 

PLACE IN RETURN Box to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K:IProj/Aco&PrelelRC/DateDuo.indd

 

LAYER-BY-LAYER MODIFICATION OF MEMBRANES FOR
ION SEPARATIONS AND CATALYSIS

By

Lu Ouyang

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Chemistry

2009

ABSTRACT

LAYER-BY-LAYER MODIFICATION OF MEMBRANES FOR
ION SEPARATIONS AND CATALYSIS

By

Lu Ouyang

Layer-by-layer (LBL) adsorption of polycations and polyanions enables the
formation of functional thin ﬁlms containing charged materials ranging from polymer
electrolytes to metal nanoparticles and viruses. This adsorption method is applicable to a
variety of substrates including ﬂat surfaces, colloids, and membrane pores. This
dissertation examines the use of LBL adsorption to modify membranes for speciﬁc
applications in ion separations and catalysis.

To simultaneously achieve high permeability and selectivity in ion separations,
composite membranes contain a thin, selective layer on a thicker porous support that
provides mechanical stability. Polyelectrolyte multilayer ﬁlms adsorbed on porous
ultraﬁltration membranes present an ultrathin, selective skin that can be tailored for
selective removal of multivalent ions in the presence of monovalent ions. Deposition
conditions such as the type of terminating layer, the number of bilayers deposited, and the
pH and ionic strength of the polyelectrolyte deposition solutions allow for optimization of
ﬁlm properties for selected ion separations. Speciﬁcally, five bilayers of
poly(styrenesulfonate)/poly(allylamine hydrochloride) (PSS/PAH) on porous alumina
supports allow a solution ﬂux of 0.85 m3/(m2 day) at 4.8 bar and exhibit 95% rejection of
Mng along with a Na+/Mg2+ selectivity of 22. Films with 4.5 bilayers of

PSS/poly(diallyldimethylammonium chloride) (PDADMAC) deposited on alumina

supports show a 98% rejection of phosphate, a chloride/phosphate selectivity of 48, and a
solution ﬂux of 2.4 m3/(m2 day). Both membranes exhibit higher ﬂuxes and selectivities
than commercially available nanoﬁltration membranes.

Metal nanoparticles are attractive catalysts due to their high surface area to
volume ratio and unique electronic properties. However, their high surface energy often
leads to aggregation, which greatly reduces catalytic activity. Alternating adsorption of
polyelectrolytes and metal nanoparticles is a simple and effective method to load ﬂat
sheet and hollow ﬁber membranes with metal nanoparticles without nanoparticle
aggregation. Catalytic, immobilized nanoparticles such as Au and Pd allow the use of
such membranes as catalytic reactors. Hollow ﬁber membrane reactors coated with ﬁhns
of Au nanoparticle/polyelectrolyte show high catalytic activity in the reduction of 4-
nitrophenol with NaBI-I4, and >99% initial conversion of 4—nitrophenol by HCOONa also
takes place in ﬂat sheet membranes containing Pd nanoparticles. However, a slight
conversion decline over time takes place in both cases. This conversion decline probably
stems from catalyst fouling by byproducts of 4-aminophenol oxidation.

Overall, LBL adsorption provides a simple, versatile method for modifying a
variety of porous substrates to create functional membranes. Deposition of
polyelectrolyte ﬁlms on the surface of porous support yields composite membranes for
ion separations, whereas adsorption of metal nanoparticle/polyelectrolyte ﬁlms in porous
supports gives catalytic membrane reactors. Further work is needed, however, to
decrease the time and processing required in the LBL method and to increase the stability

of membrane reactors prepared through LBL deposition.

To my family, for all their love and support

iv

ACKNOWLEDGEMENTS

First and foremost, I would like to thank my advisor, Dr. Merlin Bruening, for
his support and guidance during my graduate study. I learned a lot from him, not
only about how to be a good scientist, but also about how to be a nice person. I
appreciate his patience and encouragement, without which I will not able to ﬁnish this
degree or dissertation.

I would also like to thank past and present members in the Bruening group. I
am especially grateful to Dr. Malaisamy Ramamoorthy and Dr. David M. Dotzauer
for their generous help and direct involvement in my research projects. I would also
like to express my gratitude to Dr. Lei Sun, Dr. Srividhya Kidambi, F ei Xu, Dr. J inhua
Dai and Dr. Matthew D. Miller for their help and suggestions in my graduate study.

I want to thank Dr. Jean-Francois Lahitte for being my co-advisor and his
generous help when I was working in the Chemical Engineering Laboratory at
University Paul Sabatier in Toulouse, France. I would also like to thank Dr. Jorge
Macanas, who is a great scientist and remarkable colleague to work with.

Lastly, I want to give my special gratitude to my family for all their support

and especially my husband, Tao, for his love and encouragement.

TABLE OF CONTENTS

Page
List of Tables ........................................................................................ ix
List of Figures ....................................................................................... x
List of Schemes ...................................................................................... xiv
List of Abbreviations ................................................................................ xv
Chapter 1. Introduction and Background ................................................... 1
1.1. Functional Thin Films ......................................................................... 1
1.1.1. Applications of Functional Thin Films ....................................................... 1
1.2. Methods for Preparing Ultrathin Functional Films ........................................ 2
1.2.1.Langmuir-Blodgett (LB) Technique .......................................................... 2
1.2.1.1. History and Background of the LB Technique ......................................... 2
1.2.1.2. Advantages of the LB Method ........................................................... 5
1.2.2. Self-Assembled Monolayers (SAMs) ....................................................... 6
1.2.2.1. History of SAMs ........................................................................... 6
1.2.2.2. SAMs Formed via Covalent Bonding to a Surface .................................... 6
1.2.2.3. Advantages of SAMs ..................................................................... 10
1.2.3. Layer-by-Layer (LBL) Assembly .......................................................... 11
1.2.3.1. Driving Forces Involved in LBL Assembly ........................................... 12
1.2.3.1.]. LBL Assembly Based on Electrostatic Interactions..................................12
1.2.3.1.2. LBL Assembly via Hydrogen Bonding .............................................. 13
1.2.3.1.3. LBL Assembly Based on Covalent Bonding ........................................ 16
1.2.3.1.4. Other Interaction Involved in LBL Assembly ...................................... 19
1.2.3.2. LBL Assembly on Substrates of Various Geometries ......................................... 20
1.2.3.3. Optimization of the LBL Adsorption Process ........................................ 24
1.2.4. Comparison of LB Films, SAMs and LBL Films ........................................ 30
1.3. Selected Applications of LBL Films ....................................................... 31
1.3.1. Separation Membranes ....................................................................... 31
1.3.1.1. Nanoﬁltration Membranes ............................................................... 31
1.3.1.2. Membranes for Gas Separation ........................................................... 33

1.3.2. Membranes for Protein Adsorpt10n35
1.3.3. Bio- and Chemical Reactors37
1.4. Scope of This Work ........................................................................... 40
1.5. References ..................................................................................... 43

Chapter 2. Polyelectrolyte Multilayer Films as Nanofiltration Membranes for
Cation Separation ........................ 52

 

vi

2.1. Introduction ..................................................................................... 52

2.2. Experimental Section .......................................................................... 54
2.2.1. Materials ......................................................................................... 54
2.2.2. Film Deposition .............................................................................. 55
2.2.3. Characterization of Membranes ............................................................... 56
2.2.3.1. Film Thickness Estimations with SEM .............. 56
2.2.3.2. Zeta Potential measurements ............................................................ 56
2.2.4.Nanoﬁltration Stud1esS7
2.3. Results and Discussion ....................................................................... 59
2.3.1.Estimation of the Thicknesses of PEMs ................................................................. 59
2.3.2. Nanoﬁltration of Mixed-Salt Solutions with PSS/PAH on PBS Supports ........... 61
2.3.3. Nanoﬁltration of Mixed-Salt Solutions with Commercial NF Membranes and
PSS/PAH on Porous Alumina Supports ................................................. 66
2.3.4. Nanoﬁltration of Single-Salt Solutions at Two Salt Concentrations ................. 68
2.3.5. Nanoﬁltration of Mixed-Salt Solutions with PSS/PDADMAC or Hybrid
Membranes .................................................................................. 70
2.3.6. Calcium Rejection by PSS/PAH Films on Porous Alumina... ............................... 73
2.3.7. Streaming Potential Measurements .......................................................... 74

2.4. Conclu510ns79
2.5. References80

Chapter 3. Multilayer Polyelectrolyte Nanofiltration Membranes for Selective

 

Phosphate Removal ..................... 84
3.1. Introduction .................................................................................... 84
3 ..2 Experimental Section ............................................... g ........................... 87
3...21 Materials. .. 8.7
3. 2. 2. Film Formation on Porous Alumina. ..................................................................... 88
3. 2. 3. Transport Studies ............................................................................. 88
3.3. Results and Discussion ....................................................................... 89
3.3.1.Membrane Formation .......................................................................... 90
3.3.2. NF Experiments with PSS/PDADMAC Membranes ................................... 90
3.3.3. NF Experiments with Commercial NF Membranes ..................................... 94

3.4. Conclu81ons97
3.5. References98

Chapter 4. Catalytic Hollow-Fiber Membranes Prepared Using Layer-by-Layer

Adsorption of Polyelectrolytes and Metal Nanoparticles ............... 101
4..l Introduction... ...101
4. 2. Experimental section ........................................................................ 104
4. 2.1 Materials” . 1.04
4. 2. 2. Preparation of Citrate- Stabilized Au Nanopartlcles 106
4. 2. 3. Modiﬁcation of the Hollow Fiber membranes ......................................... 107
4.2.4. Characterization of the Hollow Fiber Membranes ..................................... 109

vii

4.2.5. Catalytic Reactions and Leaching Studies .............................................. 109

4.3. Results and Discussion ...................................................................... 110
4.3.1. Membrane Characterization .............................................................. 110
4.3.1.1. SEM Characterization of Immobilized Nanoparticles .............................. 110
4.3.1 ..2 Catalyst Loading... ...112
4.3.1.3. Water Permeability ....................................................................... 114
4.3.2. Catalysis with Nanoparticle-Coated Fibers ............................................. 116

4.3.2.1. Catalytic Reduction Using PES Fibers Coated With PSS/PAH/Au NP Films...117
4.3.2.2. Catalytic Reduction Using PS Fibers Coated With PSS/PAH/Au NP F ihns. . ..122

 

4.4. Conclusions. . . . . . . .. ................................................................................................ 125
4.5. References......... .................................................................................................. 127
Chapter 5. Reduction of 4-Nitrophenol with Sodium Formate Using Catalytic

Flat-Sheet Membranes Containing Pd Nanoparticles 131
5.1. Introduction ................................................................................... 131
5. 2. Experimental Section ........................................................................ 132
5.2.1. Materials. .. 132
5. 2. 2. Preparation of C1trate-Stab1llzed Pd Nanopart1cles .......................................... 134
5. 2. 3. Modiﬁcation of Flat-Sheet Membranes ................................................. 135
5.2.4. Characterization of the Flat-Sheet Membranes .......................................... 136
5.2.5. Catalytic Reactions and Leaching Studies ............................................... 137
5.3. Results and Discussion ...................................................................... 137
5.3.1. SEM Characterization of Immobilized Nanoparticles. .. ....................................... 138
5.3.2. Catalysis with Pd Nanoparticle-Loaded Alumina Membranes ....................... 139
5.3.2.1. Effect of Feed pH on 4-Nitrophenol Conversion .................................... 140
5.3.2.2. Dependence of 4-Nitrophenol Conversion on HCOONa Concentration ......... 141
5.3.2.3. SEM Characterization of Deposit Formation on Alumina Membranes ........... 142
5.3.3. Catalysis with Pd Nanoparticle-Loaded PES Membranes ............................ 143
5.3.3.1. 4-Nitrophenol Reduction in PBS Membranes Coated with PSS/PEI/PdNP

Films ...................................................................................... 144

5.3.3.2. Catalyst leaching ......................................................................... 144
5.3.3.2.1. Use of a Capping Layer to Minimize Catalyst Leaching ......................... 145
5.3.3.2.2. Effect of Solution Flux and Feed Concentration on Catalyst Leaching ........ 147
5.4 Conclusions... .... .152

5.5. Reference5154

Chapter 6. Summary and Future Work 156

viii

Table 2.1.

Table 2.2.

Table 2.3.

Table 2.4.

Table 2.5.

Table 2.6.

Table 3.1.

Table 3.2.

Table 3.3.

Table 4.1.

LIST OF TABLES

Aqueous diffusion coefﬁcients and Stokes’ radii of several cations .......... 59

Solution ﬂuxes and ion rejections from NF with PSS/PAH ﬁlms deposited
on50kDaPES supports. ................................................................. 63

Solution ﬂuxes and ion rejections from NF with PSS/PAH ﬁlms deposited
on porous alumina (0.02 yum-diameter surface pores) supports .. ............... 67

Cation rejections and ﬂux values for single-salt NF with PSS/PAH ﬁlms
deposited on different substrates .................................................. 69

Solution ﬂuxes and ion rejections from NF with PSS/PDADMAC and hybrid
(PSS/PDADMAC)4+(PSS/PAH) ﬁlms deposited on porous alumina (0.02
um-diameter surface pores) supports .............................................. 71

Solution ﬂuxes and ion rejections from NF with Membrane K (ﬁve bilayers
of PSS/PAH deposited on porous alumina supports) ........................... 74

Molecular weights (MW), Stokes’ radii (rs) and aqueous diffusion coefﬁcients
(D) of several anions ................................................................ 86

Solution ﬂuxes, rejections and selectivities from NF with
(PSS/PDADMAC)XPSS-coated alumina membranes and feed solutions
containing NaHzPO4 (1 mM) and NaCl (1 mM). 92

Solution ﬂuxes, rejections, and selectivities from NF with commercially
available membranes and feed solutions containing NaHzPO4 (1 mM) and
NaCl (1 mM)... ............................................................................................. 95

Au loading along PES ﬁbers coated with PSS/PAH/AuNP ﬁlms ............ 113

ix

LIST OF FIGURES

Figure 1.1. Formation of a monolayer of amphiphilic molecules at the air-water

interface.......... ............................................................................................. 3
Figure 1.2. Formation of Langmuir-Blodgett ﬁlms ......................................................... 4
Figure 1.3. Three possible ﬁlm architectures in Langmuir-Blodgett ﬁlms. . . . . . . . . . ..........5
Figure 1.4. Structure of a SAM ..................................................................... 11

Figure 1.5. LBL assembly of polyelectrolyte-colloid bilayers through electrostatic
interactions ........................................................................................ 13

Figure 1.6. Structure of polymers used in LBL assembly based on hydrogen bonding... 14

Figure 1.7. Modiﬁcation of a pore of an alumina membrane by LBL adsorption of two
oppositely charged polyelectrolytes and Au nanoparticles..... . . . . . . ..............22

Figure 1.8. LBL assembly on particle templates for hollow-sphere formation... ... .......23

Figure 1.9. Preparation of polymer and inorganic-organic composite nanotubes by an
LBL-colloid templating approach .................................................................. 24

Figure 1.10. Buildup of LBL ﬁlms by consecutive spinning of polycation and polyanion
molecules. . . . . . ............................................................................................... 26

Figure 1.11. Schematic diagram of the internal structure of (PAI-I/CdS)n ﬁlms prepared
by dipping and spin methods ......................................................................... 27

Figure 1.12. LBL deposition by alternating spray-coating of polycations and polyanions
on a negatively charged substrate ...................................................... 29

Figure 1.13. Schematic diagram of modiﬁcation of the interior of membranes with PAA-
terrninated PAA/PAH ﬁlms and activation of these ﬁlms for covalent protein
attachment ...................................................................................................... 36

Figure 1.14. Multi-enzyme reactor prepared by LBL deposition on an ultraﬁltration
membrane .............................................................................. 38

Figure 1.15. Schematic diagram of the reduction of 4-nitrophenol by NaBH4 during ﬂow
through a membrane modiﬁed with nanoparticle/polyelectrolyte ﬁlms. . . ....39

Figure 2.1. Structures of the polyelectrolytes used in this study ..................................... 56

Figure 2.2. Cross-sectional SEM images of a porous alumina membrane coated with a
[PSS/PAI-I]n ﬁlm ....................................................................... 61

Figure 2.3. i; potentials of [PSS/PAH]n and [PSS/PDADMAC]n ﬁlms on PBS
substrates. . . . . ........................................................................................... 75

Figure 2.4. C potentials of [PSS/PDADMAC]4+[PSS/PAH] ﬁlms as a function of the
concentration of KCl employed in the streaming potential measurements...77

