SDS Concentration ................................. 81
3.3.3.1.3. Swelling in Salt Solutions................................................................. 84
3.3.3.2.
Cationic Surfactant (DTAB) ................................................................ 86
3.3.3.2.1. Effect of Brush Thickness ................................................................ 87
3.3.3.2.2. Effect of DTAB Concentration......................................................... 88
3.3.3.3.
Dead-End Filtration of Surfactant Solutions with PMEDSAH-Modified
Membranes............................................................................................................. 90
3.4. Conclusions ........................................................................................................ 92
REFERENCES .............................................................................................................. 94
Chapter 4. Oil Coalescence on Superoleophobic Membrane Surfaces during DeadEnd and Cross-Flow Filtrations .................................................................................. 102
4.1. Introduction ...................................................................................................... 102

vii

4.2. Experimental .................................................................................................... 104
4.2.1. Materials ................................................................................................... 104
4.2.2. Preparation of Polyelectrolyte Brushes on Membranes ............................ 105
4.2.3. Contact Angle Measurement..................................................................... 105
4.2.4. Emulsion Preparation and Characterization.............................................. 105
4.2.5. Dead-End Filtration .................................................................................. 105
4.2.6. Visualization of Oil Coalescence on Membranes during Filtration.......... 107
4.3. Results and Discussion ..................................................................................... 108
4.3.1. Wettability of Membranes ........................................................................ 108
4.3.2. Dead-End Filtrations of O/W Submicron Emulsions ............................... 110
4.3.2.1.
Track-Etched Membrane ................................................................... 110
4.3.2.1.1. Bare Track-Etched Membrane........................................................ 110
4.3.2.1.2. Track-Etched Membranes Modified with PDA-BIBB ................... 113
4.3.2.1.3. Track-Etched Membrane Modified with PMEDSAH Brushes ...... 116
4.3.2.2.
Nylon Membranes ............................................................................. 117
4.3.2.2.1. Bare Nylon Membrane ................................................................... 117
4.3.2.2.2. Nylon Membranes Modified with PMEDSAH Brushes ................ 118
4.3.3. DOTM ....................................................................................................... 121
4.4. Conclusions ...................................................................................................... 122
REFERENCES ............................................................................................................ 123
Chapter 5. Summary and Future Work ..................................................................... 127
5.1. Research Summary ........................................................................................... 127
5.2. Future Work: Membrane Modification with Zwitterionic Brushes for Oil
Coalescence and Fouling Resistance........................................................................... 128
5.2.1. Membrane Characterization ...................................................................... 129
5.2.2. Dead-End and Cross-Flow Filtrations of the Same Emulsion. ................. 129
5.2.3. Compare Zwitterionic Brushes with Neutral Hydrophilic Brushes .......... 130
REFERENCES ............................................................................................................ 131

viii

LIST OF TABLES

Table 1.1. The comparison between membrane coalescence and fiber bed coalescer. .... 14
Table 1.2. Titles of the remaining chapters of this dissertation ........................................ 14
Table 2.1. Mass balance for the dead-end filtration of surfactant-stabilized emulsions
using PSPMK-modified (1-h polymerization time) and NF270 membranes, and an
impermeable transparency film. Values of n are the percentages of oil (relative to that in
the initial feed) found in the total permeate and retentate fractions, and on the cell after
the filtration. The value adsorbed on the cell was calculated by subtraction and represents
the amount of oil not accounted for in the permeate and retentate. The total filtration time
was 12 h. Initial studies with a teflon cell instead of the Amicon cell showed higher levels
of adsorption. .................................................................................................................... 32
Table 2.2. Membrane porosity determined by the mass loss of wet membranes after
drying. The PSPMK-modified membranes show lower porosity than the bare membrane.
........................................................................................................................................... 35
Table 4.1. Results of oil coalescence in dead-end filtration and crossflow (DOTM)
through bare and modified track-etched membranes. “Y” represents clear oil coalescence
and demulsification, whereas “N” indicates no obvious oil coalescence, even if a small
increase in oil droplet size appeared. .............................................................................. 121

ix

LIST OF FIGURES

Figure 1.1. Surfactant classification based on the charge of the hydrophilic head group,
and specific examples. SDS-sodium dodecyl sulfate; CTAB-cetrimonium bromide......... 3
Figure 1.2. Surfactant-stabilized oil droplets in an O/W emulsion..................................... 4
Figure 1.3. Membrane filtration classification based on the membrane pore size. The
column at the left lists the method, the range of pore diameters, and typical operating
pressures, whereas the figure shows solutes that the membrane rejects (bent red arrows)
or allows to pass (blue arrows). .......................................................................................... 7
Figure 1.4. (a) Diagram of membrane fouling when oil droplets attach to the surface. (b)
Qualitative plot of the fouling-induced decline in flux with time (curve II) during
constant-pressure filtration. In contrast, the flux through membranes that do not foul is
constant with time (b, line I). .............................................................................................. 7
Figure 1.5. Synthetic strategies for the preparation of polymer brushes: (a) grafting to
strategy, polymer synthesis occurs before attachment to the surface; (b) grafting from
strategy, monomers are polymerized from initiators anchored to the surface. ................... 8
Figure 1.6. Wetting of a silicon wafer modified with PSPMK (poly(3-sulfopropyl
methacrylate potassium salt)) brushes. (a) In air, the silicone oil droplet spreads on the
surface to give a low contact angle. (b) In water, silicone oil forms contact angles of
nearly 180°, and the oil droplet may detach from the surface. Reprinted from reference 44
with permission of the American Chemical Society......................................................... 10
Figure 1.7. The work of adhesion needed to separate oil from a surface in water. .......... 11
Figure 1.8. Proposed process of oil coalescence during flow through a porous
hydrophobic membrane. I. Oil droplets (yellow) adsorb and coalesce on the pore surface
(dashed gray). II and III. Oil droplets grow and eventually block the pores creating water
occlusions (blue). IV and V. Oil and water occlusions pass through the membrane
creating water-in-oil droplets, which are large enough to detach from membrane.18....... 12
Figure 2.1. Diagram of the apparatus for dead-end filtration. .......................................... 30
Figure 2.2. (a) GC calibration curve showing a linear relationship between the ratio of
GC peak area (hexadecane/decane) and the concentration of hexadecane at high
hexadecane concentration. Figure (b) shows the calibration curve at low concentration.
Hexadecane concentrations in all filtrations were calculated using the left curve.
Estimates of detection limits and low concentrations for determining high rejections were
obtained using the right curve. .......................................................................................... 31

x

Figure 2.3. Reflectance FTIR spectrum of (a) the monolayer formed from treatment of a
Au-coated wafer with a solution containing (BrC(CH3)2COO(CH2)11S)2 and (b) PSPMK
brushes on a Au-coated wafer. .......................................................................................... 33
Figure 2.4. “Dry” ellipsometric thicknesses (blue squares) and sulfonate IR absorbances
at 1219 cm-1 (red circles) as a function of polymerization time for PSPMK brushes grown
from an initiator monolayer on Au. Error bars represent standard deviations (n=12
determinations on 4 wafers). ........................................................................................... 344
Figure 2.5. ATR-IR spectra of a hydroxylated membrane (a) before and (b) after growth
of PSPMK brushes. Spectrum (c) results from subtracting 0.8 × spectrum (a) from
spectrum (b) and shows absorbances due primarily to the PSPMK brush. We employed
the factor of 0.8 to minimize infrared absorbances from the base nylon membrane. ....... 35
Figure 2.6. SEM images of (a, c) bare nylon membranes and (b, d) PSPMK-modified
membranes. Figures a and c are the top views of the membranes, where figures b and d
show cross sections of the membranes. ............................................................................ 36
Figure 2.7. The water-swollen thickness (blue squares) and percent swelling (red circles)
of PSPMK brushes as a function of their “dry thickness”. Error bars represent standard
deviations (n=12 determinations on 4 wafers). ................................................................. 37
Figure 2.8. Refractive indices of water-swollen PSPMK brushes as a function of their dry
thickness. When the dry thickness is <~100 nm, the refractive indices are around 1.36.
Refractive indices increase from 1.36 to 1.39 when dry thicknesses surpass 100 nm and
swelling decreases. Each point represents an average of refractive indices determined at
wavelengths ranging from 410 to 740 nm. ....................................................................... 37
Figure 2.9. Contact angles of hexadecane on PSPMK-modified wafers (a, b) and bare Aucoated wafers (c,d). The images were obtained in air (a, c) and water (b,d). ................... 38
Figure 2.10. Swollen thicknesses of PSPMK brushes on Au-coated wafers in water and in
solutions with different concentrations of SDS. Error bars represent standard deviations
(n=12 measurements on 4 wafers). ................................................................................... 39
Figure 2.11. Reflectance IR spectra of PSPMK brushes on Au-coated wafers before
(black line) and after immersion in 4 mM (red line) and 15 mM (blue line) SDS solutions.
Small changes in the intensity of sulfonate stretches with minimal change in the carbonyl
stretch may reflect a change in the film orientation or the environment around the
sulfonates after immersion in surfactant solutions and rinsing with water. However, the
spectra show no aliphatic stretches of surfactant around 2900 cm-1. ................................ 40
Figure 2.12. The water-swollen thicknesses of PSPMK brushes on Au-coated wafers
before (blue squares) and after (red circles) immersion in a 20 mM SDS solution.......... 41
Figure 2.13. Reflectance IR spectra of PSPMK brushes before (bottom) and after (top)
immersion in 2 mM CTAB for 12 h and rinsing with water. Growth of the PSPMK brush
employed 1-h polymerization. .......................................................................................... 42
xi

Figure 2.14. Percent change in swollen thickness (blue squares) and percent increase in
the CH2 IR absorbance at 2925 cm-1 (red circles) after immersion of brushes in solutions
with varying CTAB concentrations. Swelling was determined first in deionized water and
subsequently in the CTAB solution, whereas the IR absorbance was determined after
immersion in deionized water or CTAB solution and subsequent rinsing with water, and
drying with N2. All films were initially prepared using a 1-h polymerization time. Error
bars represent standard deviations (n=12 measurements on 4 wafers). ............................ 43
Figure 2.15. Contact angle of hexadecane in deionized water on a PSPMK-modified
wafer treated with CTAB. The wafer was exposed to 5 mM CTAB for 12 h and rinsed
with water before the contact angle measurement. ........................................................... 44
Figure 2.16. The change in the water-swollen thickness of PSPMK brushes in 3 mM
CTAB (blue squares) with respect to the thickness in deionized water, and the percent
increase in the –CH2 IR absorbance at 2925 cm-1 (red circles) as a function of the
sulfonate absorbance at 1219 cm-1 in the pristine film. IR absorbances were determined
after rinsing films with water and drying with N2. The increase in absorbance at 2925 cm1
stems from surfactant adsorption, and film density should increase with increasing
sulfonate absorbance. ........................................................................................................ 45
Figure 2.17. Size distributions of hexadecane submicron emulsions stabilized by SDS (a)
and CTAB (b). .................................................................................................................. 46
Figure 2.18. Specific fluxes (blue squares) in dead-end filtration of SDS-stabilized
hexadecane submicron emulsions through PSPMK-modified membranes. The SPMK
polymerization time was 1 h (a) or 45 min (b). The green horizontal lines show the pure
water flux before filtration. Initially, the emulsion contained 24.3 mM hexadecane in 0.36
mM SDS............................................................................................................................ 47
Figure 2.19. Pure water flux (after a 12-h filtration of an SDS-stabilized emulsion) for
membranes modified using (a) 1-h and (b) 45-min SI-ATRP of SPMK. ......................... 47
Figure 2.20. Replicate dead-end filtrations. (a), (b): filtration of SDS-stabilized
hexadecane emulsions with bare nylon membranes with 1.2 Îźm pores. The specific flux
(blue circles) drops at the beginning due to membrane fouling and increases quickly after
the oil permeates through the membrane in 1.5 min. The hexadecane concentration (red
circles) falls to zero because deionized water replenishes the feed cell. (c): filtration of an
SDS-stabilized hexadecane emulsion with a PSPMK-modified membrane
(polymerization time: 45min). The permeate flux (blue circles) increases after 5h of
filtration, and visible oil appears in the permeate after that time...................................... 48
Figure 2.21. Permeate hexadecane concentrations (red circles) and specific fluxes (blue
squares) during dead-end filtration of a CTAB-stabilized hexadecane submicron
emulsion through a PSPMK-modified membrane. The horizontal green line shows the
pure water flux prior to the filtration. The membrane was modified with a 1-h
polymerization. The PSPMK-modified membrane showed low oil rejection, and all of the
hexadecane permeated through the membrane in 4 h. The initial hexadecane feed

xii

concentration was 17.9 mM, but this value declined rapidly during the filtration due to
low rejection and cell replenishment with deionized water. ............................................. 50
Figure 2.22. Cartoon of filtration of CTAB-stabilized submicron emulsions with PSPMK
modified-membranes: (a) polymer brushes swell well in water prior to contact with
CTAB and oil droplets stabilized by CTAB; (b) positively charged oil droplets adsorb to
the negatively charged PSPMK to cause brush collapse. The blue, yellow and green balls
represent water, oil droplets and CTAB molecules, respectively. The brush modification
likely extends much farther into the membrane than the figure shows. ........................... 51
Figure 2.23. Evolution of specific permeate flux with time during filtration of aqueous
submicron emulsions through NF270 membranes. Blue squares and red circles represent
filtration with an SDS-stabilized and a CTAB-stabilized emulsion, respectively. The
horizontal green line shows the average pure water flux. ................................................. 52
Figure 3.1. Optical models for PMEDSAH brushes in (a) air and (b) solutions. Image (c)
shows the model we used to determine optical constants of surfactant solutions. ........... 66
Figure 3.2. Apparatus for the measurement of hexadecane contact angles on wafers in
surfactant solutions. .......................................................................................................... 68
Figure 3.3. Diagram of the apparatus for dead-end filtration. .......................................... 69
Figure 3.4. Reflectance FTIR spectrum of (a) the disulfide initiator ((BrC(CH3)2COO(CH2)11S)2) and (b) PMEDSAH brushes on a Au-coated Si wafer. The brush dry thickness
was 412 nm. ...................................................................................................................... 71
Figure 3.5. Dry brush thickness (blue circles) and IR absorbances of sulfonate groups at
1218 cm-1 (red squares) as a function of polymerization time for PMEDSAH brushes
grown on Au wafers. Error bars represent the standard deviation of measurements on
three different films........................................................................................................... 71
Figure 3.6. XPS wide scan spectra of (a) bare nylon membranes and membranes modified
with (b) PDA-BIBB initiator and (c) PMEDSAH brushes. .............................................. 72
Figure 3.7. SEM images of (a,d) bare nylon membranes and membranes modified with
(b,e) PDA-BIBB initiator and (c, f) PMEDSAH brushes. Figures a, b, c are the top-view
images while Figures d, e, f are the cross-sectional images. Modification employed 48 h
of PDA-BIBB deposition and 24 h of polymerization...................................................... 73
Figure 3.8. Water-swollen thicknesses (red squares) and percent swelling (blue circles) of
PMEDSAH brushes as a function of their dry thickness. Error bars represent standard
deviations of measurements on three films. Error bars are not always visible due to their
small values. Swollen thicknesses were determined in water after a 12-h immersion. .... 74
Figure 3.9. Reflectance IR absorbance at 1218 cm-1 (dry films) divided by the swollen
PMEDSAH brush thickness in water as a function of the dry film thickness. The ratio of
IR absorbance to swollen film thickness provides a measure of density. The plot is
xiii

divided into two regions according to dry brush thickness: I (< 20 nm), II (> 20 nm).
Error bars represent standard deviations of measurements on three films. ...................... 75
Figure 3.10. Contact angles of hexadecane on bare Au-coated wafers (a, b), PMEDSAHmodified wafers (c, d) and PMEDSAH-modified membranes (e, f). The images were
obtained in air (a, c, e) and water (b, d, f). Images in water are rotated 180°. ................. 75
Figure 3.11. Reflectance IR spectra of a PMEDSAH brush before (red) and after (blue)
immersion in a 6 mM SDS solution and rinsing with water. The brush dry thickness was
340 nm. ............................................................................................................................. 76
Figure 3.12. Reflectance IR spectra of a PMEDSAH brush after immersion in 6 mM SDS
followed by a water rinse (red), and after immersion in the same SDS solution without a
water rinse (blue). In the latter case, the film was dried with a Kimwipe to remove
residual solution. A bare Au-coated surface after immersion in 6 mM SDS without water
rinsing (dried with a Kimwipe) served as a control (green). The figure on the right
expands the region of the spectrum where CH2 stretches appear. The brush dry thickness
was 340 nm. ...................................................................................................................... 77
Figure 3.13. Ellipsometric thicknesses of swollen PMEDSAH brushes in 6 mM SDS
(blue circles), as well as in deionized water before (red squares) and after (green
triangles) immersion in SDS solution and rinsing. Thicknesses are plotted as a function of
dry brush thickness in ambient air, and dashed vertical lines depict dry thicknesses of 20
and 150 nm. Error bars represent standard deviations of measurements on three films.
The error bars are often not obvious due to their small values, and red and green symbols
typically overlap................................................................................................................ 78
Figure 3.14. PMEDSAH brush density, as indicated by IR absorbance (1218 cm-1 in dry
films) divided by the brush thickness in water (red squares) or by the brush thickness in 6
mM SDS (blue circles) as a function of dry brush thickness. The plot is divided into three
regions according to dry brush thickness: I (< 20 nm), II (20-150 nm), III (> 150 nm).
Error bars represent standard deviations of measurements on three films. Some of the
error bars are not visible due to their small values. .......................................................... 78
Figure 3.15. Percent thickness difference between PMEDSAH films in 6 mM SDS and in
water as a function of the dry brush thickness. The plot is divided into three regions
according to dry brush thickness: I (< 20 nm), II (20-150 nm) and III (> 150 nm). Error
bars represent standard deviations of measurements on three films. Some of the error
bars are not visible due to their small values. ................................................................... 79
Figure 3.16. Contact angles of hexadecane on (a) PMEDSAH brush-modified Au-wafer
in 1 mM SDS, (b) PMEDSAH brush-modified Au-wafer in 6 mM SDS, (c) bare Auwafer in 1 mM SDS and (d) bare Au-wafer in 6 mM SDS. PMEDSAH brushes in (a) and
(b) have a dry thickness of 260 nm. .................................................................................. 81
Figure 3.17. Percent thickness difference of PMEDSAH brushes in SDS solutions
(compared to in water) as a function of SDS concentration and dry thickness. Symbols in
different colors and shapes represent brushes with dry thickness of 33 nm (blue empty
xiv

squares), 24 nm (green empty circles), 33 nm (blue empty squares), 67 nm (green solid
diamonds), 136 nm (yellow solid triangles), 200 nm (red solid squares), and 267 nm
(brown solid circles). Error bars represent standard deviations of measurements on three
films. The Y error bars are not obvious due to their small values. ................................... 82
Figure 3.18. Ellipsometric thicknesses (a, c, e, g, i, k, m) and percent thickness
differences relative to in deionized water (b, d, f, h, j, l, n) of swollen PMEDSAH brushes
in 0.5, 1, 2, 4, 12, 16, 20 mM SDS as a function of dry brush thickness. In the figures on
the left, thicknesses of swollen PMEDSAH brushes in SDS, as well as in deionized water
before and after immersion in SDS solution and rinsing are plotted as blue circles, red
squares and green triangles, respectively. Error bars represent standard deviations of
measurements on three films. Some error bars are not visible due to their small values.
All the data in Figure 3.18 was measured using the same batch of samples in the
following order: 0.5, 1, 6, 12, 20, 4, 2, 16 mM SDS. ....................................................... 84
Figure 3.19. Swollen thicknesses of PMEDSAH brushes in 5 mM NaCl (blue circles),
and the deionized water-swollen thicknesses before (red squares) and after (green
triangles) immersion in NaCl (and rinsing) as a function of dry brush thickness. Error
bars represent standard deviations of measurements on three films. Some error bars are
not visible due to their small values.................................................................................. 85
Figure 3.20. Percent thickness difference between PMEDSAH films in 5 mM NaCl and
in water as a function of the dry brush thickness. Error bars represent standard deviations
of measurements on three films. ....................................................................................... 85
Figure 3.21. Swollen thicknesses of PMEDSAH brushes in 500 mM NaCl (blue circles),
and the water-swollen thickness before (red squares) and after (green triangles)
immersion in NaCl (and rinsing) as a function of dry brush thickness. Error bars represent
standard deviations of measurements on three films. Some error bars are not visible due
to their small values. ......................................................................................................... 86
Figure 3.22. Reflectance IR spectra of a PMEDSAH brush after immersion in 10 mM
DTAB and a water rinse (red), and after immersion in 10 mM DTAB without a water
rinse (blue). In the latter case, the film was dried with a Kimwipe to remove residual
solution. A bare Au-coated surface after immersion in 10 mM DTAB without water
rinsing (dried with a Kimwipe) served as a control (green). The brush dry thickness was
340 nm. ............................................................................................................................. 86
Figure 3.23. Reflectance IR spectra of a PMEDSAH brush before (red) and after (blue)
immersion in 10 mM DTAB solution and rinsing. The brush dry thickness was 340 nm.
........................................................................................................................................... 87
Figure 3.24. Swollen thickness of PMEDSAH brushes in 5 mM DTAB solution (blue
circles), and the water-swollen thickness before (red squares) and after (green triangles)
immersion in DTAB solutions and rinsing as a function of dry brush thickness. Error bars
represent standard deviations of measurements on three films. Some error bars are not
obvious due to their small values. ..................................................................................... 88

xv

Figure 3.25. Ellipsometric thicknesses (a, c, e, g) and percent thickness differences
relative to in deionized water (b, d, f, h) of swollen PMEDSAH brushes in 1, 2, 10, and
20 mM DTAB as a function of dry brush thickness. Thicknesses of swollen PMEDSAH
brushes in DTAB, as well as in deionized water before and after immersion in DTAB and
rinsing are plotted as blue circles, red squares and green triangles, respectively. Error bars
represent standard deviations of measurements on three films. Some error bars are not
visible due to their small values. ....................................................................................... 88
Figure 3.26. Dead-end filtrations of 6 mM SDS (red squares) and 2 mM DTAB (green
triangles) through (a) PMEDSAH-coated nylon membranes and (b) bare nylon
membranes. Pure water permeability (blue circles) was determined between surfactant
filtrations. .......................................................................................................................... 90
Figure 3.27. Replicate dead-end filtrations of 6 mM SDS (red squares) and 2 mM DTAB
(green triangles) through (a) PMEDSAH-coated nylon membranes and (b) bare nylon
membranes. Pure water permeability (blue circles) was determined before and after
surfactant filtrations. These experiments were performed with a different set of
membranes than Figure 3.26. ............................................................................................ 92
Figure 4.1. GC calibration curve showing a linear relationship between the ratio of GC
peak area (hexadecane/decane) and the concentration of hexadecane at high hexadecane
concentration. .................................................................................................................. 106
Figure 4.2. Apparatus for DOTM system. Reprinted from reference 26 with permission
from Elsevier................................................................................................................... 107
Figure 4.3. Contact angles of hexadecane (with red dye) in water on nylon membranes
with a pore size of 0.45 Âľm after coating with PDA-BIBB and PMEDSAH brushes. The
PMEDSAH polymerization time was 4 h, and the PDA-BIBB deposition times were (a) 1
h, (b) 2 h, (c) 3.5 h and (d) 5.5 h. .................................................................................... 109
Figure 4.4. Contact angles of hexadecane (with red dye) in water on nylon membranes
with a pore size of 0.45 Âľm after coating with PDA-BIBB and PMEDSAH brushes. The
deposition time of PDA-BIBB is 4 h. Polymerization time for PMEDSAH brushes is (a)
2 h, (b) 4 h and (c) 6 h. A bare nylon membrane (d) served as a control. In Figures (a),
(b), and (c), the droplet is not visible because it detaches from the membrane. ............. 109
Figure 4.5. Permeate flux and oil rejection during dead-end filtration of an SDS-stabilized
emulsion through a bare track-etched membrane. .......................................................... 111
Figure 4.6. Oil droplet size before (a) and after (b) dead-end filtration of an SDSstabilized emulsion through a bare track-etched membrane. ........................................ 1111
Figure 4.7. Images of an SDS-stabilized O/W submicron emulsion at different times
during filtration through a bare track-etched membrane with 0.2 Îźm pores. ................. 111
Figure 4.8. Permeate flux and oil rejection during dead-end filtration of a CTABstabilized emulsion through a bare track-etched membrane. .......................................... 112
xvi

Figure 4.9. Images of a CTAB-stabilized O/W submicron emulsion at different times
during filtration through a track-etched membrane with 0.2 Îźm pores. ......................... 112
Figure 4.10. Permeate flux (blue) and oil rejection (red) during the dead-end filtration of
an SDS-stabilized emulsion through a track-etched membrane modified with PDA-BIBB
(the PDA-BIBB deposition time is 48 h). ....................................................................... 114
Figure 4.11. Images of an SDS-stabilized O/W submicron emulsion at different times
during dead-end filtration through a track-etched membrane modified with PDA-BIBB
(the deposition time is 48 h)............................................................................................ 114
Figure 4.12. Permeate flux (blue) and oil rejection (red) during the dead-end filtration of
CTAB-stabilized emulsion using a track-etched membrane modified with PDA-BIBB
(the deposition time is 48 h)............................................................................................ 115
Figure 4.13. Oil droplet size distribution before (a) and after (b) dead-end filtration of a
CTAB-stabilized emulsion through a track-etched membrane modified with PDA-BIBB
(deposition time: 48 h). ................................................................................................... 116
Figure 4.14. Permeate flux (blue) and oil rejection (red) during the dead-end filtration of
an SDS-stabilized emulsion through a track-etched membrane modified with PMEDSAH
brushes (the polymerization time was 12 h). .................................................................. 116
Figure 4.15. Oil droplet size before (a) and after (b) dead-end filtration (10 h) of an SDSstabilized emulsion through a track-etched membrane modified with PMEDSAH brushes
(polymerization time: 12 h). ........................................................................................... 117
Figure 4.16. Permeate flux (blue) and oil rejection (red) during dead-end filtration of an
SDS-stabilized emulsion through a bare nylon membrane with a pore size of 0.45 Âľm.
The photos show the O/W submicron emulsion at different times during the filtration. 118
Figure 4.17. Permeate flux (blue) and oil rejection (red) during the dead-end filtration of
a CTAB-stabilized emulsion through a bare nylon membrane with a pore size of 0.45 Âľm.
The photo shows the O/W submicron emulsion after 2 h of filtration. .......................... 118
Figure 4.18. Permeate flux (blue) and oil rejection (red) during dead-end filtration of an
SDS-stabilized emulsion through a nylon membrane modified with PMEDSAH brushes
(24 h polymerization time). The photo shows the O/W submicron emulsion after 2 hfiltration. The pure water flux through the modified membrane was 2600 LMH/Bar. .. 119
Figure 4.19. Permeate flux (blue) and oil rejection (red) during dead-end filtration of an
SDS-stabilized emulsion through a nylon membrane modified with PSPMK brushes (the
polymerization time was 24 h). ....................................................................................... 120
Figure 4.20. Permeate flux (blue) and oil rejection (red) during dead-end filtration of a
CTAB-stabilized emulsion through a nylon membrane modification with PMEDSAH
brushes (24-h polymerization time). ............................................................................... 121

xvii

KEY TO ABBREVIATIONS

SBMA

sulfobetaine methacrylate

O/W emulsion

oil-in-water emulsion

SI-ATRP

surface-initiated atom transfer radical polymerization

CRP

controlled radical polymerization.