Figure 2.5. C potentials of [PSS/PDADMAC]4+[PSS/PAH] ﬁlms as a function of the
concentration of MgClz employed in the streaming potential

measurements ........................................................................ 78
Figure 3.1. Fractional speciation ofphosphate as a function opr87
Figure 4.1. Schematic diagram of the cross section of a porous hollow ﬁber membrane

modiﬁed with a polyelectrolyte/AuNP ﬁlm104
Figure 4.2. SEM images of PBS hollow ﬁber membranes .......................................... 105
Figure 4.3. SEM images of PS hollow ﬁber membranes.........................................106

Figure 4.4. Schematic diagram of the ﬂow conﬁguration for modifying a hollow ﬁber
membrane with a PSS/PAH/AuNP ﬁlm ...................................................... 107

Figure 4.5. SEM images of PBS hollow ﬁber membranes coated with PSS/PAH/AuNP
ﬁlms... ......................................................................................................... 111

Figure 4.6. SEM images of PS hollow ﬁber membranes coated with PSS/PAH/AuNP
ﬁlms ............................................................................................................. 112

Figure 4.7. Water permeabilities of PBS hollow ﬁber membranes before and after
adsorption ofdifferent polyelectrolyte/AuNP ﬁlmsl 15

Figure 4.8. SEM images of the shell surfaces of bare PES hollow ﬁber membranes taken
from different rolls ....................................................................................... 116

Figure 4.9. Percent reduction of 4-nitrophenol during passage of a 0.5 mM 4-nitrophenol,
25 mM NaBH4 aqueous solution through a PBS hollow ﬁber membrane
coated with a PSS/PAH/AuNP ﬁlm] 19

Figure 4.10. Schematic diagram of the shell to lumen solution ﬂow conﬁguration used in
most catalytic reactions in this study ............................................................ 120

xi

Figure 4.11. Percent reduction of 4-nitrophenol during passage of a 0.5 mM 4-
nitrophenol, 25 mM NaBH4 aqueous solution through a PBS hollow ﬁber
membrane coated with a PSS/PAH/AuNP ﬁlm ................................. 121

Figure 4.12 SEM image of the lumen surface of the catalytic hollow ﬁber after use in 4-
nitrophenol reduction ............................................................... 122

Figure 4.13. Percent reduction of 4-nitrophenol during passage of a 0.5 mM 4-
nitrophenol, 25 mM NaBl-I4 aqueous solution through a PS hollow ﬁber
membrane coated with a PSS/PAH/AuNP ﬁlm .................................. 123

Figure 4.14. Percent reduction of 4-nitrophenol during passage of a 0.5 mM 4-
nitrophenol, 25 mM NaBI-I4 aqueous solution through a PS hollow ﬁber
membrane coated with a PSS/PAH/AuNP ﬁlm .................................. 125

Figure 5.1. SEM images ofalumina and PBS ﬂat-sheet membranes........................134

Figure 5.2. TEM image of citrate-stabilized Pd nanoparticles and histogram of Pd
nanoparticle diameters ................................................................................. 135

Figure 5.3. Apparatus employed for membrane modiﬁcation and catalytic reactions... 1 36
Figure 5.4. Structures of the polyelectrolytes used in this study ................................... 138

Figure 5.5. SEM images of cross-sections of ﬂat-sheet membranes coated with
polyanion/PEI/PdNP. . . ................................................................................ 1 39

Figure 5.6. Percent reduction of 4-nitrophenol during passage of a 0.5 mM 4-nitrophenol,
50 mM HCOONa aqueous solution through an alumina membrane coated
with a PAA/PEI/PdNP ﬁlm ......................................................................... 141

Figure 5.7. Effect of HCOONa concentration on the percent reduction of 4-nitrophenol
during passage of a 0.5 mM 4-nitrophenol, HCOONa aqueous solution
through an alumina membrane coated with a PAA/PEI/PdNP ﬁlm ............ 142

Figure 5.8. SEM images of the skin surfaces of PAA/PEI/PdNP-coated alumina
membranes after their use in catalyzing the reduction of 4-nitrophenol by
HCOONa .................................................................................................... 143

Figure 5.9. Percent reduction of 4-nitrophenol during passage of a 0.5 mM 4-nitrophenol,
50 mM HCOONa aqueous solution through a PBS membrane coated with a
PSS/PEI/PdNP ﬁlm...... ............................................................................... 144

Figure 5.10. Conversion of 4-nitrophenol during passage of a 0.5 mM 4-nitrophenol, 50
mM HCOONa aqueous solution through a PBS membrane coated with a
PSS/PEI/PdNP/PEI ﬁlm ............................................................................... 146

xii

Figure 5.11. Conversion of 4-nitrophenol during passage of a 0.5 mM 4-nitrophenol, 50
mM HCOONa aqueous solution through a PBS membrane coated with a
PSS/PEI/PdNP/PAH ﬁlm .......................................................... 147

Figure 5.12. Percent reduction of 4-nitrophenol during passage of a 0.5 mM 4-
nitrophenol, 50 mM HCOONa aqueous solution through a PBS membrane
coated with a PSS/PEI/PdNP/PAH ﬁlm .......................................... 148

Figure 5.13. Percent reduction of 4-nitrophenol during passage of a 0.5 mM 4-
nitrophenol, 50 mM HCOONa aqueous solution through a PBS membrane
coated with a PSS/PEI/PdNP/PAH ﬁlm .......................................... 149

Figure 5.14. Percent reduction of 4-nitrophenol during passage of a 0.1 mM 4-
nitrophenol, 10 mM HCOONa aqueous solution through a PBS membrane
coated with a PSS/PEI/PdNP/PAH ﬁlm ........................................... 151

Figure 5.15. Percent reduction of 4-nitrophenol during passage of a 0.1 mM 4-

nitrophenol, 10 mM HCOONa aqueous solution (pH adjusted to 4.5) through
a PBS membrane coated with a PSS/PEI/PdNP/PAH ﬁlm .................... 152

Figure 6.1. Schematic diagram of a gas-liquid contactor in a hollow ﬁber membrane..158

xiii

Scheme 1.1.

Scheme 1.2.

Scheme 1.3.

Scheme 1.4.

Scheme 1.5.

Scheme 1.6.

Scheme 1.7.

Scheme 4.1.

LIST OF SCHEMES

Formation of an OTS SAM on a glass substrate... ..................................... 8

Formation of a gold-supported alkane thiol SAM with terminal functional
groups ......................................................................................................... 10

Fabrication of LBL ﬁlms via hydrogen bonding between poly(acry1ic acid)
and poly(4-vinylpyridine) ........................................................ 15

Synthethis of multilayer ﬁlms via stepwise reaction of diisocyanates and
diamines.......... ........................................................................................ 17

LBL assembly of polymer ﬁlms through “click” chemistry . . .19

F orrnation of Pd nanoparticles through reduction of Pd(II) in [PAA/PEI-

Pd(II)]nPAA ﬁlms ................................................................... 21
Imidization of a poly(amic acid)/PAH ﬁlm via heating to produce a gas-

selective material .................................................................. 34
Au—catalyzed reduction of 4—nitrophenol with NaBH4 ..................... 117

xiv

AFM

AuN P
CP

DP

EDC

F ESEM
FT-IR
GA
GOD
ICP-OES
LB

LBL

MF

NF

NPS
OTS
PAA
PAH
PAN
PDADMAC

PdNP

LIST OF ABBREVIATIONS

Atomic Force Microscopy

Atomic Absorption Spectroscopy

Au Nanoparticle

Coprecipitation

Deposition-Precipitation
N-(3-dimethylaminopropyl)-N’-ethylcarbodiimide
Field Emission Scanning Electron Microscopy
Fourier Transform-Infrared Spectroscopy
Glucoamylase

Glucose Oxidase

Inductively Coupled Plasma Optical Emission Spectroscopy
Langmuir-Blodgett

Layer-by-Layer

Microﬁltration

Nanoﬁltration

Nanoporous Polymer Spheres
N-octadecyltrichlorosilane

Poly(acrylic acid)

Poly(allylamine hydrochloride)
Poly(acrylonitrile)
Poly(diallyldimethylmnmonitun chloride)

Pd Nanoparticle

XV

PEI
PEMs
PEO
PES
PET
PMMA
PS

PSS
PSSMA
PVC
PVP
PVS
QCM
RO
SAM

TALH

Poly(ethylenimine)
Polyelectrolyte Multilayer Films
Poly(ethylene oxide)
Poly(ethersulfone)
Poly(ethylene terephthalate)
Poly(methyl methacrylate)
Polystyrene
Poly(tyrenesulfonate)
Poly(4-styrenesulfonic acid-co-maleic acid)
Poly(vinyl chloride)

Poly(vinyl pyrrolidone)
Poly(vinylsulfate)

Quartz Crystal Microbalance
Reverse Osmosis

Self assembled monolayer

Titanium (IV) bis(ammonium lactate) dihydroxide

xvi

Chapter 1 Introduction and Background

This dissertation describes the growth of polyelectrolyte multilayer ﬁlms (PEMs)
using layer-by-layer (LBL) adsorption on and in porous supports to form high-ﬂux
nanoﬁltration (NF) membranes and membrane reactors that employ immobilized metal
nanoparticles as catalysts. These studies build on previous research and recent
developments in the preparation of thin ﬁlms through LBL adsorption. To put my
research in perspective, this introduction ﬁrst describes the importance of functional thin
ﬁlms and two well-developed thin ﬁlm preparation methods that have been used for more
than 50 years. Subsequently, I focus on the ﬁlm-formation technique employed in this
dissertation, LBL adsorption. This method was developed in the 19903 and has attracted
extensive research interest during the past two decades due to its simplicity and
versatility. The discussion of the LBL method includes the driving forces involved in
assembly, the applicability of this technique to substrates of different geometries, and
recent developments in optimization of the LBL process. Because my work aims at using
LBL adsorption to create nanoﬁltration and catalytic membranes, I also present
applications of the LBL method in these areas. Finally, I give a brief outline of this

dissertation.

1.1. Functional Thin Films
1.1.1. Applications of Functional Thin Films
Functional thin ﬁlms have important practical applications in a number of areas

including separation membranes, sensors, and solar cells. Thus, optimization of the

methods for preparation of functional thin ﬁlms could potentially improve the efﬁciency
of a wide range of practical devices. Ultrathin ﬁlms with thicknesses on the 11m scale are
particularly attractive because their minimal thickness allows the use of relatively
expensive materials and may also improve device performance. In the case of sensors, a
thinner sensing layer usually results in a lower response time,1 whereas in separation
membranes, ultrathin membrane skins allow high permeation rates.2 Because fabrication
of ultrathin ﬁlm on solid surfaces is vital for creating devices ranging from membranes to

transistors, various methods have been developed for thin ﬁlm preparation.3

1.2. Methods for Preparing Ultrathin Functional Films

A partial list of strategies for thin ﬁlm fabrication includes spin-coating}9

10-16 17-21

physical/chemical vapor deposition, electrospray deposition, sol-gel processing,”

27-29 30-33

26 sputtering, and pulsed laser deposition. . Here, I focus on three methods for
forming ultrathin organic ﬁlms with thicknesses on nanometer scale. Two of them, the
Langmuir-Blodgett (LB) technique and adsorption of self-assembled monolayer (SAMs),
are well established and have been widely used for more than 50 years. The other
strategy, LBL adsorption emerged in the 19903 and progressed rapidly due to the ease of
ﬁlm formation and ﬁlm versatility.
1.2.1. Langmuir-Blodgett (LB) Technique
1.2.1.1. History and Background of the LB Technique

The Langrnuir-Blodgett (LB) technique yields ordered monomolecular ﬁlms with

a densely packed structure and precisely controlled thickness.34 Irving Langmuir initiated

work in this area in the early 1900s while studying the structure of monolayers of

amphiphilic species, molecules with a polar head group and a long hydrocarbon tail, at
the air-water interface.35 Speciﬁcally, be dissolved amphiphilic molecules in a water-
immiscible organic solvent and carefully cast this solution onto a clean water surface. A
disordered layer of amphiphiles formed on the water after the solvent evaporated.
Compression of the molecules through the movement of a barrier across the surface
(Figure 1.1) forced the molecules to pack closely and form a monolayer with the

hydrophobic tails of the molecules aligned in the same direction.

——0
Hydrophobic tail Polar headgroup
(ex: alkyl chain) (ex: -OH,-COOH)

ll/—\\./.|

ﬂComprasion to form

Air-water ordered monolayer

interface

Figure 1.1. Formation of a monolayer of amphiphilic molecules at the air-water interface.

Blodgett, together with Langmuir, found that once the amphiphilic molecules
attain the desired organization, the monolayer can be transferred from the air-water
interface to solid substrates (e.g., glass, mica, silica or quartz) by simply immersing the
substrate vertically through the organized monolayer.36 In the case of a hydrophobic

substrate, the monolayer transfers to the solid surface via hydrophobic interactions

between the alkyl chain and the substrate. During withdrawal of the hydrophobic
substrate from the subphase, hydrophilic interactions drive the formation of the second

layer (Figure 1.2). Further dipping can lead to deposition of multilayer ﬁlms.37

/

:0—

 

 

<_—_1L arensqns I

(a)

  

0')
C
U'
0!
0".-
ﬂ
0)
4-0-
CD

 

A
C"
V

Figure 1.2. Formation of Langmuir-Blodgett ﬁlms. (a) a single monolayer being
transferred to a hydrophobic support on a down-trip. (b) transfer of a second monolayer
on an up-trip to form a bilayer. (The dashed lines represent the air-water interface.)

(Redrawn from Biosens. Bioelectron. 2005, 21, 1-20.)

Different dipping schemes result in the three types of Langmuir-Blodgett

multilayer ﬁlms shown in Figure 1.3.38 Y-type multilayer ﬁlms are the most common

and form on either hydrophilic or hydrophobic substrates. Because head-head or tail—tail
interactions drive the ﬁlm formation, Y-type ﬁlms are typically more stable than X- or Z-
type ﬁlms. In contrast, X- and Z-type multilayer ﬁlms can only form on hydrophobic and
hydrophilic substrates, respectively. In some cases, the spacing between the hydrophilic

11111111le111111111111111111111
iiiiiiliii iliiiii m 1111111111
111111“ 4W

X-type Y-tYPe Z-tYPe

 

 

 

 

Figure 1.3. Three possible ﬁlm architectures in Langmuir-Blodgett ﬁlms. From left to
right: X-type ﬁlms on a hydrophobic substrate, Y-type ﬁlms on a hydrophilic substrate
and Z-type ﬁlms on a hydrophilic substrate. (Redrawn from Chem. Rev. 2004, 104, 5479-

5501.)

1.2.l.2. Advantages of the LB Method

Although it is one of the earliest strategies for fabricating a supramolecular
assembly, the LB approach offers several advantages over other techniques. First, the LB
method allows precise control over ﬁlm thickness at the molecular level.39 Second, this

technique affords homogeneous deposition of a monolayer over relatively large surface

areas. Finally, the LB method can provide multilayered ﬁlm structures with a deﬁned
sequence of composition along with ordered molecular arrangement and orientation.40
Although this technique was traditionally limited to depositing organic amphiphilic
molecules, LB ﬁlms composed of a wide variety of inorganic complexes with interesting
magnetic and conducting properties can be prepared.38’39 This increased versatility
expanded the potential applications of the LB method to areas such as biosensors,
semiconductors, optics and electronics.l’38’39’4"42
1.2.2. Self-Assembled Monolayers (SAMs)

Although the LB technique is more than 70 years old, it still requires special
equipment for ﬁlm fabrication, and the weak interactions between the ﬁlm and the
substrate limit the chemical and mechanical stability of LB ﬁlm3.3’4345 Self-assembled
monolayers (SAMs) attach to solid surfaces through relatively strong interactions
(physisorption, chemisorption, or covalent bond formation), and are therefore generally
more stable than LB ﬁlms.34’45'47
1.2.2.1. History of SAMs

Bigelow ﬁrst noted the self—assembly of organic monolayers in the 19503 when he
found that long chain alkylalcohols adsorb on clean glass surfaces.48 A few years later,
Shafrin and Zisman discovered similar adsorption of long-chain alkylamines on Pt

9

substrates.4 However, these layers formed via physisorption and could be easily

removed from the substrate.1

Allara and Nuzzo later formed more robust ﬁlms through
chemisorption, where deprotonated fatty acids bound strongly to alumina surfaces,50 and
soon thereafter SAMs formed by covalent bonding appeared.

1.2.2.2. SAMs Formed via Covalent Bonding to a Surface

There are two primary groups of SAMs based on strong chemisorption. The ﬁrst
group, initially developed by Sagiv and coworkers, utilizes silane chemistry to link the
monolayer to the surfaces"52 Well-packed monolayers form on various substrates (such
as clean glass, poly(vinyl alcohol), oxidized polyethylene, and evaporated alumina)
through simple immersion of the substrate in an organic solution containing n-
octadecyltrichlorosilane (OTS). Scheme 1.1 shows the chemistry behind the ﬁlm
formation, which includes hydrolysis of the Si-Cl bonds, adsorption on the substrate via —

OH groups, and condensation polymerization of the silanol groups.

H20

 

 

 

 

Cl—Si—CI > HO—Si—OH
Hydrolysis
Cl OH
Adsorption
H.
/
HO—SI—O. O--S1—OH
I /
OS\ iO/ WHWO 0 H O
Sl/ WPolymerization H/ \H H/ \H
?/ (Si/S n Condensation (‘3/ ?/
SI 8' Si
U Glass U Glass \J

 

 

 

 

 

 

Scheme 1.1. Formation of an OTS SAM on a glass substrate. (Redrawn from J. Am.

Chem. Soc. 1980, 102, 92-98.)

The other extensively studied method for formation of covalently immobilized
SAMs is adsorption of sulfur-containing compounds (alkane thiols and alkyl sulﬁdes) on
gold substrates.53'56 Sulfur compounds are highly reactive towards noble metals, and
alkane thiols have high afﬁnity for a variety of substrates including gold,54'56 silver,”59
copper,58 palladiumf’o’“ and platinum.62 The strong interaction between thiols and these
metals makes it possible to generate well-deﬁned and relatively stable SAMs.

Monolayers formed by adsorption of alkane thiols on gold (Scheme 1.2) comprise
the most extensively studied SAMs. Gold is the most commonly used substrate for
several reasons. First, binding to the surface occurs through interactions between “soft”
gold and sulfur atoms, whereas many other ftmctional groups (acids, amines, etc.) are
relatively “hard” and do not have a strong afﬁnity for the gold surface.1 This allows the
incorporation of other functional groups into these SAMs without disruption of the gold-
sulfur interaction. Second, gold is easy to clean and reasonably stable under standard
laboratory conditions. The interaction between gold and sulfur is strong enough to
displace weakly bound impurities. In addition, gold-coated substrates are amenable to a
wide range of SAM characterization techniques including ellipsometry, reﬂectance

fourier transform-infrared spectroscopy (FT-IR), quartz crystal microbalance (QCM)

gravimetry, and electrochemical methods.63

 

 

 

 

= :1
Adsorption
HSM on gold
X
" a 3
Gold
(X=CH3, COOH, NHZ, etc.)

Scheme 1.2. Formation of a gold-supported alkane thiol SAM with terminal functional

groups.

Formation of alkane thiol SAMs typically occurs by immersion of a cleaned gold
substrate in an alcoholic solution of thiol and removal of weakly adsorbed molecules by
rinsing with alcohol.45 Adsorption times vary from minutes to days, depending on the
nature and concentration of the thiol solution. Kinetic studies of this type of SAM
formation indicate that there are two adsorption steps.“55 The initial step, which is well
described by diffusion-controlled Langmuir adsorption, depends on the thiol
concentration. This step takes place in several minutes and includes an oxidative addition
of S-H to the gold surface followed by a reductive elimination of the hydrogen.64 In the
second step, which lasts several hours, the alkyl side chains assemble together to
maximize Van der Waal’s interactions and form two-dimensional crystalline structures.
Molecules with long alkyl chains such as hexadecanethiol lead to well-packed quasi-
crystalline monolayers while molecules with short alkyl chain form liquid-like
monolayers."64
1.2.2.3. Advantages of SAMs

As Figure 1.4 shows, the formation of SAMs is driven by strong interactions such

as covalent bond formation between the functionalized headgroup and the substrate

10

surface, as well as hydrophobic interactions between the hydrocarbon backbones of the
absorbent molecules. Thus, SAMs are reasonably stable.40 The ease of preparation and
highly ordered ﬁlm structure make SAMs promising candidates for modiﬁcation of
surface properties including wetting, frictional coefﬁcient, and susceptibility to

65-68

corrosion. More importantly, SAMs can incorporate a wide range of functional

groups both in the alkyl chain and at the chain endgroups. This enables the formation of

surfaces capable of speciﬁc interactions.64’69

All these unique features make SAMs
attractive in the areas of sensors, wetting and lubrication control, and biomolecular and
molecular electronic devices.45’46’63’69’7O In some cases, the combination of LB and SAM
techniques yields highly ordered and thermodynamically stable thin ﬁlms that take

advantage of assets of both methods.”44

Endgroup

<— Alkyl chain

(— Headgroup
<— Substrate

 

Figure 1.4. Structure of a SAM. The shaded and open circles indicate chemisorbing
headgroups and free endgroups, respectively. (Reproduced with permission from Prog.

Surf. Sci. 2000, 65, 151-257.)

1.2.3. Layer-by-Layer (LBL) Assembly
Layer-by-layer (LBL) ﬁlm formation technique involves alternating adsorption of

interacting materials. In the early 19903, Decher developed electrostatic methods for the

formation of polymeric multilayer ﬁlms,”73 but other interactions can also facilitate LBL
ﬁlm formation. Below I discuss ﬁlms prepared using different interactions.
1.2.3.1. Driving Forces Involved in LBL Assembly
1.2.3.1.]. LBL Assembly Based on Electrostatic Interactions

LBL adsorption of polyelectrolytes is one of the easiest surface modiﬁcation
methods because it simply involves exposure of substrates to oppositely charged species,

with water rinses between each deposition step. Many water-soluble and multiply

7 1,73-76 77-80

charged species such as polymeric electrolytes, proteins, colloidal
nanoparticles,8"84 DNA,85'87 and dye moleculesg‘s’gs‘g9 can be incorporated in LBL ﬁlms.
Figure 1.5 illustrates the formation of polyelectrolyte-colloid bilayers on a
negatively charged substrate. Speciﬁcally, the negatively charged solid substrate is
initially placed in a dilute solution of polycations, and multiple polyelectrolyte-surface
interactions result in the formation of a thin layer of polycation. This process is
entropically favored because the polyvalent attachment of a single polycation molecule to
the surface releases many counter-ions (both cations and anions) into the solution.90
Overcompensation of the surface charge by the polycation leads to a positive surface,”
and the modiﬁed substrate is then rinsed with water and subsequently immersed in a
solution of negatively charged colloids. Adsorption of the colloid takes place and again
reverses the surface charge. After another water rinse, one “bilayer” of the polycation-

colloid ﬁlm is present on the substrate, and this process can be repeated until the desired

number of bilayers or ﬁlm thickness has been reached."3"90‘92‘93

12

Repeat

 

  
   
 

     

 

Adsorption
‘e

 

Two bilayers of
polyelectrolyte-colloid ﬁlm

 

 

 

I Charge Overcompensation -—> Charge Reversal]

 

Figure 1.5. LBL assembly of polyelectrolyte-colloid bilayers through electrostatic
interactions. (Reproduced by permission of the PCCP Owner Societies from Phys. Chem.

Chem. Phys. 2007, 9, 2319-2340.)

1.2.3.1.2. LBL Assembly via Hydrogen Bonding

Stockton and Rubner94 and Zhang and coworkers95 introduced hydrogen bonding
as a driving force for LBL assembly in 1997. In this method, alternating adsorption of
polymers with hydrogen bond donors and polymers with hydrogen bond acceptors results
in ﬁlm formation. Stockton and Rubner examined the LBL adsorption of polyaniline and
a variety of neutral water-soluble polymers, including poly(vinyl pyrrolidone) (PVP),

poly(vinyl alcohol), poly(acrylamide), and poly(ethylene oxide) (PEO).94 Figure 1.6

shows the structures of these polymers. F T-IR studies conﬁrm that hydrogen bonding
drives the ﬁlm formation. The non-hydrogen-bonded N-H stretch of pure polyaniline
appears at about 3380 cm", while the hydrogen-bonded N-H stretch has an energy of
about 3310 cm". The appearance of pronounced peaks at 3304 and 3300 cm", for
polyaniline/PVP and polyaniline/PEO systems, respectively, suggests that a large fraction

of the N-H groups in polyaniline is involved in the hydrogen bonding.94

*0 W10 "Olin

Poly(vinyl pyrrolidone) Poly(vinyl alcohol)

“2
c
f \31 H2
" c o
x li/ \C/ \1\
HzN o “2 "

Poly(acrylamide) Poly(ethylene oxide)

Figure 1.6. Structure of polymers used in LBL assembly based on hydrogen bonding.93’94

Zhang and coworkers demonstrated the fabrication of LBL ﬁlms through
hydrogen bonding using poly(acrylic acid) (PAA) as a hydrogen bond donor and poly(4-

vinylpyridine) as a hydrogen bond acceptor as shown in Scheme 1.3.92 Adsorption of

14

PAA layers was performed using solutions at pH 3.28 to suppress ionization of the
carboxylic acid groups. UV-vis spectrophotometry suggests linear ﬁlm growth, and FT-
IR spectroscopy conﬁrms that the ﬁlm growth occurs due to hydrogen bonding.

However, ﬁlms only form when the carboxylic groups of PAA are mostly protonated to

2

allow PAA to serve as a hydrogen bond donor.9

 

Scheme 1.3. Fabrication of LBL ﬁlms via hydrogen bonding between poly(acrylic acid)
and poly(4-vinylpyridine). (Reproduced by permission of The Royal Society of

Chemistry ﬁ'om Chem. Commun. 2007, 1395-1405.)

Hydrogen bonds are more sensitive to changes in their environment than
electrostatic interactions, and LBL ﬁlms formed via hydrogen bonding may only be
stable under certain conditions. Sukhishvili and Granick studied the stability of

hydrogen-bonded LBL ﬁlms of poly(methacrylic acid) (PMAA) and poly(vinyl

15

pyrrolidone) at different pH values, and found that these ﬁlms dissolve at pH values
higher than 6.9.91"97 Dissolution occurs due to ionization of the carboxylic groups in
PMAA, which introduces an unacceptably large electric charge within the ﬁlm.96 Similar
phenomena appear in other LBL assemblies formed through hydrogen bonding,97 and
dissolution under speciﬁc conditions enables the formation of responsive ﬁlms that might
prove useful in biomedical applications.93
1.2.3.1.3. LBL Assembly Based on Covalent Bonding

The use of covalent interactions as a basis for the LBL method should
signiﬁcantly improve ﬁlm stability, which is advantageous for most applications. In
1997, Crooks and coworkers reported one of the ﬁrst examples of multilayer polymer
ﬁlms formed through covalent bonding.”99 LBL reactions of poly(maleic anhydride)-
co-poly(methyl vinyl ether) (GrantrezTM) with amine-terminated dendrimers gave ﬁlms
on glass, Si, and Au-coated Si substrates. These covalently assembled ﬁlms had a
thickness of ~15 nrn per bilayer. Due to the presence of ionizable amine and carboxylic
groups, these materials exhibited pH-switchable perrnselectivity that allowed exclusion of
cations and anions at low and high pH, respectively.98’99

Kohli and Blanchard reported the covalent growth of multilayer ﬁlms of
diphenylmethane derivatives using urea interlayer linking chemistry.loo Alternating
addition of diisocyanates and diamines (Scheme 1.4) afforded a controlled linear ﬁlm
growth, with a total thickness of 9.8 nm for 7-bilayer ﬁlms.100 Kohli and Blanchard later
demonstrated the compatibility of covalent and ionic interactions in LBL growth, which

might prove useful in the formation of ultrathin conducting and insulating layers.""

16

 

Me
OH MW\ 3“ O
\ NH
OEt ‘ s
7 Me Me
OH Toluene, 24 hrs. 20°C 0
\
W<\/\NH2
Me Me

 

DMF, 4 hrs, 45°C under Ar Me Me H O H
wash with DMF l
O\ J\
SN N N nco
H

DMF, 4 hrs, 45°C
H2N NH2 wash with DMF, H20

W‘s/\AN

s\ | l I II4

Me Me H\O/ H\O/

FCC '11.th O 0 ”“
MW; Me H H

Scheme 1.4. Synthethis of multilayer ﬁlms via stepwise reaction of diisocyanates and

diamines. Further layers can be deposited similarly. (Redrawn from Langmuir 2000, 16,

4655-4661.)

Caruso and coworker recently described covalent LBL assembly through “click”

chemistry.102 “Click” chemistry refers to a set of high-yield reactions that can be

17

performed under simple conditions.103 One of the most commonly used click reaction is
the Cu(I)-catalyzed variant of the Huisgen 1,3-dipolar cycloaddition of azides and
alkynes to form 1,2,3-triazoles.103 '104 Caruso and coworkers demonstrated covalent LBL
assembly on quartz, silicon or gold through alternating immersion in Cu(I)-containing
solutions of poly(acrylic acid) copolymerized with either azide or alkyne-containing
monomers (Scheme 1.5).l02 Film thickness increased linearly with the number of layers
when using alternating poly(acrylic acids) with ~16% azide and ~14% alkyne. This
method yields stable ﬁlms prepared from only one type of polyelectrolyte, which cannot
be realized using conventional LBL assembly. A subsequent study applied this ﬁlm-
forrnation method on particle substrates. Dissolution of the particles gave polymer

capsules with a pH-dependent size.105

18

ICll Ill
N3 N3
Cu(I)
H H N3 N3
| I
C C

0
O

CEO—I
CEO—I

 

C C
III III
i <.=
H H
/ [N / N
I I
N~N N~Nl
N—N N— CU“)
\\ 11
\ N \ N

Scheme 1.5. LBL assembly of polymer ﬁlms through “click” chemistry. (Redrawn from

J. Am. Chem. Soc. 2006, 128, 9318-9319.)

1.2.3.1.4. Other Interactions Involved in LBL Assembly
Extensive recent research shows that the driving force for LBL assembly is not

limited to electrostatic interactions and hydrogen or covalent bonding. Hydrophobic

06 107,108 109-ll]

interactions,1 bio-speciﬁc recognition, charge transfer interactions,
supramolecular interactions,43 and n-rt interactions112 can all facilitate LBL ﬁlm
formation. The versatility of the driving forces in LBL assembly enables the utilization
of a wide range of molecules in ﬁlm formation, and thus, expands the application of this

technique.

19

1.2.3.2. LBL Assembly on Substrates of Various Geometries
One of the most attractive features of LBL assembly is its applicability to not only
planar substrates, but also to porous membranes or even colloids and

nanorods.1t90,92.93.l 13,1 14

Our group demonstrated the deposition of metal
nanoparticle/polyelectrolyte multilayer ﬁlms on colloidal particles and in porous
membranesns"120 Kidambi and Bhattacharjee deposited Pd nanoparticles in PEMs by
alternating immersion of alumina particles in solutions of poly(acrylic acid) (PAA) and
poly(ethylenimine) (PEI)-Pd(II) followed by reduction of Pd(II) with NaBH4 (Scheme
1.6).l '5 '1 ‘7 Dotzauer immobilized Au nanoparticles in alumina membranes by alternating
adsorption of PAA, poly(allylamine hydrochloride) (PAH), and citrate-stabilized Au

colloids (Figure 1.7). The high surface area of 0.2 rim-diameter pores of alumina

membranes gave rise to a high catalyst loading.l '9

20

 
 
 
  

20 mM PAA
10 min

(1). (2). (1)

Alumina

 

I333 PEI-Pd(ll)
..8: PAA

. Pd particle

 

 

 

Scheme 1.6. Formation of Pd nanoparticles through reduction of Pd(II) in [PAA/PEI-

Pd(II)]nPAA ﬁlms. (Reproduced with permission from Chem. Mater. 2005, I 7, 301-307.)

21

 

W

          

 

 

 

« f Polyanion

i Polycation

: deposition 2, W
23,321 ' 1 Polycation
e
a :. . “' Au nanoparticle
e . . . it».
Positively charged Polyanlon-coated Polycatlon-coated AU-COIIOId coated

pore surface pore surface pore surface pore surface

Figure 1.7. Modiﬁcation of a pore of an alumina membrane by LBL adsorption of two
oppositely charged polyelectrolytes and Au nanoparticles. (Reproduced with permission

from Nano Lett. 2006, 6, 2268-2272.)