SPMK

3-sulfopropyl methacrylate potassium salt

MEDSAH

[2-(methacryloyloxy)ethyl]dimethyl-(3-sulfopropyl)
ammonium hydroxide

PDA

polydopamine

FTIR

Fourier-transform infrared spectroscopy

SDS

sodium dodecyl sulfate

CTAB

cetrimonium bromide

DTAB

dodecyltrimethylammonium bromide

BSA

bovine serum albumin

BIBB

Îą-bromoisobutyryl bromide

PPEGMA

poly(poly(ethylene glycol) methacrylate)

PHEMA

poly(2-hydroxyethyl methacrylate)

XPS

X-ray photoelectron spectroscopy

xviii

SEM

scanning electron microscope

AFM

atomic force microscope

xix

Chapter 1. Introduction
This dissertation aims to develop polyanionic and polyzwitterionic brushes as
superoleophobic coatings that resist fouling during filtration of surfactant-stabilized oil-inwater (O/W) emulsions. Polyanionic brushes swell greatly in water and do not adsorb
anionic surfactants due to electrostatic repulsion. Remarkably, microfiltration membranes
modified with these brushes reject >99.9% of the oil from emulsions stabilized by anionic
surfactants, and the permeate flux does not decline over time, which demonstrates excellent
resistance to fouling. In contrast, electrostatic attraction induces strong adsorption of
cationic surfactants or cationic surfactant-stabilized oil droplets to polyanionic brushes.
Thus, I examined zwitterionic poly(sulfobetaine methacrylate) (PSBMA) brushes as
coatings that may resist fouling in solutions containing either cationic or anionic
surfactants. However, these brushes adsorb both anionic and cationic surfactants, and this
sorption strongly depends on thickness of the swollen brushes, surfactant charge and
surfactant concentration. The swelling of PSBMA brushes in surfactant solutions results
from a combination of charge-screening, disordering of water molecules around
zwitterions and the surfactant hydrophobicity. While developing these brushes, I obtained
preliminary filtration data that suggest that oil droplets coalesce on some membrane
surfaces, which could provide another approach for combatting fouling.
This chapter first introduces O/W submicron emulsions and describes the
importance of removing oil from these mixtures. The second section of this introduction
reviews common approaches for oil-water separation, emphasizing membrane filtration as
an effective method, especially for submicron O/W emulsions. I then present background
on polyelectrolyte brushes as effective anti-fouling materials, and briefly explore their

1

synthesis and underwater superoleophobicity. Finally, the fourth section gives a brief
description of oil coalescence on membranes, which is a much less common strategy than
membrane filtration but may provide an effective alternative mechanism for separating
O/W emulsions. At the end of this chapter, I briefly outline the rest of this dissertation.
1.1.

O/W Emulsions in Produced Water

1.1.1. Oil and Grease in Produced Water
Produced water, the wastewater that stems from oil and gas extraction from underground
deposits, is a massive, unwanted byproduct in the oil industry.1 According to a 2012 market
study, fossil fuel extraction creates more than 115 billion barrels of produced water each
year. Generation of each barrel of oil leads to around 3 barrels of produced water. This
ratio was 8/1 (barrels of produced water/barrels of oil) for the United States in 2012 and
will likely reach 12/1 by 2027. In the worst cases, this ratio may increase to 50/1.2 Oil and
grease are among the major components in produced water, with a wide concentration
range from 20-200000 mg/L.3 As discharged oil and grease may significantly damage the
environment, the US Environmental Protection Agency issues regulations governing their
discharge. For example, the maximum concentration of oil and grease in the effluents of
any new oil and gas facilities cannot exceed 42 mg/L at any time. Meanwhile, the average
concentration in effluents for 30 successive days cannot exceed 29 mg/L.4 These tight
regulations of oil discharge along with the growth of oil and gas industries require new
advances in efficient oil-water separations.
1.1.2. Surfactant-Stabilized Oil-in-Water (O/W) Emulsions
Based on the sizes of oil droplets, oil and grease in produced water exist in three categories:
free oil (> 150 Âľm), dispersed oil (20-150 Âľm) and emulsified oil (< 20 Âľm).5 Emulsions

2

with oil droplets smaller than 1 Âľm are termed submicron emulsions6 and are common in
drug delivery,7,

8

pesticide formulations,9 food processing,10 cosmetics11 and emulsion

polymerization.12, 13 Separation of these tiny oil droplets from water is very challenging,
particularly because surfactants stabilize these droplets.
Surfactants are amphiphilic compounds with both hydrophilic and hydrophobic
moieties. According to the charge on the hydrophilic groups, surfactants fall into four
groups: cationic, anionic, nonionic and zwitterionic (Figure 1.1). The amphiphilicity of
surfactants allows them to reside at the oil-water interface, increasing oil solubility in water
and preventing coalescence of oil droplets (Figure 1.2).

Figure 1.1. Surfactant classification based on the charge of the hydrophilic head group, and specific
examples. SDS-sodium dodecyl sulfate; CTAB-cetrimonium bromide.

3

Figure 1.2. Surfactant-stabilized oil droplets in an O/W emulsion.

1.2.

Separation of O/W Emulsions

1.2.1. Common Oil-Water Separation Methods
For produced water, oil-water separation typically occurs in three stages.14, 15 The primary
treatment relies on the density difference between oil and water and uses instruments such
as the American Petroleum Institute gravity separator. In general, solid particles will settle
to the bottom and the oil will rise to the surface, leaving the wastewater in the middle phase.
However, the primary treatment can only isolate free oil (droplets larger than 150 Âľm).16
The remaining oil droplets are too small to separate from water using density differences.
Stokes’ equation (Eq. 1.1) illustrates this point as it describes droplets’ vertical velocity, V,
V=

d2 g(ρw −ρo )
18Âľw

Eq. 1.1

as a function of the droplet diameter, d, the densities of water and oil, ρw and ρo respectively,
and the dynamic viscosity of water, Îźw. (The variable g represents the acceleration due to

gravity.) Although oil droplets will float to the surface due to their low density relative to
water, small droplet diameters lead to a low flotation rate and a low separation efficiency.
Thus, the secondary separation stage aims to coalesce the oil droplets. Common
methods include gas flotation and the use of plate separators. Gas flotation induces gas

4

bubbles to contact and coalesce with oil droplets, and the oil-gas aggregates rise to the
surface rapidly due to their low densities and large sizes.17 As their name implies, plate
separators employ plates to provide more surface area for oil attachment and coalescence,
leading to more separation than in simple gravity separators. However, these conventional
coalescing methods focus on oil droplets between 50 and 100 Âľm in the absence of
surfactants.18, 19
Treatment of surfactant-stabilized emulsified oil occurs primarily at the tertiary
stage. One common method in this stage is membrane filtration. This technique effectively
separates oil from emulsions, even submicron O/W emulsions, and is becoming more
common in both lab-scale and industrial applications. Section 1.2.3 discusses membrane
filtration in more detail.
1.2.2. Conventional Methods for Demulsification
Due to the stability of emulsified oil, breaking emulsions into oil and water phases, termed
demulsification, is often indispensable before separation. The traditional demulsification
methods mainly include chemical addition and thermal treatment (evaporation and
incineration). Both of these techniques are effective under certain conditions but suffer
from shortcomings.20 Chemical demulsification is very effective when a proper demulsifier
is available. For example, addition of HCl to emulsions stabilized by ester surfactants will
destabilize the system due to the surfactant decomposition.21 However, chemical
demulsification is highly susceptible to changes in the influents. Analysis of the emulsion
chemical composition is vital for choosing the right demulsifier. Other limitations include
mechanical problems due to clogging of chemical feeding lines, sludge generation and
corrosion. Thermal treatment is more universal since it doesn’t rely on specific chemical

5

interactions. However, heating leads to high energy costs, and oil evaporation may produce
a contaminated condensate.
1.2.3. Membrane Filtration
Compared to conventional oil-water separation methods, membrane filtration has several
advantages:20, 21 (1) It doesn’t require extraneous chemicals in most cases. (2) Energy costs
are much lower than in thermal treatment. (3) The operation is amenable to automation. (4)
Low weight and small space requirements give rise to low installation costs. (5) Because
of the sieving-based barrier technology, the quality of the treated stream is not highly
susceptible to the properties of the influent stream, especially when compared with
chemical demulsification. (6) Membrane filtration has the potential to combine with other
separation methods such as hydrocyclones.
Membrane filtration can be divided into four categories according to the membrane
pore size: microfiltration (MF), ultrafiltration (UF), nanofiltration (NF) and reverse
osmosis (RO). As Figure 1.3 shows, as the pore size decreases, membrane filtration
requires higher operating pressures but the minimum size of the rejected species
decreases.22, 23
Like all the other methods for oil-water separation, membrane filtration has its
limitations. Most importantly, fouling greatly decreases the water-permeability of
membranes. Oil droplets will attach to the membrane and block the pores during filtration,
leading to decreased hydrophilicity and a permeate flux that declines with filtration time
(Figure 1.4). Thus, fouling requires membrane cleaning or even replacement, which leads
to a high maintenance cost. Several approaches may reduce fouling, including the use of

6

vibratory or centrifugal devices,24, 25 membrane modification,26-28 and pretreatment of feed
Water

Monovalent Multivalent Viruses Bacteria
Ions
Ions

Suspended
Solids

Microfiltration (MF)
0.05-2.0 Âľm
10-100 psig
Ultrafiltration (UF)
0.003-0.1 Âľm
30-150 psig
Nanofiltration (NF)
0.001-0.01 Âľm
150-400 psig
Reverse Osmosis (RO)
< 0.001 Âľm
200-1200 psig

Figure 1.3. Membrane filtration classification based on the membrane pore size. The column at the
left lists the method, the range of pore diameters, and typical operating pressures, whereas the figure
shows solutes that the membrane rejects (bent red arrows) or allows to pass (blue arrows).

solutions.29 Among these methods, membrane modification may prove the least expensive
if modifications are stable. This work focuses on modifying membranes with
polyelectrolyte brushes, so the next section introduces methods for preparing and
characterizing such coatings.
a

b

I

II

Figure 1.4. (a) Diagram of membrane fouling when oil droplets attach to the surface. (b) Qualitative
plot of the fouling-induced decline in flux with time (curve II) during constant-pressure filtration.
In contrast, the flux through membranes that do not foul is constant with time (b, line I).

7

1.3.

Polyelectrolyte Brushes

1.3.1. Synthesis via Surface-Initiated Atom Transfer Radical Polymerization (SI-ATRP)
Polyelectrolyte brushes are among the most popular and well-studied anti-fouling
coatings.30-33 They consist of polyelectrolytes with one end attached to an interface,34 and
both grafting to and grafting from strategies can yield such structures (Figure 1.5).35 The
grafting to strategy refers to the attachment of prefabricated polymers via either
physisorption or covalent bonds to a substrate. Although this strategy is feasible, it cannot
produce thick and dense polymer brushes because steric repulsion between the coating and
incoming chains limits the chain attachment.36,

37

In the grafting from method,

polymerization occurs directly from immobilized initiators. This enables the growth of
thick, dense brushes because small incoming monomers can readily diffuse to initiators or
growing chain ends. Controlled radical polymerization (CRP) techniques are particularly
attractive for the grafting from strategy because they allow control over the thickness,
composition and architecture of polymer brushes and are compatible with many functional
groups in both aqueous and organic media.34 Among different CRP techniques for brush
fabrication, surface-initiated atom transfer radical polymerization (SI-ATRP) is especially
versatile as it provides controlled polymerization of a wide range of functional monomers.
b

a

Figure 1.5. Synthetic strategies for the preparation of polymer brushes: (a) grafting to strategy,
polymer synthesis occurs before attachment to the surface; (b) grafting from strategy, monomers
are polymerized from initiators anchored to the surface.

8

ATRP relies on the reversible redox activation of a dormant alkyl halide (R-X),
which is either the initiator or the end-group on a growing polymer chain, using a transition
metal complex (Mtn-Y/Ligand, Y is the counterion or another ligand and Mt is a transition
metal ion). Scheme 1.1 shows a general mechanism of ATRP.38 Homolytic cleavage of the

Scheme 1.1. Transition-metal-catalyzed ATRP.38

R-X bond through a one-electron oxidation of the transition metal complex reversibly
produces a reactive radical. The rate constants for the forward and reverse reactions are kact
and kdeact, respectively. Meanwhile, in addition to deactivation the generated radicals may
undergo both polymerization with a rate constant kp and termination with a rate constant kt.
To achieve well-controlled ATRP, the concentration of radicals should be low to minimize
radical termination, and this requires a rapid reversible deactivation. Moreover, fast
initiation is also necessary for uniform polymer growth. A successful SI-ATRP reaction
with uniform and adjustable brush growth needs optimization of the various constituents
of the reaction such as the initiator, catalyst and solvent. Among these factors, the catalyst
is the key to ATRP since it determines the equilibrium state of the radical formation and
deactivation reactions.39
Despite excellent control of polymerization under many conditions, SI-ATRP
suffers two major problems in the fabrication of polyelectrolyte brushes on membranes.
Considering the chemical compatibility of membrane substrates, water is the preferred
solvent. However, prior studies report that the ATRP catalyst deactivates by forming

9

complexes with charged monomers in water (Scheme 1.2).40 Moreover, water will
accelerate the disproportionation of Cu (I), leading to the loss of catalyst.41 These side
reactions may lead to a less controlled polymerization and a relatively rough brush surface.

Scheme 1.2. Catalyst deactivation by charged monomers in water.40

1.3.2. Wettability of Polyelectrolyte Brushes
Several studies reported that polyelectrolyte brushes form underwater superoleophobic
surfaces,42-44 which are defined as surfaces with an oil contact angle larger than 150° with
contact angle hysteresis < 10°.45, 46 For example, as Figure 1.6 shows, silicone oil spreads
on polyelectrolyte brushes in air, whereas in water oil will bead up and detach from same
b

a

Figure 1.6. Wetting of a silicon wafer modified with PSPMK (poly(3-sulfopropyl methacrylate
potassium salt)) brushes. (a) In air, the silicone oil droplet spreads on the surface to give a low
contact angle. (b) In water, silicone oil forms contact angles of nearly 180°, and the oil droplet may
detach from the surface. Reprinted from reference 44 with permission of the American Chemical
Society.

surface. This phenomenon results from a small work of adhesion, which is the reversible
work needed to separate a unit area of one phase from another.47, 48 As shown in Figure 1.7,
the work of adhesion for oil and a surface in water can be expressed as:

10

Wows = γsw + γow − γso

Eq. 2

where Wows is the work of adhesion needed to separate oil from a solid surface. The terms
Îłsw, Îłow and Îłso represent the surface free energies of a solid-water interface, an oil-water
interface and a solid-oil interface, respectively. Combining Eq. 2 and Young’s Equation
(Eq. 3) yields a more useful expression relating the work of adhesion with the contact angle
(Eq. 4).
γsw − γso = γow cosθ

Wows = γow (1 + cos θ)

Eq. 3
Eq. 4

where θ is the contact angle between oil and the solid surface under water.
Because of their strong affinity for water, polyelectrolyte brushes become
superoleophobic with a large contact angle of oil under water.42 According to Eq. 4, a large
oil contact angle leads to a small work of adhesion, indicating that oil can detach from the
surface easily. Thus, polyelectrolyte brushes are good choices as materials that resist
fouling with oil.
Water

Îłso

Oil
Surface

Îłow

Îłsw

Oil

Surface

Figure 1.7. The work of adhesion needed to separate oil from a surface in water.

11

1.4.

Membrane Coalescence for Separation of O/W Emulsions

As discussed above, conventional coalescence techniques can hardly separate surfactantstabilized O/W emulsions, especially submicron emulsions. However, Nitsch and
coworkers reported that emulsified oil droplets will coalesce as they are pumped through a
hydrophobic membrane.18, 49 They proposed a mechanism for this process assuming that
oil droplets
oil
water

water
Step II

Step I

Step III

oil
water
water

oil

Step V

Step IV

Figure 1.8. Proposed process of oil coalescence during flow through a porous hydrophobic
membrane. I. Oil droplets (yellow) adsorb and coalesce on the pore surface (dashed gray). II and
III. Oil droplets grow and eventually block the pores creating water occlusions (blue). IV and V.
Oil and water occlusions pass through the membrane creating water-in-oil droplets, which are large
enough to detach from membrane.18

12

the porous membrane is a network of capillaries. As Figure 1.8 shows, emulsified oil
droplets first attach to the hydrophobic membrane surface (Step I), followed by contact and
coalescence with nearby oil droplets (Step II). As the oil aggregates become large enough
to cover the membrane pores, water occlusion occurs (Step III) and leads to the formation
of W/O (water-in-oil) droplets at the membrane outlet (Steps IV, V). Compared with the
original tiny oil droplets, these W/O aggregates are much larger. Therefore, they can float
to the surface rapidly according to Stokes’ equation (Eq. 1).
Using several filtration membranes, Agarwal and co-workers investigated the
factors that affect oil coalescence including filter pore size, surface roughness and surface
wettability.50, 51 They found that an effective coalescing filter should meet the following
criteria. Firstly, the filter should be hydrophobic to enable oil attachment and coalescence.
Secondly, flow through the membrane should occur vertically since in this direction oil
droplets will possess high mobility. This configuration may also enhance oil droplet contact
and coalescence. Thirdly, the filter should have a rough surface to promote oil adhesion
that induces oil coalescence. Based on these fundamental studies, several groups
successfully fabricated hydrophobic coalescing membranes for emulsion separation.52-55
Nevertheless, as a technique in development, in-membrane coalescence still faces several
challenges, such as low coalescence of submicron oil droplets or droplets stabilized by
mixed surfactants. Still, membrane coalescence was already more effective in coalescing
oil than a common fiber bed coalescer, as shown in Table 1.1.18 For example, membrane
coalescence can separate surfactant-stabilized oil droplets whereas a fiber bed coalescer is
only effective for oil droplets with not enough or even no surfactants. In addition,

13

membrane coalescence provides a higher flow rate and better separation efficiency, which
is defined as

where Vorg,

ω

and Vorg,

Îą

Ρ=

Vorg,ω
Vorg,Îą

× 100%

Eq. 5

represent separated and initial volume of organic phase,

respectively.
Table 1.1. The comparison between membrane coalescence and fiber bed coalescer.
Membrane Coalescence

1.5.

Fiber bed Coalescer

Surfactant

yes

partial

No

O/W conc. (vol, %)

0.1-7

2-5

0.1-0.2

Droplet size (Âľm)

1-10

< 100

10-100

Flow rate (m3/m2h)

3-660

2.5-6.5

7-47

Separation efficiency (%)

80-98

complete/partial

50-98

Outline of This Dissertation
As shown in Table 1.2., Chapter 2 describes (a) the controlled growth of

polyanionic brushes composed of poly(3-sulfopropyl methacrylate salts) (PSPMK), (b)
brush swelling and surfactant adsorption in solutions containing either anionic or cationic
Table 1.2. Titles of the remaining chapters of this dissertation.
Chapter 2

Adsorption of Anionic or Cationic Surfactants in Polyanionic Brushes and Its
Effect on Brush Swelling and Fouling Resistance during Emulsion Filtration

Chapter 3

Aqueous Swelling of Zwitterionic Poly(Sulfobetaine Methacrylate) Brushes in
the Presence of Ionic Surfactants

Chapter 4

Oil Coalescence on Porous Membrane Surfaces during Dead-End Filtration

Chapter 5

Summary and Future Work

14

surfactant and (c) separation of O/W submicron emulsion using microfiltration membranes
modified with PSPMK brushes. In short, the polyanionic brushes show minimal surfactant
sorption or oil fouling when using an anionic surfactant, but high sorption and fouling when
using a cationic surfactant. In an effort to create brushes that resist sorption of both cationic
and anionic surfactants, I began investigating the properties of polyzwitterionic brushes.
Specifically, chapter 3 discusses the swelling of poly[2-(methacryloyloxy)ethyl dimethyl(3-sulfopropyl)ammonium hydroxide] (PMEDSAH) films in the presence of anionic and
cationic surfactants. These brushes show sorption of both cationic and anionic surfactants,
leading a more swelling or collapsed brush layer, depending on the surfactant charge and
concentration, as well as brush thickness. Chapter 4 shows preliminary oil coalescence data
using membranes with different coatings, focusing on oil coalescence on less hydrophilic
membranes. Preliminary data suggest oil coalescence can be adjusted by changing
membrane substrate, surface modification, pore size and filtration pressure. Finally,
Chapter 5 summarizes my PhD research and its potential application in the separation of
O/W submicron emulsions.

15

REFERENCES

16

REFERENCES

1.
Stephenson, M. A Survey of Produced Water Studies. In Produced Water; Springer,
1992; pp 1-11.
2.
BCC Research The North American Market for Produced Water Treatment
Equipment.
https://www.bccresearch.com/pressroom/env/north-american-marketproduced-water-treatment-equipment-reach-nearly-$1.2-billion-2017 (accessed Aug 9,
2017).
3.
Duong, P. H.; Chung, T.-S. Application of Thin Film Composite Membranes with
Forward Osmosis Technology for the Separation of Emulsified Oil–Water. J. Membr. Sci.
2014, 452, 117-126.
4.
Environmental
Protection
Angency
40
Cfr
435.15.
https://www.gpo.gov/fdsys/pkg/CFR-2010-title40-vol29/pdf/CFR-2010-title40-vol29sec435-15.pdf (accessed Aug 9, 2017).
5.
Rhee, C.; Martyn, P.; Kremer, J. In Removal of Oil and Grease in the Hydrocarbon
Processing Industry, Proceedings 42 nd Industrial Waste Conference. Purdue University,
1987; pp 143-150.
6.
Calvo, P.; Remunan-Lopez, C.; Vila-Jato, J.; Alonso, M. Development of Positively
Charged Colloidal Drug Carriers: Chitosan-Coated Polyester Nanocapsules and
Submicron-Emulsions. Colloid Polym.r Sci. 1997, 275, 46-53.
7.
Lovelyn, C.; Attama, A. A. Current State of Nanoemulsions in Drug Delivery.
J.Biomater. Nanobiotechnol. 2011, 2, 626.
8.
Benita, S.; Levy, M. Submicron Emulsions as Colloidal Drug Carriers for
Intravenous Administration: Comprehensive Physicochemical Characterization. J. Pharm.
Sci. 1993, 82, 1069-1079.
9.
Wang, L.; Li, X.; Zhang, G.; Dong, J.; Eastoe, J. Oil-in-Water Nanoemulsions for
Pesticide Formulations. J. Colloid Interface Sci. 2007, 314, 230-235.
10.
Quintanilla-Carvajal, M. X.; Camacho-DĂ­az, B. H.; Meraz-Torres, L. S.; ChanonaPĂŠrez, J. J.; Alamilla-BeltrĂĄn, L.; JimenĂŠz-Aparicio, A.; GutiĂŠrrez-LĂłpez, G. F.
Nanoencapsulation: A New Trend in Food Engineering Processing. Food Eng. Rev. 2010,
2, 39-50.
11.
Bernardi, D. S.; Pereira, T. A.; Maciel, N. R.; Bortoloto, J.; Viera, G. S.; Oliveira,
G. C.; Rocha-Filho, P. A. Formation and Stability of Oil-in-Water Nanoemulsions
Containing Rice Bran Oil: In Vitro and in Vivo Assessments. J. Nanobiotechnol. 2011, 9,
44.

17

12.

Asua, J. M. Miniemulsion Polymerization. Prog. Polym. Sci. 2002, 27, 1283-1346.

13.
Bechthold, N.; Tiarks, F.; Willert, M.; Landfester, K.; Antonietti, M. In
Miniemulsion Polymerization: Applications and New Materials, Macromolecular
Symposia, 2000; Wiley Online Library: 2000; pp 549-555.
14.
Duhon, H. Produced Water Treatment: Yesterday, Today, and Tomorrow. Oil and
Gas Facilities 2012, 1, 29-30.
15.
Frankiewicz, T. In Understanding the Fundamentals of Water Treatment, the Dirty
Dozen-12 Common Causes of Poor Water Quality, 11th produced water seminar, Houston,
TX, Jan, 2001; 2001.
16.
Rhee, C. H.; Martyn, P. C.; Kremer, J. Removal of Oil and Grease in Oil Processing
Wastewater. Sanitation District of Los Angeles County, USA 1989.
17.
Moosai, R.; Dawe, R. A. Gas Attachment of Oil Droplets for Gas Flotation for Oily
Wastewater Cleanup. Sep. Purif. Technol. 2003, 33, 303-314.
18.
Hoffmann, S.; Nitsch, W. Membrane Coalescence for Phase Separation of Oil‐in‐
Water Emulsions Stabilized by Surfactants and Dispersed into Smallest Droplets. Chem.
Eng. Technol. 2001, 24, 22-27.
19.

Beychok, M. R. Aqueous Wastes from Petroleum and Petrochemical Plants. 1967.

20.
Cheryan, M.; Rajagopalan, N. Membrane Processing of Oily Streams. Wastewater
Treatment and Waste Reduction. J. Membr. Sci. 1998, 151, 13-28.
21.
Katnik, K. E.; Pavilcius, A. M. In Novel Chemical Approach for the Treatment of
Oily Wastewaters, 33rd Ind. Waste Conf. Proc.;(United States), 1978; Nalco Chemical Co.,
Northbrook, IL: 1978.
22.
Fane, A. Membranes for Water Production and Wastewater Reuse. Desalination
1996, 106, 1-9.
23.
General Electric Water & Process Technologies, Filtration and Separation Spectrum.
http://products.pargreen.com/Asset/Filtration-spectrum.pdf (accessed Aug 9, 2017).
24.
Petala, M.; Zouboulis, A. Vibratory Shear Enhanced Processing Membrane
Filtration Applied for the Removal of Natural Organic Matter from Surface Waters. J.
Membr. Sci. 2006, 269, 1-14.
25.
Shi, W.; Benjamin, M. M. Membrane Interactions with Nom and an Adsorbent in a
Vibratory Shear Enhanced Filtration Process (Vsep) System. J. Membr. Sci. 2008, 312, 2333.
26.
Rana, D.; Matsuura, T. Surface Modifications for Antifouling Membranes. Chem.
Rev. 2010, 110, 2448-2471.

18

27.
Zhou, R.; Ren, P.-F.; Yang, H.-C.; Xu, Z.-K. Fabrication of Antifouling Membrane
Surface by Poly (Sulfobetaine Methacrylate)/Polydopamine Co-Deposition. J. Membr. Sci.
2014, 466, 18-25.
28.
Kang, G.-d.; Cao, Y.-m. Development of Antifouling Reverse Osmosis Membranes
for Water Treatment: A Review. Water Research 2012, 46, 584-600.
29.
Gryta, M. Chemical Pretreatment of Feed Water for Membrane Distillation. Chem.
Pap. 2008, 62, 100-105.
30.
Gleason, K. K.; Zafarullah, K.; Shafi, H. Z.; Yang, R., Development of Zwitterionic
Coatings That Confer Ultra Anti-Biofouling Properties to Commercial Reverse Osmosis
Membranes. U.S. Patents: 9,598,598, Mar. 21, 2017.
31.
Wang, Z.; van Andel, E.; Pujari, S. P.; Feng, H.; Dijksman, J. A.; Smulders, M. M.;
Zuilhof, H. Water-Repairable Zwitterionic Polymer Coatings for Anti-Biofouling Surfaces.
J. Mater. Chem. B 2017.
32.
Yang, W. J.; Neoh, K.-G.; Kang, E.-T.; Teo, S. L.-M.; Rittschof, D. Polymer Brush
Coatings for Combating Marine Biofouling. Prog. Polym. Sci. 2014, 39, 1017-1042.
33.
Yang, Z.; Tarabara, V. V.; Bruening, M. L. Adsorption of Anionic or Cationic
Surfactants in Polyanionic Brushes and Its Effect on Brush Swelling and Fouling
Resistance During Emulsion Filtration. Langmuir 2015, 31, 11790-11799.
34.
Barbey, R.; Lavanant, L.; Paripovic, D.; SchĂźwer, N.; Sugnaux, C.; Tugulu, S.;
Klok, H.-A. Polymer Brushes Via Surface-Initiated Controlled Radical Polymerization:
Synthesis, Characterization, Properties, and Applications. Chem. Rev. 2009, 109, 54375527.
35.
Zhao, B.; Brittain, W. J. Polymer Brushes: Surface-Immobilized Macromolecules.
Prog. Polym. Sci. 2000, 25, 677-710.
36.
Balazs, A. C.; Singh, C.; Zhulina, E.; Chern, S.-S.; Lyatskaya, Y.; Pickett, G. Theory
of Polymer Chains Tethered at Interfaces. Prog. Surf. Sci. 1997, 55, 181-269.
37.
RĂźhe, J.; Knoll, W. Functional Polymer Brushes. J. Macromol. Sci. Polym. Rev.
2002, 42, 91-138.
38.
Matyjaszewski, K.; Xia, J. Atom Transfer Radical Polymerization. Chem. Rev. 2001,
101, 2921-2990.
39.
Pintauer, T.; Matyjaszewski, K. Structural Aspects of Copper Catalyzed Atom
Transfer Radical Polymerization. Coord. Chem. Rev. 2005, 249, 1155-1184.
40.
Sankhe, A. Y.; Husson, S. M.; Kilbey, S. M. Direct Polymerization of Surface‐
Tethered Polyelectrolyte Layers in Aqueous Solution via Surface‐Confined Atom Transfer
Radical Polymerization. J. Polym. Sci. Polym. Chem. 2007, 45, 566-575.