As shown in Figure 1.8, dissolution or calcination of the colloidal particles that
sometimes serve as LBL substrates yields hollow materialsm’122 In the case of
calcination, the LBL method should include the deposition of inorganic materials that are
stable under these conditions. Hollow capsules prepared by LBL methods can potentially
be used in the ﬁelds of medicine, pharmaceutics, and material science for applications

involving encapsulation or controlled release.12”123"24'127

22

Polymer

     

AdSOl’ptiOfl .1 W W _. (2) 3.02
(1) . . (3) PDADMAC

Colloidal W ' ' 1
Particle
(2), (3)...

Calclnatlon

   

Hollow

Hollow Silica Spheres "Maggi?“ Inorganic-Hybrid Spheres

Figure 1.8. LBL assembly on particle templates for hollow-sphere formation. The
substrates are polystyrene (PS) latex particles with a diameter of 640 nm. (Reproduced

with pennission from Science 1998, 282, 1111-1114.)

A similar synthetic approach with nanotube substrates yields hollow nanotubular
structures that may be of interest in optical, electronic, and magnetic devices.m"28"29
Polymer nanotubes were prepared by alternating adsorption of
poly(diallyldimethylarnmonium chloride) (PDADMAC) and poly(styrene sulfonate) (PSS)
on nickel nanorods followed by core dissolution (route A in Figure 1.9). Titania-based
nanotubes were obtained with a similar approach (route B in Figure 1.9) where nickel
nanorods were coated with PDADMAC and an inorganic precursor, titanium (IV)
bis(ammonium lactate) dihydroxide (TALH). Hydrolysis of the titania precursor under
reﬂux was followed by nickel core dissolution. This method is a water-based technique,
and it allows for formation of nanotubes of various compositions, including protein

128,130

nanotubes, and control over the wall thickness of the nanotubes.122

23

‘.

Nickel Nanorods

C 1.PDADMAC c ;-:2L\:MAC
2.PSS .
/A B\

 

1.Reﬂux
Dissolve Ni . .
2.Dlssolve Nl
T“: .. 75% 2133 t
;. ‘:‘:&
Polymer Nanotubes Titania-based Nanotubes

Figure 1.9. Preparation of polymer and inorganic-organic composite nanotubes by an
LBL-colloid templating approach. (Reproduced with permission from Nano Lett. 2001, I ,
727-730.)

1.2.3.3. Optimization of the LBL Adsorption Process
Although LBL assembly is relatively simple, the tedious dipping and rinsing steps

limit the potential application of this technique.131 To overcome this challenge, Shiratori

24

and coworkers developed an automatic LBL ﬁlm formation system with precise control

over the mass of materials depositedm’I33

In this case, a QCM with an integrated
oscillation circuit was attached to the arm of a robot to monitor the frequency shift during
the adsorption or rinsing process. Using the frequency change as feedback, the mass of
materials deposited in each step was controlled precisely. Transmission electron
microscopy and atomic force microscopy (AFM) showed that the surface roughness of
the ﬁlms formed by this method was much smaller than the roughness of ﬁlms prepared
by the conventional time-controlled dipping method. Not only polyelectrolytes, but also
inorganic materials such as TiOz can form uniform structures with this method.133
However, automation of the LBL method does not reduce the time required for
ﬁlm formation, which usually ranges between several minutes and several hours,
depending on the absorbing systems and the target number of bilayers.134 This adsorption
process is time consuming because the polyelectrolyte molecules must diffuse toward the
substrate, interact with the surface, and then rearrange to pack denselym'136 The spin-
coating method developed by Lee and coworkers (Figure 1.10)dramatically reduces the

preparation time per layer from minutes to seconds.134

This signiﬁcant time reduction
stems from several sources. First, the adsorption and rearrangement of adsorbed chains
on the substrate and the elimination of weakly bound molecules are achieved almost
simultaneously in a short time. Second, the polyelectrolyte concentration increases
signiﬁcantly due to the rapid elimination of water during the spinning process, and thus
thick layers are formed. Finally, the electrostatic interactions between oppositely charged

polymers are enhanced because the adsorbed water molecules, which generally screen the

electrostatic attraction, may be partially forced out of the ﬁlm during spinning.I35

25

Air ﬂow
Air shear
force

Centrifugal I
force

Preadsorbed
cwater molecules

After spinning
including washing steps

Polyanion

   

Polycation

 

Figure 1.10. Buildup of LBL ﬁlms by consecutive spinning of polycation and polyanion
molecules. (Reproduced by permission of the PCCP Owner Societies from Phys. Chem.

Chem. Phys. 2007, 9, 2319-2340.)

Another advantage associated with the spin-coating method is that it yields a
highly ordered internal structure, which cannot be easily obtained with the conventional
dipping method.135 Figure 1.11 clearly shows this effect by comparing the structures of
the (PAH/CdS)n ﬁlms obtained with dipping and spin-coating methods. (The structures
in the ﬁgure are based on X-ray reﬂectometry and AFM imaging.) This difference in
ﬁlm structures presumably occurs due to the mechanical effects of the air shear force

during the spinning process.3 Later research showed that the spin-coating method is

26

effective in LBL assembly of a wide variety of polyelectrolytes including polymers,

dendrimers, and dye molecules.'37"39

Dipping (PAHICdS)n Films Spin (PAHICdS)n Films
PAH CdS

 
    

l
\\\\ \\\\ \\\\ \\\\\\. \\\\'\.‘ \\\ \\\\\ \V.‘ \\\\,\\\ '\\ \v a \\ \vnAx‘. \\\\\\'\\.\\.\\.‘~.‘»\\\\\\\~.\'\\ \\‘~ \\.\\\\‘n\\\~

v.
-‘\m\'\'\‘- .‘\‘~«\\\‘- -\ .‘\\\\\\\~;\\\\\\'\\.‘ .‘\'\\'«\‘o.\\\ \x‘ xx»‘e\\\\\\~.~.\\\~.~.\\\ \‘Lx. \ \~\\\

\M\‘\

requinN latteua to 9323mm

IIIEIIIIIIIIIlelIlIlele I
IXIjIIIII It XIIIXILI I

\"\ \\'\\‘» \ -‘-‘\‘~\ ‘C’\\\ \‘M‘t\.“

 

   

V
Figure 1.11. Schematic diagram of the internal structure of (PAH/CdS)n ﬁlms prepared

by dipping and spin methods. (Reproduced with permission from Adv. Mater. 2001, 13,

1076-1078.)

Although the spin-coating method may reduce the time required for LBL
assembly, it requires special equipment and will likely not be applicable to large
substrates. In 2000, Schlenoff and coworkers ﬁrst reported the deposition of PEMs by
alternating spraying of the polycation and polyanion solutions on the substrate (Figure
1.12), which is faster, more convenient, and generally more applicable for modiﬁcation

(1.140

of large surfaces than the spin-coating metho Speciﬁcally, Schenoffs group

27

prepared PEMs of PDADMAC/PSS on silica wafers by the spray-coating method. These
ﬁlms had a similar composition and thickness as ﬁlms formed by the conventional
dipping method.140 In 2005, Decher and coworkers examined spray coating of PSS/PAH
ﬁlms to investigate the inﬂuence of various deposition parameters (spraying time,
polyelectrolyte concentration, and the effect of drying during ﬁlm formation) on ﬁlm
growth.I41 Film thickness increased linearly with the number of layer pairs, which is
similar to the conventional dipping method, and AF M and X-ray reﬂectometry data
showed that ﬁlms with small surface roughnesses formed in spraying times as short as 6 3
per layer. In “dipping methods, homogeneous ﬁlms cannot be formed in such short
contact times. Moreover, because drainage constantly removes a certain quantity of the
excess materials on the substrate, the rinsing step can be eliminated to further speed up

the ﬁhn formation process!“

28

 

 

 

Polycation

Figure 1.12. LBL deposition by alternating spray-coating of polycations and polyanions
on a negatively charged substrate. (The rinsing between deposition steps is not shown.)

(Reproduced with permission from Nat. Mater. 2009, 8, 449-450.)

Frey and coworkers later demonstrated the applicability of the spray-coating
method to inorganic/organic nanocomposite ﬁlms consisting of Au nanoparticles and the

photosensitive polycation nitrodiazoresin.I42

Hammond and coworkers developed an
automated spray system and extended the versatility of the spray-coating method to
deposition of ﬁlms of strong and weak polyelectrolytes, hydrogen-bonded ﬁlms, and
ﬁlms containing colloidal nanoparticles. They also reported a 25-fold decrease in process
time compared to the conventional dipping method and coating of colloidal

3

nanoparticles.l4 In contrast to the fast spin-coating method, spray-coating is not

restricted with respect to substrate size.

29

1.2.4. Comparison of LB Films, SAMs and LBL Films

The above discussions show that the LB method provides highly ordered
monolayer or multilayer structures, but it is a relatively time-consuming process and
requires use of expensive instruments and precise control over surface pressure during

dipping and ﬁlm transfer.36

The LB method is also not generally applicable to many
kinds of non-amphiphilic materials, and ﬁlms suffer from instability?“40 SAMs are more
robust than ﬁlms prepared with the LB method,40 but SAMs provide ﬁlm with a very
limited thickness.3 Compared to LB and SAM methods, LBL adsorption technique is a
more versatile technique with many advantages for multilayer ﬁlm formation.

Among the advantages of LBL assembly, ﬁlm fabrication is relatively simple and
sophisticated instruments are not necessary for the assembly process.”92 Because LBL
assembly can occur via different intermolecular interactions, it allows for controlled

incorporation of a wide range of materials including polyelectrolytes,“’73'7‘5‘I32

biomaterial3,30’87‘88’128‘130 dye molecules, 88’89’l38’144’145 and inorganic substancesW’W’121
Another outstanding property of LBL adsorption is the ability to coat substrates of
various geometries including ﬂat surfaces, membranes, and colloidal particles that can
sometimes be removed to give hollow capsule materials.“3 ,121,123.124,146 In addition, the
ﬁlm thickness can be controlled at the nrn scale by varying the number of layers
deposited and adsorption conditions such as the pH and ionic strength of the

1’90’93’l47 Due to these prominent assets, the LBL

polyelectrolyte deposition solutions.
method has been explored for development of functional materials ranging from

separation membranes to nano-sized reactors and drug delivery systemsz’3 These

applications will be discussed in more detail in the following section.

30

1.3. Selected Applications of LBL Films
1.3.1. Separation Membranes

Different from highly ordered LB ﬁlms, LBL ﬁlms tend to be amorphous
structures with the polyelectrolytes penetrating into adjacent layers.72 This kind of
structure allows the transport of small molecules and ions through the ﬁlms, making LBL
ﬁlms interesting barriers for selective passage of small molecules. Below I discuss
nanoﬁltration and gas separation with membranes prepared by LBL deposition.
1.3.1.1. Nanofiltration Membranes

Nanoﬁltration is a pressure-driven, membrane-based ﬁltration process that is
widely used in water softening and pollutant removal.I48 Different from reverse osmosis,
nanoﬁltration occurs at a lower pressure and provides lower rejection of monovalent

ions. '49

These factors provide substantial energy savings compared to reverse osmosis in
applications such as water softening that do not require the removal of monovalent ions.
Highly permeable membranes generally consist of a thin selective layer deposited

on a thicker porous support that provides mechanical strength.15 0"52

Tieke’s group
reported that PEMs comprised of 60 layer pairs of poly(vinylarnine) and
poly(vinylsulfate) (PVS) on a poly(acrylonitrile) (PAN )/poly(ethy1ene terephthalate)
(PET) support exhibited a 95% rejection of MgClz at an operating pressure of 10 bar with
a feed concentration of 0.01 M MgClz.153 However, the use of such a large number of

layers makes fabrication impractical and yields very low ﬂux. We showed that 5-bilayer

PSS/PAH ﬁlms on porous alumina supports exhibit a 95% rejection of Mng along with

31

a NaW/MgZW selectivity of 22 under a pressure of 4.8 bar.154 Chapter 2 presents the details
of that work.

PEMs also proved useful in the selective removal of multivalent anions from
solutions containing monovalent anions. El-Hashani and coworkers prepared 60-bilayer
ﬁlms of the dicopper complex of l8-azocrown-N6 (hexacyclen, 1,4,7,10,13,16-
hexaazaoctadecane, aza6) and PVS for separation of chloride and sulfate.‘55 These ﬁlms
exhibited a chloride rejection of 45.7% and a sulfate rejection of 91.9% using an applied
pressure of 25 bar.155 Previous work in our group also investigated the separation of
sulfate from chloride using different polyelectrolyte systems. Stanton and coworkers
reported that 4.5-bilayer PSS/PAH ﬁlms showed a Cl'/SO42' selectivity of about 30 when
the top layer was deposited from a solution of high ionic strength.152 By changing the top
layer from PSS to PAA, the Cl'/SO42' selectivity was increased to as high as 80, albeit
with a 50% drop in solution ﬂux.”2 Alumina membranes coated with
(PSS/PDADMAC)3PSS ﬁlms are also proved to be useful in chloride/sulfate separation.
The Cl'/SO42‘ selectivity of these membranes was comparable to that of PSS/PAH
systems, but the ﬂux was about 50% higher.156

Deng and coworkers developed a hybrid NF membrane using PSS, PAH and a
copolymer polyelectrolyte poly(4-styrenesulfonic acid-co-maleic acid) sodium salt
(PSSMA).I48 Films containing only two bilayers of (PAH/PSS/PAH/PSSMA) on PAN
ultraﬁltration substrates exhibited a Na2SO4 rejection as high as 91.4% at an applied
pressure of 2 bar.148 Other work reported that PEMs on porous supports are effective in
NF removal of salt from reactive-dye solutions and selective removal of ﬂuoride or

phosphate from aqueous solutionsm'l59

32

1.3.1.2. Membranes for Gas Separation

Fabrication of gas separation membranes using LBL adsorption is more difﬁcult
than formation of NF membranes, because gas separations are more susceptible to defects
in the ﬁlm. Nevertheless, McCarthy reported selective gas permeation through PAH/PSS

160 The poly(4—methyl-1-pentene)

ﬁlms deposited on oxidized poly(4-methyl-1-pentene).
provides a continuous support that minimizes defects. Enhanced selectivities in the
permeation of H2, N2 and 02 occurred upon increasing the number of PSS/PAH layers
from 20 to 200, and 200-layer ﬁlms gave an Oz/Nz selectivity as high as 15.I60 Tieke’s
group also tested the gas separation performance of PAH/PSS ﬁlms on a PAN/PET
support, but even 60-bilayer ﬁlms showed essentially no selectivity in the permeation of
N2, 02 and Aal‘”

Sullivan created selective gas-selective membranes by LBL adsorption of
poly(amic acid) and PAH followed by heat-induced imidization (Scheme 1.7).162 Films
containing 9.5 bilayers of this ﬁlm on alumina exhibited Olez selectivities of 6.9 and
COleH4 selectivities of 69.”2 These membranes are promising for gas separation

because of their stability and the selectivity of polyimides, and their formation does not

involve the tedious deposition of hundreds of layers.

33

 

assesses

heat
0 ac CF3 0-
o
' n
0 o n
NH, NH; rm,+ NH:

M c ..8
weeeaawa

Scheme 1.7. Imidization of a poly(amic acid)/PAH ﬁlm via heating to produce a gas-

selective material. (Reproduced from Langmuir, 2008, 24, 7663-7673.)

34

1.3.2. Membranes for Protein Adsorption
As discussed in section 1.2.3.2, LBL adsorption is useful for modifying the
interior of porous substrates to exploit the high internal surface area of these supports in

2 Dai coated

applications such as catalysis and protein puriﬁcation through adsorption.
the interior surface of a porous alumina support (pore size of 0.2 pm) with
(PAA/PAH).,PAA ﬁlms and subsequently activated the free —COOH groups of the
terminating PAA layer with N-(3-dimethylaminopropyl)-N’-ethy1carbodiirnide (EDC)
and N-hydroxysuccinimide to allow for covalent attachment of antibodies (Figure
1.13).I63 This method allowed development of a ﬂow-through system for antigen

analysis, with minimal nonspeciﬁc protein adsorption on the PAA terminated ﬁlms.164

35

 

  
 

1. anti-IgG

2. wash
with buffer

 

Figure 1.13. Schematic diagram of modiﬁcation of the interior of membranes with PAA-
terminated PAA/PAH ﬁlms and activation of these ﬁlms for covalent protein attachment.

(Reproduced with permission from Langmuir, 2008, 24, 7663-7673.)

In 2006, Caruso and coworkers reported the use of PAA/PAH nanoporous

polymer spheres (NPS) for protein immobilization.“55

After incubation in a lysozyme
solution for 60 min, the cross-linked (PAA/PAH)2PAA NPS contained ~90 wt%
lysozyme, which corresponds to a lysozyme loading of 470 mg/mL in the NPS. The
loading can be reversed by changing the solution pH to mediate the electrostatic

interactions between the proteins and the NPS. These materials are potentially useful in

biosensing, enzyme catalysis and controlled drug delivery.165

36‘

1.3.3. Bio- and Chemical Reactors

Because LBL adsorption can easily incorporate materials such as enzymes and
nanoparticles, this method has been widely used in the development of enzyme and
chemical reactors. Kunitake and coworkers developed a multi-enzyme reactor containing
glucose oxidase (GOD) and glucoamylase (GA) by alternating deposition of
polyelectrolytes and these two enzymes on an ultraﬁltration membrane (Figure 1.14).166
Starch passing through these modiﬁed membranes was converted to glucose through the
hydrolysis of the glycoside bond by GA, and the glucose was subsequently converted to
gluconolactone by GOD with H202 as a byproduct. Unreacted starch was removed by
the ultraﬁltration membrane due to its high molecular weight. Optimization of ﬂow rate

and the arrangement and number of ﬁlm layers can lead to development of highly

speciﬁc enzyme reactors. '66

37

Substrate
Solution

 

  

Giuoono-
54actone

  
  

H202

Figure 1.14. Multi-enzyme reactor prepared by LBL deposition on an ultraﬁltration
membrane. (Reproduced by permission of the PCCP Owner Societies from Phys. Chem.

Chem. Phys. 2007, 9, 2319-2340.)

Dotzauer and coworkers prepared catalytic membranes through LBL adsorption
of polyelectrolytes and Au nanoparticles in porous alumina (Figure 1.15).”9 These
membranes catalyzed the reduction of 4-nitrophenol by NaBH4 with remarkably high and
stable conversions of 4-nitrophenol at a ﬂow rate of 0.17 mL/(cm2 3). These results
indicate a turnover number > 1000 mol of 4-nitrophenol per mol of Au in the membrane.
In addition to high stability and turnover numbers, the LBL method is also applicable to
deposition of catalytic nanoparticles in polymeric membranes such as track-etched

P01ycarbonate and porous nylon,l 19.120

38

BH" BH"
N02 no2

3H;

H H

 

5‘ Metal nanoparticle
m Polycation
Mn Polyanion

  

 

 

 

 

' . . NHZ. . . ,, . ‘
NH2 ? NH2
H
EH H
Figure 1.15. Schematic diagram of the reduction of 4-nitrophenol by NaBH4 during ﬂow

through a membrane modiﬁed with nanoparticle/polyelectrolyte ﬁlms. (Reproduced with

permission from Langmuir, 2008, 24, 7663-7673.)

As discussed in section 1.2.3.2, the LBL method can be used to form PEM
capsules. Shchukin and Mohwald prepared PSS/PAH capsules containing TiO2
nanoparticles that catalyze the photosynthesis of urea from inorganic precursors (CO2 and
N03") in aqueous solution.167 Poly(vinyl alcohol), which served as an electron donor to
facilitate the reaction, was encapsulated inside the capsules together with TiO2
nanoparticles. The effect of the capsule size was investigated and the highest yield in

urea photosynthesis was achieved by Cu-modiﬁed T102 nanoparticles encapsulated inside

39

2.2-um PSS/PAH capsules.

1.4. Scope of This Work

This dissertation focuses on the use of LBL adsorption for preparation of NF
membranes tailored for speciﬁc NF separations and for formation of catalytic membranes.
Chapter 2 investigates a series of polyelectrolyte systems for cation separations and
shows that of the systems examined, PSS/PAH ﬁhns provide the highest Mg2+ rejection
and NaVMng selectivity. This chapter also studies the effect of deposition parameters
(number of bilayers, pH and ionic strength of the polyelectrolyte deposition solution) on
selectivity and ﬂux. Streaming potential measurements demonstrate that the magnitude
of positive surface charge increases with increasing concentrations of Mg2+ in feed
solutions or when the outer polycation layer is deposited from a solution of high ionic
strength. These membranes are potentially useful in water softening and wastewater
treatment in the pulp and paper industry.

Chapter 3 describes the use of (PSS/PDADMAC),.PSS ﬁlms for removal of
phosphate in the presence of chloride. The phosphate rejection depends on the pH of the
feed solution, and the highest rejection occurs when the feed pH is 8.4, as opposed to pH
5.6 or 7.0, presumably due to the fact that a large fraction of phosphate is HPO42' rather
than H2P04' at the higher pH. At a feed pH of 8.4, (PSS/PDADMAC)4PSS ﬁlms on
porous alumina supports exhibit a phosphate rejection of 98%, a chloride/phosphate
selectivity of 48 and a solution ﬂux of 2.4 m3/(m2 day) at 4.8 bar. The low chloride
rejection and high solution ﬂux make these membranes attractive for chloride/phosphate

separations.

4O

In chapters 4 and 5, I present the formation of catalytic membrane reactors by
modifying the interior pores of porous hollow ﬁber or ﬂat-sheet membranes through
alternating adsorption of polyelectrolytes and metal nanoparticles, which serve as
catalysts in chemical reactions. The LBL method affords control over the amount of
catalyst deposited through varying the number of bilayers in the ﬁlm. Moreover, the
deposition yields a high density of metal nanoparticles in the ﬁlm, and scanning electron
microscopy (SEM) images show that the colloids are well separated with no visible
aggregation.

Chapter 4 focuses primarily on catalytic hollow ﬁber membranes. Immobilization
of metal nanoparticles in hollow ﬁber pores via LBL adsorption of polyelectrolytes and
negatively charged Au nanoparticles yields catalytic reactors with high surface areas.
Catalytic reduction of 4-nitrophenol with NaBI-I4, which can be easily monitored by UV-
Vis spectrophotometry, demonstrates that the Au nanoparticles in the hollow ﬁber
membrane are highly catalytically active. In a single pass through the membrane, >99%
of the 4-nitrophenol is reduced to 4-aminophenol, but this conversion decreases over time,
perhaps because of catalyst fouling by byproducts of 4-amin0phenol oxidation.

In chapter 5, I describe catalytic ﬂat-sheet membrane reactors that contain Pd
nanoparticles immobilized in the membrane pores. Films of PSS/PEI/PdNP have a
higher Pd loading than PSS/PAH/PdNP coatings, presumably because of strong
interactions between PEI and Pd. These membranes catalyze 4-nitrophenol reduction by
HCOONa. High (>99%) 4-nitrophenol conversion occurs, but deposits appear on the
membrane surface, perhaps again due to 4-aminophenol oxidation.

The last chapter presents conclusions and some proposed future work, particularly

41

with regard to stabilizing membranes and facilitating their manufacture.

42

1.5. References
(1) Davis, F .; Higson, S. P. J. Biosens. Bioelectron. 2005, 21, 1-20.

(2) Bruening, M. L.; Dotzauer, D. M.; Jain, P.; Ouyang, L.; Baker, G. L. Langmuir
2008, 24, 7663-7673.

(3) Ariga, K.; Hill, J. P.; Ji, Q. Phys. Chem. Chem. Phys. 2007, 9, 2319-2340.

(4) Seon, J .-B.; Lee, S.; Kim, J. M.; Jeong, H.-D. Chem. Mater. 2009, 21 , 604-611.
(5) Yin, L.-T.; Hu, C.-Y.; Chang, C.-H. Sens. Actuators, B 2008, 130, 374-378.

(6) Darren, S.; Anthony, H.; Robert, S. J. Appl. Polym. Sci. 2007, 103, 1287-1294.
(7) Park, M. S.; Lee, Y.; Kim, J. K. Chem. Mater. 2005, 1 7, 3944-3950.

(8) Hoggan, E. N.; Flowers, D.; Wang, K.; DeSimone, J. M.; Carbonell, R. G. Ind.
Eng. Chem. Res. 2004, 43, 2113-2122.

(9) Muller-Buschbaum, P.; Stamm, M. Macromolecules 1998, 31, 3686-3692.

(10) Kumar, N.; Yanguas-Gil, A.; Daly, S. R.; Girolami, G. S.; Abelson, J. R. J. Am.
Chem. Soc. 2008, 130, 17660-17661.

(11) Binions, R.; Carrnalt, C. J .; Parkin, I. P.; Pratt, K. F. E.; Shaw, G. A. Chem. Mater.
2004, 16, 2489-2493.

(12) Montero, L.; Baxamusa, S. H.; Borros, S.; Gleason, K. K. Chem. Mater. 2008, 21,
399-403.

(13) Manning, T. D.; Parkin, I. P.; Pemble, M. E.; Sheel, D.; Vemardou, D. Chem.
Mater. 2004, 16, 744-749.

(14) Berry, A. D.; Holm, R. T.; Mowery, R. L.; Turner, N. H.; Fatemi, M. Chem.
Mater. 2002, 3, 72-77.

(15) Shirrnan, T.; Freeman, D.; Posner, Y. D.; Feldman, 1.; Facchetti, A.; van der
Boom, M. E. J. Am. Chem. Soc. 2008, 130, 8162-8163.

(16) Li, L.; Yang, Y.; Yang, G.; Chen, X.; Hsiao, B. S.; Chu, B.; Spanier, J. E.; Li, C.
Y. Nano Lett. 2006, 6, 1007-1012.

(17) Rietveld, I. B.; Kobayashi, K.; Yamada, H.; Matsushige, K. J. Phys. Chem. B
2006, 110, 23351-23364.

43

(18) NcNeal, C. J.; Macfarlane, R. D.; Thurston, E. L. Anal. Chem. 1979, 51, 2036-
2039.

(19) Rajeshwar, K.; de Tacconi, N. R.; Chenthamarakshan, C. R. Chem. Mater. 2001,
13, 2765-2782.

(20) Bagheri-Tar, F.; Sahimi, M.; Tsotsis, T. T. Ind. Eng. Chem. Res. 2007, 46, 3348-
3357.

(21) Tatoulian, M.; Areﬁ-Khonsari, F .; Tatoulian, L.; Amouroux, J .; Borra, J. P. Chem.
Mater. 2006, 18, 5860-5863.

(22) Armelao, L.; Bottaro, G.; Bruno, G.; Losurdo, M.; Pascolini, M.; Soini, E.;
Tondello, E. J. Phys. Chem. C2008, 112, 14508-14512.

(23) Haruvy, Y.; Webber, S. E. Chem. Mater. 1992, 4, 89-94.

(24) F ireman-Shoresh, S.; Popov, 1.; Avnir, D.; Marx, S. J. Am. Chem. Soc. 2005, 127,
2650-2655.

(25) Caruso, R. A.; Antonietti, M. Chem. Mater. 2001, 13, 3272-3282.

(26) Xiao, L.; Zhang, H.; Scanlon, E.; Rarnanathan, L. S.; Choc, E.-W.; Rogers, D.;
Apple, T.; Benicewicz, B. C. Chem. Mater. 2005, 1 7, 5328-5333.

(27) Sreenivas, G.; Ang, S. S.; Fritsch, I.; Brown, W. D.; Gerhardt, G. A.; Woodward,
D. J. Anal. Chem. 1996, 68, 1858-1864.

(28) Yang, G. H.; Zhang, Y.; Kang, E. T.; Neoh, K. G. J. Phys. Chem. B 2003, 107,
2780-2787.

(29) Biederrnan, H.; Zeuner, M.; Zalman, J .; Bilkova, P.; Slayinska, D.; Stelmasuk, V.;
Boldyreva, A. Thin Solid Films 2001, 392, 208-213.

(30) Lu, Z. 0.; Lo, M. F.; Chung, C. Y. J. Phys. Chem. C 2008, 112, 7069-7078.

(31) Patel, N.; Femandes, R.; Guella, G.; Kale, A.; Miotello, A.; Patton, B.; Zanchetta,
C. J. Phys. Chem. C 2008, 112, 6968-6976.

(32) Kim, l.-D.; Rothschild, A.; Hyodo, T.; Tuller, H. L. Nano Lett. 2006, 6, 193-198.
(33) Nam, H.-J.; Sasaki, T.; Koshizaki, N. J. Phys. Chem. B 2006, 110, 23081-23084.
(34) Tsai, P.-S.; Yang, Y.-M.; Lee, Y.-L. Langmuir 2006, 22, 5660-5665.

(35) Langmuir, I. J. Am. Chem. Soc. 1917, 39, 1848-1906.

44

(36) McCullough, D. H. 1.; Regen, S. L. Chem. Commun. 2004, 2787-2791.
(37) Blodgett, K. B. J. Am. Chem. Soc. 1935, 57, 1007-1022.
(38) Talham, D. R. Chem. Rev. 2004, 104, 5479-5501.

(39) Clemente-Leon, M.; Coronado, E.; Soriano-Portillo, A.; Mingotaud, C.;
Dominguez-Vera, J. M. Adv. Colloid Interface Sci. 2005, 116, 193-203.

(40) Ariga, K.; Nakanishi, T.; Michinobu, T. J. Nanosci. Nanotechnol. 2006, 6, 2278-
2301.

(41) Talham, D. R.; Yarnarnoto, T.; Meisel, M. W. J. Phys. .° Condens. Matter 2008, 20,
184006 (13pp).

(42) Major, S. S.; Talwar, S. S.; Srinivasa, R. S. Pramana 2006, 67, 121-134.

(43) Crespo-Biel, O.; Dordi, B.; Reinhoudt, D. N.; Huskens, J. J. Am. Chem. Soc. 2005,
12 7, 7594-7600.

(44) Jaiswal, A.; Rajagopal, D.; Lakshmikantham, M. V.; Cava, M. P.; Metzger, R. M.
Phys. Chem. Chem. Phys. 2007, 9, 4007-4017.

(45) Fendler, J. H. Chem. Mater. 2001, 13, 3196-3210.

(46) Arya, S. K.; Solanki, P. R.; Datta, M.; Malhotra, B. D. Biosens. Bioelectron. 2009,
24, 2810-2817.

(47) Kondo, T.; Uosaki, K. J. Photochem. Photobiol., C 2007, 8, 1-17.

(48) Bigelow, W. C.; Pickett, D. L.; Zisman, W. A. J. Colloid Sci. 1946, I, 513-538.
(49) Shafrin, E. G.; Zisman, W. A. J. Colloid Sci. 1949, 4, 571-590.

(50) Allara, D. L.; Nuzzo, R. G. Langmuir 1985, 1, 45-52.

(51) Pomerantz, M.; Segmiiller, A.; Netzer, L.; Sagiv, J. Thin Solid Films 1985, 132,
153-162.

(52) Sagiv, J. J. Am. Chem. Soc. 1980, 102, 92-98.

(53) . Nuzzo, R. G.; Zegarski, B. R.; Dubois, L. H. J. Am. Chem. Soc. 1987, 109, 733-
740.

(54) Bain, C. D.; Whitesides, G. M. J. Am. Chem. Soc. 1988, 110, 3665-3666.

45

(55) Bain, C. D.; Troughton, E. B.; Tao, Y. T.; Evall, J.; Whitesides, G. M.; Nuzzo, R.
G. J. Am. Chem. Soc. 1989, 111, 321-335.

(56) Poirier, G. E.; Pylant, E. D. Science 1996, 272, 1145-1148.

(57) Kang, J. F.; Ulman, A.; Liao, S.; Jordan, R.; Yang, G.; Liu, G.-Y. Langmuir 2001,
17, 95-106.

(58) Laibinis, P. E.; Whitesides, G. M.; Allara, D. L.; Tao, Y.-T.; Parikh, A. N.; Nuzzo,
R. G. J. Am. Chem. Soc. 1991, 113, 7152-7167.

(59) Walczak, M. M.; Chung, C.; Stole, S. M.; Widrig, C. A.; Porter, M. D. J. Am.
Chem. Soc. 1991, 113, 2370-2378.

(60) Love, J. C.; Wolfe, D. B.; Haasch, R.; Chabinyc, M. L.; Paul, K. E.; Whitesides,
G. M.; Nuzzo, R. G. J. Am. Chem. Soc. 2003, 125, 2597-2609.

(61) Carvalho, A.; Geissler, M.; Schmid, H.; Michel, B.; Delamarche, E. Langmuir
2002, 18, 2406-2412.

(62) Li, Z.; Chang, S.-C.; Williams, R. S. Langmuir 2003, 19, 6744-6749.