19

41.
Tsarevsky, N. V.; Braunecker, W. A.; Matyjaszewski, K. Electron Transfer
Reactions Relevant to Atom Transfer Radical Polymerization. J.Organomet. Chem. 2007,
692, 3212-3222.
42.
Kobayashi, M.; Terayama, Y.; Yamaguchi, H.; Terada, M.; Murakami, D.; Ishihara,
K.; Takahara, A. Wettability and Antifouling Behavior on the Surfaces of Superhydrophilic
Polymer Brushes. Langmuir 2012, 28, 7212-7222.
43.
Zhu, Y.; Zhang, F.; Wang, D.; Pei, X. F.; Zhang, W.; Jin, J. A Novel Zwitterionic
Polyelectrolyte Grafted Pvdf Membrane for Thoroughly Separating Oil from Water with
Ultrahigh Efficiency. J. Mater. Chem. A 2013, 1, 5758-5765.
44.
Liang, S.; Qi, G.; Xiao, K.; Sun, J.; Giannelis, E. P.; Huang, X.; Elimelech, M.
Organic Fouling Behavior of Superhydrophilic Polyvinylidene Fluoride (Pvdf)
Ultrafiltration Membranes Functionalized with Surface-Tailored Nanoparticles:
Implications for Organic Fouling in Membrane Bioreactors. J. Membr. Sci. 2014, 463, 94101.
45.
Feng, X.; Jiang, L. Design and Creation of Superwetting/Antiwetting Surfaces. Adv.
Mater. 2006, 18, 3063-3078.
46.
Drelich, J.; Chibowski, E. Superhydrophilic and Superwetting Surfaces: Definition
and Mechanisms of Control. Langmuir 2010, 26, 18621-18623.
47.
StrĂśm, G.; Fredriksson, M.; Stenius, P. Contact Angles, Work of Adhesion, and
Interfacial Tensions at a Dissolving Hydrocarbon Surface. J. Colloid Interface Sci. 1987,
119, 352-361.
48.
Wang, H.; Shi, H.; Li, Y.; Wang, Y. The Effects of Leaf Roughness, Surface Free
Energy and Work of Adhesion on Leaf Water Drop Adhesion. PLOS ONE 2014, 9, e107062.
49.
Daiminger, U.; Nitsch, W.; Plucinski, P.; Hoffmann, S. Novel Techniques for
Iol/Water Separation. J. Membr. Sci. 1995, 99, 197-203.
50.
Agarwal, S.; von Arnim, V.; Stegmaier, T.; Planck, H.; Agarwal, A. Role of Surface
Wettability and Roughness in Emulsion Separation. Sep. Purif. Technol. 2013, 107, 19-25.
51.
Agarwal, S.; von Arnim, V.; Stegmaier, T.; Planck, H.; Agarwal, A. Effect of
Fibrous Coalescer Geometry and Operating Conditions on Emulsion Separation. Ind. Eng.
Chem. Res. 2013, 52, 13164-13170.
52.
Hu, D.; Li, X.; Li, L.; Yang, C. Designing High-Caliber Nonwoven Filter Mats for
Coalescence Filtration of Oil/Water Emulsions. Sep. Purif. Technol. 2015, 149, 65-73.
53.
Hu, D.; Li, L.; Li, Y.; Yang, C. Restructuring the Surface of Polyurethane Resin
Enforced Filter Media to Separate Surfactant Stabilized Oil-in-Water Emulsions Via
Coalescence. Sep. Purif. Technol. 2017, 172, 59-67.

20

54.
Kulkarni, P. S.; Patel, S. U.; Patel, S. U.; Chase, G. G. Coalescence Filtration
Performance of Blended Microglass and Electrospun Polypropylene Fiber Filter Media.
Sep. Purif. Technol. 2014, 124, 1-8.
55.
Tan, K. Y.; Hughes, T. L.; Nagl, M.; Huck, W. T. Nonfouling Capture–Release
Substrates Based on Polymer Brushes for Separation of Water-Dispersed Oil Droplets. ACS
Appl. Mater. Interfaces 2012, 4, 6403-6409.

21

Chapter 2. Adsorption of Anionic or Cationic Surfactants in Polyanionic Brushes
and Its Effect on Brush Swelling and Fouling Resistance during Emulsion Filtration
Portions of this chapter are reprinted from our published paper in Langmuir (Yang, Z.,
Tarabara, V. V., and Bruening, M. L. Langmuir 2015, 31 (43), 11790–11799).
2.1. Introduction
Oil and grease are common pollutants in wastewater from processes such as oil production,
food processing, can manufacturing, and wool scouring,1 and limitations on the release of
oil in wastewater necessitate separation of oil-water mixtures prior to water discharge.
According to U.S. regulations, new oil and gas facilities should limit the maximum
concentration of oil and grease in wastewater to 42 mg/L, and the average of daily values
over 30 consecutive days should not exceed 29 mg/L.2 Density-based separation
technology, e.g. flotation and hydrocyclones, can readily isolate free oil (droplets with
diameters >150 Îźm) and dispersed oil (20-150 Îźm droplets) from water, but emulsified oil
(droplets with diameters <20 Îźm) is more difficult to remove.3 Demulsification of such
surfactant-stabilized droplets through chemical or thermal treatments can increase droplet
sizes under certain conditions, but these techniques have limitations.1 Chemical treatment
requires customization for a given feed and may introduce corrosion problems. Although
thermal treatment is more universal, it incurs high energy costs, and evaporated oil may
present another disposal issue.
Compared to other separation methods for removal of emulsified oil, membrane
filtration offers several advantages including a relatively low energy cost, possible
automation, and no addition of extraneous chemicals.1,

4

However, fouling due to oil

accumulation at the membrane surface or in membranes pores often rapidly decreases

22

permeate flux or increases filtration pressures. To solve this problem, a number of groups
modified membrane surfaces to achieve both high flux and high oil rejection.5-10 However,
only a few studies examined the separation of micron and submicron-sized emulsions
because the small oil droplets are difficult to separate from water.11-12 Submicron emulsions
are common in drug delivery,13-14 pesticide application,15 food processing,16 cosmetics17-18
and emulsion polymerization.19-20 Considering the increasing popularity of submicron
emulsions and tighter regulations on oil discharge, their separation from water will likely
become increasingly important.
Two recent studies indicate that polyelectrolyte brushes on membranes limit
fouling by oil emulsions while increasing oil rejection.12, 21 Water sorption in these brushes
creates extremely hydrophilic surfaces to which oil does not adhere. However, the fouling
resistance of charged brushes may vary greatly with the surfactant that stabilizes the
emulsion. Katleen and co-workers found that cationic surfactants readily foul unmodified
negatively charged nanofiltration membranes.22 At higher concentrations, even negatively
charged surfactants foul these membranes, suggesting the need for higher surface charge
density, such as that in polyelectrolyte brushes, to prevent fouling. Daniel and co-workers
increased both the concentration of initiator and polymerization time during membrane
modification to create a dense, neutral brush that shows strong antifouling ability.23 A few
experimental24-26 and computational27-28 studies examined adsorption of charged
surfactants to polyelectrolyte brushes and suggest that a relatively low surfactant
concentration and low brush density promote brush collapse. Moreover, surfaces modified
with polyelectrolyte brushes become more hydrophobic after absorbing cationic
surfactants.29 However, these studies focused on adsorption of ionic surfactants to

23

oppositely charged or nonionic polymer brushes.24, 30 A very recent paper (submitted at
approximately the same time as the paper describing this research) focused on the
adsorption of ionic surfactants to oppositely charged polymer brushes and the lubrication
behavior after surfactant adsorption.31
This work examines interactions of cationic and anionic surfactants with brushes
composed of poly(3-sulfopropyl methacrylate salts) (PSPMK, Scheme 2.1). We chose
these brushes because Kobayashi et al. showed they are superoleophobic in water, although
that study did not exam surfactant-stabilized oil droplets. Negatively charged polymer
brushes are attractive because they may resist adsorption of anionic surfactants, which are
the most common class of surfactants in the oilfield market.32 We employ surface-initiated
atom-transfer radical polymerization (SI-ATRP)33 to grow these brushes on both Aucoated Si wafers and porous nylon membranes. The wafers enable studies of film thickness,
swelling in water and surfactant solutions, surfactant adsorption, and wettability (contact
angles), whereas dead-end filtration with membranes examines fouling, oil rejection, and
a

disulfide initiator
O

b

Îą-bromoisobutyryl bromide
(BIBB)
O
hydroxylated
H 3C
C
Br
nylon membrane Br
O
CH3
N CH2O C
N CH2OH

Br
S CH2 O
11

Au-coated S CH O
2 11
wafer
O

O

Br

Br
S CH2 O
11

O
H 2C

O

O

S
O
C
OK
O
CH3
SPMK
O

S CH2 O
11

CH3
CH3
C CH2C n
CH3
O
O

CuCl/CuCl2/Bpy
Catalyst

O

S

H 2C

O

S
O
C
OK
O
CH3
SPMK

O CH3 CH3
N CH2O C C CH2C n
CH3
O
O

O
O K

H 3C

CH3
Br

CuCl/CuCl2/Bpy
Catalyst

O

S

O
O K

Scheme 2.1. SI-ATRP of SPMK on Au-coated wafers (a) and hydroxylated nylon membranes (b)

permeate flux as functions of surfactant composition and brush thickness. For comparison,
we also perform dead-end filtration with commercial NF270 membranes.34

24

2.2.

Experimental

2.2.1. Materials
Hydroxylated (LoProdyne LP) nylon membranes with 1.2 Îźm diameter pores were
obtained from Pall Corporation, and NF270 membranes were a gift from Dow Chemical.
All membranes were cut into 2-cm discs, and nylon membranes were exposed to UV/ozone
for 15 min prior to modification. Au-coated Si wafers were prepared at LGA Thin Films
(Santa Clara, CA) by sputtering 200 nm of gold on 20 nm of Cr on Si wafers. 3-sulfopropyl
methacrylate potassium salt (98%, SPMK), 11-mercapto-1-undecanol (97%), ammonium
chloride (≥99.5%), α-bromoisobutyryl bromide (98%, BIBB), 2,2’-bipyridine (≥99%,
Bpy), decane (≥99%), hexadecane (99%), sodium dodecyl sulfate (98.5%, SDS), and
cetrimonium bromide (≥99%, CTAB) were obtained from Sigma-Aldrich. Bromine
(100.0%) and triethylamine (100%) were purchased from J.T. Baker, and copper (I)
chloride (≥ 99.995%), copper (II) chloride dihydrate (≥ 99.0%) and dimethylformamide (≥
99.8%, DMF) were acquired from Jade Scientific. Anhydrous dichloromethane (Macron
Fine Chemicals) was used as received. Chloroform (Macron Fine Chemicals) and
triethylamine were dried with molecular sieves (4 Å, Jade Science) before use. Aqueous
solutions were prepared using deionized water (Milli-Q, 18.2 MΊ cm), and disulfide
initiator, (BrC(CH3)2COO(CH2)11S)2, was synthesized according to a literature
procedure.35
2.2.2. Preparation of PSPMK Brushes on Au-Coated Wafers and Nylon Membranes
2.2.2.1.Initiator Immobilization
The protocol for immobilizing the disulfide initiator on Au-coated wafers included washing
the substrate with ethanol, drying with N2, cleaning with UV/ozone for 15 min, immersion

25

in a 1 mM ethanolic solution of (BrC(CH3)2COO(CH2)11S)2 for 24 h, rinsing with ethanol
and drying under a stream of N2. Initiator immobilization on hydroxylated nylon
membranes employed an esterification reaction with BIBB as described in the literature.36
In this procedure, 30 mL of chloroform and 2.4 mL of triethylamine were mixed under N2
and kept at 0 °C in a sealed round-bottom flask for 1 h. In a N2-filled glovebag, this mixture
was introduced into a 50-mL petri dish containing 6 hydroxylated nylon membrane discs,
and 0.5 mL of BIBB was added to the solution. The reaction mixture was gently stirred
with a magnetic stir bar for 20 min, and the BIBB-esterified membranes were removed
from the glove bag and washed with copious amounts of acetone and then with a
methanol/water (1/1, v/v) mixture prior to drying under a stream of N2. The membranes
were further dried under vacuum overnight before polymerization.
2.2.2.2.Polymerization of SPMK
The SI-ATRP is similar to a literature procedure.37-38 In a typical synthesis, SPMK (36
mmol) was dissolved in a DMF/water mixture (1/1 v/v, 30 mL) in a round-bottom flask.
The mixture was degassed via four freeze-pump-thaw cycles. A freshly prepared Cu-Bpy
stock solution containing CuCl (0.002 mmol), CuCl2 (0.002 mmol), and Bpy (0.01 mmol)
in a DMF/water mixture (1/1 v/v, 5 mL) was degassed via four freeze-pump-thaw cycles
and mixed with the monomer solution in a petri dish in a N2-filled glove bag.
Polymerization occurred after immersion of initiator-modified membranes or wafers in the
mixed solution at room temperature. After the desired polymerization time, the modified
substrates were removed from the glove bag, rinsed with 100 mL of water, and dried with
N2.

26

2.2.3. Characterization of Modified Au-Coated Wafers and Nylon Membranes
Wafers coated with disulfide initiator and PSPMK brushes were characterized by
reflectance Fourier transform infrared spectroscopy (FTIR) (Nicolet 6700 FT-IR
spectrometer, Thermo Scientific, with a Pike grazing angle (80 Âş) attachment). A
UV/ozone-cleaned Au-coated wafer served as a background. Infrared spectra of
membranes modified with BIBB and PSPMK brushes were obtained using a PerkinElmer
FT-IR spectrometer (Spectrum One, PerkinElmer) with an attenuated total reflectance
(ATR) accessory (Universal ATR-1 Reflection Top-Plate). The spectrum of an unmodified
nylon membrane was subtracted from the spectrum of modified membranes to highlight
absorbances due to newly grown films. To examine surfactant adsorption, PSPMK brushmodified wafers were transferred directly from water to surfactant-containing solutions.
After an overnight immersion in the surfactant solution, the wafer was rinsed with 100 mL
of water from a wash bottle for ~ 2 min prior to drying with N2 and obtaining a reflectance
FTIR spectrum.
The “dry” thickness and refractive indices of the polymer brushes on Au-coated
wafers were determined in air using a rotating analyzer ellipsometer (model M-44, J.A.
Woollam) at an incident angle of 75°. Similar thickness determinations for films immersed
in water employed a homemade trapezoidal cell with glass windows.39-40 In all
experiments, the refractive index was fit using a Cauchy model. Ellipsometric
measurements were performed on at least three spots on each wafer, with many different
films.
Contact angles on surfaces were acquired using a dynamic contact angle analyzer
(FTA 200, Firsttenangstroms). Hexadecane drops were placed on the modified and

27

unmodified Au-coated wafers both before and after immersion of substrates in water to
determine contact angles. We attempted to measure a hexadecane contact angle directly in
solutions containing either 0.36 mM SDS or 0.36 mM CTAB, but hexadecane droplets
would not form at the end of the syringe needle due to the surfactants. As an alternative,
the wafers were pretreated by immersion in the surfactant solutions overnight, followed by
washing with 100 mL of water and drying with N2. A drop of hexadecane was placed on
these dry wafers, and they were immersed in surfactant solutions prior to determining the
contact angle. The morphologies of membranes were also observed using a field-emission
scanning electron microscope (JSM-7500F, JEOL). The membrane porosity was
determined from the mass loss of wet membranes after drying. After polymerization,
polymers physically adsorbed on the membrane were washed off by pumping pure water
through the membrane for 5 h. Membranes were then patted dry with a Kimwipe to remove
water on their surface, and immediately weighed (wet mass). To obtain the mass of a dry
membrane, the wet membranes were dried in a vacuum oven for 24 h and weighed. The
membrane porosity was calculated using Eq. 2.141:
Porosity = ďż˝

(Ww −Wd )/ρwater

(Ww −Wd )/ρwater +Wd /ρpolymer

� × 100%

Eq. 2.1

where Ww and Wd are the masses of wet and dry membranes, respectively, ρwater is the
density of water (1 g/cm3), and ρpolymer represents the density of nylon, which is 1.15
g/cm3.42

2.2.4. Emulsion Preparation and Characterization
Hexadecane emulsions stabilized with SDS or CTAB were prepared as described
previously.43 Preparation of stable hexadecane-in-water emulsions with oil droplet

28

diameters smaller than 1 μm requires an oil content ≤10% (volume fraction).1 SDS (0.0144
g) or CTAB (0.0182 g) was dissolved along with NaCl (0.0058 g) in a hexadecane/water
mixture (1/9, v/v, 100 mL), and this mixture was sonicated for 3 min (1 min on, 0.5 min
off) using a horn ultrasonicator (power setting: 6.5, power output: 57 W, Sonicator 3000,
Misonix). The size distribution of the oil droplets was determined by light scattering using
a Malvern Instruments particle size analyzer (Mastersizer 2000, Malvern Instruments) with
a manually controlled pump and sample dispersion unit (Hydro SM, Malvern Instruments).
These stock emulsions were stored for up to two weeks prior to dilution and filtration. At
longer times phase separation occurs with CTAB emulsions, and even for shorter times,
some phase separation may decrease the oil concentration.
2.2.5. Dead-End Filtration
The specific flux (L/(m2¡h¡bar)), referred to as LMH/bar) and observed oil rejection (R, %)
of the membranes were determined in dead-end filtration using a 15-mL Amicon cell
(Model 8010, Millipore) with a suspended stir bar rotating at 45 rpm. The cell was
connected to a stainless steel feed tank (standard 2 gallon pressure vessel, Pope Scientific)
that contained deionized water (Figure 2.1),44 and the applied pressure was always 1.4 bar.
A stainless steel frit with nominal 1 Îźm pores (316LSS 1 Îźm Discs, Mott Corporation)
supported the membrane. Initially, pure water flux was determined using deionized water
as the feed. Permeate fractions were collected over specific times and weighed to determine
the flux. In filtration of emulsions, the Amicon cell was filled with 13 mL of deionized

29

water and 1 mL of the 10% hexadecane emulsion. The solution was mixed with the hanging

Filtration Cell
Stir Bar

O Ring
Membrane
Frit
Holder
Bottom

Reservoir (water)
Pressure Source N2

Magnetic Stirrer

Permeate

Figure 2.1. Diagram of the apparatus for dead-end filtration.

magnetic stir bar for 30 sec. The cap was placed on the Amicon cell, and pressurized N2
was applied to the water-containing feed tank so deionized water entered the stirred
Amicon cell to replace the permeate. Permeate fractions were collected over specific times
and weighed to determine flux, and rejection values, R, were calculated using Eq. 2.2:
R = �1 −

Cp
Cf

� × 100%

Eq. 2.2

where Cf and Cp are the hexadecane concentrations in the feed and the permeate,
respectively. During the filtration, the volume of feed solution decreases slightly due to a
small amount of air entering the cell so the feed concentration increases ~20% during
experiments where the rejection is essentially 100%. For rejection calculations, Cf was the
average of the feed concentration before and after filtration. Hexadecane concentrations
were determined by gas chromatography (GC-17a, Shimadzu) using a hexadecane
calibration curve with decane as an internal standard. To prepare GC samples, hexadecane
in permeate solutions was extracted by mixing 1 mL of saturated NaCl solution, 5 drops of
30

1 M HCl, 1 mL of dichloromethane and 1 mL of the permeate solution for 30 s on a vortex
mixer (BV1000, Benchmark Scientific). After phase separation, 0.5 mL of the organic
phase was added to a GC vial containing 10 ÎźL of decane as an internal standard. Figure
2.2 shows the calibration curve. Several blank solutions without hexadecane were analyzed
to determine the minimum detectable signal (average of the blank signal plus three times
the standard deviation of this signal.) Based on this minimum detectable signal and the
calibration curve in Figure 2.2b, the detection limit of these analyses is 9 x 10-6 M
hexadecane.

Figure 2.2. (a) GC calibration curve showing a linear relationship between the ratio of GC peak
area (hexadecane/decane) and the concentration of hexadecane at high hexadecane concentration.
Figure (b) shows the calibration curve at low concentration. Hexadecane concentrations in all
filtrations were calculated using the left curve. Estimates of detection limits and low concentrations
for determining high rejections were obtained using the right curve.

To examine whether oil adsorbed significantly to the filtration cell, a control
experiment employed an impermeable transparency film (3M, PP2500) in place of a
membrane. Because the film is impermeable, any change in the amount of hexadecane after
the experiment represents oil adsorption on the Amicon cell or film. Table 2.1 describes
these results and shows <20% adsorption to the cell. All of the filtration experiments were
repeated with 2 or 3 membranes.

31

Table 2.1. Mass balance for the dead-end filtration of surfactant-stabilized emulsions using
PSPMK-modified (1-h polymerization time) and NF270 membranes, and an impermeable
transparency film. Values of n are the percentages of oil (relative to that in the initial feed) found
in the total permeate and retentate fractions, and on the cell after the filtration. The value adsorbed
on the cell was calculated by subtraction and represents the amount of oil not accounted for in the
permeate and retentate. The total filtration time was 12 h. Initial studies with a teflon cell instead
of the Amicon cell showed higher levels of adsorption.

2.3.

Membrane

Surfactant

n (Permeate)

n (Retentate)

n (Cell)

PSPMK-modified membrane

SDS

1.7%

78.9%

19.4%

PSPMK-modified membrane

CTAB

81.0%

0.4%

18.6%

NF270

SDS

0.4%

87.0%

12.6%

NF270

CTAB

0.03%

84.7%

15.3%

Impermeable transparency
film (3M, PP2500)

SDS

N/A

77.9%

22.1%

Results & Discussion

2.3.1. Brush Synthesis on Au-Coated Wafers and Nylon Membranes
Reproducible growth of polymer brushes is the first step in investigating both the
interaction of brushes with surfactants and their use as membrane skins for emulsion
filtration, and Scheme 2.1 shows the SI-ATRP strategy we employ to grow brushes on Aucoated wafers and nylon membranes. The Au surfaces enable brush characterization with
reflectance IR spectroscopy and ellipsometry. IR spectra confirm both initiator adsorption
and brush growth on the Au surface (Figure 2.3). In particular, sulfonate stretches at 1219
and 1049 cm-1 and a C=O stretch at 1737 cm-1 confirm the presence of PSPMK brushes.
Figure 2.4 shows the peak height for the sulfonate stretch at 1219 cm-1 and the “dry”
ellipsometric thickness of PSPMK brushes (on Au-coated wafers) as a function of

32

polymerization time. Both the peak height and the dry thickness increase approximately
linearly with polymerization time, indicating continuous growth of PSPMK brushes.
a
Absorbance

0.0005

C=O str.
1738 cm-1

C-O-C sym. str.
1170 cm-1
C-O-C asym. str.
1278 cm-1

CH2 scissor
1468 cm-1

2000

1800

1600

1400

Wavenumnbers
b

Absorbance

-SO3- sym. str.
1049 cm-1

C=O str.
1730 cm-1

1800

1000

-SO3- asym. str.
1219 cm-1

0.05

2000

1200

(cm-1)

1600

1400

Wavenumnbers

1200

1000

800

(cm-1)

Figure 2.3. Reflectance FTIR spectrum of (a) the monolayer formed from treatment of a Au-coated
wafer with a solution containing (BrC(CH3)2COO(CH2)11S)2 and (b) PSPMK brushes on a Aucoated wafer.

33

Dry Thickness (nm)

0.25

200

0.2

150

0.15
100

0.1

50

0.05

0

IR Absorbance (-SO3-)

0.3

250

0
0

10

20

30

40

50

60

70

Polymerization Time (min)

Figure 2.4. “Dry” ellipsometric thicknesses (blue squares) and sulfonate IR absorbances at 1219
cm-1 (red circles) as a function of polymerization time for PSPMK brushes grown from an initiator
monolayer on Au. Error bars represent standard deviations (n=12 determinations on 4 wafers).

SI-ATRP on nylon membranes is difficult to characterize, but ATR-IR spectra
clearly show the presence of C=O and sulfonate stretches after growth of PSPMK (Figure
2.5). The mass loss of wet membranes after drying also indicates a decrease in porosity
from 66% for a bare membrane to around 56% after modification with PSPMK brushes
(Table 2.2). Top-view SEM images (Figure 2.6a, b) suggest a lower porosity and smaller
pore size at the membrane surface after the growth of polymer brushes. Cross-sectional
SEM images (Figure 2.6c, d) also reveal a smaller pore size after polymerization, especially
in the top layer of the membrane. Based on these images, pore size appears to decline after
brush growth. However, pores in these spongy membranes are difficult to interrogate with
microscopy. Additionally, SEM images show dry membranes whereas during filtration
brushes will swell.

34

Absorbance

c

b

-SO3- asym. str.
1191 cm-1
-SO3- sym. str.
1045 cm-1

C=O str.
1722 cm-1

a
2000

1800

1600

1400

Wavenumbers

1200

1000

(cm-1)

Figure 2.5. ATR-IR spectra of a hydroxylated membrane (a) before and (b) after growth of PSPMK
brushes. Spectrum (c) results from subtracting 0.8 × spectrum (a) from spectrum (b) and shows

absorbances due primarily to the PSPMK brush. We employed the factor of 0.8 to minimize
infrared absorbances from the base nylon membrane.
Table 2.2. Membrane porosity determined by the mass loss of wet membranes after drying. The
PSPMK-modified membranes show lower porosity than the bare membrane.
Polymerization

Dry Weight (g)

Wet Weight (g)

Porosity

1

0.0374

0.0756

54.0%

2

0.0362

0.0745

54.9%

1

0.0372

0.0860

60.1%

2

0.0372

0.0766

54.9%

1

0.0312

0.0850

66.5%

2

0.0332

0.0876

65.3%

3

0.0332

0.0884

65.6%

Average Porosity

Time
45 min

54.4%

1h

57.5%

0
(bare)

35

65.8%

a

b

50 Îźm

50 Îźm

c

d

10 Îźm

10 Îźm

Figure 2.6. SEM images of (a, c) bare nylon membranes and (b, d) PSPMK-modified membranes.
Figures a and c are the top views of the membranes, where figures b and d show cross sections of
the membranes.

2.3.2. Film Swelling and Contact Angles
PSPMK brushes should resist oil adsorption in large part because they swell extensively in
water. Figure 2.7 shows the thickness of water-swollen PSPMK brushes (on Au-coated
wafers) and their percent swelling (percent increase in thickness after immersion in
deionized water for 12 h) as a function of their “dry” thickness. Notably, films swell by up
to 280%, and the maximum swollen thickness is around 340 nm. Some films increase in
thickness as much as 240 nm due to water sorption. The swelling data are consistent with
changes in the refractive index of the PSPMK brush after immersion in water (see Figure
2.8, higher swelling gives lower refractive indices). When the “dry” thickness of the brush
is < ~100 nm, the percent swelling is essentially constant. Lower swelling at the higher

36

“dry” thicknesses may occur due to a lower chain density at the exterior of thicker films.

400

320%

300

240%

200

160%

100

80%

0

Swelling (%)

Swollen Thickness (nm)

(Some chain termination likely occurs during the polymerization.)

0%
0

30

60

90

120 150 180 210

Dry Thickness (nm)

Figure 2.7. The water-swollen thickness (blue squares) and percent swelling (red circles) of
PSPMK brushes as a function of their “dry thickness”. Error bars represent standard deviations
(n=12 determinations on 4 wafers).

Refractive Index

1.4
1.39
1.38
1.37
1.36
1.35
1.34
0

50

100

150

200

250

Dry Thickness (nm)

Figure 2.8. Refractive indices of water-swollen PSPMK brushes as a function of their dry thickness.
When the dry thickness is <~100 nm, the refractive indices are around 1.36. Refractive indices
increase from 1.36 to 1.39 when dry thicknesses surpass 100 nm and swelling decreases. Each point
represents an average of refractive indices determined at wavelengths ranging from 410 to 740 nm.