(63) Love, J. C.; Estroff, L. A.; Kriebel, J. K.; Nuzzo, R. G.; Whitesides, G. M. Chem.
Rev. 2005, 105, 1103-1170.

(64) Ulman, A. Chem. Rev. 1996, 96, 1533-1554.
(65) Genzer, J.; Eﬁmenko, K. Science 2000, 290, 2130-2133.

(66) Rangger, G. M.; Romaner, L.; Heimel, G.; Zojer, E. Surf Interface Anal. 2008, 40,
371-378.

(67) Jennings, G. K.; Laibinis, P. E. Colloids Surf, A 1996, 116, 105-114.

(68) Ulman, A.; Evans, S. D.; Shnidman, Y.; Sharma, R.; Eilers, J. E.; Chang, J. C. J.
Am. Chem. Soc. 1991, 113, 1499-1506.

(69) Schreiber, F. Prog. Surf Sci. 2000, 65, 151-257.
(70) Bent, S. F. ACS Nano 2007, 1, 10-12.
(71) Decher, G.; Hong, J. D.; Schmitt, J. Thin Solid Films 1992, 210-211, 831-835.

(72) Decher, G. Science 1997, 277, 1232-1237.

46

(73) Lvov, Y.; Decher, G.; Moehwald, H. Langmuir 1993, 9, 481-486.
(74) Decher, G.; Lvov, Y.; Schmitt, J. Thin Solid Films 1994, 244, 772-777.

(75) Hoogeveen, N. G.; Cohen Stuart, M. A.; Fleer, G. J.; Bdhmer, M. R. Langmuir
1996, 12, 3675-3681.

(76) Chen, W.; McCarthy, T. J. Macromolecules 1997, 30, 78-86.

(77) Lvov, Y.; Ariga, K.; Ichinose, 1.; Kunitake, T. J. Am. Chem. Soc. 1995, 117,
6117-6123.

(78) Sultana, N.; Schenkman, J. B.; Rusling, J. F. J. Am. Chem. Soc. 2005, 127, 13460-
13461.

(79) Beissenhirtz, M. K.; Scheller, F. W.; Lisdat, F. Anal. Chem. 2004, 76, 4665-4671.

(80) Caruso, F.; Furlong, D. N.; Ariga, K.; Ichinose, 1.; Kunitake, T. Langmuir 1998,
14, 4559-4565.

(81) Lvov, Y.; Ariga, K.; Onda, M.; Ichinose, 1.; Kunitake, T. Langmuir 1997, 13,
6195-6203.

(82) Cassagneau, T.; Mallouk, T. E.; Fendler, J. H. J. Am. Chem. Soc. 1998, 120,
7848-7859.

(83) Ostrander, J. W.; Mamedov, A. A.; Kotov, N. A. J. Am. Chem. Soc. 2001, 123,
1101-1110.

(84) Tian, S.; Liu, J .; Zhu, T.; Knoll, W. Chem. Mater. 2004, 16, 4103-4108.

(85) Johnston, A. P. R.; Mitomo, H.; Read, E. S.; Caruso, F. Langmuir 2006, 22, 3251-
3258.

(86) Chen, X.; Lang, J .; Liu, M. Thin Solid Films 2002, 409, 227-232.
(87) Lvov, Y.; Decher, G.; Sukhorukov, G. Macromolecules 1993, 26, 5396-5399.
(88) Ariga, K.; Lvov, Y.; Kunitake, T. J. Am. Chem. Soc. 1997, 119, 2224-2231.

(89) Tedeschi, C.; Caruso, F .; Mohwald, H.; Kirstein, S. J. Am. Chem. Soc. 2000, 122,
5841-5848.

(90) Bergbreiter, D. E.; Liao, K.-S. Soft Matter 2009, 5, 23-28.

(91) Bemdt, P.; Kurihara, K.; Kunitake, T. Langmuir 1992, 8, 2486-2490.

47

(92) Zhang, X.; Chen, H.; Zhang, H. Chem. Commun. 2007, 1395-1405.

(93) Quinn, J. F.; Johnston, A. P. R.; Such, G. K.; Zelikin, A. N.; Caruso, F. Chem.
Soc. Rev. 2007, 36, 707-718.

(94) Stockton, W. B.; Rubner, M. F. Macromolecules 1997, 30, 2717-2725.

(95) Wang, L.; Wang, Z.; Zhang, X.; Shen, J.; Chi, L.; Fuchs, H. Macromol. Rapid
Commun. 1997, 18, 509-514.

(96) Sukhishvili, S. A.; Granick, S. J. Am. Chem. Soc. 2000, 122, 9550-9551.
(97) Sukhishvili, S. A.; Granick, S. Macromolecules 2002, 35, 301-310.

(98) Liu, Y.; Bruening, M. L.; Bergbreiter, D. E.; Crooks, R. M. Angew. Chem. Int. Ed.
Engl. 1997, 36, 2114-2116.

(99) Liu, Y.; Zhao, M.; Bergbreiter, D. E.; Crooks, R. M. J Am. Chem. Soc. 1997, 119,
8720-8721.

(100) Kohli, P.; Blanchard, G. J. Langmuir 2000, 16, 4655-4661.
(101) Kohli, P.; Blanchard, G. J. Langmuir 2000, 16, 8518-8524.

(102) Such, G. K.; Quinn, J. F.; Quinn, A.; Tjipto, E.; Caruso, F. J. Am. Chem. Soc.
2006, 128, 9318-9319. '

(103) Kolb, H. C.; Finn, M. G.; Sharpless, K. B. Angew. Chem. Int. Ed. 2001, 40, 2004-
2021.

(104) Bock, V. D.; Hiemstra, H.; van Maarseveen, J. H. Eur. J. Org. Chem. 2006, 51-68.

(105) Such, G. K.; Tjipto, E.; Postma, A.; Johnston, A. P. R.; Caruso, F. Nano Lett.
2007, 7, 1706-1710.

(106) Kotov, N. A. Nanostruct. Mater. 1999, 12, 789-796.
(107) Anzai, J .-i.; Kobayashi, Y. Langmuir 2000, 16, 2851-2856.
(108) Inoue, H.; Anzai, J .-i. Langmuir 2005, 21 , 8354-8359.

(109) Shimazaki, Y.; Mitsuishi, M.; Ito, S.; Yarnamoto, M. Langmuir 1997, 13, 1385-
1387.

48

(110) Wang, X.; Naka, K.; Itoh, H.; Uemura, T.; Chujo, Y. Macromolecules 2003, 36,
533-535.

(111) Shimazaki, Y.; Nakamura, R.; Ito, S.; Yamarnoto, M. Langmuir 2001, 17, 953-
956.

(112) Tang, T.; Qu, J .; Mullen, K.; Webber, S. E. Langmuir 2006, 22, 26-28.

(113) Sukhorukov, G. B.; Donath, E.; Lichtenfeld, H.; Knippel, E.; Knippel, M.; Budde,
A.; Mohwald, H. Colloids Surf, A 1998, 13 7, 253-266.

(114) Caruso, F. Adv. Mater. 2001, 13, 11-22.

(115) Kidambi, S.; Dai, J.; Li, J.; Bruening, M. L. J. Am. Chem. Soc. 2004, 126, 2658-
2659.

(116) Kidambi, S.; Bruening, M. L. Chem. Mater. 2005, 17, 301-307.
(117) Bhattacharjee, S.; Bruening, M. L. Langmuir 2008, 24, 2916-2920.

(118) Bhattacharjee, S.; Dotzauer, D. M.; Bruening, M. L. J. Am. Chem. Soc. 2009, 131,
3601-3610.

(119) Dotzauer, D. M.; Dai, J .; Sun, L.; Bruening, M. L. Nano Lett. 2006, 6, 2268-2272.

(120) Dotzauer, D. M.; Bhattacharjee, S.; Wen, Y.; Bruening, M. L. Langmuir 2009, 25,
1865-1871.

(121) Caruso, F .; Caruso, R. A.; Mohwald, H. Science 1998, 282, 1111-1114.

(1122) Mayya, K. S.; Gittins, D. 1.; Dibaj, A. M.; Caruso, F. Nano Lett. 2001, 1, 727-730.
(123) Caruso, R. A.; Susha, A.; Caruso, F. Chem. Mater. 2001, 13, 400-409.

(124) Caruso, F.; Caruso, R. A.; Mohwald, H. Chem. Mater. 1999, 11, 3309-3314.

(125) Lvov, Y.; Antipov, A. A.; Mamedov, A.; Mohwald, H.; Sukhorukov, G. B. Nano
Lett. 2001, 1, 125-128.

(126) Schneider, G.; Decher, G. Nano Lett. 2004, 4, 1833-1839.
(127) Liang, Z.; Susha, A.; Caruso, F. Chem. Mater. 2003, 15, 3176-3183.
(128) Tian, Y.; He, Q.; Cui, Y.; Li, J. Biomacromolecules 2006, 7, 2539-2542.

(129) Liang, Z.; Susha, A. S.; Yu, A.; Caruso, F. Adv. Mater. 2003, 15, 1849-1853.

49

(130) Hou, S.; Wang, J.; Martin, C. R. Nano Lett. 2005, 5, 231-234.

(131) Bruening, M.; Dotzauer, D. Nat. Mater. 2009, 8, 449-450.

(132) Shiratori, S. S.; Yamada, M. Polym. Adv. Technol. 2000, 11, 810-814.

(133) Shiratori, S. S.; Ito, T.; Yamada, T. Colloids Surf, A 2002, 198-200, 415-423.

(134) Lee, S.-S.; Hong, J.-D.; Kim, C. H.; Kim, K.; Koo, J. R; Lee, K.-B.
Macromolecules 2001, 34, 5358-5360.

(135) Cho, J.; Char, K.; Hong, J. D.; Lee, K.-B. Adv. Mater. 2001, 13, 1076-1078.

(136) Ladarn, G.; Schaad, P.; Voegel, J. C.; Schaaf, P.; Decher, G.; Cuisinier, F.
Langmuir 1999, 16, 1249-1255.

(137) Chiarelli, P. A.; Johal, M. S.; Holmes, D. J .; Casson, J. L.; Robinson, J. M.; Wang,
H.-L. Langmuir 2002, 18, 168-173.

(138) Campbell, V. E.; Chiarelli, P. A.; Kaur, S.; Johal, M. S. Chem. Mater. 2005, 17,
186-190.

(139) Lefaux, C. J .; Zimberlin, J. A.; Dobrynin, A. V.; Mather, P. T. J. Polym. Sci, Part
B: Polym. Phys. 2004, 42, 3654-3666.

(140) Schlenoff, J. B.; Dubas, S. T.; F arhat, T. Langmuir 2000, 16, 9968-9969.

(141) Izquierdo, A.; Ono, S. S.; Voegel, J. C.; Schaaf, P.; Decher, G. Langmuir 2005,
21 , 7558-7567.

(142) Lu, C.; Donch, 1.; Nolte, M.; F ery, A. Chem. Mater. 2006, 18, 6204-6210.

(143) Krogman, K. C.; Zacharia, N. S.; Schroeder, S.; Hammond, P. T. Langmuir 2007,
23, 3137-3141.

(144) Jiang, S.; Liu, M. Chem. Mater. 2004, 16, 3985-3987.

(145) Cooper, T. M.; Campbell, A. L.; Crane, R. L. Langmuir 1995, 11, 2713-2718.
(146) Caruso, F.; Mohwald, H. J. Am. Chem. Soc. 1999, 121, 6039-6046.

(147) Dubas, S. T.; Schlenoff, J. B. Macromolecules 1999, 32, 8153-8160.

(148) Deng, H.-Y.; Xu, Y.-Y.; Zhu, B.-K.; Wei, X.-Z.; Liu, F.; Cui, Z.-Y. J. Membr. Sci.
2008, 323, 125-133.

50

(149) Hilal, N.; Al-Zoubi, H.; Darvvish, N. A.; Mohamma, A. W.; Abu Arabi, M.
Desalination 2004, 1 70, 281-308.

(150) Liu, X.; Bruening, M. L. Chem. Mater. 2004, 16, 351-357.
(151) Miller, M. D.; Bruening, M. L. Langmuir 2004, 20, 11545-11551.

(152) Stanton, B. W.; Harris, J. J.; Miller, M. D.; Bruening, M. L. Langmuir 2003, 19,
7038-7042.

(153) Jin, W.; Toutianoush, A.; Tieke, B. Langmuir 2003, 19, 2550-2553.

(154) Ouyang, L.; Malaisamy, R.; Bruening, M. L. J. Membr. Sci. 2008, 310, 76-84.
(155) El-Hashani, A.; Toutianoush, A.; Tieke, B. J. Membr. Sci. 2008, 318, 65-70.
(156) Hong, S. U.; Malaisamy, R.; Bruening, M. L. J. Membr. Sci. 2006, 283, 366—372.

(157) Hong, S. U.; Miller, M. D.; Bruening, M. L. Ind. Eng. Chem. Res. 2006, 45, 6284-
6288.

( l 58) Hong, S. U.; Malaisamy, R.; Bruening, M. L. Langmuir 2007, 23, 1716-1722.
(1 59) Hong, S. U.; Ouyang, L.; Bruening, M. L. J. Membr. Sci. 2009, 32 7, 2-5.
(1 60) Levasalmi, J .-M.; McCarthy, T. J. Macromolecules 1997, 30, 1752-1757.

;1 61) van Ackem, F.; Krasemann, L.; Tieke, B. Thin Solid Films 1998, 327-329, 762-
66.

(1 62) Sullivan, D. M.; Bruening, M. L. Chem. Mater. 2003, 15, 281-287.
(I 63) Dai, J.; Baker, G. L.; Bruening, M. L. Anal. Chem. 2006, 78, 135-140.
(1 64) Salloum, D. 3.; Schlenoff, J. B. Biomacromolecules 2004, 5, 1089-1096.
(1 65) Wang, Y.; Caruso, F. Chem. Mater. 2006, 18, 4089-4100.

(51066) Onda, M'; Lvov, Yd Ariga, K4 Kunitake, T. J. Ferment. Bioeng. 1996, 82, 502-
6 -

(1 6‘7) Shchukin, D. G.; Mohwald, H. Langmuir 2005, 21, 5582-5587.

51

Chapter 2
Polyelectrolyte Multilayer Films as Nanoﬁltration Membranes for

Cation Separations

This work was recently published in Journal of Membrane Science.1

2.1. Introduction
Hard water is common throughout the world, and the United States Geological

Survey reports that 85% of US homes have hard water that usually contains high levels of

2

Ca W or Mng.2 Such water is not harmful, but it causes undesirable effects such as

clogging of water faucets and deterioration of equipment, which is a costly problem in
both residential and industrial settings.3 The conventional water softening method, which
is still popular around the world, is ion exchange using cation exchange resins in their
sodium form.4 However, the resin must be regenerated periodically in a process that
consumes time, salt, and energy. Membrane-based separations provide an alternative to
conventional treatments of hard water, and nanoﬁltration (NF) is probably the most
promising membrane-based technology for alleviating water hardness because of its

relatively low energy costs.“'7 Unlike reverse osmosis (RO), NF allows the passage of
small monovalent ions such as Na+ whose rejection is not necessary for softening

water.4’8’9 Passage of these monovalent ions reduces the osmotic pressure that must be
overcome for water softening, and the high permeability of NF membranes (relative to

RO membranes) also leads to low-pressure separations.7 Hence, NF is attractive for

52

removal of multivalent ions in water softening and other applications including
puriﬁcation of wastewater in the pulp and paper industry, where environmental
legislation requires the removal of elements such as Ca, Fe or Mg.lo
Several research groups performed fundamental studies to understand the
mechanisms behind multivalent ion removal in NF and showed that separation of charged
solutes is based on both size and Donnan exclusion.4'5 ,11 Thus, cation separations require
a positively charged membrane with a relatively small effective pore size. Unfortunately,
most commercial NF membranes are negatively charged}1 "'3 although several
experimental studies show that positively charged membranes can be fabricated.”16
This chapter examines the properties of a series of positively charged membranes
prepared by layer-by-layer adsorption of polyanions and polycations. The layer-by-layer
deposition method was developed in the 19903,”18 and a number of studies examined its
utility for creating pervaporation, NF, and gas-separation membranes.”23 This method is
attractive for preparing the ultrathin skin of composite membranes because ﬁlm thickness
can be controlled at the nanometer scale to allow for high ﬂux, and ﬁlm properties can be
optimized by varying polyelectrolytes or deposition conditions.24'29 Additionally, the
Charge near the surface of the ﬁlm can be either positive or negative depending on
Whether ﬁlms are terminated with a polycation or a polyanion.
Although a number of papers examined the use of polyelectrolyte multilayers

(FEMS) in anion separations,20’30'34 very few studies investigated cation separations.

Ti'i‘ke’s group reported that PEMs comprised of 60 layer pairs of poly(vinylarnine) and
P0137 (vinylsulfate) exhibit a 95% rejection of MgCl2 at an operating pressure of 10 bar,”

but the use of such a large number of layers makes membrane fabrication time consuming

53

and yields very low ﬂux. We studied NF of MgSO4 with much thinner PEMs that allow

high ﬂux,35 but ultrathin, highly permeable PEMs have not been examined for the

separation of monovalent and divalent cations, which is important in water softening.

This work focuses on NF of mixed-salt solutions containing NaW and Mg2+ or NaW

and Ca2+, the three typical components of hard water. Notably, four- or ﬁve-bilayer
poly(styrenesulfonate)/poly(allylamine hydrochloride) (PSS/PAH) ﬁlms deposited on
ultraﬁltration membranes or porous alumina substrates provide ~95% rejection of Mg2+
and a NaW/Mg2+ selectivity of around 20. Ca2+ rejections are similar to those of Mng.

Variation of polyelectrolytes and deposition conditions allows optimization of cation
rejections by controlling ﬁlm properties such as ionic crosslink density,33 zeta potential,36
and swelling.” Streaming potential investigations demonstrate that the use of higher salt

concentrations in polycation deposition solutions leads to higher charge near the ﬁlm

surface and, hence, higher Mng rejections. Moreover, C potentials obtained using
different concentrations of KCl and MgCl2 suggest that adsorption of Mg2+ on the

membrane surface leads to high Mg2+ rejections even at high MgCl2

concentrations. '2’38’39

2.2. Experimental Section

2.2.1. Materials

Poly(sodium 4-styrenesulfonate) (PSS, MW = 70,000 Da), poly(allylamine
hydrochloride) (PAH, MW = 70,000 Da), and poly(diallyldimethylammonium chloride)

(PDADMAC, MW = 100,000-200,000 Da, 20 wt% in water) were obtained from Sigma-

54

Aldrich and used without further puriﬁcation. NaCl and NaBr (Jade Scientiﬁc), MnCl2
(Alfa Aesar), MgCl2-6H2O (Columbus Chemical), and CaCl2-2H2O (Sigrna-Aldrich)

were also used as received. Deionized water (Milli-Q puriﬁcation system, 18.2 MQ-cm)
was employed for rinsing membranes and preparing polyelectrolyte and feed solutions.
The pH of polyelectrolyte solutions was adjusted with dilute HCl or NaOH.

Two types of substrates were used in the experiments: porous alumina supports
(0.02-um Whatrnan Anodisk ﬁlters, average membrane thickness of 60 um) and
polyethersulfone (PES) ultraﬁltration membranes (Millipore, cat. number PBQK 02510,
molecular weight cutoff of 50 000 Da, effective pore size of 4-5 nm, membrane thickness
of 100-150 urn). Prior to ﬁlm deposition, the porous alumina supports were UV/O3
cleaned with the feed side up (Boekel UV-Clean Model 135500) for 15 min. To remove
the wetting agent on the surface of the PES ultraﬁltration membranes, these materials
were soaked in deionized water for about one hour, during which the water was replaced
two or three times.

2.2.2. Film Deposition

The pretreated substrate was loaded in a home-made membrane holder that
exposes only the feed side of the membrane to the polyelectrolyte solutions. Deposition
started with immersion of the membrane surface in 0.02 M PSS. (Polyelectrolyte
concentrations are always given with respect to the repeating unit.) PSS was deposited
ﬁrst clue to potential electrostatic attractions with the positively charged alumina support
or hydrophobic interactions with PES.3 "40 The substrate was then rinsed with deionized
water for 1 min before immersion in a polycation solution (PAH or PDADMAC)

followed by another l-min rinse. Additional layers were deposited similarly to give the

55

desired number of bilayers. The time allowed for polyelectrolyte adsorption varied
according to the type of multilayer deposited. For PSS/PAH, the substrate was dipped for
2 min in PSS solutions and 5 min in PAH solutions, while for PSS/PDADMAC, the
deposition time was 3 min for both polyelectrolytes. Different deposition conditions
(type and concentration of supporting electrolyte and pH value) were chosen based on

35

prior work with anion separations. Figure 2.1 shows the structures of the

polyelectrolytes investigated in this study.

[I
6 ~ .. ..
Ni-lgcr N’Cl'
SO‘WNW P 1 (dW/IIIIW th I
Pol all lamine 0y '3 Y line Y
Poly(styrenesulfonatel “$26,220,“, ammonium chloride)
PSS PAH PDADMAC

Figure 2.1. Structures of the polyelectrolytes used in this study.

2.2.3. Characterization of Membranes
2.2.3.1. Film Thickness Estimations with SEM

Cross-sectional images of porous alumina supports coated with PEMs were
obtained with a Hitachi $4700 11 ﬁeld-emission scanning electron microscope (FESEM).
Membranes were fractured immediately after exposure to liquid nitrogen and the top,
bottom, and cross-sectional surfaces of the membrane were subsequently sputter-coated
(Pelco model SC-7 auto sputter coater) with 5 nm of gold before imaging.

2.2.3.2. Zeta Potential measurements

56

An asymmetric clamping cell (Anton Paar, Graz, Austria) attached to a streaming
potential analyzer (BI-EKA, Brookhaven Instruments, Holtsville, NY) was employed to
measure the C potentials of surfaces coated with PEMs. The test sample was placed
against a grooved poly(methyl methacrylate) (PMMA) spacer, and the streaming
potential was measured between two Ag/AgCl electrodes on opposite ends of the

clamping cell. Solutions of KCl (1 mM, 5 mM, or 10 mM) or MgCl2 (0.333 mM, 1.666

mM, and 3.33 mM) served as the electrolytes for the measurements. Polyelectrolyte
multilayer ﬁlms on PBS membranes were prepared in the same way as for transport
studies, but different PES substrates (50 kDa membranes from Pall, Port Washington, NY)

were used due to the large surface area requirement of the clamping cell. The i; potentials

of the PEMs were calculated using equation 1, where CT,“ is the C potential of the
membrane, CAvg is the effective 1; potential measured with the membrane pressed against

the spacer, and Cspace, is the C potential of the PMMA spacer surface.4|

(Test : 24A vg _ (Spacer (1)

To ﬁnd Cspacer, or CpMMA, the C potential of PMMA clamped to the PMMA spacer was

determined prior to the streaming potential measurement of the membranes coated with
PEM ﬁlms. The reported values are the average of t; potentials determined in two ﬂow
directions with at least two membranes.
2.2.4. Nanoﬁltration Studies

NF experiments were performed at room temperature (23 °C) using a customized

cross-ﬂow system described previously.35 The system was pressurized with Ar to 4.8 bar,

57

and the feed tank was ﬁlled with a 2 L solution containing NaCl, MgCl2, or CaCl2, or a
combination of these salts. Approximate osmotic pressures of lg/L NaW (from NaCl),

Mg2+ (from MgC12), and Ca2+ (from CaC12) solutions are 2.1 bar, 3.0 bar, and 1.8 bar,

respectively, assuming 100% rejection. A stainless steel frit (Mott Corp., Farmington,
CT) was used as a preﬁlter to remove rust or particulate matter from the feed solution.
The ﬂow meter, which was located between the pump and the membrane cells, was set to
18 mL/min, which is high enough to minimize concentration polarization.35 The
membrane was housed in an O-ring cell that exposes an active surface area of 1.7 cm2.
After an 18 h-stabilization period, permeate samples were collected every 30 min for 2 h,
and the reported ﬂuxes are the steady-state permeate ﬂuxes after 18 h of ﬁltration. The
feed solution was sampled at the end of the ﬁltration to take into account any
concentration changes during the experiment. Cation concentrations were determined by
atomic absorption spectroscopy (AAS) using a Varian Spectra Atomic Absorption-200
instrument, and the solute rejection, R, was calculated using equation (2), where CI, and
Cf are solute concentrations in the permeate and feed, respectively. Rejection and ﬂux
values are averages from at least two membranes, and :1: values represent standard

deviations.

R=(1—Cp/Cf)x100% (2)

Equation (3) deﬁnes the transport selectivity of solute A over solute B, a 1: , which can be

simply expressed in terms of rejections.

58

A C... ’03,, =100—8..(%)

a :4 -- -
3 CH C, IOO—RB(%) (3)

 

 

P

2.3. Results and Discussion

Table 2.1 lists the diffusion coefﬁcients and Stokes’ radii of hydrated NaW,
MgZWand Ca2+ ions. (Stokes’ radii are determined from diffusion coefﬁcients and are
similar to hydrated radii.) The Stokes’ radius of Mng is almost twice that of NaW, which
is reﬂected by the smaller diffusivity for Mng. Although the Stokes’ radius of Ca2+ is
smaller than that of MgZW, it is still 0.125 nm larger than that of NaW. Given these radii,

size-selective nanoﬁltration of hydrated Mng or Ca2+ in the presence of NaW will require

membranes with pore sizes on the sub-nm scale. Previous studies of neutral molecule
transport suggest that the effective pore diameters of PSS/PAH42 and PSS/PDADMAC33
membranes are about 0.4-0.5 nm and 0.7-1.0 nm, respectively. Thus, we investigated

these polyelectrolyte systems for selective removal of MgZW or Ca2W in the presence of

+

Na.

Table 2.1. Aqueous diffusion coefﬁcients (D) and Stokes’ radii (rs) of several cations.5‘43

 

 

 

 

Ions D (mW/s) rs (11m)
NaW 1.333x 10'9 0.183
Mg2+ 0.706x 10'9 0.345
CazW 0.792x 10'9 0.308

 

 

 

 

 

2.3.1. Estimation of the Thicknesses of PEMs

59

One of the major assets of PEMs is their minimal thickness, which should allow
high ﬂuxes in NF. However, quantifying the thicknesses of ﬁlms on porous substrates is
difﬁcult. Under typical deposition conditions (0.5 M NaCl in PSS solutions and 1 M
NaCl in PAH solutions), [PSS/PAH]5 ﬁlms deposited on aluminum-coated Si wafers
have an ellipsometric thickness of 222t1 11m, and previous work suggests that the
thicknesses of PEMs grown on ﬂat surfaces are similar to those of ﬁlms formed on
porous alumina membranes.20 Figure 2.2-(a) shows a cross-sectional SEM image of a
porous alumina membrane coated with a [PSS/PAH]5 ﬁlm. Although it is difﬁcult to tell
exactly where the polyelectrolyte multilayer begins and ends, the image shows that the
ﬁlm is generally less than 50 nm thick. In Figure 2.2-(b) of alumina coated with
[PSS/PAHLO ﬁlms, the polyelectrolyte coating is somewhat easier to see and has a
thickness of about 50 nm. lmportantly, the interior of the alumina membrane is clearly
open. The structure of porous alumina, which includes 20 nm surface pores and larger

(0.2 rim-diameter) underlying pores in the bulk, facilitates complete coverage of these

supports.

60

 

Figure 2.2 Cross-sectional SEM images of a porous alumina membrane coated with a

[PSS/PAH]n ﬁlm. (a) [PSS/PAH]5 and (b) [PSS/PAHLO.

2.3.2. Nanofiltration of Mixed-Salt Solutions with PSS/PAH on PBS Supports

To examine the separation of cations, we perform NF of solutions containing 1
g/L Na+ (added in the form of NaCl) and 1 g/L Mng (added in the form of MgC12),
Previous research showed that [PSS/PAH]4PSS and [PSS/PAH]5 ﬁlms on porous alumina
exhibit a rejection of MgSO4 as high as 96% in single salt experiments,35 30 we begin our
studies with membranes containing PSS/PAH ﬁlms.

Table 2.2 presents NF results obtained with PSS/PAH ﬁlms deposited on 50 kDa
PES supports and shows how deposition conditions affect NF performance. Membrane

A is produced using the same solution compositions that previously yielded 96% MgSO4

rejection,35 but the rejection of Mg2+ from a solution containing MgC12 and NaCl is only

61

82%. This result suggests that the 96% MgSOa rejection observed previously in single
salt experiments occurred in large part because of high exclusion of SO42' ions. To
enhance the rejection of Mng from chloride solutions, we change the top layer of
PSS/PAH membranes from PSS to PAH, which is positively charged. A 6% increase in
Mng rejection results from this change (compare membranes A and B, Table 2.2),
suggesting that Donnan exclusion plays a role in the separation process. However, the
ﬂux decreass by about 25% on going from [PSS/PAH]4PSS to [PSS/PAH]4 because
membranes capped with PAH swell less than those capped with PSS, and less swelling

should also lead to higher size-based Mng rejection.““"45

62

Table 2.2. Solution ﬂuxes and ion rejections from NF with PSS/PAH ﬁlms deposited on

50 kDa PES supports. NF experiments were performed at 4.8 bar using mixed feed

solutions containing 1 g/L Na+ and 1 g/L Mng (chloride salts).

 

(supporting salt concentration

Membrane

and pH value
in deposition solutions)

Rejection of
Cations (%)

Flux

 

Na“ Mg2+

(m3/m2 day)

Selectivity
(Nd/M820

 

A

[PSS/PAH]4PSS
PSS: 0.5 M MnC12, pH=2.l
PAH: 0.5 M NaBr, pH=2.3

-35 i1 82.3 i0.9

0.52i0.06

7.6i0.5

 

[PSS/PAH},
PSS: 0.5 M MnCl2, pH=2.1
PAH: 0.5 M NaBr, pH=2.3

-27i3 88.6i0.7

0.38i0.02

11.1i0.4

 

[PSS/PAH].
PSS: 0.5 M NaCl, pH=2.l
PAH: 0.5 M NaCl, pH=2.3

-20i1 87.8 i0.8

0.38i0.02

9.9i0.6

 

[PSS/PAH]4
PSS: 0.5 M NaCl,pH=2.1
PAH: 0.5 M NaCl, pH=2.3

-13i4 92.0il.1

0.33 i002

14.2i1.3

 

[PSS/PAH]...
PSS: 0.5 M NaCl, pH=2.l
PAH: l M NaCl, pH=2.3

92.0i0.2

0.56i0.02

l3.6i'0.3

 

[PSS/PAH},
PSS: 0.5 M NaCl, pH=2.3
PAH: l M NaCl, pH=2.3

-20:l:2 94.1 i0.8

0.52i0.01

l8.3i2.1

 

[PSS/PAH},
PSS: 0.5 M NaCl, pH=4.5
PAH: l M NaCl, pH=2.3

-13i3 92.4i0.4

0.50-1.20.09

14.8i0.6

 

[PSS/PAH},
PSS: 0.5 M NaCl, pH=2.3
PAH: l M NaCl, pH=4.5

-13i'3 89.8:tl.5

0.55 $0.05

11.2i1.2

 

 

[PSS/PAH];
PSS: 0.5 M NaCl, pH=2.3
PAH: l M NaCl, pH=2.3

 

-l7i2 83.5i1.1

 

0721-003

 

7.1 i0.4

 

 

 

* The top PAH layer was formed in a solution with 1 M NaCl at pH of 2.3.

Interestingly, negative Na+ rejection appears in all the NF experiments with mixed
NaCl and MgCl2 solutions in this study. This phenomenon occurs due to the strong Mg2+
rejection and low Cl' rejection of these membranes. Because the feed concentration of

Cl’ is higher than that of Na+ and little Mg2+ passes through the membrane, anions (Cl')

63

momentarily cross the membrane faster than cations to create a diﬁusion potential that
enhances the transport of NaW. (This diffusion potential also accelerates Mg2+ transport,
but Mng is highly excluded from the membrane.) In other words, with a high Mng
rejection, the co-transport of Na+ with Cl' results in a NaW concentration in the permeate
that exceeds the Na+ concentration in the feed, i.e., negative rejection.46 (An equivalent
explanation can be made using Donnan potentials“)

In early studies of polyelectrolyte ﬁlms, we and others employed MnCl2 and
NaBr as supporting electrolytes during deposition of PSS and PAH, respectively.”’48’49
However, NaCl is a more economical and environmentally ﬁiendly reagent, so we also
examin membranes prepared with 0.5 M NaCl in PSS and PAH deposition solutions.
Comparison of membranes B and C in Table 2.2 shows that ﬂux and rejection are
comparable for [PSS/PAH}, ﬁlms deposited using either NaCl or MnC12/NaBr as
supporting electrolytes. Thus, all the PEMs subsequently described here are produced
using NaCl as supporting electrolyte.