Consistent with high aqueous swelling, the hexadecane contact angles on PSPMKcoated wafers increase from 50Âş in air to ~180Âş in water (Figures 2.9a and 2.9b). Kobayashi

37

and coworkers obtained similar results with hexadecane on PSPMK and other
polyelectrolyte brushes.45 Bare gold surfaces show hexadecane contact angles <5 Âş in both
air and water (Figures 2.9c and 2.9d). Eq. 2.2, often known as Young’ s equation, shows
the surface tensions that affect hexadecane contact angles in air, θHD−Air , and in water,
𝜃𝜃𝐻𝐻𝐻𝐻−𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊𝑊 , where γFAir , γFHD , γHDAir , γFW , and γHDW are the surface tensions between

film and air, film and hexadecane, hexadecane and air, film and water, and hexadecane and
water, respectively.
cos θHD−Air =

γFAir −γFHD

cos θHD−Water =

ÎłHDAir

γFW −γFHD
ÎłHDW

Eq. 2.2

Because ÎłHDAir < ÎłHDW , for values <90 Âş contact angles are often greater in water than air.
In contrast, for contact angles >90 Âş, if it acted independently, the high value of ÎłHDW

would decrease the contact angle value compared to the contact angle in air. However, for
a highly swollen, water-like film, 𝛾𝛾𝐹𝐹𝐹𝐹 approaches zero and γFHD ≈ γHDW , so

cos θHD−water approaches -1 (Figure 2.9b). Thus, oil droplets should dewet and detach

easily on PSPMK surfaces in water. This superoleophobicity also occurs on modified nylon
membranes, leading to a strong resistance to fouling.

a

c

b

~180°

0° d

<5°

50°

Figure 2.9. Contact angles of hexadecane on PSPMK-modified wafers (a, b) and bare Au-coated
wafers (c,d). The images were obtained in air (a, c) and water (b,d).

38

2.3.3. Surfactant Interactions with PSPMK Brushes
Charged surfactants may adsorb to PSPMK brushes through hydrophobic interactions with
brush alkyl backbones, but electrostatic forces between PSPMK side chains and surfactants
will be repulsive for anionic surfactants and attractive for cationic surfactants. Thus, this
section examines adsorption of anionic and cationic surfactants in PSPMK films and the
effect of adsorption on brush swelling.
2.3.3.1.Anionic Surfactant
For studies of SDS adsorption to PSPMK brushes, we first polymerized SPMK for 1 h
from many initiator-modified Au-coated wafers to create films with water-swollen
thicknesses around 340 nm. The CMC of SDS is 8 mM.46 Thus, to examine brush swelling
450

Swollen Thickness (nm)

400
350
300

water

250

2 mM SDS

200

4 mM SDS

150

6 mM SDS

100

10 mM SDS
15 mM SDS

50

20 mM SDS

0
0

1

2

3

4

5

6

7

Sample

Figure 2.10. Swollen thicknesses of PSPMK brushes on Au-coated wafers in water and in solutions
with different concentrations of SDS. Error bars represent standard deviations (n=12 measurements
on 4 wafers).

at SDS concentrations above and below the CMC, we immersed the different films in
aqueous solutions with SDS concentrations ranging from 2 to 20 mM and determined the

39

water-swollen thicknesses with in situ ellipsometry. In all cases the swollen thickness was
around 340 nm (Figure 2.10) both in water and in the SDS solutions, suggesting that SDS
does not affect the film architecture. Moreover, after a brief rinse with water, the
reflectance IR spectra of these films showed no increase in the absorbance from CH2

Absorbance

stretches, which indicates minimal SDS adsorption (Figure 2.11).

3500

3000

2500

2000

Wavenumbers

1500

1000

500

(cm-1)

Figure 2.11. Reflectance IR spectra of PSPMK brushes on Au-coated wafers before (black line)
and after immersion in 4 mM (red line) and 15 mM (blue line) SDS solutions. Small changes in the
intensity of sulfonate stretches with minimal change in the carbonyl stretch may reflect a change in
the film orientation or the environment around the sulfonates after immersion in surfactant solutions
and rinsing with water. However, the spectra show no aliphatic stretches of surfactant around 2900
cm-1.

Shorter polymerization times lead to lower “dry” film thicknesses and more
swelling of PSPMK brushes (see Figure 2.4 and 2.7), so we also examined SDS adsorption
in thinner films that have a lower polymer density in water. However, as Figure 2.12 shows,
films with water-swollen thicknesses ranging from 75 to 350 nm exhibited no substantial
change in their swollen thickness upon immersion in 20 mM SDS solutions.

40

Swollen Thickness (nm)

400

300

2 red circles

200

100

2 red circles and 2 blue squares

0
0

10

20

30

40

50

60

70

Polymerization Time (min)

Figure 2.12. The water-swollen thicknesses of PSPMK brushes on Au-coated wafers before (blue
squares) and after (red circles) immersion in a 20 mM SDS solution.

Water contact angles also show no change after treatment of films with SDS. The
hexadecane contact angle in water was ~180 Âş, even after exposing the film to 20 mM SDS
and briefly rinsing with water. Moreover, the hexadecane contact angle in a 20 mM SDS
solution was also ~180Âş, and the oil easily detaches from the PSPMK-modified surface.
Taken together, these results imply that negatively charged brushes resist SDS adsorption
and should resist fouling in the filtration of emulsions stabilized with this surfactant.
Because surfactant should coat oil droplets in stabilized emulsions, we expect the fouling
tendencies of such emulsions to resemble those of the surfactant itself.
2.3.3.2.Cationic Surfactant
2.3.3.2.1. Effect of CTAB Concentration
Unlike SDS, the cationic surfactant CTAB will experience electrostatic attraction to
PSPMK brushes. After immersion in a 3 mM CTAB solution and rinsing with water, the
absorbances of the CH2 symmetric and asymmetric stretches at 2925 and 2854 cm-1
increase dramatically in the reflectance IR spectra of PSPMK films (Figure 2.13),

41

indicating CTAB adsorption. The position and intensity of sulfonate absorbances also shift,
and the intensity of the C=O peak decreases suggesting a change in the “dry” PSPMK
conformation and in the counterions of the sulfonate groups after CTAB exposure.47-49

0.05
Absorbance

CH2 sym. str.
2854 cm-1

CH2 asym. str.
2925 cm-1
3500

2500

Wavenumbers

1500

500

(cm-1)

Figure 2.13. Reflectance IR spectra of PSPMK brushes before (bottom) and after (top) immersion
in 2 mM CTAB for 12 h and rinsing with water. Growth of the PSPMK brush employed 1-h
polymerization.

Figure 2.14 shows the increase in the absorbance at 2925 cm-1 as a function of the
CTAB concentration in solution (films were rinsed with water prior to obtaining the
spectra). At CTAB concentrations from 0.4 to 1.5 mM, the absorbance at 2925 cm-1
increases slightly with CTAB concentration, but upon increasing the CTAB concentration
from 1.5 to 5 mM, the amount of CTAB adsorption, as quantified by the increase in
absorbance at 2925 cm-1, decreases 76%. Nevertheless, the overall swollen film thickness
increases continuously with CTAB concentration (Figure 2.14). At the lowest CTAB
concentrations, surfactants exist as single molecules (the CMC is 0.9 mM50) that should
diffuse throughout the film and adsorb to brushes. The surfactants likely replace water in
the film, so despite the brush becoming more hydrophobic it does not collapse significantly.

42

With increasing CTAB concentration, adsorption of micelles may dominate, and their
added mass apparently increases the thickness of the swollen film. However, extensive
adsorption of micelles at the brush surface may hinder further sorption of CTAB micelles
and single molecules (Figure 2.14) despite the increase in swollen thickness. CTAB
adsorption likely neutralizes much of the negative charge near the surface of the PSPMK
brushes and makes them more oleophilic, as indicated by only a 58Âş hexadecane contact

1600%

Swollen Thickness Change (%)

50%
40%

1200%

30%
20%

800%

10%
0%

400%

-10%
-20%

0%
0

1

2

3

4

5

Increase in CH2 Absorbance at 2925
cm-1

angle in water (Figure 2.15) after CTAB adsorption.

6

CTAB Concentration (mM)

Figure 2.14. Percent change in swollen thickness (blue squares) and percent increase in the CH2
IR absorbance at 2925 cm-1 (red circles) after immersion of brushes in solutions with varying CTAB
concentrations. Swelling was determined first in deionized water and subsequently in the CTAB
solution, whereas the IR absorbance was determined after immersion in deionized water or CTAB
solution and subsequent rinsing with water, and drying with N2. All films were initially prepared
using a 1-h polymerization time. Error bars represent standard deviations (n=12 measurements on
4 wafers).

43

58°

Figure 2.15. Contact angle of hexadecane in deionized water on a PSPMK-modified wafer treated
with CTAB. The wafer was exposed to 5 mM CTAB for 12 h and rinsed with water before the
contact angle measurement.

2.3.3.2.2. Effect of the Density of PSPMK Brushes on CTAB Adsorption and Changes in
Swelling
In early studies, we used some older batches of the disulfide initiator that did not always
give similar thicknesses for a given polymerization time, presumably because some of the
initiator had decomposed and the densities of chains were not the same on all wafers.
However, these films show similar swollen thicknesses, so they offer a chance to examine
the effect of brush density on CTAB adsorption. We employ the magnitude of the –SO3absorbance at 1219 cm-1 as an indicator of brush density for films with similar swollen
thicknesses of 340 nm. Figure 2.16 shows that with higher density (greater -SO3absorbance), PSPMK brushes collapse less and even increase in thickness upon immersion
in 3 mM CTAB. However, the -CH2 absorbance due to adsorbed surfactant is similar for
all of the films after rinsing with water. As mentioned above, adsorption of CTAB micelles
apparently stretches polymer chains for films with a high density. In this case, the micelle
does not simply replace water. For the lowest density films, the micelle adsorption causes
chain collapse, perhaps as the chains conform to the micelle.

44

1000%

40%

800%

20%

600%

0%
400%

-20%

200%

-40%

0%

-60%
0.08

0.11

0.14

0.17

0.2

IR Absorbance of SO3

0.23

Increase in CH2 Absorbance at
2925 cm-1 (%)

Swollen Thickness Change (%)

60%

0.26

-

Figure 2.16. The change in the water-swollen thickness of PSPMK brushes in 3 mM CTAB (blue
squares) with respect to the thickness in deionized water, and the percent increase in the –CH2 IR
absorbance at 2925 cm-1 (red circles) as a function of the sulfonate absorbance at 1219 cm-1 in the
pristine film. IR absorbances were determined after rinsing films with water and drying with N2.
The increase in absorbance at 2925 cm-1 stems from surfactant adsorption, and film density should
increase with increasing sulfonate absorbance.

2.3.4. Emulsion Filtration with PSPMK-Modified Membranes
2.3.4.1.Filtration of SDS-Stabilized Emulsions
We examined dead-end filtration of hexadecane submicron emulsions stabilized with either
SDS or CTAB. Hexadecane serves a representative oil with a low surface energy10 and is
widely used in fundamental research21, 45, 51. Light-scattering measurements show that the
oil droplet size distributions (Figure 2.17) are similar for emulsions stabilized with SDS
and CTAB. In both cases, the droplets have a size range from 40 to 1000 nm, and the
diameters of 90% (by volume) of the oil droplets are smaller than 500 nm. The volumeweighted mean diameters are 280 and 229 nm for SDS and CTAB-stabilized oil droplets,
respectively. All these data correspond to the definition of a submicron emulsion,14 and the
droplet size distribution was approximately the same at the beginning and end of filtrations.
The filtration employs initial oil volume fractions of 0.7%. The oil content in wastewater

45

from different industries varies from 0.0013% to 25.97% (volume fraction), and most oily
wastes contain 0.1-1% oil.1 During filtrations of SDS-stabilized emulsions with PSPMKmodified and NF270 membranes, the hexadecane concentration in the permeate
approaches or is less than the limit of detection (9 x 10-6 M). Thus, the oil rejection is
usually >99.9% and always >99.8%.

b

a

Figure 2.17. Size distributions of hexadecane submicron emulsions stabilized by SDS (a) and
CTAB (b).

Figure 2.18a shows the specific flux (flux divided by the transmembrane pressure
drop) during filtration of an SDS-stabilized emulsion through a PSPMK-modified (1-h
polymerization) membrane. Because the oil rejection is around 100% over the entire
filtration, the feed oil concentration is nearly constant. The flux during the filtration (~18
LMH/bar) is about 75% of the pure water flux (~23 LMH/bar), which suggests minimal
membrane fouling. More importantly, the flux does not decline over 12 h, further
confirming the fouling resistance of the brushes. The pure water flux after filtration was
29.9 LMH/bar (Figure 2.19a), 30% higher than that before filtration. This increase in pure

46

water flux may result from a change in the brush confirmation, but the high water flux after

Figure 2.18. Specific fluxes (blue squares) in dead-end filtration of SDS-stabilized hexadecane
submicron emulsions through PSPMK-modified membranes. The SPMK polymerization time was
1 h (a) or 45 min (b). The green horizontal lines show the pure water flux before filtration. Initially,
the emulsion contained 24.3 mM hexadecane in 0.36 mM SDS.

b

a

150

Pure Water Flux
(LMH/bar)

Pure Water Flux
(LMH/bar)

40
30
20
10
0
0

0.5

1

1.5

2

100
50
0
0

Time (h)

0.2

0.4

0.6

0.8

Time (h)

Figure 2.19. Pure water flux (after a 12-h filtration of an SDS-stabilized emulsion) for membranes
modified using (a) 1-h and (b) 45-min SI-ATRP of SPMK.

filtration confirms the strong anti-fouling nature of the PSPMK-modified membrane. The
highly hydrated, negatively charged brush clearly resists adsorption of oil droplets. We
should note that a control nylon membrane with no modification shows much higher flux
(2 to 3 orders of magnitude) but little oil rejection due to the large pores (nominally 1.2
Îźm) (Figure 2.20a, b).

47

15000

0.02

12500

0.015

10000

0.01

7500

0.005

5000

Hexadecane
Concentration (M)

Permeate Flux
(LMH/bar)

a

0
0

20

40

60

80

100

Permeate Flux
(LMH/bar)

b

10000

0.025

9000

0.02

8000

0.015

7000

0.01

6000

0.005

5000

0
0

20

40

60

80

Hexadecane
Concentration (M)

Time (min)

100

Time (min)

Permeate Flux
(LMH/bar)

c

30
25
20
15
10
5
0
0

1

2

3

4

5

6

7

Time (h)

Figure 2.20. Replicate dead-end filtrations. (a), (b): filtration of SDS-stabilized hexadecane
emulsions with bare nylon membranes with 1.2 Îźm pores. The specific flux (blue circles) drops at
the beginning due to membrane fouling and increases quickly after the oil permeates through the
membrane in 1.5 min. The hexadecane concentration (red circles) falls to zero because deionized
water replenishes the feed cell. (c): filtration of an SDS-stabilized hexadecane emulsion with a
PSPMK-modified membrane (polymerization time: 45min). The permeate flux (blue circles)
increases after 5h of filtration, and visible oil appears in the permeate after that time.

48

Decreasing the polymerization time from 1 h to 45 min during membrane
modification should lead to polymer brushes with a similar swollen thickness (Figure 2.4)
but a lower charge density (Figure 2.7). Figure 2.18b shows the permeate flux for a
membrane modified with 45 min of SPMK polymerization. Compared to filtration through
a membrane modified with 1 h of polymerization, the fluxes and oil rejections are similar.
However, for the membrane modified with 45 min of polymerization the flux increases
gradually during the experiment, and the pure water flux after filtration increases
dramatically (Figure 2.19b), suggesting some collapse of the low-density brushes. A
replicate membrane (Figure 2.20c) prepared with 45 min of PSPMK polymerization was
not hydrophilic enough to maintain high oil rejection longer than 5 h. Thus, decreasing the
polymerization time may increase flux, but the low density of polymer chains will likely
lead to eventual brush collapse.
2.3.4.2.Filtration of CTAB-Stabilized Emulsions
In dramatic contrast to SDS, filtration of CTAB-stabilized submicron emulsions through
PSPMK-modified membranes leads to oil breakthrough and an initial dramatic increase in
flux (Figure 2.21). Because we add the emulsion to the Amicon cell and replenish the cell
with deionized water from a feed tank during the filtration, the concentrations of oil and
surfactant in the feed rapidly decline when oil rejection is low. Thus, the hexadecane
concentration in the permeate first increases with time as the membrane becomes more
permeable in the presence of the cationic surfactant and then declines because of less oil in
the feed. Permeate flux also initially increases with time as the brush collapses and then
becomes constant when there is minimal surfactant in the feed. However, the flux never
returns to its initial low value, suggesting the brush remains collapsed.

49

0.014

50

0.012
0.01

40

0.008

30

0.006

20

Pure Water Flux: 13.5 LMH/bar

0.004

10

0.002

0

Hexadecane Concentration (M)

Permeate Flux (LMH/bar)

60

0
0

2

4

6

8

10

12

Time (h)

Figure 2.21. Permeate hexadecane concentrations (red circles) and specific fluxes (blue squares)
during dead-end filtration of a CTAB-stabilized hexadecane submicron emulsion through a
PSPMK-modified membrane. The horizontal green line shows the pure water flux prior to the
filtration. The membrane was modified with a 1-h polymerization. The PSPMK-modified
membrane showed low oil rejection, and all of the hexadecane permeated through the membrane
in 4 h. The initial hexadecane feed concentration was 17.9 mM, but this value declined rapidly
during the filtration due to low rejection and cell replenishment with deionized water.

As Figure 2.22 illustrates, we speculate that for low-density chains, polymer
brushes likely collapse as they wrap around CTAB-stabilized oil micelles. Even for
membranes modified with a 1-h polymerization, brush deswelling is sufficient to give a
permeate flux that is 4 times the pure water flux prior to exposure to surfactant. Moreover,
the membranes show minimal oil rejection after brush collapse. This mechanism is
consistent with the data in Figure 2.16, where brush thickness decreases in the presence of
CTAB for films with the lowest density. The relatively low density of functional groups
on the nylon surface likely leads to less initiator and significantly less dense brushes than
polymerization from a monolayer of initiator on the Au-coated wafer. The concentration

50

of CTAB in the initial feed solution is also low, 0.36 mM, which tends to lead to brush
collapse (Figure 2.14).

a

b

Figure 2.22. Cartoon of filtration of CTAB-stabilized submicron emulsions with PSPMK
modified-membranes: (a) polymer brushes swell well in water prior to contact with CTAB and oil
droplets stabilized by CTAB; (b) positively charged oil droplets adsorb to the negatively charged
PSPMK to cause brush collapse. The blue, yellow and green balls represent water, oil droplets and
CTAB molecules, respectively. The brush modification likely extends much farther into the
membrane than the figure shows.

2.3.5. Comparison of Filtration with PSPMK-Modified Membranes and NF270
Nanofiltration Membranes
We also performed dead-end emulsion filtration through commercial NF270 membranes.
We chose these membranes because their pure water specific flux (~19.6 LMH/bar) is
similar to that of brush-modified membranes, which enables a fair comparison of fouling
tendencies. Moreover, these membranes have a negatively charged surface (zeta potential
of -20 mV around neutral pH34, 52-53), and their small effective pore size (0.49 nm54) should
lead to nearly 100% oil rejection. As expected, the NF270 membranes exhibited oil
rejections >99.9%, regardless of surfactant. Initially, flux declines most rapidly for
solutions containing the cationic surfactant (Figure 2.23), presumably because of favorable
electrostatic interactions with the negatively charged surface.54 Even with an anionic
surfactant, however, the negatively charged surface is not sufficient to prevent fouling.

51

With SDS-stabilized emulsions, the permeate flux declines 98.7% over the course of the
12-h filtration. Given time, micelle adsorption to the membranes occurs regardless of
surfactant charge. Interestingly, the flux at the end of the filtration is higher for the CTABstabilized emulsion than for the SDS-stabilized system. After formation of an oil coating
on the membrane, fouling will likely depend more on the emulsion than on the properties
of the underlying brush.

Permeate Flux (LMH/bar)

24
Pure Water Flux = 19.6 LMH/bar

20
16
12
8
4
0
0

2

4

6

8

10

12

Time (h)

Figure 2.23. Evolution of specific permeate flux with time during filtration of aqueous submicron
emulsions through NF270 membranes. Blue squares and red circles represent filtration with an
SDS-stabilized and a CTAB-stabilized emulsion, respectively. The horizontal green line shows the
average pure water flux.

Compared to the brush-modified membranes prepared with a 1-h polymerization
(Figure 2.18a), the NF270 membranes show similar oil rejection (almost 100%), but the
brush-modified membranes exhibit no significant permeate flux declines with time. For
example, 12 h into filtration of SDS-stabilized emulsions, the permeate flux was ~18
LMH/bar for the PSPMK-modified membrane and only ~0.2 LMH/bar for the NF270
membrane. With CTAB-stabilized emulsions, the NF270 membrane is clearly superior to
the brush-modified membrane because of higher rejection. Nevertheless, the NF270

52

membrane still shows a strong fouling tendency in the cationic surfactant. We should also
note that a recent study suggests that growth of zwitterionic polymer brushes only on the
surface of membranes may lead to permeate fluxes up to 6000 LMH/bar,12 which could
make such membranes much more attractive for oil-water separations.
2.4.

Conclusions

This research investigated the effect of surfactants (a) on the swelling of polyanionic
brushes on planar surfaces and (b) on emulsion filtration through porous membranes
modified with similar brushes. Anionic surfactants do not alter the aqueous swelling of
PSPMK brushes, nor do they significantly decrease flux and rejection during filtration of
surfactant-stabilized oil emulsions through brush-modified membranes. In contrast,
cationic surfactants adsorb to the polyanionic brushes, presumably due to electrostatic
attraction. When polymer chain density and surfactant concentration are low, adsorption
of CTAB causes brush collapse, and on microporous membranes this collapse leads to
greatly enhanced flux and minimal rejection of oil droplets with diameters <1 Îźm. On
planar surfaces, high-density brushes do not collapse in the presence of CTAB, although
they do adsorb the surfactant. Unfortunately, synthesis of high-density brushes on
microporous membranes is a challenge due to a low functional group density on the
membrane surface. However, knowledge of the surfactants present in a given solution may
allow design of specific brushes (polyanionic or polycationic) that resist fouling.

53

REFERENCES

54

REFERENCES

1.
Cheryan, M.; Rajagopalan, N. Membrane Processing of Oily Streams. Wastewater
Treatment and Waste Reduction. J. Membr. Sci. 1998, 151, 13-28.
2.
40 CFR 435.15, Standards of Performance for New Sources (NSPS),
Environmental Protection Agency: Washington, DC, 1994.
3.
Duhon, H. Produced Water Treatment: Yesterday, Today, and Tomorrow. Oil Gas
Facil. 2012, 1, 29-31.
4.
Frankiewicz, T. In Understanding the Fundamentals of Water Treatment, the Dirty
Dozen — 12 Common Causes of Poor Water Quality, 11th Produced Water Seminar,
Houston, TX, Jan. 17-19; 2001.
5.
Duong, P. H. H.; Chung, T. S. Application of Thin Film Composite Membranes
with Forward Osmosis Technology for the Separation of Emulsified Oil-Water. J. Membr.
Sci. 2014, 452, 117-126.
6.
Hu, X.; Yu, Y.; Zhou, J.; Wang, Y.; Liang, J.; Zhang, X.; Chang, Q.; Song, L. The
Improved Oil/Water Separation Performance of Graphene Oxide Modified Al2O3
Microfiltration Membrane. J. Membr. Sci. 2015, 476, 200-204.
7.
Kota, A. K.; Kwon, G.; Choi, W.; Mabry, J. M.; Tuteja, A. Hygro-Responsive
Membranes for Effective Oil-Water Separation. Nat. Commun. 2012, 3, 1025-1032.
8.
Pan, Y. Q.; Wang, T. T.; Sun, H. M.; Wang, W. Preparation and Application of
Titanium Dioxide Dynamic Membranes in Microfiltration of Oil-in-Water Emulsions. Sep.
Purif. Technol. 2012, 89, 78-83.
9.
Tuteja, A.; Choi, W.; Mabry, J. M.; McKinley, G. H.; Cohen, R. E. Robust
Omniphobic Surfaces. P. Natl. Acad. Sci. USA 2008, 105 (47), 18200-18205.
10.
Kwon, G.; Kota, A. K.; Li, Y. X.; Sohani, A.; Mabry, J. M.; Tuteja, A. On-Demand
Separation of Oil-Water Mixtures. Adv. Mater. 2012, 24, 3666-3671.
11.
Solomon, B. R.; Hyder, M. N.; Varanasi, K. K. Separating Oil-Water
Nanoemulsions Using Flux-Enhanced Hierarchical Membranes. Sci. Rep. 2014, 4, 55045509.
12.
Gao, S. J.; Zhu, Y. Z.; Zhang, F.; Jin, J. Superwetting Polymer-Decorated SWCNT
Composite Ultrathin Films for Ultrafast Separation of Oil-in-Water Nanoemulsions. J.
Mater. Chem. A 2015, 3, 2895-2902.
13.

Lovelyn, C.; Attama, A. A. Current State of Nanoemulsions in Drug Delivery. J.

Biomater. Nanobiotechnol. 2011, 2, 626-639.
55

14.
Benita, S.; Levy, M. Submicron Emulsions as Colloidal Drug Carriers for
Intravenous Administration: Comprehensive Physicochemical Characterization. J. Pharm.
Sci. 1993, 82, 1069-1079.
15.
Wang, L. J.; Li, X. F.; Zhang, G. Y.; Dong, J. F.; Eastoe, J. Oil-in-Water
Nanoemulsions for Pesticide Formulations. J. Colloid Interf. Sci. 2007, 314, 230-235.
16.
Quintanilla-Carvajal, M. X.; Camacho-Diaz, B. H.; Meraz-Torres, L. S.; ChanonaPerez, J. J.; Alamilla-Beltran, L.; Jimenez-Aparicio, A.; Gutierrez-Lopez, G. F.
Nanoencapsulation: A New Trend in Food Engineering Processing. Food Eng. Rev. 2010,
2, 39-50.
17.
Bernardi, D.; Pereira, T.; Maciel, N.; Bortoloto, J.; Viera, G.; Oliveira, G.; RochaFilho, P. Formation and Stability of Oil-in-Water Nanoemulsions Containing Rice Bran
Oil: in vitro and in vivo Assessments. J. Nanobiotechnol. 2011, 9, 44-52.
18.
Talegaonkar, S.; Negi, L. Nanoemulsion in Drug Targeting. In Targeted Drug
Delivery : Concepts and Design, Devarajan, P. V.; Jain, S., Eds.; Springer International
Publishing, 2015, pp 433-459.
19.

Asua, J. M. Miniemulsion Polymerization. Prog. Polym. Sci. 2002, 27, 1283-1346.

20.
Bechthold, N.; Tiarks, F.; Willert, M.; Landfester, K.; Antonietti, M. Miniemulsion
Polymerization: Applications and New Materials. Macromol. Symp. 2000, 151, 549-555.
21.
Zhu, Y.; Zhang, F.; Wang, D.; Pei, X. F.; Zhang, W.; Jin, J. A Novel Zwitterionic
Polyelectrolyte Grafted PVDF Membrane for Thoroughly Separating Oil from Water with
Ultrahigh Efficiency. J. Mater. Chem. A 2013, 1, 5758-5765.
22.
Boussu, K.; Kindts, C.; Vandecasteele, C.; Van der Bruggen, B. Surfactant Fouling
of Nanofiltration Membranes: Measurements and Mechanisms. Chemphyschem 2007, 8,
1836-1845.
23.
Wandera, D.; Himstedt, H. H.; Marroquin, M.; Wickramasinghe, S. R.; Husson, S.
M. Modification of Ultrafiltration Membranes with Block Copolymer Nanolayers for
Produced Water Treatment: The Roles of Polymer Chain Density and Polymerization Time
on Performance. J. Membr. Sci. 2012, 403, 250-260.
24.
de Vos, W. M.; Biesheuvel, P. M.; de Keizer, A.; Kleijn, J. M.; Stuart, M. A. C.
Adsorption of Anionic Surfactants in a Nonionic Polymer Brush: Experiments,
Comparison with Mean-Field Theory, and Implications for Brush-Particle Interaction.
Langmuir 2009, 25, 9252-9261.
25.
Konradi, R.; RĂźhe, J. Binding of Oppositely Charged Surfactants to
Poly(methacrylic acid) Brushes. Macromolecules 2005, 38, 6140-6151.

56

26.
Dunlop, I. E.; Thomas, R. K.; Titmus, S.; Osborne, V.; Edmondson, S.; Huck, W.
T. S.; Klein, J. Structure and Collapse of a Surface-Grown Strong Polyelectrolyte Brush
on Sapphire. Langmuir 2012, 28, 3187-3193.
27.
Cao, Q. Q.; Zuo, C. C.; Li, L. J. Electrostatic Binding of Oppositely Charged
Surfactants to Spherical Polyelectrolyte Brushes. Phys. Chem. Chem. Phys. 2011, 13,
9706-9715.
28.
Cao, Q. Q.; Zuo, C. C.; Li, L. J.; Gao, M. F. Interactions of Polyelectrolyte Brushes
with Oppositely Charged Surfactants. Colloid Polym. Sci. 2011, 289, 1089-1102.
29.
DĂśbbelin, M.; Arias, G.; Loinaz, I.; Llarena, I.; Mecerreyes, D.; Moya, S. Tuning
Surface Wettability of Poly(3-sulfopropyl methacrylate) Brushes by Cationic SurfactantDriven Interactions. Macromol. Rapid Commun. 2008, 29, 871-875.
30.
Wang, H.; Zhang, H.; Yuan, S. L.; Liu, C. B.; Xu, Z. Molecular Dynamics Study
of the Adsorption of Anionic Surfactant in a Nonionic Polymer Brush. J. Mol. Model. 2014,
20, 2267-2275.
31.
Zhang, R.; Ma, S.; Wei, Q.; Ye, Q.; Yu, B.; van der Gucht, J.; Zhou, F. The Weak
Interaction of Surfactants with Polymer Brushes and Its Impact on Lubricating Behavior.
Macromolecules 2015, 48, 6186-6196.
32.
IndustryARC. Oilfield Surfactants Market Analysis. 2017. http://industryarc.com/Report/243/oil-field-surfactants-market-analysis.html (accessed Aug 12, 2017).
33.
Matyjaszewski, K.; Xia, J. Atom Transfer Radical Polymerization. Chem. Rev.
2001, 101, 2921-2990.
34.
Manttari, M.; Pekuri, T.; Nystrom, M. NF270, a New Membrane Having Promising
Characteristics and Being Suitable for Treatment of Dilute Effluents from the Paper
Industry. J. Membr. Sci. 2004, 242, 107-116.
35.
Shah, R. R.; Merreceyes, D.; Husemann, M.; Rees, I.; Abbott, N. L.; Hawker, C.
J.; Hedrick, J. L. Using Atom Transfer Radical Polymerization to Amplify Monolayers of
Initiators Patterned by Microcontact Printing into Polymer Brushes for Pattern Transfer.
Macromolecules 2000, 33, 597-605.
36.
Xu, F. J.; Zhao, J. P.; Kang, E. T.; Neoh, K. G.; Li, J. Functionalization of Nylon
Membranes via Surface-Initiated Atom-Transfer Radical Polymerization. Langmuir 2007,
23, 8585-8592.
37.
Masci, G.; Bontempo, D.; Tiso, N.; Diociaiuti, M.; Mannina, L.; Capitani, D.;
Crescenzi, V. Atom Transfer Radical Polymerization of Potassium 3-sulfopropyl
Methacrylate: Direct Synthesis of Amphiphilic Block Copolymers with Methyl
Methacrylate. Macromolecules 2004, 37, 4464-4473.

57

38.
Unsal, E.; Elmas, B.; Caglayan, B.; Tuncel, M.; Patir, S.; Tuncel, A. Preparation of
an Ion-Exchange Chromatographic Support by a "Grafting From" Strategy Based on Atom
Transfer Radical Polymerization. Anal. Chem. 2006, 78, 5868-5875.
39.
Huang, W.; Baker, G. L.; Bruening, M. L. Controlled Synthesis of Cross-Linked
Ultrathin Polymer Films by Using Surface-Initiated Atom Transfer Radical
Polymerization. Angew. Chem. 2001, 113, 1558-1560.
40.
Harris, J. J.; Bruening, M. L. Electrochemical and in Situ Ellipsometric
Investigation of the Permeability and Stability of Layered Polyelectrolyte Films. Langmuir
1999, 16, 2006-2013.
41.
Jasiewicz, K.; Pietrzak, R. Metals Ions Removal by Polymer Membranes of
Different Porosity. Scientific World J. 2013, 2013, 1-7.
42.
Density of Plastics. http://www.dotmar.com.au/density.html (accessed Sep 29
2015).
43.
Staples, E.; Penfold, J.; Tucker, I. Adsorption of Mixed Surfactants at the Oil-Water
Interface. J. Phys. Chem. B 2000, 104, 606-614.
44.
Palencia, M.; Rivas, B. L.; Pereira, E.; Hernandez, A.; Pradanos, P. Study of
Polymer-Metal Ion-Membrane Interactions in Liquid-Phase Polymer-Based Retention
(LPR) by Continuous Diafiltration. J. Membr. Sci. 2009, 336, 128-139.
45.
Kobayashi, M.; Terayama, Y.; Yamaguchi, H.; Terada, M.; Murakami, D.;
Ishihara, K.; Takahara, A. Wettability and Antifouling Behavior on the Surfaces of
Superhydrophilic Polymer Brushes. Langmuir 2012, 28, 7212-7222.
46.
Zhang, X. C.; Jackson, J. K.; Burt, H. M. Determination of Surfactant Critical
Micelle Concentration by a Novel Fluorescence Depolarization Technique. J. Biochem.
Bioph. Meth. 1996, 31, 145-150.
47.
Meaurio, E.; Zuza, E.; Lopez-Rodriguez, N.; Sarasua, J. R. Conformational
Behavior of Poly(L-lactide) Studied by Infrared Spectroscopy. J. Phys. Chem. B 2006, 110,
5790-5800.
48.
Kujawski, W.; Nguyen, Q.; Neel, J. Infrared Investigations of Sulfonated Ionomer
Membranes. I. Water–Alcohol Compositions and Counterions Effects. J. Appl. Polym. Sci.
1992, 44, 951-958.
49.
Leyte, J. C.; Zuiderwe.Lh; Vledder, H. J. An IR Investigation of PolyionCounterion Interactions. Spectrochim. Acta, Part A 1967, A 23, 1397-1407.
50.
Li, W.; Zhang, M.; Zhang, J.; Han, Y. Self-Assembly of Cetyl
Trimethylammonium Bromide in Ethanol-Water Mixtures. Front. Chem. China 2006, 1,
438-442.
51.
Solomon, B. R.; Hyder, N.; Varanasi, K. K. Separating Oil-Water Nanoemulsions
Using Flux-Enhanced Hierarchical Membranes. Sci. Rep. 2014, 4, 1-6.
52.
Nghiem, L. Removal of Emerging Trace Organic Contaminants by Nanofiltration
and Reverse Osmosis. Ph.D. Thesis [Online], University of Wollongong, Australia, 2005.
http://ro.uow.edu.au/theses/377/ (accessed Aug 12, 2017)

58

53.
Malaisamy, R.; Talla-Nwafo, A.; Jones, K. L. Polyelectrolyte Modification of
Nanofiltration Membrane for Selective Removal of Monovalent Anions. Sep. Purif.
Technol. 2011, 77, 367-374.
54.
Lee, J. H.; Chung, J. Y.; Chan, E. P.; Stafford, C. M. Correlating Chlorine-Induced
Changes in Mechanical Properties to Performance in Polyamide-Based Thin Film
Composite Membranes. J. Membr. Sci. 2013, 433, 72-79.

59

Chapter 3. Aqueous Swelling of Zwitterionic Poly(Sulfobetaine Methacrylate)
Brushes in the Presence of Ionic Surfactants
Portions of this chapter are being submitted to Macromolecules (Yang, Z., Tarabara, V.
V., and Bruening, M. L.).
3.1. Introduction
Polyelectrolyte brushes, dense assemblies of charged polymers attached through one end
to a surface,1 can be extremely hydrophilic and provide both enthalpic and entropic barriers
to nonspecific adsorption of a wide range of analytes.2-4 Thus, a number of studies
employed such coatings as anti-biofouling materials,5-10 and these brushes may also serve
as lubricating films with extremely low friction coefficients in water.11-14 Additionally,
several groups derivatized polyelectrolyte brushes with affinity reagents to capture specific
proteins.15-18 In certain cases, the properties of specific brushes can change in response to
temperature,19, 20 pH21-23 and solvent,24-27 creating “smart” surfaces.
However, small ions may diffuse into polyelectrolyte brushes and change their
swelling and anti-fouling properties. For example, polyanionic and polycationic brushes
collapse in solutions with increasing ionic strength due to charge screening and a decrease
in electrostatic repulsion.28-30 This work focuses on the impact of ionic surfactants on brush
swelling in aqueous solutions. Due to their amphiphilic nature that results from both
hydrophilic and hydrophobic moieties, surfactants are vital in a wide range of applications
such as protein solubilization,31, 32 emulsification,33-35 cleaning,36-38 and fabric softening.3941

Even low concentrations of surfactants may dramatically change the anti-fouling

properties of polyelectrolyte films. Our previous work showed that polyanionic brushes
resist adsorption of anionic surfactants,42 and polyanionic brush-coated membranes

60

completely repel emulsified oil droplets stabilized with these surfactants. However,
surfactants and surfactant-stabilized oil droplets with a charge opposite to that of the
polyelectrolyte brushes adsorb strongly to these coatings because of electrostatic
attraction.42, 43 As a result, in the presence of oppositely charged surfactants, polyelectrolyte
brushes become hydrophobic and collapse in polar solvents.
Polyzwitterionic brushes contain both positive and negative charges, and as with
polyanionic or polycationic brushes, the strong affinity of these charged groups for water
leads to a superhydrophilic film.9, 44-48 Nevertheless, the strong inter- and intra-molecular
associations between opposite charges in polyzwitterionic brushes lead to unusual trends
in swelling. For example, in contrast to polyanionic or polycationic brushes,
poly(sulfobetaine methacrylate) (PSBMA) brushes show enhanced swelling in aqueous
solutions with increasing ionic strength.29, 49-51 Thus, the presence of both cationic and
anionic groups in the polyzwitterions may lead to a different response to ionic surfactants
than in simple polyelectrolyte brushes. Although some studies examined surfactant
interactions with zwitterionic groups,52-58 very few of them focused on surfactantpolyzwitterionic brush interactions,53, 56 especially for brushes with varying thicknesses.
Zhou and co-workers recently showed no effect of surfactants on the lubrication behavior
of polyzwitterionic brushes, suggesting no surfactant uptake.58 However, here we show
that surfactant sorption is a function of surfactant charge and concentration, as well as brush
thickness.
This research examines the swelling of zwitterionic PSBMA brushes in the
presence of both anionic and cationic surfactants. We chose PSBMA because among the
common zwitterionic polymers poly(sulfobetaine methacrylate), poly(carboxybetaine

61

methacrylate) and poly(phosphobetaine methacrylate), the swelling behavior of PSBMA
in both water and salt solutions is better understood,29, 50 which may prove helpful in
explaining PSBMA swelling in the presence of ionic surfactants. We employ surfaceinitiated

atom

transfer

radical

polymerization

(SI-ATRP)

to

grow

poly[2-

(methacryloyloxy)ethyl dimethyl-(3-sulfopropyl)ammonium hydroxide] (PMEDSAH)
brushes on both Au-coated Si wafers and porous nylon membranes (Scheme 3.1). Films on
Au-coated wafers facilitate characterization of brush thickness, swelling, and surfactant
adsorption, whereas membranes enable examination of the effect of swelling on filtration.
a

b

Scheme 3.1. Surface-initiated atom transfer radical polymerization of MEDSAH on (a) Au-coated
wafers and (b) nylon membranes.

3.2.

Experimental

3.2.1. Materials
Hydrophilic nylon membranes with nominal 0.45 Îźm diameter pores were obtained from
Sterlitech (25 mm diameter, NY4525100). Si (111) wafers were purchased from University
Wafer (Boston, MA) and coated with Au at LGA Thin Films (Santa Clara, CA) by

62

sputtering 200 nm of gold on 20 nm of Cr on Si wafers. [2-(Methacryloyloxy) ethyl]
dimethyl-(3-sulfopropyl)

ammonium

hydroxide

(97%,

MEDSAH),

dopamine

hydrochloride (98%), 11-mercapto-1-undecanol (97%), ammonium chloride (≥99.5%), αbromoisobutyryl bromide (98%, BIBB), 2,2’-bipyridine (≥99%, bpy) and sodium dodecyl
sulfate

(98.5%,

SDS)

were

obtained

from

Sigma-Aldrich.

Tris(hydroxymethyl)aminomethane hydrochloride (≥99%, Tris hydrochloride) was
purchased from Invitrogen, and dodecyltrimethylammonium bromide (99%, DTAB) was
acquired from Acros Organics. Bromine (100%) and triethylamine (100%) were purchased
from J.T. Baker, whereas copper (I) chloride (≥99.995%), copper (II) chloride dihydrate
(≥99.0%) and dimethylformamide (≥99.8%, DMF) were acquired from Jade Scientific.
Acetone (99.7%) and ethyl acetate were purchased from Fisher Scientific and anhydrous
dichloromethane, hexane (analytical reagent grade, AR) and chloroform (AR) were
obtained from Macron Fine Chemicals. Dichloromethane was used as received, and
chloroform, toluene and triethylamine were dried with molecular sieves (3 Å, SigmaAldrich) before use. Aqueous solutions were prepared using deionized water (Milli-Q, 18.2
MΊ¡cm). The disulfide initiator, (BrC(CH3)2COO(CH2)11S)2 was synthesized according to
literature procedures.59
3.2.2. Preparation of PMEDSAH Brushes on Au-Coated Wafers and Nylon Membranes
3.2.2.1.Initiator Immobilization
Immobilization of the disulfide initiator on Au-coated wafers included cleaning the
substrate with ethanol, drying with N2, cleaning with UV/ozone for 15 min, immersion in
a 1 mM ethanolic solution of (BrC(CH3)2COO(CH2)11S)2 for 24 h, rinsing with ethanol and
drying under a stream of N2. The protocol for immobilizing the initiator on nylon

63

membranes combined two literature procedures.60, 61 Nylon membranes were extracted
with acetone for 12 h and then dried with N2. Dopamine hydrochloride (2.1 mmol) was
dissolved in DMF (20 mL), and this mixture was degassed via three freeze-pump-thaw
cycles before transfer into a glove bag and mixing with triethylamine (1.1 mmol). BIBB
(1.1 mmol) was then added dropwise to the mixture with gentle stirring, followed by
stirring for 3 h. The solution was subsequently well mixed with 100 mL of Tris
hydrochloride solution (2.8 M), and the pH was adjusted to 8.5 using NaOH. Clean nylon
membranes were immersed in this dopamine solution for 48 h, removed from the glove
bag and the solution, and rinsed with ~100 mL of water and dried with N2, forming
polydopamine initiators (PDA-BIBB) on the surface.
3.2.2.2.Polymerization of MEDSAH
PMEDSAH brushes on Au-coated wafers were prepared using the parameters of a
procedure for the synthesis of polyanionic brushes.42 MEDSAH (35 mmol) was dissolved
in a DMF/water mixture (1/1 v/v, 30 mL) in a round-bottom flask. The mixture was
degassed via four freeze-pump-thaw cycles. The catalyst was prepared by dissolving CuCl
(0.016 mmol), CuCl2 (0.016 mmol) and bpy (0.08 mmol) in a DMF/water mixture (1/1 v/v,
40 mL) and degassing via four freeze-pump-thaw cycles. Then 5 mL of the Cu-bpy solution
was mixed with the monomer solution in a N2-filled glove bag. The polymerization
solution was injected into vials, and wafers or membranes modified with initiators were
submerged in these solutions. After various polymerization times, the wafers or membranes
were removed from the glove bag, rinsed with ~100 mL of water and dried with N2. All
membrane samples in this work were prepared with a polymerization time of 24 h.
Polymerization times on Au-coated wafers ranged from 0.5 to 8 h.

64

3.2.3. Characterization of Modified Au-Coated Wafers and Nylon Membranes
3.2.3.1.Fourier Transform Infrared Spectroscopy (FTIR)
Au-coated wafers modified with disulfide initiator and PMEDSAH brushes were
characterized by reflectance FTIR spectroscopy (Nicolet 6700 FT-IR spectrometer,
Thermo Scientific, with a Pike grazing angle (80°) attachment and p-polarization). A
UV/ozone-cleaned Au-coated wafer served as a background. To examine surfactant
adsorption to swollen brushes, brush-modified wafers were transferred directly from water
to surfactant solutions without drying. After a 12-h immersion in the surfactant solution,
the wafer was rinsed with ~100 mL of water from a wash bottle for ~2 min prior to drying
with N2 and obtaining a reflectance FTIR spectrum. To determine whether surfactant
adsorbs to brushes in solution and can be removed by water rinsing, wafers were not rinsed
by water after removal from the surfactant solutions. Instead, the solution on wafers was
wiped off with a KimWipe to remove surfactants in solutions, followed by IR
characterization. A bare Au-coated wafer treated in the same way served as a control. IR
data are reported as a pseudo absorbance, which is the logarithm of the ratio of the reflected
intensities from the bare wafer and from the film-coated wafer.
3.2.3.2.Ellipsometry
Both dry and swollen thicknesses and refractive indices of the polymer brushes on Aucoated wafers were determined using a rotating analyzer ellipsometer (model M-44, J.A.
Woollam) at an incident angle of 75°. For determining the swollen thickness, brush-coated
wafers were immersed in solution in a homemade trapezoidal cell with glass windows.62,
63

To determine brush thickness and optical constants, the refractive index of the brush was

65

fit using an effective medium approximation (EMA) model, as shown in Figure 3.1. For
samples in air, the EMA model contains a Cauchy material for PMEDSAH brushes along
with air-filled voids. The Cauchy material employed Eq. 3.1 to fit its refractive indices,
n (Îť) = A +

B

Eq. 3.1

Îť2

where n is the refractive index, Îť is the wavelength of incident light in Âľm, and A and B
are constants. To characterize brushes, initial estimate values of A and B were set as 1.5
and 0.01, respectively. Initial guesses for the fractions of both polymers and air were 50%.
When samples were immersed in solutions, an ambient layer representing the solution was
set as the top material above the EMA layer (Figure 3.1b). The refractive index of the
ambient layer was determined as a Cauchy material in an experiment with reflection from
a Si wafer covered by a 25-nm thick SiO2 layer (Figure 3.1c). Since all solutions are
aqueous, the initial estimates of the parameters A and B in the Cauchy equation for the
ambient layer were set as 1.33 and 0.01, respectively. In the EMA model, initial guesses
for the fractions of polymers and solutions were 50%. The actual values of all parameters
were determined by fitting calculated data to experimental data. Ellipsometric
measurements were performed on three spots on each wafer.
a

ambient: air

b

ambient: solution

c

ambient: solution

EMA
(Cauchy + void)

EMA
(Cauchy + Ambient)

SiO2

Au

Au

Si

Figure 3.1. Optical models for PMEDSAH brushes in (a) air and (b) solutions. Image (c) shows
the model we used to determine optical constants of surfactant solutions.

3.2.3.3.X-Ray Photoelectron Spectroscopy (XPS)

66

XPS was carried out with a Perkin-Elmer PHI 5600 ESCA system with an Al KÎą X-ray
source (1486.6 eV) running at a power of 350 W at a take-off angle of 45°.
3.2.3.4.Scanning Electron Microscopy (SEM)
The surface morphology of membranes was observed with field-emission SEM (JSM7500F, JEOL). The membranes were mounted on the sample studs with double-sided
conductive adhesive tape. A thin layer of osmium (8 nm) was coated on the sample surface
prior to the SEM imaging, which was performed at an accelerating voltage of 5 kV. To
characterize their cross-sections, membranes were frozen in liquid nitrogen and after
removal from the nitrogen, they were quickly cracked with a razor blade to create a sharp
edge with minimal membrane deformation.
3.2.3.5.Contact Angle Measurements
Contact angles of water or hexadecane on surfaces were determined using a phone camera
and contact angle calculation software (VCA2000, AST Products). Contact angles of water
in air were determined by placing a drop of water on dry samples. To determine the surface
hydrophilicity in water and surfactant solutions, the samples were first immersed in these
solutions for 12 h. Then the samples were placed facing downward in a funnel capped with
a rubber stopper and filled with water or surfactant solutions. A drop of hexadecane was
released from beneath the samples using a syringe (Figure 3.2). We employed this method
because hexadecane droplets tend to float away from needles and surfaces in these
solutions.

67

Figure 3.2. Apparatus for the measurement of hexadecane contact angles on wafers in surfactant
solutions.

3.2.4. Determination of Membrane Permeability and Brush Swelling on Membranes
The specific fluxes through membranes (L/(m2¡h¡bar)), LMH/bar) were determined in
dead-end filtration using a 15-mL Amicon cell (Model 8010, Millipore) with a suspended
stir bar rotating at 45 rpm. The cell was connected to a stainless steel feed tank (standard 2
gallon pressure vessel, Pope Scientific) (Figure 3.3), and the applied pressure was constant
at 0.7 bar. Pure water permeability was determined using deionized water as the feed,
whereas the permeability in surfactant and salt solutions was measured by filling the feed
tank with those solutions. Changes in the swelling of PMEDSAH brushes on membranes
were estimated via the Hagen-Poiseuille equation (Eq. 3.2),64
𝐽𝐽

ΔP

=

nπr4

8ητΔx

68

Eq. 3.2

where J is the permeate flux (L/(m2¡h)), r is the radius of membrane pores, n is the number

of pores per unit area, ΔP is the transmembrane pressure (0.7 bar), η is the dynamic

viscosity, which is almost the same for water and the solutions examined in this work,65-67
Δx is the membrane thickness (~100 µm) and τ is the pore tortuosity. Thus, by comparing
water permeabilities of modified membranes and bare nylon membranes and assuming
constant τ and n, the pore size of modified membranes can be calculated using Eq. 3.3 and
a value for the radii of pores in the unmodified membrane.
J1
J2

𝑟𝑟

= ďż˝ 1ďż˝
𝑟𝑟2

4

Eq. 3.3

The swollen thickness of brushes is equal to the difference between the pore radius of bare
membranes and brush-modified membranes. (This is a crude approximation given the
spongy nature of these membranes.) All of the filtration experiments were repeated with 2
or 3 membranes.

Figure 3.3. Diagram of the apparatus for dead-end filtration.

3.3.

Results and Discussion

3.3.1. Brush Synthesis on Au-Coated Wafers and Nylon Membranes
69

Scheme 3.1 shows the SI-ATRP strategy we employ to grow polyzwitterionic brushes on
Au-coated wafers. The reflective Au surface enables brush characterization with
reflectance IR spectroscopy and ellipsometry. IR spectra confirm both the disulfide
initiator adsorption (Figure 3.4a) and PMEDSAH brush growth (Figure 3.4b). In particular,
the presence of sulfonate stretches at 1218 and 1041 cm-1, a C=O stretch at 1731 cm-1 and
a C-N stretch at 1486 cm-1 are consistent with the presence of brushes on the surface. Figure
3.5 shows that both the peak height for the sulfonate stretch at 1218 cm-1 and the dry
thickness of PMEDSAH brushes increase with polymerization time, indicating continuous
brush growth. With the exception of the longest polymerization time, the sulfonate
stretching peak height and thickness increase proportionally, giving confidence in the
characterization techniques. The relatively large standard deviations in Figure 3.5 and
disagreement in IR and ellipsometry data at the longest polymerization time may suggest
a rough surface and some uncontrolled polymerization. This could stem in part from the
insolubility of PMEDSAH in common organic solvents (such as DMF) despite full
dissolution of the monomer MEDSAH. Thus, polymerization may proceed in a
heterogeneous system that gives rise to a broad molecular weight distribution and large
thickness variations.68
a
Absorbance

0.0005

2000

C-O-C
asym.
str.

C=O
str.
1738

CH2
scissor
1468

C-O-C
sym. str.
-1
1172 cm

-1

1800

1600

1400

Wavenumber

70

(cm-1)

1200

1000

Figure 3.4 (cont’d)

b

-

0.1

Absorbance

-

-SO3
asym.
str.

C=O
str.

-SO3
sym.
str.

C-N
str.

2000

1800

1600

1400

1200

Wavenumber

1000

800

(cm-1)

Figure 3.4. Reflectance FTIR spectrum of (a) the disulfide initiator ((BrC(CH3)2COO(CH2)11S)2)
and (b) PMEDSAH brushes on a Au-coated Si wafer. The brush dry thickness was 412 nm.
0.8
0.6

400

0.4
200

0.2

IR Absorbance (-SO3-)

Dry Thickness (nm)

600

0

0
0

100

200

300

400

500

600

Polymerization Time (min)

Figure 3.5. Dry brush thickness (blue circles) and IR absorbances of sulfonate groups at 1218 cm1

(red squares) as a function of polymerization time for PMEDSAH brushes grown on Au wafers.

Error bars represent the standard deviation of measurements on three different films.

We employed XPS to characterize the growth of the PDA-BIBB initiator and
zwitterionic PMEDSAH brushes on non-reflective nylon membranes. Figure 3.6 shows
XPS wide scan spectra of the surfaces of a bare nylon membrane, a PDA-BIBB-coated
nylon membrane, and a PMEDSAH-coated nylon membrane. After the polymerization of
dopamine-BIBB on the membrane, signals representing Br 3d (76 eV), Br 3p (188 eV) and
71

Br 3s (261.1 eV) appear, which is consistent with the presence of initiating groups on the
PDA-BIBB coated surface (Figure 3.6b). The strong S 2p (172 eV) and S 2s (236 eV)
signals only detected on PMEDSAH-coated membranes (Figure 3.6c) confirm the growth
of PMEDSAH brushes. Signals from Br remain after polymerization, suggesting that some
initiators are unreacted or some of the chain ends remain active after 24 h of polymerization.

a

N 1s
O 1s

Intensity (Arb. Units)

C 1s

b
Br 3d

Br 3p

Br 3s

c
Br 3d
S 2p

Br 3p
S 2s
Br 3s

0

300

600

900

Binding Energy (eV)

Figure 3.6. XPS wide scan spectra of (a) bare nylon membranes and membranes modified with (b)
PDA-BIBB initiator and (c) PMEDSAH brushes.

SEM images show the differences in morphology for bare, PDA-BIBB- and
PMEDSAH-modified nylon membranes. Top-view SEM images (Figure 3.7a, b, c) suggest
a decreasing porosity and pore size after coating with PDA-BIBB and PMEDSAH. Crosssectional images (Figure 3.7d, e, f) show that the membrane remains porous throughout the

72

modification steps, but due to the spongy structure, we cannot tell from these images
(Figure 3.7d, e, f) whether the pore size decreases in the membrane interior after coating.
a

c

b

10 Âľm

d

10 Âľm

e

10 Âľm

f

Figure 3.7. SEM images of (a,d) bare nylon membranes and membranes modified with (b,e) PDABIBB initiator and (c, f) PMEDSAH brushes. Figures a, b, c are the top-view images while Figures
d, e, f are the cross-sectional images. Modification employed 48 h of PDA-BIBB deposition and
24 h of polymerization.

3.3.2. Brush Swelling and Wettability
Figure 3.8 shows the swollen thicknesses of PMEDSAH brushes on Au-coated wafers in
deionized water along with their percent swelling (compared to a dry film in air) as a
function of the dry brush thickness. In the plot, the swollen thickness appears to be a linear
function of the dry thickness, with ~150% swelling. However, the thinnest brushes (dry
thickness <20 nm) tend to swell more (~250%). This agrees with another measure of the
brush density, the ratio of the sulfonate group IR absorbance (1218 cm-1) to the swollen
brush thickness.69, 70 This ratio has its lowest values in films with dry thicknesses <20 nm
(Figure 3.9). Previous research shows that the formation of intra-group, inter-chain, and
intra-chain associations in polyzwitterionic brushes yields crosslinked networks.29, 68 Very

73

thin brushes with short polymer chains and low chain densities likely possess fewer interand intra-chain associations, leading to higher swelling in water. As brushes grow thicker,
longer chains give more chances for crosslinking and formation of denser networks. After
the dry brush thickness surpasses 20 nm, the brush density in water is relatively constant.
However, we should note that determination of swollen thickness is more prone to error

800

400%

600

300%

400

200%

200

100%

Swelling %

Swollen Thickness (nm)

for thinner films.

0%

0
0

100

200

300

400

Dry Thickness (nm)

Figure 3.8. Water-swollen thicknesses (red squares) and percent swelling (blue circles) of
PMEDSAH brushes as a function of their dry thickness. Error bars represent standard deviations
of measurements on three films. Error bars are not always visible due to their small values. Swollen

IR Absorbance / Swollen Thickness
(nm-1)

thicknesses were determined in water after a 12-h immersion.
0.0008

I

0.0007

II

0.0006
0.0005
0.0004
0.0003
0.0002
0

100

200

Dry Thickness (nm)

74

300

400

Figure 3.9. (cont’d) Reflectance IR absorbance at 1218 cm-1 (dry films) divided by the swollen
PMEDSAH brush thickness in water as a function of the dry film thickness. The ratio of IR
absorbance to swollen film thickness provides a measure of density. The plot is divided into two
regions according to dry brush thickness: I (< 20 nm), II (> 20 nm). Error bars represent standard
deviations of measurements on three films.

We characterized surface wettability by measuring hexadecane contact angles in
both air and water. As Figure 3.10 shows, on PMEDSAH-coated wafers and membranes,
the contact angle of hexadecane increases from ~10° to 150-170° after changing the
ambient region from air to water. In most cases, hexadecane will bead up and eventually
detach from the surface on both wafers and membranes. In control experiments, bare Aucoated wafers show hexadecane contact angles smaller than 20° in both air and water.
These results show that PMEDSAH brushes on both gold and modified nylon become
superoleophobic underwater, consistent with previous research.30, 61
a

~ 0°

b

16°

c

12°

d

174°

e

~10°

f

~150°

Figure 3.10. Contact angles of hexadecane on bare Au-coated wafers (a, b), PMEDSAH-modified
wafers (c, d) and PMEDSAH-modified membranes (e, f). The images were obtained in air (a, c, e)
and water (b, d, f). Images in water are rotated 180°.

3.3.3. Interactions between Ionic Surfactants and PMEDSAH Brushes
PSBMA brushes are well known as anti-biofouling materials,71-73 and they also show
significant adsorption of inorganic salts, which increases brush swelling.29,

75

50

Ionic

surfactants contain both charged groups and alkyl backbones and possess properties of both
ions and hydrophobic molecules. Thus, surfactants may affect swelling very differently
than simple salts, so this section examines interactions between PSBMA brushes and ionic
surfactants, as well as the resulting brush swelling.
3.3.3.1.Anionic Surfactant (SDS)
3.3.3.1.1. Effect of PMEDSAH Brush Thickness on Swelling
To study SDS adsorption as a function of PMEDSAH brush thickness, we prepared an
extensive series of films with dry thicknesses varying from 2 to 700 nm. As Figure 3.11

Absorbance

0.05

4000

3000

2000

Wavenumber

1000

(cm-1)

Figure 3.11. Reflectance IR spectra of a PMEDSAH brush before (red) and after (blue) immersion
in a 6 mM SDS solution and rinsing with water. The brush dry thickness was 340 nm.

shows, the IR spectra of PMEDSAH-coated wafers before and after immersion in SDS
solutions are essentially identical. The only difference is that the peak heights after
immersion in SDS solutions decrease by about 5%, which may indicate a small amount of
polymer degradation.74, 75 In addition, the swollen thickness in pure water is the same
before and after immersion in SDS solutions (see Figure 3.13 below), indicating that any
SDS adsorption on PMEDSAH brushes and its effect on brush swelling disappear after

76

water rinsing. To characterize surfactant adsorption for films while immersed in surfactant
solutions, we dried brush-coated wafers with a KimWipe to remove surfactant solutions
prior to IR characterization but did not rinse with water. The spectra of films treated in this
way show a clear increase in the intensities of the CH2 symmetric and asymmetric stretches
at 2854 and 2927 cm−1 (Figure 3.12), indicating SDS adsorption. To demonstrate that this
adsorption stems from the SDS in PMEDSAH brush layers, rather than the SDS in residual
solution left on the surface, we performed the same experiment with a bare Au-coated
wafer. As Figure 3.12 shows, this control experiment showed no absorbance from CH2
stretches, suggesting that the new CH2 peaks in SDS-exposed PMEDSAH films stem from
SDS adsorbed to PMEDSAH brushes. This result is consistent with previous research
where quartz crystal microbalance with dissipation monitoring (QCM-D) showed SDS
adsorption in PMEDSAH brushes.58

4000

Absorbance

Absorbance

0.05

3000

Wavenumber

2000

3200

1000

CH2 asym. str.
2927 cm-1

0.005

CH2 sym. str.
2854 cm-1
3000

2800

2600

Wavenumber (cm-1)

(cm-1)

Figure 3.12. Reflectance IR spectra of a PMEDSAH brush after immersion in 6 mM SDS followed
by a water rinse (red), and after immersion in the same SDS solution without a water rinse (blue).
In the latter case, the film was dried with a Kimwipe to remove residual solution. A bare Au-coated
surface after immersion in 6 mM SDS without water rinsing (dried with a Kimwipe) served as a
control (green). The figure on the right expands the region of the spectrum where CH2 stretches
appear. The brush dry thickness was 340 nm.

77

Figure 3.13 shows the swelling of PMEDSAH brushes in both water and 6 mM SDS
solutions. We divide the data into three regions according to their dry thickness: I (<20
nm), II (20-150 nm) and III (>150 nm). For PMEDSAH brushes in region I, a significant
thickness increase occurs upon transfer of brushes from water into 6 mM SDS. As
discussed above, crosslinked networks are likely present in the PMEDSAH brush, and SDS
may diffuse into the brush and screen charges on polymer chains, interrupting the
crosslinks and allowing polymer chains to stretch. Weaker polymer networks in thinner
films probably lead to a lower brush density in SDS solution than in water (Figure 3.14).

Swollen Thickness (nm)

800

I

II

III

600

400

200

0
0

100

200

300

400

Dry Thickness (nm)

Figure 3.13. Ellipsometric thicknesses of swollen PMEDSAH brushes in 6 mM SDS (blue circles),
as well as in deionized water before (red squares) and after (green triangles) immersion in SDS
solution and rinsing. Thicknesses are plotted as a function of dry brush thickness in ambient air,
and dashed vertical lines depict dry thicknesses of 20 and 150 nm. Error bars represent standard
deviations of measurements on three films. The error bars are often not obvious due to their small
values, and red and green symbols typically overlap.

78

IR Absorbance / Swollen Thickness
(nm-1)

0.0008

I

II

III

0.0006
0.0004
0.0002
0
0

100

200

300

400

Dry Thickness (nm)

Figure 3.14. PMEDSAH brush density, as indicated by IR absorbance (1218 cm-1 in dry films)
divided by the brush thickness in water (red squares) or by the brush thickness in 6 mM SDS (blue
circles) as a function of dry brush thickness. The plot is divided into three regions according to dry
brush thickness: I (< 20 nm), II (20-150 nm), III (> 150 nm). Error bars represent standard
deviations of measurements on three films. Some of the error bars are not visible due to their small
values.

In region II, where the dry thickness of PMEDSAH brushes is between 20 and 150
nm, the swollen brush thickness in SDS is essentially constant at around 400 nm (Figure
3.13), so the brush density in 6 mM SDS increases with brush dry thickness (Figure 3.14)
and the percent thickness difference between PMEDSAH films in 6 mM SDS and in water
declines (Figure 3.15). Remarkably, even in region III in Figure 3.13, the brush swollen
thickness in 6 mM SDS still remains constant, so the film density becomes greater than in
deionized water (Figure 3.14) and the percent thickness difference between brushes in
deionized water and in 6 mM SDS becomes negative (Figure 3.15). Chen and co-workers
reported that binding of salts like NaCl disturbs the ordering of strongly hydrogen-bonded
water molecules in the zwitterionic brush layer to disrupt superhydrophilicity.76 We think
SDS has the same effect as high concentrations of salts. In addition, the hydrophobic tail

79

Percent Thickness Difference (%)

800%

I

II

III

600%
400%
200%
0%
-200%
0

100

200

300

400

Dry Thickness (nm)

Figure 3.15. Percent thickness difference between PMEDSAH films in 6 mM SDS and in water as
a function of the dry brush thickness. The plot is divided into three regions according to dry brush
thickness: I (< 20 nm), II (20-150 nm) and III (> 150 nm). Error bars represent standard deviations
of measurements on three films. Some of the error bars are not visible due to their small values.

of SDS will reduce the hydrophilicity of PMEDSAH brushes. Unlike thin brushes with
short polymer chains that tend to swell upon breaking of crosslinks, thick brushes consist
of longer and more extended chains. Therefore, as brushes grow thicker, the disordered
water binding and increased hydrophobicity after SDS adsorption will lead to less brush
swelling despite disruption of crosslinking. However, even for thick films hexadecane
cannot attach to these brush-coated surfaces in 6 mM SDS, whereas a control experiment
shows that hexadecane quickly sticks to a bare Au-coated wafer (Figure 3.16). As the
contact angle only reflects the wettability of the exposedsurface,77-79 this result indicates
that the outermost brush layer is still superhydrophilic, although the brush swells less than
in deionized water.

80

a

c

b

45°

d

54°

Figure 3.16. Contact angles of hexadecane on (a) PMEDSAH brush-modified Au-wafer in 1 mM
SDS, (b) PMEDSAH brush-modified Au-wafer in 6 mM SDS, (c) bare Au-wafer in 1 mM SDS
and (d) bare Au-wafer in 6 mM SDS. PMEDSAH brushes in (a) and (b) have a dry thickness of
260 nm.

3.3.3.1.2. Swelling as a Function of SDS Concentration
Figure 3.17 shows the percent thicknesses difference of PMEDSAH brushes in SDS
solutions (compared to in water) as a function of SDS concentrations. To examine the effect
of concentrations both above and below the SDS critical micelle concentration (8 mM),80
we set the surfactant concentration range from 0.5 to 20 mM. As Figure 3.17 reveals, for
thin brushes (dry thickness < 150 nm) increasing the SDS concentration from 0.5 to 6 mM
leads to higher brush swelling, presumably due to more charge screening. With a brush dry
thickness >150 nm, increasing the SDS concentration leads to a thickness decrease. In this
case due to the more flexible and extended chains, the disordering of water molecules and
the hydrophobicity of SDS may dominate thickness changes, and thus the increase in SDS
concentration will lead to more SDS adsorption and a less swollen polymer network.
Above 6 mM, further increasing the SDS concentration has minimal effect on the swelling
of PMEDSAH brushes, probably because SDS adsorption in brush layers reaches its upper
limit near the critical micelle concentration (8 mM). Figure 3.18 shows additional data on
the swollen thicknesses and percent swelling differences (relative to in deionized water) of
81

PMEDSAH brushes in 0.5, 1, 2, 4, 12, 16, 20 mM SDS. These data follow the trends in

Percent Thickness Difference (%)

Figure 3.17.
700%

500%

300%

100%

-100%
0

5

10

15

20

SDS Concentration (mM)

Figure 3.17. Percent thickness difference of PMEDSAH brushes in SDS solutions (compared to in
water) as a function of SDS concentration and dry thickness. Symbols in different colors and shapes
represent brushes with dry thickness of 33 nm (blue empty squares), 24 nm (green empty circles),
33 nm (blue empty squares), 67 nm (green solid diamonds), 136 nm (yellow solid triangles), 200
nm (red solid squares), and 267 nm (brown solid circles). Error bars represent standard deviations
of measurements on three films. The Y error bars are not obvious due to their small values.

82

Figure 3.18 (cont’d)

83

Figure 3.18 (cont’d)

Figure 3.18. Ellipsometric thicknesses (a, c, e, g, i, k, m) and percent thickness differences relative
to in deionized water (b, d, f, h, j, l, n) of swollen PMEDSAH brushes in 0.5, 1, 2, 4, 12, 16, 20
mM SDS as a function of dry brush thickness. In the figures on the left, thicknesses of swollen
PMEDSAH brushes in SDS, as well as in deionized water before and after immersion in SDS
solution and rinsing are plotted as blue circles, red squares and green triangles, respectively. Error
bars represent standard deviations of measurements on three films. Some error bars are not visible
due to their small values. All the data in Figure 3.18 was measured using the same batch of samples
in the following order: 0.5, 1, 6, 12, 20, 4, 2, 16 mM SDS.

3.3.3.1.3. Swelling in Salt Solutions
To examine if the changes in brush swelling upon SDS addition are simple chargescreening effects, we also characterized the swelling behavior of PMEDSAH brushes in
NaCl solutions. As Figures 3.19 and 3.20 show, low NaCl concentrations (5 mM) lead to
small increases in brush swelling below 50% (compared to swelling in deionized water),
whereas brushes in 500 mM NaCl show a swelling trend similar to that in 6 mM SDS
(compare Figure 3.13 and Figure 3.21). This suggests that disordering of water molecules
in thick PMEDSAH films decreases brush thickness in both SDS and 500 mM NaCl
84

solutions. Compared with NaCl, SDS starts to induce the unusual brush swelling trends
from a much lower concentration, presumably because the hydrophobic tail of SDS favors

Swollen Thickness (nm)

partitioning into the film.
1200
900
600
300
0
0

100

200

300

400

500

Dry Thickness (nm)

Figure 3.19. Swollen thicknesses of PMEDSAH brushes in 5 mM NaCl (blue circles), and the
deionized water-swollen thicknesses before (red squares) and after (green triangles) immersion in
NaCl (and rinsing) as a function of dry brush thickness. Error bars represent standard deviations of

Percent Thickness Difference (%)

measurements on three films. Some error bars are not visible due to their small values.
50%
40%
30%
20%
10%
0%
0

100

200

300

400

500

Dry Thickness (nm)

Figure 3.20. Percent thickness difference between PMEDSAH films in 5 mM NaCl and in water
as a function of the dry brush thickness. Error bars represent standard deviations of measurements
on three films.

85

Swollen Thickness (nm)

1000
800
600
400
200
0
0

100

200

300

400

Dry Thickness (nm)

Figure 3.21. Swollen thicknesses of PMEDSAH brushes in 500 mM NaCl (blue circles), and the
water-swollen thickness before (red squares) and after (green triangles) immersion in NaCl (and
rinsing) as a function of dry brush thickness. Error bars represent standard deviations of
measurements on three films. Some error bars are not visible due to their small values.

3.3.3.2.Cationic Surfactant (DTAB)
To examine the effect of surfactant charge on zwitterionic brushes, we selected DTAB as a cationic
surfactant because it has the same hydrophobic tail as SDS. Similar to experiments with SDS, IR
spectra show that DTAB adsorbs to PMEDSAH brushes immersed in surfactant solution (Figure

4000

0.02

Absorbance

Absorbance

3.22), but rinsing readily removes the surfactant (Figure 3.23).

3000

2000

3200

1000

CH2 asym. str.
2923 cm-1

0.01

CH2 sym. str.
2854 cm-1

3000

2800

2600

Wavenumber (cm-1)

Wavenumber (cm-1)

Figure 3.22. Reflectance IR spectra of a PMEDSAH brush after immersion in 10 mM DTAB and
a water rinse (red), and after immersion in 10 mM DTAB without a water rinse (blue). In the latter
case, the film was dried with a Kimwipe to remove residual solution. A bare Au-coated surface
after immersion in 10 mM DTAB without water rinsing (dried with a Kimwipe) served as a control
(green). The brush dry thickness was 340 nm.

86

Absorbance

0.05

4000

3000

2000

Wavenumber

1000

(cm-1)

Figure 3.23. Reflectance IR spectra of a PMEDSAH brush before (red) and after (blue) immersion
in 10 mM DTAB solution and rinsing. The brush dry thickness was 340 nm.

3.3.3.2.1. Effect of Brush Thickness
As Figure 3.24 shows, when dry PMEDSAH brushes are thinner than 200 nm, their
swelling in 5 mM DTAB is slightly higher than that in water (from a 70% to 0% increase
in swollen thickness as brushes grow thicker), presumably due to charge screening and
breaking of crosslinks. However, for these films the percent increase in swelling (compared
to in water) is around 10-fold lower than with 6 mM SDS. This low swelling in the solution
of cationic surfactant may occur because the quaternary ammonium cation of DTAB has
hydrophobic methyl groups that decrease partitioning into the superhydrophilic brush layer
or make films more hydrophobic. The HLB (hydrophilic-lipophilic balance) values
calculated by the Davies method for SDS and DTAB are 40 and 23.3, respectively.81, 82 For
PMEDSAH brushes with dry thicknesses around 300 nm, immersion in 5 mM DTAB
results in collapsed brushes (compared to in deionized water) with a water-swollen
thickness around 400 nm, which is the same swelling behavior as in concentrated SDS

87

solutions. This probably results from the disordering of water molecules and the
hydrophobic tail of surfactants.

Swollen Thickness (nm)

1000
800
600
400
200
0
0

100

200

300

400

Dry Thickness (nm)

Figure 3.24. Swollen thickness of PMEDSAH brushes in 5 mM DTAB solution (blue circles), and
the water-swollen thickness before (red squares) and after (green triangles) immersion in DTAB
solutions and rinsing as a function of dry brush thickness. Error bars represent standard deviations
of measurements on three films. Some error bars are not obvious due to their small values.

3.3.3.2.2. Effect of DTAB Concentration
In solutions with DTAB concentrations of 1 and 2 mM, the swelling of PMEDASH brushes
is similar with that shown in Figure 3.24 (see Figure 3.25). When the DTAB concentration
is 10 or 20 mM, however, no decrease in PMEDSAH brush swelling (relative to in
deionized water) occurs regardless of brush thickness. Instead, there is a steady brush
swelling 50% higher than in water, except for very thin films which show swelling >50%
higher than in water (Figure 3.25). Instead of moving freely on PMEDSAH brush surfaces
in 6 mM SDS and in 5 mM DTAB, hexadecane droplets attach to the brush surface in 10
mM DTAB, indicating reduced hydrophilicity of the outermost brush layers. As DTAB
molecules diffuse into PMEDSAH brushes, some of them will form ion pairs with
zwitterions, but the binding is not strong due to the hydrophobicity of DTAB. Thus, we
suppose when DTAB concentration is close to its critical micelle concentration, which is
14 mM,83 the bonded DTAB molecules may form micelles with free DTAB molecules in

88

the outermost brush region, decreasing the brush superhydrophilicity and perhaps
preventing further DTAB diffusion into the brushes.

Figure 3.25. Ellipsometric thicknesses (a, c, e, g) and percent thickness differences relative to in
deionized water (b, d, f, h) of swollen PMEDSAH brushes in 1, 2, 10, and 20 mM DTAB as a
function of dry brush thickness. Thicknesses of swollen PMEDSAH brushes in DTAB, as well as
in deionized water before and after immersion in DTAB and rinsing are plotted as blue circles, red
squares and green triangles, respectively. Error bars represent standard deviations of measurements
on three films. Some error bars are not visible due to their small values.

89

3.3.3.3.Dead-End Filtration of Surfactant Solutions with PMEDSAH-Modified Membranes
To examine the implications of ionic surfactant adsorption to PSBMA brushes, we
performed dead-end filtrations of surfactant solutions through PMEDSAH-modified
membranes. As discussed above, thick PMEDSAH brushes decrease in thickness only in

Permeability (LMH/Bar)

a
10000
7500
5000

water

water

SDS

DTAB

water

2500
0
0

5

10

15

20

Time (h)

Permeability (LMH/Bar)

b
20000
15000
10000

water

SDS

water

DTAB

water

5000
0
0

5

10

15

20

25

Time (h)

Figure 3.26. Dead-end filtrations of 6 mM SDS (red squares) and 2 mM DTAB (green triangles)
through (a) PMEDSAH-coated nylon membranes and (b) bare nylon membranes. Pure water
permeability (blue circles) was determined between surfactant filtrations.

SDS solutions with surfactant concentrations of 6 mM and above or in DTAB solutions
with surfactant concentrations of 5 mM and below. Thus, we prepared 6 mM SDS and 2
mM DTAB to examine the swelling of brushes on membranes. Figure 3.26 shows that pure
water permeabilities of the PMEDSAH-modified membrane and the bare nylon membrane
with a pore size of 0.45 Âľm are ~7500 and ~13000 LMH/Bar, respectively. With the crude

90

assumption of a uniform distribution of cylindrical pores with a diameter of 0.45 Âľm, and
using Equation 3, these filtration data suggest a swollen brush thickness of ~30 nm in
deionized water. This low thickness implies that PMEDSAH brushes are much thinner than
brushes on Au-coated wafers. Based on the permeabilities of PMEDSAH-modified
membranes in surfactant solutions, the calculated swollen brush thicknesses in 6 mM SDS
and 2 mM DTAB are 180 and 130 nm, respectively. This is consistent with the high
swelling of thin brushes in 6 mM SDS (Figure 3.13) and relatively low swelling in 2 mM
DTAB (Figure 3.25). The adsorption of SDS to PMEDSAH brushes on nylon induces
charge screening and brush swelling, leading to blockage of membrane pores and
extremely low permeability. Less adsorption of DTAB, presumably due to the
hydrophobicity of its head group, induces lower brush swelling and a higher permeability.
As discussed above, rinsing with water removes SDS and DTAB from polymer brushes.
Thus, the membrane permeability returns to its original value after rinsing with water, as
Figure 26 shows. Figure 3.26b also suggests significant adsorption of multilayers of DTAB
in bare nylon membranes, probably due to the hydrophobicity of DTAB. Given the small
length of DTAB (< 2 nm) compared with the membrane pore size (0.45 Âľm), the low fluxes
imply multilayer adsorption. In contrast, SDS decreases flux through nylon membranes by
only 30%. Figure 3.27 shows replicate data for filtrations.

91

Permeability (LMH/Bar)

a
10000
7500
5000

water SDS water DTAB

water

2500
0
0

5

10

15

20

Time (h)

Permeability (LMH/Bar)

b
20000
15000
10000

water SDS

water DTAB water

5000
0
0

5

10

15

20

Time (h)

Figure 3.27. Replicate dead-end filtrations of 6 mM SDS (red squares) and 2 mM DTAB (green
triangles) through (a) PMEDSAH-coated nylon membranes and (b) bare nylon membranes. Pure
water permeability (blue circles) was determined before and after surfactant filtrations. These
experiments were performed with a different set of membranes than Figure 3.26.

3.4.

Conclusions

This research investigated the sorption of ionic surfactants in zwitterionic PSBMA brushes
along with the resulting brush swelling. For thin brushes with dry thicknesses <20 nm, the
presence of SDS leads to a swelling increase of about 400% with respect to swelling in
water. Presumably the surfactant diffuses into the brush layer and screens charges to disrupt
ionic crosslinks and enhance aqueous swelling. In contrast, for thicker films the surfactant
decreases aqueous swelling, probably because of the hydrophobicity of the surfactant tail

92

and the disordering of water molecules around zwitterions. Similar trends in swelling as a
function of dry film thickness occur in 500 mM NaCl. The more hydrophobic cationic
surfactant DTAB has less effect on brush swelling than SDS for thin brushes. However,
with low surfactant concentrations (≤ 5 mM) and thick (>200 nm) brushes DTAB has the
same effect as SDS: a decrease in aqueous swelling. Filtration data with brush-modified
membranes are consistent with high swelling of thin brushes in SDS, as permeability
decreases by 99.8% in 6 mM SDS. Meanwhile, the membrane permeability in 2 mM DTAB
is more than 20 times higher than that in 6 mM SDS, consistent with lower swelling of
PMEDSAH brushes in the cationic surfactant.

93

REFERENCES

94

REFERENCES

1.
Barbey, R.; Lavanant, L.; Paripovic, D.; Schuwer, N.; Sugnaux, C.; Tugulu, S.;
Klok, H. A. Polymer Brushes Via Surface-Initiated Controlled Radical Polymerization:
Synthesis, Characterization, Properties, and Applications. Chem. Rev. 2009, 109, 54375527.
2.
Li, L. Y.; Chen, S. F.; Zheng, J.; Ratner, B. D.; Jiang, S. Y. Protein Adsorption on
Oligo(Ethylene Glycol)-Terminated Alkanethiolate Self-Assembled Monolayers: The
Molecular Basis for Nonfouling Behavior. J. Phys. Chem. B 2005, 109, 2934-2941.
3.
Chen, S. F.; Li, L. Y.; Zhao, C.; Zheng, J. Surface Hydration: Principles and
Applications toward Low-Fouling/Nonfouling Biomaterials. Polymer 2010, 51, 52835293.
4.
Jeon, S.; Lee, J.; Andrade, J.; De Gennes, P. Protein—Surface Interactions in the
Presence of Polyethylene Oxide: I. Simplified Theory. J. Colloid Interface Sci. 1991, 142,
149-158.
5.
Bernards, M. T.; Cheng, G.; Zhang, Z.; Chen, S. F.; Jiang, S. Y. Nonfouling
Polymer Brushes Via Surface-Initiated, Two-Component Atom Transfer Radical
Polymerization. Macromolecules 2008, 41, 4216-4219.
6.
Chang, Y.; Liao, S. C.; Higuchi, A.; Ruaan, R. C.; Chu, C. W.; Chen, W. Y. A
Highly Stable Nonbiofouling Surface with Well-Packed Grafted Zwitterionic
Polysulfobetaine for Plasma Protein Repulsion. Langmuir 2008, 24, 5453-5458.
7.
Feng, W.; Brash, J. L.; Zhu, S. P. Non-Biofouling Materials Prepared by Atom
Transfer Radical Polymerization Grafting of 2-Methacryloloxyethyl Phosphorylcholine:
Separate Effects of Graft Density and Chain Length on Protein Repulsion. Biomaterials
2006, 27, 847-855.
8.
Feng, W.; Zhu, S. P.; Ishihara, K.; Brash, J. L. Adsorption of Fibrinogen and
Lysozyme on Silicon Grafted with Poly(2-Methacryloyloxyethyl Phosphorylcholine) Via
Surface-Initiated Atom Transfer Radical Polymerization. Langmuir 2005, 21, 5980-5987.
9.
Zhang, Z.; Chen, S. F.; Jiang, S. Y. Dual-Functional Biomimetic Materials:
Nonfouling Poly(Carboxybetaine) with Active Functional Groups for Protein
Immobilization. Biomacromolecules 2006, 7, 3311-3315.
10.
Higaki, Y.; Kobayashi, M.; Murakami, D.; Takahara, A. Anti-Fouling Behavior of
Polymer Brush Immobilized Surfaces. Polymer Journal 2016, 48, 325-331.
11.
Kobayashi, M.; Takahara, A. Synthesis and Frictional Properties of Poly(2,3Dihydroxypropyl Methacrylate) Brush Prepared by Surface-Initiated Atom Transfer
Radical Polymerization. Chem. Lett. 2005, 34, 1582-1583.
95

12.
Kobayashi, M.; Terayama, Y.; Hosaka, N.; Kaido, M.; Suzuki, A.; Yamada, N.;
Torikai, N.; Ishihara, K.; Takahara, A. Friction Behavior of High-Density Poly(2Methacryloyloxyethyl Phosphorylcholine) Brush in Aqueous Media. Soft Matter 2007, 3,
740-746.
13.
Sakata, H.; Kobayashi, M.; Otsuka, H.; Takahara, A. Tribological Properties of
Poly(Methyl Methacrylate) Brushes Prepared by Surface-Initiated Atom Transfer Radical
Polymerization. Polymer Journal 2005, 37, 767-775.
14.
Wei, Q. B.; Cai, M. R.; Zhou, F.; Liu, W. M. Dramatically Tuning Friction Using
Responsive Polyelectrolyte Brushes. Macromolecules 2013, 46, 9368-9379.
15.
Wang, W.; Li, L.; Henzler, K.; Lu, Y.; Wang, J.; Han, H.; Tian, Y.; Wang, Y.;
Zhou, Z.; Lotze, G. Protein Immobilization onto Cationic Spherical Polyelectrolyte
Brushes Studied by Small Angle X-Ray Scattering. Biomacromolecules 2017, 18, 15741581.
16.
Levin, A.; Czeslik, C. Effect of High Pressure on Protein Binding to Poly (Acrylic
Acid) Brushes. Biophys. J. 2017, 112, 590a.
17.
Khalaf, R.; de Neuville, B. C.; Morbidelli, M. Protein Adsorption in Polyelectrolyte
Brush Type Cation-Exchangers. J. Chromatogr. A 2016, 1471, 126-137.
18.
Iwasaki, Y.; Omichi, Y.; Iwata, R. Site-Specific Dense Immobilization of Antibody
Fragments on Polymer Brushes Supported by Silicone Nanofilaments. Langmuir 2008, 24,
8427-8430.
19.
Xia, F.; Feng, L.; Wang, S.; Sun, T.; Song, W.; Jiang, W.; Jiang, L. Dual‐
Responsive Surfaces That Switch between Superhydrophilicity and Superhydrophobicity.
Adv. Mater. 2006, 18, 432-436.
20.
Azzaroni, O.; Brown, A. A.; Huck, W. T. S. UCST Wetting Transitions of
Polyzwitterionic Brushes Driven by Self-Association. Angew. Chem. Int. Ed. 2006, 45,
1770-1774.
21.
Zhang, J.; Kou, R.; Liu, G. Effect of Salt Concentration on the pH Responses of
Strong and Weak Polyelectrolyte Brushes. Langmuir 2017.
22.
Liu, G. M.; Zhang, G. Z. Periodic Swelling and Collapse of Polyelectrolyte Brushes
Driven by Chemical Oscillation. J. Phys. Chem. B 2008, 112, 10137-10141.
23.
Zhou, F.; Shu, W. M.; Welland, M. E.; Huck, W. T. S. Highly Reversible and MultiStage Cantilever Actuation Driven by Polyelectrolyte Brushes. J. Am. Chem. Soc. 2006,
128, 5326-5327.
24.
Gao, K.; Kearney, L. T.; Howarter, J. A. Planar Phase Separation of Weak
Polyelectrolyte Brushes in Poor Solvent. J.Polymer Sci. Part B: Polymer Phys. 2017, 55,
370-377.
96

25.
Li, L.; Li, J.; Lukehart, C. M. Graphitic Carbon Nanofiber-Poly(Acrylate) Polymer
Brushes as Gas Sensors. Sensor Actuat. B-Chem 2008, 130, 783-788.
26.
Santer, S.; Kopyshev, A.; Donges, J.; Yang, H. K.; RĂźhe, J. Dynamically
Reconfigurable Polymer Films: Impact on Nanomotion. Adv. Mater. 2006, 18, 2359-2362.
27.
Li, D. J.; Sheng, X.; Zhao, B. Environmentally Responsive "Hairy" Nanoparticles:
Mixed Homopolymer Brushes on Silica Nanoparticles Synthesized by Living Radical
Polymerization Techniques. J. Am. Chem. Soc. 2005, 127, 6248-6256.
28.
Terayama, Y.; Arita, H.; Ishikawa, T.; Kikuchi, M.; Mitamura, K.; Kobayashi, M.;
Yamada, N.; Takahara, A. In Chain Dimensions in Free and Immobilized Brush States of
Polysulfobetaine in Aqueous Solution at Various Salt Concentrations, J. Phys. Conf. Ser.,
2011; IOP Publishing: 2011; p 012010.
29.
Hoshino, T.; Tanaka, Y.; Jinnai, H.; Takahara, A. Surface and Interface Analyses
of Polymer Brushes by Synchrotron Radiation. J. Phys. Soc. Jpn. 2013, 82.
30.
Kobayashi, M.; Terayama, Y.; Yamaguchi, H.; Terada, M.; Murakami, D.;
Ishihara, K.; Takahara, A. Wettability and Antifouling Behavior on the Surfaces of
Superhydrophilic Polymer Brushes. Langmuir 2012, 28, 7212-7222.
31.
Moore, P. N.; Puvvada, S.; Blankschtein, D. Role of the Surfactant Polar Head
Structure in Protein-Surfactant Complexation: Zein Protein Solubilization by SDS and by
SDS/C12En Surfactant Solutions. Langmuir 2003, 19, 1009-1016.
32.
Deo, N.; Jockusch, S.; Turro, N. J.; Somasundaran, P. Surfactant Interactions with
Zein Protein. Langmuir 2003, 19, 5083-5088.
33.
Rosen, M. J., Emulsification by Surfactants. In Surfactants and Interfacial
Phenomena, John Wiley & Sons, Inc., 2004; p 303-331.
34.
Bouchemal, K.; Briancon, S.; Perrier, E.; Fessi, H. Nano-Emulsion Formulation
Using Spontaneous Emulsification: Solvent, Oil and Surfactant Optimisation. Int. J.
Pharm. 2004, 280, 241-251.
35.

Walstra, P. Principles of Emulsion Formation. Chem. Eng. Sci. 1993, 48, 333-349.

36.
Franklin, V. J. Cleaning Efficacy of Single-Purpose Surfactant Cleaners and MultiPurpose Solutions. Contact Lens and Anterior Eye 1997, 20, 63-68.
37.
Kaplan, M. C.; Jegou, A.; Chaufer, B.; Rabiller-Baudry, M.; Michalsky, M. C.
Adsorption of Lysozyme on Membrane Material and Cleaning with Non-Ionic Surfactant
Characterized through Contact Angle Measurements. Desalination 2002, 146, 149-154.
38.
Jurado, E.; Bravo, V.; Luzon, G.; Fernandez-Serrano, M.; Garcia-Roman, M.;
Altmajer-Vaz, D.; Vicaria, J. M. Hard-Surface Cleaning Using Lipases: EnzymeSurfactant Interactions and Washing Tests. J. Surfactants Deterg. 2007, 10, 61-70.

97

39.
Giolando, S. T.; Rapaport, R. A.; Larson, R. J.; Federle, T. W.; Stalmans, M.;
Masscheleyn, P. Environmental Fate and Effects of Deedmac - a New Rapidly
Biodegradable Cationic Surfactant for Use in Fabric Softeners. Chemosphere 1995, 30,
1067-1083.
40.
Miao, Z. C.; Yang, J. Z.; Wang, L.; Liu, Y.; Zhang, L.; Li, X. L.; Peng, L. Synthesis
of Biodegradable Lauric Acid Ester Quaternary Ammonium Salt Cationic Surfactant and
Its Utilization as Calico Softener. Mater. Lett. 2008, 62, 3450-3452.
41.
Saraiva, S. A.; Abdeinur, P. V.; Catharino, R. R.; Nunes, G.; Eberlin, M. N. Fabric
Softeners: Nearly Instantaneous Characterization and Quality Control of Cationic
Surfactants by Easy Ambient Sonic-Spray Ionization Mass Spectrometry. Rapid Commun.
Mass Spec. 2009, 23, 357-362.
42.
Yang, Z. F.; Tarabara, V. V.; Bruening, M. L. Adsorption of Anionic or Cationic
Surfactants in Polyanionic Brushes and Its Effect on Brush Swelling and Fouling
Resistance During Emulsion Filtration. Langmuir 2015, 31, 11790-11799.
43.
Dunlop, I. E.; Thomas, R. K.; Titmus, S.; Osborne, V.; Edmondson, S.; Huck, W.
T. S.; Klein, J. Structure and Collapse of a Surface-Grown Strong Polyelectrolyte Brush
on Sapphire. Langmuir 2012, 28, 3187-3193.
44.
Li, X.; Hu, X.; Cai, T. Construction of Hierarchical Fouling Resistance Surfaces
onto Poly (Vinylidene Fluoride) Membranes for Combating Membrane Biofouling.
Langmuir 2017, 33, 4477-4489.
45.
Higaki, Y.; Fröhlich, B.; Yamamoto, A.; Murakami, R.; Kaneko, M.; Takahara, A.;
Tanaka, M. Ion-Specific Modulation of Interfacial Interaction Potentials between Solid
Substrates and Cell-Sized Particles Mediated Via Zwitterionic, Super-Hydrophilic Poly
(Sulfobetaine) Brushes. J. Phys. Chem. B 2017, 121, 1396-1404.
46.
Ezzat, M.; Huang, C. J. Zwitterionic Polymer Brush Coatings with Excellent AntiFog and Anti-Frost Properties. RSC. Adv. 2016, 6, 61695-61702.
47.
Kobayashi, M.; Terayama, Y.; Kikuchi, M.; Takahara, A. Chain Dimensions and
Surface Characterization of Superhydrophilic Polymer Brushes with Zwitterion Side
Groups. Soft Matter, 2014, 9, 5138-5148.
48.
Sun, Y.; Chen, C.; Xu, H.; Lei, K.; Xu, G.; Zhao, L.; Lang, M. Surface Modification
of Silicon Wafer by Grafting Zwitterionic Polymers to Improve Its Antifouling Property.
Appl. Surf. Sci. 2017, 419, 642-649.
49.
Polzer, F.; Heigl, J.; Schneider, C.; Ballauff, M.; Borisov, O. V. Synthesis and
Analysis of Zwitterionic Spherical Polyelectrolyte Brushes in Aqueous Solution.
Macromolecules 2011, 44, 1654-1660.

98

50.
Kobayashi, M.; Ishihara, K.; Takahara, A. Neutron Reflectivity Study of the
Swollen Structure of Polyzwitterion and Polyeletrolyte Brushes in Aqueous Solution. J.
Biomat. Sci.-Polym. E 2014, 25, 1673-1686.
51.
Xiang, T.; Luo, C. D.; Wang, R.; Han, Z. Y.; Sun, S. D.; Zhao, C. S. Ionic-StrengthSensitive Polyethersulfone Membrane with Improved Anti-Fouling Property Modified by
Zwitterionic Polymer Via in Situ Cross-Linked Polymerization. J. Membr. Sci. 2015, 476,
234-242.
52.
Che, Y.-J.; Tan, Y.; Ren, X.; Xin, H.; Meng, F. Solution Properties of
Hydrophobically Modified Acrylamide-Based Polysulfobetaines in the Presence of
Surfactants. Colloid Polym. Sci. 2012, 290, 1237-1245.
53.
Chen, H.; Kelley, M.; Guo, C.; Yarger, J. L.; Dai, L. L. Adsorption and Release of
Surfactant into and from Multifunctional Zwitterionic Poly(NIPAm-co-DMAPMA-coAAc) Microgel Particles. J. Colloid Interface Sci. 2015, 449, 332-340.
54.
Costa, T.; de Azevedo, D.; Stewart, B.; Knaapila, M.; Valente, A. J. M.; Kraft, M.;
Scherf, U.; Burrows, H. D. Interactions of a Zwitterionic Thiophene-Based Conjugated
Polymer with Surfactants. Polym. Chem. 2015, 6, 8036-8046.
55.
Johnson, K. M.; Fevola, M. J.; Lochhead, R. Y.; McCormick, C. L.
Hydrophobically Modified Acrylamide-Based Polybetaines. II. Interaction with
Surfactants in Aqueous Solution. J. Appl. Polym. Sci. 2004, 92, 658-671.
56.
Johnson, K. M.; Poe, G. D.; Lochhead, R. Y.; McCormick, C. L. The Synthesis of
Hydrophobically Modified Water ‐ Soluble Polyzwitterionic Copolymers and
Responsiveness to Surfactants in Aqueous Solution. J. Macromol. Sci., Part A 2004, 41,
587-611.
57.
Kudaibergenov, S.; Jaeger, W.; Laschewsky, A. In Supramolecular Polymers
Polymeric Betains Oligomers; Abe, A., Kobayashi., S., Eds.; Springer, 2006; pp 157-224.
58.
Zhang, R.; Ma, S.; Wei, Q.; Ye, Q.; Yu, B.; van der Gucht, J.; Zhou, F. The Weak
Interaction of Surfactants with Polymer Brushes and Its Impact on Lubricating Behavior.
Macromolecules 2015, 48, 6186-6196.
59.
Shah, R. R.; Merreceyes, D.; Husemann, M.; Rees, I.; Abbott, N. L.; Hawker, C.
J.; Hedrick, J. L. Using Atom Transfer Radical Polymerization to Amplify Monolayers of
Initiators Patterned by Microcontact Printing into Polymer Brushes for Pattern Transfer.
Macromolecules 2000, 33, 597-605.
60.
Li, C. Y.; Wang, W. C.; Xu, F. J.; Zhang, L. Q.; Yang, W. T. Preparation of pHSensitive Membranes Via Dopamine-Initiated Atom Transfer Radical Polymerization. J.
Membr. Sci. 2011, 367, 7-13.
61.
Ye, G.; Lee, J.; Perreault, F.; Elimelech, M. Controlled Architecture of DualFunctional Block Copolymer Brushes on Thin-Film Composite Membranes for Integrated
99

“Defending” and “Attacking” Strategies against Biofouling. ACS Appl. Mater. Inter. 2015,
7, 23069-23079.
62.
Harris, J. J.; Bruening, M. L. Electrochemical and in Situ Ellipsometric
Investigation of the Permeability and Stability of Layered Polyelectrolyte Films. Langmuir
2000, 16, 2006-2013.
63.
Huang, W.; Baker, G. L.; Bruening, M. L. Controlled Synthesis of Cross-Linked
Ultrathin Polymer Films by Using Surface-Initiated Atom Transfer Radical
Polymerization. Angew. Chem. Int. Ed. 2001, 40, 1510-1512.
64.
Mulder, J., Basic Principles of Membrane Technology, 2nd Ed.; Springer, 1996;
Vol. 4, p 170.
65.
Hossain, M. F.; Biswas, T. K.; Islam, M.; Huque, M. E. Volumetric and
Viscometric Studies on Dodecyltrimethylammonium Bromide in Aqueous and in Aqueous
Amino Acid Solutions in Premicellar Region. Monatsh. Chem. Chem. Mon. 2010, 141,
1297-1308.
66.
Ito, T.; Mizutani, H. The Viscosity of Aqueous Sodium Dodecyl Sulfate Solution
and the Effect of Inorganic Electrolytes on It. J. Japan Oil Chemists' Soc. 1968, 17, 246248.
67.
Kestin, J.; Khalifa, H. E.; Correia, R. J. Tables of the Dynamic and Kinematic
Viscosity of Aqueous NaCl Solutions in the Temperature Range 20–150 C and the Pressure
Range 0.1–35 Mpa. J. Phys. Chem. Ref. Data 1981, 10, 71-88.
68.
Terayama, Y.; Kikuchi, M.; Kobayashi, M.; Takahara, A. Well-Defined Poly
(Sulfobetaine) Brushes Prepared by Surface-Initiated ATRP Using a Fluoroalcohol and
Ionic Liquids as the Solvents. Macromolecules 2010, 44, 104-111.
69.
Crompton, T. R., Practical Polymer Analysis. Springer Science & Business Media:
2012.
70.
Milosevic, M.; King, S. W. Validation of a Correction Procedure for Removing the
Optical Effects from Transmission Spectra of Thin Films on Substrates. J. Appl. Phys.
2012, 112, 093514.
71.
Bengani, P.; Kou, Y.; Asatekin, A. Zwitterionic Copolymer Self-Assembly for
Fouling Resistant, High Flux Membranes with Size-Based Small Molecule Selectivity. J.
Membr. Sci. 2015, 493, 755-765.
72.
Li, Q.; Bi, Q.-Y.; Liu, T.-Y.; Wang, X.-L. Resistance to Protein and Oil Fouling of
Sulfobetaine-Grafted Poly (Vinylidene Fluoride) Hollow Fiber Membrane and the
Electrolyte-Responsive Behavior in NaCl Solution. Appl. Surf. Sci. 2012, 258, 7480-7489.
73.
Xiang, T.; Luo, C.-D.; Wang, R.; Han, Z.-Y.; Sun, S.-D.; Zhao, C.-S. IonicStrength-Sensitive Polyethersulfone Membrane with Improved Anti-Fouling Property
100

Modified by Zwitterionic Polymer Via in Situ Cross-Linked Polymerization. J. Membr.
Sci. 2015, 476, 234-242.
74.
Tugulu, S.; Klok, H.-A. Stability and Nonfouling Properties of Poly (Poly
(Ethylene Glycol) Methacrylate) Brushes under Cell Culture Conditions.
Biomacromolecules 2008, 9, 906-912.
75.
Sheiko, S. S.; Sun, F. C.; Randall, A.; Shirvanyants, D.; Rubinstein, M.; Lee, H.-i.;
Matyjaszewski, K. Adsorption-Induced Scission of Carbon-Carbon Bonds. Nature 2006,
440, 191.
76.
Leng, C.; Han, X.; Shao, Q.; Zhu, Y.; Li, Y.; Jiang, S.; Chen, Z. In Situ Probing of
the Surface Hydration of Zwitterionic Polymer Brushes: Structural and Environmental
Effects. J. Phys. Chem. C 2014, 118, 15840-15845.
77.
Hänni-Ciunel, K.; Findenegg, G. H.; von Klitzing, R. Water Contact Angle on
Polyelectrolyte-Coated Surfaces: Effects of Film Swelling and Droplet Evaporation. Soft
Mater. 2007, 5, 61-73.
78.
Park, I. J.; Lee, S.-B.; Choi, C. K.; Kim, K.-J. Surface Properties and Structure of
Poly (Perfluoroalkylethyl Methacrylate). J. Colloid Interface Sci. 1996, 181, 284-288.
79.
Zisman, W. A. In Contact Angle, Wettability, and Adhesion, Fowkes. M. F., Ed.;
ACS, 1964; Vol. 1, p 22.
80.
Khan, A. M.; Shah, S. S. Determination of Critical Micelle Concentration (CMC)
of Sodium Dodecyl Sulfate (SDS) and the Effect of Low Concentration of Pyrene on Its
CMC Using Origin Software. J. Chem. Soc. Pakistan 2008, 30, 186.
81.
Davies, J. A Quantitative Kinetic Theory of Emulsion Type, I. Physical Chemistry
of the Emulsifying Agent. Proc. 2nd Intern. Congr. Surface Activity, Butterworths
Scientific Publication, London 1957, 426.
82.
Proverbio, Z.; Bardavid, S.; Arancibia, E.; Schulz, P. Hydrophile–Lipophile
Balance and Solubility Parameter of Cationic Surfactants. Colloids Surf. A: Physicochem.
Eng. Asp. 2003, 214, 167-171.
83.
Bahri, M. A.; Hoebeke, M.; Grammenos, A.; Delanaye, L.; Vandewalle, N.; Seret,
A. Investigation of SDS, DTAB and CTAB Micelle Microviscosities by Electron Spin
Resonance. Colloids Surf. A: Physicochem. Eng. Asp. 2006, 290, 206-212.

101

Chapter 4. Oil Coalescence on Superoleophobic Membrane Surfaces during DeadEnd and Cross-Flow Filtrations
4.1.

Introduction

Membrane filtration plays an important role in oil-water separations, especially for removal
of submicron and surfactant-stabilized oil droplets from water.1-4 However, fouling
presents a major challenge in these and other membrane applications.5-7 Foulants attach to
the surface and block membrane pores during filtration to greatly reduce membrane
permeability. As a result, frequent membrane cleaning or replacement may make filtration
unacceptably expensive.
To solve this problem, research groups around the world are creating new
membranes with excellent fouling resistance. For example, Yan and co-workers modified
ultrafiltration membranes with nano-sized alumina to resist fouling in wastewater from an
oil field.8 Chang produced a poly(ethylene glycol)-coated poly (vinylidene fluoride)
membrane that resists fouling by bovine serum albumin (BSA).9 Rahimpour and coworkers prepared anti-fouling nano-porous PVDF membranes using UV photo-grafting to
coat these substrates with both acrylate and amino monomers.10 However, these modified
membranes will most likely show fouling resistance only with diluted solutions or in the
initial filtration stage.11 In addition, the complex or expensive protocols for membrane
modification are probably not suitable for large-scale applications.1
An equally viable separation strategy could aim to use membranes not as filters but
as substrates that catalyze the coalescence of oil droplets. Coalescence has been part of
effective oil-removal methods for years,12 because tiny oil droplets do not readily separate
from water. Over time many improvements to coalescence have appeared. For example,

102

plate separators employ parallel plates to prompt oil contact and coalescence.13 Induced
gas flotation relies on oil coalescing with injected air bubbles to form large and light oilair bubbles that float to the surface.14-16 Nevertheless, these traditional techniques can only
separate relatively large oil droplets (> 50 Âľm).17
In 2001, Nitsch and co-workers reported membrane-induced oil coalescence and
proposed a mechanism for this process.17 Many subsequent studies aimed to improve this
technique. Huck’s research group coated surfaces with polymer brushes to capture oil
droplets and release them after they coalescence.18 Yang and co-workers designed
nonwoven filter mats for the coalescence of emulsified oil droplets,19 and Agarwal and coworkers investigated several factors that affect membrane-induced oil coalescence,
including surface roughness and orientation.20, 21 The Fane group also studied factors that
affect oil coalescence on membranes, especially the pore size and imposed in-pore shear
rate.22
Most studies employ hydrophobic membranes for oil coalescence, but this chapter
describes my attempts to use a superoleophobic membrane to induce droplet aggregation.20
As reported previously, hydrophobic membranes may lose their functionality after
adsorption of a large amount of oil.19 Thus, even for membranes that act as coalescers, too
much oil adsorption is not productive for droplet coalescence. I hypothesized that a balance
between oleophobicity and hydrophobicity may create a surface that induces oil
coalescence and simultaneously resists oil adsorption. To initially examine this hypothesis,
I prepared zwitterionic poly(sulfobetaine methacrylate) (PSBMA) brushes on different
membranes and examined whether oil will coalesce on these substrates. Although PSBMA

103

brushes are superoleophobic in water,23, 24 their net uncharged surface may permit weak
adsorption of charged oil droplets and subsequent coalescence.
4.2.

Experimental

4.2.1. Materials
Hydrophilic nylon membranes with nominal 0.45 Îźm diameter pores (25 mm diameter,
NY4525100) was obtained from Sterlitech. Hydrophilic track-etched polycarbonate
membranes with 0.2 Âľm diameter pores were purchased from Sigma-Aldrich (25 mm
diameter, WHA110606). Si (111) wafers were obtained from University Wafer (Boston,
MA) and coated with Au at LGA Thin Films (Santa Clara, CA) by sputtering 200 nm of
gold on 20 nm of Cr on Si wafers. [2-(Methacryloyloxy) ethyl] dimethyl-(3-sulfopropyl)
ammonium hydroxide (97%, MEDSAH), 3-sulfopropyl methacrylate potassium salt (98%,
SPMK), dopamine hydrochloride (98%), 11-mercapto-1-undecanol (97%), ammonium
chloride (≥99.5%), α-bromoisobutyryl bromide (98%, BIBB), 2,2’-bipyridine (≥99%, bpy)
and sodium dodecyl sulfate (98.5%, SDS) were purchased from Sigma-Aldrich.
Tris(hydroxymethyl)aminomethane hydrochloride (≥99%, Tris hydrochloride) was
obtained from Invitrogen, and dodecyltrimethylammonium bromide (99%, DTAB) was
acquired from Acros Organics. Bromine (100%) and triethylamine (100%) were purchased
from J.T. Baker, whereas copper (I) chloride (≥99.995%), copper (II) chloride dihydrate
(≥99.0%) and dimethylformamide (≥99.8%, DMF) were acquired from Jade Scientific.
Acetone (99.7%) and ethyl acetate were purchased from Fisher Scientific. Anhydrous
dichloromethane, hexane (analytical reagent grade, AR) and chloroform (AR) were
obtained from Macron Fine Chemicals. Dichloromethane was used as received, and
chloroform, toluene and triethylamine were dried with molecular sieves (3 Å, Sigma-

104

Aldrich) before use. Araldite 2020 two-component epoxy adhesive system was obtained
from Huntsman. Aqueous solutions were prepared using deionized water (Milli-Q, 18.2
MΊ¡cm), and the disulfide initiator, (BrC(CH3)2COO(CH2)11S)2 was synthesized
according to literature procedures.25
4.2.2. Preparation of Polyelectrolyte Brushes on Membranes
The initiator employed in this chapter is polydopamine modified by Îą-bromoisobutyryl
bromide (PDA-BIBB). The protocols of initiator immobilization and polymerization of
both SPMK and MEDSAH on membranes are the same as those described in Chapter 3.
4.2.3. Contact Angle Measurement
Contact angles of water or hexadecane on membrane surfaces were determined using a
phone camera and contact angle calculation software (VCA2000, AST Products). Contact
angles in air were determined by placing a drop of water or hexadecane on dry membranes.
To determine the surface hydrophilicity in water, a drop of hexadecane was first placed on
dry membranes, followed by immersing membranes in water.
4.2.4. Emulsion Preparation and Characterization
Hexadecane submicron emulsions stabilized with SDS or CTAB were prepared and
characterized as described in Chapter 2. These emulsions were used in dead-end filtrations.
As submicron oil droplets are invisible in the direct observation though the membrane
(DOTM) system, another emulsion was prepared with a protocol described in the
literature.26 Briefly, 0.1 mmol of surfactant (SDS or CTAB) was dissolved in 1 L of water,
followed by addition of 1 mL of hexadecane. Then the mixture was stirred for 20 min at
1000 rpm with a digital stand mixer (RW20 digital dual-range mixer, IKA).
4.2.5. Dead-End Filtration

105

The specific flux (L/(m2¡h¡bar)), referred to as LMH/bar) and observed oil rejection (R, %)
of the membranes were determined in the same apparatus using the same method as
described in Chapter 2 except for a few changes. Because of the differences in mechanical
strengths of different membranes, the applied pressure was 0.7 bar, which is half of the
pressure applied in Chapter 2, to avoid breaking or deforming membranes. To achieve more
precise control of the concentration of feed solutions, 0.68 mL of the 10% hexadecane
emulsion was added to the Amicon cell, followed by the addition of 9.32 mL of water to
keep the total volume of the feed solution at 10 mL.
Oil rejection was calculated based on the calibration curve in Figure 4.1. Several
blank solutions without hexadecane were analyzed to determine the minimum detectable
signal (average of the blank signal plus three times the standard deviation of this signal.)
Based on this minimum detectable signal and the calibration curve, the detection limit of
these analyses is 8.3 ×10-6 M hexadecane. Hexadecane concentration in the feed solution
was measured before every filtration and compare with the calculated value to ensure the
accuracy of calibration curve.

Area Ratio
(Hexadecane/Decane)

1.4

y = 23.418x + 4E-05

1.2

R² = 0.9961

1
0.8
0.6
0.4
0.2
0
0

0.01

0.02

0.03

0.04

0.05

0.06

Hexadecane Conc. (M)

Figure 4.1. GC calibration curve showing a linear relationship between the ratio of GC peak area
(hexadecane/decane) and the concentration of hexadecane at high hexadecane concentration.
106

All of the filtration experiments were repeated with 2 or 3 membranes. Preliminary
data only include filtration through modified nylon and track-etched membranes.
4.2.6. Visualization of Oil Coalescence on Membranes during Filtration
Oil coalescence was observed with the DOTM system shown in Figure 4.2.26 The
hydrophilic track-etched polycarbonate membrane with 0.2 Âľm diameter pores was used

Figure 4.2. Apparatus for DOTM system. Reprinted from reference 26 with permission from
Elsevier.

for DOTM because it is sufficiently transparent when wet. Using Araldite 2020 adhesive,
the membrane disk with a diameter of 25 mm was framed between two pieces of printing
paper. A 2.4 cm2 square cutout was created on both printing papers for optical observation.
Then the framed membrane was placed in the cross-flow filtration cell and secured on the
microscope stage with the permeate side facing upwards. Flow rates of the feed and
permeate were controlled using a gear pump (drive model 75211-15, Cole-Parmer) and a
peristaltic pump (Minipuls 3, Gilson), respectively. The pressures of the feed, retentate and
permeate streams were monitored by pressure transducers (Cole-Parmer) interfaced with a
107

computer. The permeate solution was collected in a beaker on a balance to record the mass
and calculate the permeate flux. During the whole filtration and observation, both pumps
were maintained at a constant setting to keep uniform flow rates.
4.3.

Results and Discussion

4.3.1. Wettability of Membranes
Membrane modification includes two steps: initiator (PDA-BIBB) immobilization and
PMEDSAH brush growth, and the reaction times of both steps may affect the wettability.
Figure 4.3 shows contact angles of hexadecane on nylon membranes modified with
PMEDSAH brushes. To examine the effect of PDA-BIBB deposition on wettability, I kept
the PMEDSAH polymerization time constant at 4 h. As Figure 4.3 shows, longer PDABIBB deposition times lead to higher hexadecane contact angles, indicating more
oleophobic surfaces. We suppose this is due to higher PDA-BIBB coverage on the
membrane, which results in thicker PMEDSAH brushes or higher brush coverage. Lee
reported that PDA coverage reaches a maximum after immersing surfaces in dopamine
solution for about 24 h.27 Thus, all modified membranes for emulsion filtration employed
a 48-h deposition of PDA-BIBB. Figure 4.4 shows that with 4-h PDA-BIBB deposition,
membrane surfaces with PMEDSAH brushes polymerized for over 2 h become
superoleophobic when immersed in water, leading to oil detachment from membrane, as
shown in Figure 4.4a, b and c. To ensure the surface superoleophobicity, we employed a
24 h-polymerization to obtain preliminary filtration results. Future filtrations could
examine the fouling resistance of PMEDSAH brush-modified membranes prepared with
shorter polymerization times. As a control, the bare nylon membrane shows an underwater

108

contact angle of hexadecane around 80° (Figure 4.4d), indicating relatively poor fouling
resistance to oil adsorption.
a

b

c

d

Figure 4.3. Contact angles of hexadecane (with red dye) in water on nylon membranes with a pore
size of 0.45 Âľm after coating with PDA-BIBB and PMEDSAH brushes. The PMEDSAH
polymerization time was 4 h, and the PDA-BIBB deposition times were (a) 1 h, (b) 2 h, (c) 3.5 h
and (d) 5.5 h.

a

b

c

d

Figure 4.4. Contact angles of hexadecane (with red dye) in water on nylon membranes with a pore
size of 0.45 Âľm after coating with PDA-BIBB and PMEDSAH brushes. The deposition time of
PDA-BIBB is 4 h. Polymerization time for PMEDSAH brushes is (a) 2 h, (b) 4 h and (c) 6 h. A
bare nylon membrane (d) served as a control. In Figures (a), (b), and (c), the droplet is not visible
because it detaches from the membrane.

109

4.3.2. Dead-End Filtrations of O/W Submicron Emulsions
4.3.2.1.Track-Etched Membrane
4.3.2.1.1. Bare Track-Etched Membrane
As Figure 4.5 shows, during the filtration of an SDS-stabilized submicron emulsion
through bare track-etched membranes, the permeate flux is around 60 LMH/Bar, which is
only 2% of the pure water flux before filtration. This implies that many membrane pores
are at least partially blocked by the oil droplets. On the other hand, the hexadecane
concentration in the permeate is always lower than the limit of detection, indicating nearly
100% oil rejection. However, the bare track-etched membrane has a pore size of 0.2 Âľm,
whereas only about half of the oil droplets (volume fraction) have a diameter greater than
0.2 Âľm (see Figure 4.6a). The very high oil rejection may result from excluded volume, a
decrease in the pore size due to oil fouling, adsorption of small droplets, or droplet
coalescence. Comparison of the oil droplet diameters in the feed (Figure 4.6a) and retentate
(Figure 4.6b) solutions shows some oil coalescence, and the median droplet diameter,
termed as D50,28 increases from 0.196 to 0.325 Âľm. Nevertheless, this small increase in
droplet size does not change the milky look of the emulsion (Figure 4.7).
101.0%
100.5%

400

100.0%

200
0

99.5%
99.0%
0

2

4

6

Time (h)

110

8

10

Rejection (%)

Flux (LMH/Bar)

600

Figure 4.5. (cont’d) Permeate flux and oil rejection during dead-end filtration of an SDS-stabilized
emulsion through a bare track-etched membrane.

a

b

Figure 4.6. Oil droplet size before (a) and after (b) dead-end filtration of an SDS-stabilized
emulsion through a bare track-etched membrane.

1h

6h

3h

10 h

Figure 4.7. Images of an SDS-stabilized O/W submicron emulsion at different times during
filtration through a bare track-etched membrane with 0.2 Îźm pores.

When we employed the bare track-etched membrane for the filtration of CTABstabilized emulsions, the oil rejection is still almost 100%. (Figure 4.8) However,
demulsification begins from the start of filtration, making the emulsion gradually look like

111

101.0%

100.5%
400
100.0%
200
99.5%

0

Rejection (%)

Flux (LMH/Bar)

600

99.0%
0

2

4

6

8

10

Time (h)
Figure 4.8. Permeate flux and oil rejection during dead-end filtration of a CTAB-stabilized
emulsion through a bare track-etched membrane.

milky and finally clear after 3 h of filtration (Figure 4.9). Meanwhile, the permeate flux
gradually increased during the first 3 h of filtration and then remained constant. The low

0h

1.5 h

3h

10 h

Figure 4.9. Images of a CTAB-stabilized O/W submicron emulsion at different times during
filtration through a track-etched membrane with 0.2 Îźm pores.

initial permeate flux may stem from membrane fouling. The combination of the flux
increase and the emulsion clarification suggest that oil droplets attached to the membrane
eventually coalesce with other incoming droplets. Larger droplets likely float away from

112

the membrane to the top of the solution. Stoke’s equation (Eq. 1), describes the droplet
vertical velocity, V,
V=

d2 g(ρw −ρo )
18Âľw

Eq. 1

where d is the diameter of oil droplet, ρw and ρo are the densities of water and oil,

respectively, and Îźw

is

the dynamic viscosity of water. (The variable g represents the

acceleration due to gravity.). Larger droplets will experience stronger buoyant forces and
detach from membrane and float to the surface with relatively high velocities. Alternatively,
Nitsch and co-workers reported that as oil droplets become large enough to block
membrane pores, water occlusions may create W/O droplets, which can also float to the
surface.17 The detachment of oil from the membrane leads to less pore blocking and an
increasing permeate flux, as Figure 4.8 shows. Because we keep adding water from the
feed tank into the filtration cell, there is no new oil entering the system. After oil coalesces
and floats to the solution surface, the permeate flux remains constant around 300 LMH/Bar.
This flux is much lower than the pure water flux before emulsion filtration (2500
LMH/Bar), indicating that oil on or in the membrane still partially blocks the pores.
4.3.2.1.2. Track-Etched Membranes Modified with PDA-BIBB
As Figures 4.10 and 4.11 imply, after coating a track-etched membrane with PDA-BIBB,
the ATRP initiator, the modified track-etched membrane induces coalescence of droplets

113

600

101.0%

400
100.0%
200

Rejection (%)

Flux (LMH/Bar)

100.5%

99.5%

0

99.0%
0

2

4

6

8

10

Time (h)

Figure 4.10. Permeate flux (blue) and oil rejection (red) during the dead-end filtration of an SDSstabilized emulsion through a track-etched membrane modified with PDA-BIBB (the PDA-BIBB
deposition time is 48 h).

stabilized by SDS, leading to a flux increase from 200 to 500 LMH/Bar and demulsification
within 1-h of filtration. This may stem from the increased roughness after the attachment
of PDA-BIBB. As Agarwal reported, higher surface roughness prompts oil coalescence by
providing strong adhesion with oil and increasing coalescence efficiency.20 However, this
coating seems to prevent the coalescence of oil droplets stabilized by CTAB, as the flux

0h

0.25 h

0.5 h

Figure 4.11. Images of an SDS-stabilized O/W submicron emulsion at different times during deadend filtration through a track-etched membrane modified with PDA-BIBB (the deposition time is
48 h).

114

does not greatly increase over 9 h or filtration (Figure 4.12). Additionally, there is only a
small difference in oil droplet sizes before (D50 = 0.176 Âľm) and after filtration (D50 =
0.205 Âľm) of the CTAB-stabilized emulsion, as Figure 4.13 shows.
101.0%

600

400
100.0%
200

Rejection (%)

Flux (LMH/Bar)

100.5%

99.5%

99.0%

0
0

2

4

6

8

10

Time (h)

Figure 4.12. Permeate flux (blue) and oil rejection (red) during the dead-end filtration of CTABstabilized emulsion using a track-etched membrane modified with PDA-BIBB (the deposition time
is 48 h).

a

115

Figure 4.13 (cont’d)

b

Figure 4.13. Oil droplet size distribution before (a) and after (b) dead-end filtration of a CTABstabilized emulsion through a track-etched membrane modified with PDA-BIBB (deposition time:
48 h).

4.3.2.1.3. Track-Etched Membrane Modified with PMEDSAH Brushes
The data in Figures 4.14 and 4.15 indicate that track-etched membranes modified with
500

101.0%

100.5%

300
100.0%
200

Rejection (%)

Flux (LMH/Bar)

400

99.5%

100
0

99.0%
0

2

4

6

8

10

Time (h)

Figure 4.14. Permeate flux (blue) and oil rejection (red) during the dead-end filtration of an SDSstabilized emulsion through a track-etched membrane modified with PMEDSAH brushes (the
polymerization time was 12 h).

superoleophobic PMEDSAH brushes do not induce rapid demulsification of submicron
emulsions stabilized by SDS. However, the permeate flux initially increases with time and
oil droplets also become marginally larger (D50 increase from 0.213 to 0.268 Âľm) after
116

10-h filtration. This is probably because the superoleophobicity of PMEDSAH brushes
resists oil adsorption, which is the first step for membrane-induced oil coalescence.17
a

b

Figure 4.15. Oil droplet size before (a) and after (b) dead-end filtration (10 h) of an SDS-stabilized
emulsion through a track-etched membrane modified with PMEDSAH brushes (polymerization
time: 12 h).

4.3.2.2.Nylon Membranes
4.3.2.2.1. Bare Nylon Membrane
Demulsification occurs for both SDS- (Figure 4.16) and CTAB-stabilized emulsions
(Figure 4.17) during filtration through bare nylon membranes. Since the pore size (0.45
Âľm) of bare nylon membranes is larger than around 80% of the oil droplets, much of the
oil should pass through unfouled membranes. However, the negligible oil concentration in
the permeate, which is lower than the detection limit, suggesting that oil adsorbs to the
membrane and decreases pore sizes.

Increases in flux at long times and solution

clarification indicate eventual oil droplet coalescence. The pure water flux through a bare
nylon membrane was ~13000 LMH/Bar.

117

101.0%

1500

100.5%

1000

100.0%

0h

500

99.5%

3h

Rejection (%)

Flux (LMH/Bar)

2000

99.0%

0
0

2

4

6

8

10

Time (h)

Figure 4.16. Permeate flux (blue) and oil rejection (red) during dead-end filtration of an SDSstabilized emulsion through a bare nylon membrane with a pore size of 0.45 Âľm. The photos show

2000

101.0%

1500

100.5%

1000

100.0%

500

99.5%

2h

Rejection (%)

Flux (LMH/Bar)

the O/W submicron emulsion at different times during the filtration.

99.0%

0
0

2

4

6

8

10

Time (h)

Figure 4.17. Permeate flux (blue) and oil rejection (red) during the dead-end filtration of a CTABstabilized emulsion through a bare nylon membrane with a pore size of 0.45 Âľm. The photo shows
the O/W submicron emulsion after 2 h of filtration.

4.3.2.2.2. Nylon Membranes Modified with PMEDSAH Brushes
For emulsions stabilized with SDS, oil coalescence occurs during the filtration with nylon
membranes modified with PMEDSAH Brushes and leads to demulsification after 5 h

118

(Figure 4.18). As discussed above, oil coalescence leads to less pore blocking and a
resulting increase in permeate flux. The bare nylon membrane shows a drastic flux increase
101.0%

5h

1500

100.5%

1000

100.0%

500

99.5%

0

99.0%
0

2

4

6

8

Rejection (%)

Flux (LMH/Bar)

2000

10

Time (h)

Figure 4.18. Permeate flux (blue) and oil rejection (red) during dead-end filtration of an SDSstabilized emulsion through a nylon membrane modified with PMEDSAH brushes (24 h
polymerization time). The photo shows the O/W submicron emulsion after 2 h-filtration. The pure
water flux through the modified membrane was 2600 LMH/Bar.

after 3 h of filtration, but after modification with PMEDSAH brushes, the permeate flux
starts to increase dramatically after 5 h of filtration. The superoleophobic coating likely
resists some oil attachment in the beginning and delays the demulsification process. As a
comparison, the filtration of an SDS-stabilized emulsion using PSPMK-modified
membranes (Figure 4.19) leads to a constant permeate flux. As discussed in Chapter 2, this
is due to the electrostatic repulsion between negatively charged oil and polyanionic brushes.
Thus, we suppose that although the zwitterionic PMEDSAH brushes are superoleophobic,
without the strong electrostatic repulsion between oil droplets and the surface, these
brushes cannot resist all oil adsorption in dead-end filtration.

119

101.0%

30
25

100.0%

15
10

Rejection (%)

Flux (LMH/Bar)

100.5%
20

99.5%
5
99.0%

0
0

2

4

6

8

10

Time (h)

Figure 4.19. Permeate flux (blue) and oil rejection (red) during dead-end filtration of an SDSstabilized emulsion through a nylon membrane modified with PSPMK brushes (the polymerization
time was 24 h).

For CTAB-stabilized emulsions, we did not observe obvious oil coalescence and
demulsification during filtration through the PMEDSAH-modified membranes. Figure
4.20 shows the permeate flux and oil rejection during filtration. The small flux increase
may stem from either brush collapse after fouling or the coalescence of a small part of oil.
The latter may be more convincing since brush collapse may induce a much higher flux
increase. According to the Hagen-Poiseuille equation, which I described in Chapter 2, a
small change in pore radius will make a significant change in permeate flux. However, this
needs to be confirmed by comparing droplet sizes before and after filtration.

120

101.0%

150

100.5%

100

100.0%

50

99.5%

Rejection (%)

Flux (LMH/Bar)

200

99.0%

0
0

2

4

6

8

10

Time (h)

Figure 4.20. Permeate flux (blue) and oil rejection (red) during dead-end filtration of a CTABstabilized emulsion through a nylon membrane modification with PMEDSAH brushes (24-h
polymerization time).

4.3.3. DOTM
As Figure 4.2 shows, the DOTM system employs cross-flow filtration, and the flow
tangential to the membrane may result in less chances for oil contact and coalescence than
in dead-end filtration. Because the DOTM pressure and flux data are still not available, I
only show qualitatively whether droplet coalescence was visible and compare these results
with the dead-end filtration data (Table 4.1). For some types of membranes, the
disagreement in oil coalescence between cross-flow and dead-end filtration may result
from the different emulsions used in the two system or types of filtration (dead-end and
cross-flow). The particle size in the emulsions employed in cross-flow was generally over
10 Âľm to permit visualization.
Table 4.1. Results of oil coalescence in dead-end filtration and crossflow (DOTM) through bare
and modified track-etched membranes. “Y” represents clear oil coalescence and demulsification,

121

(cont’d) whereas “N” indicates no obvious oil coalescence, even if a small increase in oil droplet
size appeared.

SDS-emulsion

CTAB-emulsion

dead-end

dead-end

DOTM

filtration

DOTM

filtration

bare track-etched

N

N

Y

Y

PDA-BIBB-modified

Y

N

N

Y

N

N

N

N

track-etched
PMEDSAH-modified
track-etched

4.4.

Conclusions

Despite its wide use as an anti-fouling material, the zwitterionic PSBMA brush does not
show strong resistance to fouling with submicron oil droplets in dead-end filtrations. For
emulsions stabilized with either SDS or CTAB, oil droplets adsorb on this brush surface
and block pores, leading to a low permeate flux. However, this oil adsorption may induce
oil coalescence, which can break the emulsion and increase the permeate flux. Previous
studies about membrane coalescence mainly focused on hydrophobic membranes since oil
can attach and coalesce more on these surfaces. Our work shows that a less anti-fouling
coating such as the PSBMA brush can also induce oil coalescence, but it does not foul and
saturate with oil as easily as hydrophobic membranes. However, more experiments are
needed to better understand oil coalescence on PSBMA-modified membranes and design
a membrane that can better induce oil coalescence and resist fouling.

122

REFERENCES

123

REFERENCES

1.
Cheryan, M.; Rajagopalan, N. Membrane Processing of Oily Streams. Wastewater
Treatment and Waste Reduction. J. Membr. Sci. 1998, 151, 13-28.
2.
Eriksson, P. Nanofiltration Extends the Range of Membrane Filtration. Environ.
Prog. Sustain. Energy 1988, 7, 58-62.
3.
Gupta, A. K. S. Process for Refining Crude Glyceride Oils by Membrane Filtration.
U.S. Patent 4,062,882, December 13, 1977.
4.
Kajitvichyanukul, P.; Hung, Y.-T.; Wang, L. K. Membrane Technologies for Oil–
Water Separation. In Membrane and Desalination Technologies; Wang, L. K.; Chen, J. P.;
Hung, Y.-T.; Shammas, N. K., Eds.; Springer, 2011; pp 639-668.
5.
Gao, W.; Liang, H.; Ma, J.; Han, M.; Chen, Z.-l.; Han, Z.-s.; Li, G.-b. Membrane
Fouling Control in Ultrafiltration Technology for Drinking Water Production: A Review.
Desalination 2011, 272, 1-8.
6.
Guo, W.; Ngo, H.-H.; Li, J. A Mini-Review on Membrane Fouling. Bioresour.
Technol. 2012, 122, 27-34.
7.
Meng, F.; Chae, S.-R.; Drews, A.; Kraume, M.; Shin, H.-S.; Yang, F. Recent
Advances in Membrane Bioreactors (MBRs): Membrane Fouling and Membrane Material.
Water Research 2009, 43, 1489-1512.
8.
Yan, L.; Li, Y. S.; Xiang, C. B. Preparation of Poly (Vinylidene Fluoride) (PVDF)
Ultrafiltration Membrane Modified by Nano-Sized Alumina (Al2O3) and Its Antifouling
Research. Polymer 2005, 46, 7701-7706.
9.
Chang, Y.; Ko, C.-Y.; Shih, Y.-J.; QuĂŠmener, D.; Deratani, A.; Wei, T.-C.; Wang,
D.-M.; Lai, J.-Y. Surface Grafting Control of Pegylated Poly (Vinylidene Fluoride)
Antifouling Membrane Via Surface-Initiated Radical Graft Copolymerization. J. Membr.
Sci. 2009, 345, 160-169.
10.
Rahimpour, A.; Madaeni, S.; Zereshki, S.; Mansourpanah, Y. Preparation and
Characterization of Modified Nano-Porous PVDF Membrane with High Antifouling
Property Using UV Photo-Grafting. Appl. Surf. Sci. 2009, 255, 7455-7461.
11.
Rana, D.; Matsuura, T. Surface Modifications for Antifouling Membranes. Chem.
Rev. 2010, 110, 2448-2471.
12.
Lewandowski, A. F., Collant Filter and Oil Coalescer. U.S. Patents 5,454,937,
October 3, 1995.

124

13.
Li, J.; Huang, S.; Wang, X. Experimental Research of Separation Efficiency on
Steam-Water Separator with Corrugated Plates. Nuclear Power Engineering 2007, 28, 9497.
14.
Cairo Jr, J. A.; Young, J. A., Induced Gas Liquid Coalescer and Flotation Separator.
U.S. Patent 5,080,802, January 14, 1992.
15.
Moosai, R.; Dawe, R. A. Gas Attachment of Oil Droplets for Gas Flotation for Oily
Wastewater Cleanup. Sep.Purif. Technol. 2003, 33, 303-314.
16.
Leech III, C. A.; Radhakrishnan, S.; Hillyer, M. J.; Degner, V. Performance
Evaluation of Induced Gas Flotation Machine through Mathematical Modeling. J. Petrol.
Technol. 1980, 32, 48-58.
17.
Hoffmann, S.; Nitsch, W. Membrane Coalescence for Phase Separation of Oil‐in‐
Water Emulsions Stabilized by Surfactants and Dispersed into Smallest Droplets. Chem.
Eng. Technol. 2001, 24, 22-27.
18.
Tan, K. Y.; Hughes, T. L.; Nagl, M.; Huck, W. T. Nonfouling Capture–Release
Substrates Based on Polymer Brushes for Separation of Water-Dispersed Oil Droplets.
ACS. App. Mater. Interfaces 2012, 4, 6403-6409.
19.
Hu, D.; Li, X.; Li, L.; Yang, C. Designing High-Caliber Nonwoven Filter Mats for
Coalescence Filtration of Oil/Water Emulsions. Sep. Purif. Technol. 2015, 149, 65-73.
20.
Agarwal, S.; von Arnim, V.; Stegmaier, T.; Planck, H.; Agarwal, A. Role of Surface
Wettability and Roughness in Emulsion Separation. Sep. Purif. Technol. 2013, 107, 19-25.
21.
Agarwal, S.; von Arnim, V.; Stegmaier, T.; Planck, H.; Agarwal, A. Effect of
Fibrous Coalescer Geometry and Operating Conditions on Emulsion Separation. Ind. Eng.
Chem. Res. 2013, 52, 13164-13170.
22.
Hong, A.; Fane, A.; Burford, R. Factors Affecting Membrane Coalescence of Stable
Oil-in-Water Emulsions. J. Membr. Sci. 2003, 222, 19-39.
23.
Kobayashi, M.; Terayama, Y.; Yamaguchi, H.; Terada, M.; Murakami, D.; Ishihara,
K.; Takahara, A. Wettability and Antifouling Behavior on the Surfaces of Superhydrophilic
Polymer Brushes. Langmuir 2012, 28, 7212-7222.
24.
Kobayashi, M.; Ishihara, K.; Takahara, A. Neutron Reflectivity Study of the
Swollen Structure of Polyzwitterion and Polyelectrolyte Brushes in Aqueous Solution. J.
Biomat. Sci.-Polym. E 2014, 25, 1673-1686.
25.
Shah, R. R.; Merreceyes, D.; Husemann, M.; Rees, I.; Abbott, N. L.; Hawker, C. J.;
Hedrick, J. L. Using Atom Transfer Radical Polymerization to Amplify Monolayers of
Initiators Patterned by Microcontact Printing into Polymer Brushes for Pattern Transfer.
Macromolecules 2000, 33, 597-605.

125

26.
Tummons, E. N.; Tarabara, V. V.; Chew, J. W.; Fane, A. G. Behavior of Oil
Droplets at the Membrane Surface During Crossflow Microfiltration of Oil–Water
Emulsions. J. Membr. Sci. 2016, 500, 211-224.
27.
Lee, H.; Dellatore, S. M.; Miller, W. M.; Messersmith, P. B. Mussel-Inspired
Surface Chemistry for Multifunctional Coatings. Science 2007, 318, 426-430.
28.
HMKTest. What Is Particle Size Distribution D50, D50 Particle Size Distribution.
http://www.aimsizer.com/faqs-What-is-D50.html (accessed Aug 13, 2017).

126

Chapter 5. Summary and Future Work
5.1.

Research Summary

This dissertation summarizes my Ph.D. research, which includes an in-depth study of the
swelling of polyelectrolyte brushes in both water and surfactant solutions, fabrication of
polyanionic brush-modified membranes with nearly 100% oil rejection and strong fouling
resistance, and development of a superoleophobic membrane that can potentially resist
fouling and prompt oil coalescence simultaneously.
Chapter 2 describes the superoleophobicity of polyanionic brushes and their
application as an anti-fouling layer in the filtration of O/W submicron emulsions. The
poly(3-sulfopropyl methacrylate potassium salt) (PSPMK) brush swells by up to 280% and
is superoleophobic in water. The density of PSPMK brushes created through surfaceinitiated atom transfer radical polymerization increases with polymerization time. Due to
electrostatic repulsion, the anionic PSPMK brushes completely resist the adsorption of
anionic surfactants. Moreover, microfiltration membranes modified with PSPMK brushes
show almost 100% oil rejection and minimal fouling during dead-end filtration of
emulsions stabilized with anionic surfactants. Nevertheless, strong electrostatic attraction
induces adsorption of cationic surfactants as well as positively charged oil droplets on the
brush layer, leading to brush collapse and membrane fouling. Ellipsometric data further
show that less dense brushes at low cationic surfactant concentrations are more likely to
collapse.
Chapter 3 focuses on the swelling of zwitterionic poly(solfobetaine methacrylate)
(PSBMA) brushes in ionic surfactant solutions. Similar to the effect of inorganic salts, both
anionic and cationic surfactants adsorb to PSBMA brushes and change swelling

127

significantly. The extent of PSBMA swelling or collapse depends on the surfactant charge
and concentration, as well as the brush thickness. These changes of brush swelling may
stem from a combination of charge-screening, disordering of water molecules around
zwitterions and the surfactant hydrophobicity. Filtration of surfactant solutions with
PMEDSAH brush-modified membranes confirms high swelling of thin brushes with
anionic surfactants and less swelling with cationic surfactants.
Chapter 4 presents my attempts to modify surfaces with zwitterionic PSBMA
brushes to achieve fouling resistance and oil coalescence simultaneously. Both dead-end
and cross-flow filtrations of submicron emulsions with different membranes show the
effect of the membrane substrate on oil coalescence. As the preliminary data demonstrate,
oil coalescence occurs on PSBMA brush-modified nylon membranes, whereas tracketched membranes with the same coating do not induce oil coalescence. This may result
from the higher roughness on the nylon membrane surface. It also implies zwitterionic
PSBMA brushes cannot completely resist oil adsorption due to the lack of strong
electrostatic repulsion. Nevertheless, this weaker fouling resistance allows oil droplets to
attach to the membrane, coalesce with other droplets and finally detach from the membrane.
5.2.

Future Work: Membrane Modification with Zwitterionic Brushes for Oil

Coalescence and Fouling Resistance
As described in Chapter 4, membrane-induced oil coalescence depends on various factors
such as surface roughness, pore size and filtration pressure.1-3 To fabricate a surface coating
for both oil coalescence and fouling resistance, all of these factors may need to be
considered, which requires a deeper understanding of oil coalescence, especially on

128

membrane surfaces. In addition, more experiments are needed to optimize the surface
coating and evaluate the resistance to fouling and oil coalescence.
5.2.1. Membrane Characterization
The preliminary data mainly focus on membrane filtration, but an explanation for these
filtration results requires more information about the membrane. Scanning electron
microscope (SEM) and atomic force microscope (AFM) can provide morphology
information, such as the roughness and the pore size and porosity changes after membrane
modification. Determination of zeta potentials of membranes is also important for this
research, since it can identify the sign and estimate the magnitude of the charge that the
zwitterionic PSBMA brushes carry. Such information will help us to understand the
interaction between oil droplets and membrane surfaces. I hypothesize that a nearly neutral
membrane surface may be a good choice for providing some fouling resistance while
enhancing coalescence. Firstly, less charge ensures that the surface will not completely
repel oil droplets with the same charge. This is vital since weak oil attachment on
membrane surface is the first step of coalescence.4 In contrast, when the oil and the surface
are oppositely charged, a nearly neutral surface will avoid extensive oil adsorption or even
oil saturation.5
5.2.2. Dead-End and Cross-Flow Filtrations of the Same Emulsion.
Due to the invisibility of submicron oil droplets, we employed another emulsion with much
larger oil droplets (~ 10 Âľm) for DOTM experiments, as Chapter 4 describes. To compare
the effects of dead-end and cross-flow filtration on oil coalescence on membranes,
emulsions employed in the two filtrations should be the same. Thus, dead-end filtrations
of emulsions with larger oil droplets are necessary. From another point of view, this

129

experiment is also helpful to observe the coalescence of oil droplets in different sizes
during dead-end filtrations.
5.2.3. Compare Zwitterionic Brushes with Neutral Hydrophilic Brushes
As discussed earlier, the nearly zero net charge on zwitterionic brush surfaces may allow
both anti-fouling and oil coalescing purposes. However, neutral hydrophilic brushes such
as poly(poly(ethylene glycol) methacrylate) (PPEGMA) and poly(2-hydroxyethyl
methacrylate) (PHEMA) also have that potential.6 Unlike zwitterionic brushes, these
neutral brushes don’t have the strong intra-group, inter-or intra-chain interactions,7, 8 and
this may lead to a less hydrated brush layer. Thus, these brushes may foul more easily, or
even saturate with oil like traditional hydrophobic membranes.5 This hypothesis needs to
be verified by experiments.

130

REFERENCES

131

REFERENCES

1.
Hong, A.; Fane, A.; Burford, R. Factors Affecting Membrane Coalescence of Stable
Oil-in-Water Emulsions. J. Membr. Sci. 2003, 222, 19-39.
2.
Agarwal, S.; von Arnim, V.; Stegmaier, T.; Planck, H.; Agarwal, A. Role of Surface
Wettability and Roughness in Emulsion Separation. Sep. Purif. Technol. 2013, 107, 19-25.
3.
Agarwal, S.; von Arnim, V.; Stegmaier, T.; Planck, H.; Agarwal, A. Effect of
Fibrous Coalescer Geometry and Operating Conditions on Emulsion Separation. Ind. Eng.
Chem. Res. 2013, 52, 13164-13170.
4.
Hoffmann, S.; Nitsch, W. Membrane Coalescence for Phase Separation of Oil‐in‐
Water Emulsions Stabilized by Surfactants and Dispersed into Smallest Droplets. Chem.
Eng. Technol. 2001, 24, 22-27.
5.
Hu, D.; Li, L.; Li, Y.; Yang, C. Restructuring the Surface of Polyurethane Resin
Enforced Filter Media to Separate Surfactant Stabilized Oil-in-Water Emulsions Via
Coalescence. Sep. Purif. Technol. 2017, 172, 59-67.
6.
Barbey, R.; Lavanant, L.; Paripovic, D.; SchĂźwer, N.; Sugnaux, C.; Tugulu, S.;
Klok, H.-A. Polymer Brushes via Surface-Initiated Controlled Radical Polymerization:
Synthesis, Characterization, Properties, and Applications. Chem. Rev. 2009, 109, 54375527.
7.
Kobayashi, M.; Ishihara, K.; Takahara, A. Neutron Reflectivity Study of the
Swollen Structure of Polyzwitterion and Polyelectrolyte Brushes in Aqueous Solution. J.
Biomater. Sci. Polym. Ed. 2014, 25, 1673-1686.
8.
Kudaibergenov, S.; Jaeger, W.; Laschewsky, A., Polymeric Betaines: Synthesis,
Characterization, and Application. In Supramolecular Polymers Polymeric Betains
Oligomers, Springer: 2006; pp 157-224.

132