In addition to changing the sign of the charge near the surface of PEMs by
changing the top layer from a polyanion to a polycation, we can alter the magnitude of
this charge by varying the ionic strength of polyelectrolyte deposition solutions. Previous
work in our group showed that increasing the ionic strength employed during deposition
of the top PSS layer in PSS/PAH membranes greatly enhances Cl'/SO.42' selectivity in
NF.35 Thus, we increase the concentration of NaCl from 0.5 M to 1.0 M in the solutions
used to deposit the top PAH layer of [PSS/PAH]4 ﬁlms in an attempt to increase

NaW/Mg2+ selectivity. Due to this change, rejection of Mng increase from 88% to 92%

while ﬂux decrease slightly from 0.38 to 0.33 m'W/(m2 day) (compare membranes C and D

64

in Table 2.2). In an effort to further improve the rejection of multivalent cations, all the
PAH layers are deposited from solutions containing 1 M NaCl, while 0.5 M NaCl is
employed during deposition of PSS layers (membrane E). Rejection remains the same as
for membrane D, presumably because the composition of the outer layer is most critical
in determining the charge near the surface of PEMs. At the same time, ﬂux increases by
70% when using 1 M NaCl during deposition of all PAH layers rather than just the top
layer. This suggests that the ﬁlms are more highly swollen when all PAH layers are
deposited from solutions of high ionic strength.”5 1 Higher swelling could occur because
when high ionic strengths are employed during deposition, some of the charged groups in
the polyelectrolyte ﬁlms will have NaW or Cl' counterions.51’52 This should decrease the
electrostatic interactions between polymer chains and enhance swelling.

The charge densities of weak polyelectrolytes vary with solution pH, so the pH of
weak polyelectrolyte deposition solutions plays an important role in the formation of
PEM3.28’53 Krasemann and Tieke suggested that the density of ionic cross-links in PEMs
reaches a maximum value when the deposition pH corresponds to the average of the pKa
values of the polycation and polyanion.19 PSS is a strong polyelectrolyte and thus,
changes in pH should not signiﬁcantly alter its charge density. However, the pH of PSS
deposition solution could affect the degree of PAH protonation in the ﬁlm to which PSS
is adsorbing. Comparison of membranes E and F (Table 2.2) shows that Mg2+ rejection
increases slightly from 92 to 94% when changing the PSS deposition pH from 2.1 to 2.3.
Additionally, NaW rejection decreases from -9% to -20% to give an overall NaW/Mg2+
selectivity of 18. Increasing pH of the PSS deposition solution further to 4.5 slightly

decreases Mg2+ rejection and NaW/MgZW selectivity (membrane G).

65

PAH is a weak polyelectrolyte whose charge density varies greatly with
deposition pH.28‘54 When pH of the PAH deposition solution is 4.5 (membrane H) rather
than 2.3 (membrane F), rejection of Mg2+ decreases. This presumably occurs because of
a lower charge density in the PEMs, which results from PAH being less protonated at pH
4.5.

Although the best PEM membranes exhibit a 94% Mg2+ rejection (membrane F in
Table 2.2), ﬂux is 30% less than the value of ~0.75 m3/(m2 day) at 4.8 bar, that is
representative of pure water ﬂux through commercial NF membranes. Therefore, we
investigate the possibility of increasing ﬂux by decreasing the number of PSS/PAH
bilayers. On going from [PSS/PAH}; to [PSS/PAH]3 ﬁlms (compare membranes F and I
in Table 2.2), ﬂux increases by 40%, but Mg2+ rejection drops to 84%, suggesting that
three bilayers of PSS/PAH may not be thick enough to fully cover the 50 kDa support.
Unfortunately, decreasing the number of deposited layers is not an effective means for
increasing ﬂux without reducing Mng rejection.

2.3.3. Nanoﬁltration of Mixed-Salt Solutions with Commercial NF Membranes and

PSS/PAH on Porous Alumina Supports

In addition to the polyelectrolyte ﬁlm, the underlying support can provide
signiﬁcant resistance to water ﬂow if it is not highly permeable. Indeed, the ﬂux through
unmodiﬁed 50 kDa membranes at 4.8 bar after 18 h of equilibration was only 5.3 m3/(m2
day),31 so the resistance of the support could be signiﬁcant. To enhance ﬂux, we deposite
PSS/PAH ﬁlms on porous alumina supports that have a surface pore diameter of 0.02 pm
and an internal pore diameter of 0.2 pm. Pure water ﬂux measurements show that the

resistance of bare porous alumina membranes is approximately half of the resistance of

66

uncoated 50 kDa PES substrates after 18 h of ﬁltration. Table 2.3 demonstrates that both
[PSS/PAH]4 and [PSS/PAH]5 ﬁlms on porous alumina give rejections of Mg” that are
greater than 92%, and that the ﬂux through these composite membranes exceeds 0.8
m3/(m2 day) at a pressure of 4.8 bar. In the case of porous alumina, the resistance of the
support to water ﬂow is likely negligible compared to the resistance of the PSS/PAH
ﬁlms. Optimization of polymer supports could potentially enhance ﬂux to the levels

found with alumina substrates.

Table 2.3. Solution ﬂuxes and ion rejections from NF with PSS/PAH ﬁlms deposited on
porous alumina (0.02 pm—diameter surface pores) supports. Results with a commercial
NF270 membrane are provided for comparison. NF experiments were performed at 4.8

bar using mixed feed solutions containing 1 g/L Na+ and 1 g/L Mng (chloride salts).

 

 

 

 

 

Membrane Rejection of Cations
(supporting salt concentration (%) Flux Selectivity
and pH value + 2+ (m3/m2 day) (NaW/Mng)
. . . . Na Mg
1n deposrtron solutions)
J [PSS/PAH]4
PSS: 0.5 M NaCl, pH=2.3 -3 i 2 92.5 i 0.5 1.05 i 0.03 13.9 i- 1.0
PAH: 1 M NaCl, pH=2.3
K [PSS/PAH]5
PSS: 0.5 M NaCl, pH=2.3 -11-l_-2 95.04; 0.5 0.85 i 0.02 22.5 :1:2.3
PAH: 1 M NaCl, pH=2.3
L NF 270 -15:l:2 68.1:tl.1 0.85i0.07 3.6101

 

 

 

 

 

 

 

We also investigate the separation performance of a commercial NF membrane,
NF270, provided by Dow-FILMTEC. NF 270 is a polyamide thin-ﬁlm composite
membrane with a negative surface charge, which is not beneﬁcial for cation rejection. As

Table 2.3 shows, NF 270 exhibits a Mg2+ rejection less than 70% along with a NaW/Mg2+

67

selectivity of only 3.6. In contrast, as mentioned above, composite membranes consisting
of [PSS/PAH}; and [PSS/PAH]5 on porous alumina allow ﬂuxes comparable to that
through NF 270, while exhibiting >92% rejection of Mng and a NaW/Mg2+ selectivity as
high as 22 (Table 2.3).

2.3.4. Nanoﬁltration of Single-Salt Solutions at Two Salt Concentrations

To better understand the factors affecting salt rejection, we perform single-salt NF
with both NF270 membranes and PSS/PAH ﬁlms deposited on porous alumina or 50 kDa
PES substrates (Table 2.4). Consistent with the mixed solution results, MgCl2 rejections
are remarkably higher with PSS/PAH membranes than with NF270. As expected, the
ﬂuxes through all membranes decline at higher salt concentrations due to an increased
osmotic pressure, and this effect is less severe in the case of NaCl because of lower
rejections. For reasons that we do not understand, however, the ﬂux through NF270
membranes is a much stronger function of feed ionic strength than would be expected

based on osmotic pressure arguments.

68

Table 2.4. Cation rejections (%) and ﬂux values (listed in parentheses in n13 n1"2 day")

for single-salt NF with PSS/PAH ﬁlms deposited on different substrates. Results with a

commercial NF270 membrane are provided for comparison.

NF experiments were

 

 

 

 

 

performed at 4.8 bar.
Membrane Rejection of Na+ Rejection of Mg2W
(Slipporting. salt (NaCi)a (MgC12)a
concentratlon
and pH value in 0.1 g/L 1 g/L 0.1 g/L 1 g/L
deposition solutions)
F [PSS/PAH]4 on a 50
kDa PES membrane 74 $ 3 40 $2 93.5 $0.9 93.6 $ 1.6
PSS: 0.5 M NaC1,pH=2.3 (0.98$0.07) (0.80$0.12) (1.13 220.12) (O.47$0.07)
PAH: l M NaCl, pH=2.3
K [PSS/PAH]5 on 0.02
pm porous alumina 62 $ 5 48 $ 7 97.3 $1.1 96.3 $0.4
PSS: 0.5 M NaCl,pH=2_3 (1.51 $0.15) (1.37$0.10) (1.84$0.03) (0.82$0.02)
PAH: 1 M NaCl, pH=2.3
59$1 4l$2 47.4$l.0 52.4$l.9
L NW 270 (1.53 $0.09) (0.81 $0.14) (1.79$0.10) (0.47$0.07)

 

 

 

 

 

 

 

aAs elsewhere in this paper, concentrations of feed solutions refer to the cation, not the

salt.

The drop in NaCl rejection with increasing NaCl concentrations (Table 2.4) is
consistent with a Donnan exclusion mechanism for cation rejection, but the absence of
such an effect with MgCl2 is not. At high salt concentrations, Donnan exclusion should
decrease considerably because the concentration of ﬁxed charge in the membrane is
closer to the concentration of ions in solution. The insensitivity of MgCl2 rejection to
MgCl2 concentration could result from greater adsorption of Mg2+ on membranes at high
5,39,55

MgCl2 concentrations, which would give a more highly charged surface layer.

Greater charge near the surface may overcome the effects of increased charge screening

69

due to a higher ionic strength. However, if MgCl2 rejection were primarily due to the
large size of hydrated MgZW, rejection might also be a weak function of concentration.
Streaming potential measurements described in section 2.3.7 suggest that Mg2+ adsorbs to
the membrane to increase positive charge near the surface.

2.3.5. Nanofrltration of Mixed-Salt Solutions with PSS/PDADMAC or Hybrid

Membranes

To enhance the permeability of the PEMs, we employ PSS/PDADMAC ﬁhns
because this polyelectrolyte system swells more in water than PSS/PAH due to the
relatively low charge density of PDADMAC and the accompanying lower density of
ionic cross links.33’37 Table 2.5 shows the results of NF of mixed solutions containing 1
g/L Na+ and 1 g/L Mg2+ (chloride salts) using membranes composed of PSS/PDADMAC

ﬁlms on alumina supports.

70

Table 2.5. Solution ﬂuxes and ion rejections from NF with PSS/PDADMAC and hybrid

(PSS/PDADMAC)4+(PSS/PAH) ﬁlms deposited on porous alumina (0.02 um-diameter

surface pores) supports. NF experiments were performed at 4.8 bar using mixed feed

solutions containing 1 g/L Na+ and 1 g/L Mng (chloride salts).

 

Membrane

(supporting salt concentration

and pH value
in deposition solutions)

Rejection of Cations

(%)

 

Na+

Mg2+

Flux
(m3/m2 day)

Selectivity
(N a+ Mg”)

 

M*

[PSS/PDADMAC]4PSS
PSS: 0.5 M NaCl
PDADMAC: 0.5 M NaCl

38.3 $0.8

1.74$0.03

1.6$0.1

 

[PSS/PDADMAC]5
PSS: 0.5 M NaCl
PDADMAC: 0.5 M NaCl

42.0$0.4

1.95 $ 0.05

1.8$O.l

 

[PSS/PDADMAC]7
PSS: 0.1 M NaCl
PDADMAC: 0.1 M NaCl

24.4$ 1.3

3.58$0.11

1.2$0.1

 

[PSS/PDADMAC]4PSS/PAH
PSS: 0.5 M NaCl
PDADMAC: 0.5 M NaCl
PSS: 0.5 M NaCl, pH=2.3
PAH: 0.5 M NaCl, pH=2.3

-29$1

76.1 $0.5

1.40$O.l3

5.4$0.l

 

[PSS/PDADMAC]4PSS/PAH
PSS: 0.5 M NaCl
PDADMAC: 0.5 M NaCl
PSS: 0.5 M NaCl, pH=2.3
PAH: l M NaCl,pH=2.3

-28$2

875$ 1.6

1.l4$0.13

11.1 $0.7

 

 

[PSS/PDADMAC]4PSS/PAH
PSS: 0.5 M NaCi
PDADMAC: 0.5 M NaCl
PSS: 0.5 M NaCl, pH=2.3
PAH: 2.5 M NaCl, pH=2.3

 

-19$2

 

92.6$0.8

 

l.04$0.15

 

16.2d:1.7

 

 

* The top PSS layer was formed in a solution with 1 M NaCl.

Because both PSS and PDADMAC are strong polyelectrolytes, varying the pH

value of the deposition solutions should not change the structure of the PEMs

signiﬁcantly. Thus, PSS/PDADMAC membranes were prepared without adjusting the

pH of the polyelectrolyte solutions, and unadjusted pH values is 6.2 for PSS solutions
containing 0.1 or 0.5 M NaCl, 4.5 for PDADMAC solutions containing 0.1 M NaCl, and
5.0 for PDADMAC containing 0.5 M NaCl. NF with [PSS/PDADMAC]4PSS
membranes deposited from 0.5 M NaCl solutions (membrane M, Table 2.5) shows a
solution ﬂux of 1.7 m3/(m2 day), but a Mng rejection of only 40%. Termination of these
membranes with a layer of polycation, PDADMAC, does not greatly enhance MgZW
rejection (membrane N, Table 2.5).

Our previous research showed that PSS/PDADMAC ﬁlms prepared from 0.1 M
NaCl solutions swell 65% less than ﬁlms prepared from 0.5 M NaCl.37 Therefore, we
also examine NF with PSS/PDADMAC membranes prepared using solutions containing
0.1 M NaCl. Even though we use 7 PSS/PDADMAC bilayers in membranes prepared
using 0.1 M NaCl deposition solutions, ﬂux through these systems is still almost twice
that through [PSS/PDADMAC]5 ﬁlms prepared using solutions containing 0.5 M NaCl
(compare membranes 0 and N in Table 2.5). The high ﬂux through membrane 0
presumably occurs because ﬁlms prepared in the presence of 0.1 M NaCl are very thin.
(Such ﬁlms typically have bilayers thicknesses that are about 50% of those for ﬁlms
prepared in 0.5 M NaCl.”) Consistent with a thin ﬁlm that does not fully cover the
support, Mg2+ rejection is only 24% for membrane 0.

One could try to fully cover the support by depositing more PSS/PDADMAC
bilayers, but this would result in an impractically long deposition process. Rather than
coating the substrate with many bilayers of PSS/PDADMAC, we deposite one capping
bilayer of PSS/PAH on four bilayers of PSS/PDADMAC deposited from 0.5 M NaCl.

Three different membranes were prepared with the outer PAH layer deposited from

72

solutions containing 0.5 M, l M, or 2.5 M NaCl (membranes P, Q, and R in Table 2.5).
In all cases, ﬂux decreases and rejection of Mg2+ increases compared to
[PSS/PDADMAC]5 membranes, presumably because deposition of the PSS/PAH bilayers
yields a more dense surface.37 However, only deposition of the outer PAH layer from 2.5
M NaCl gives a Mg2+ rejection greater than 90%. In this case, the membrane
performance is essentially equivalent to that of a [PSS/PAH}; ﬁlm deposited on porous
alumina (membrane J, Table 2.3), so there is no major advantage to formation of hybrid
membranes.

2.3.6. Calcium Rejection by PSS/PAH Films on Porous Alumina

Because Ca2+ is also an important component in hard water, we examine the
ability of membrane K (Table 2.4, [PSS/PAH]5 on porous alumina) to reject Ca2+. We
choose this membrane because it gives the highest Mg2+ rejections and NaW/MgZW
selectivities. When the feed solution contains 1 g/L of Na+ and 1 g/L of Ca2+ (chloride
salts), rejection of Ca2+ is 98% and the NaW/Ca2+ selectivity is as high as 50 (Table 2.6).
Higher rejection of Ca2+ than Mg2+ is probably due to more extensive adsorption of Ca2+
on the ﬁlm, which results in a more positively charged membrane.56 The higher positive
charge may also be responsible for the relatively high rejection of Na+ in the NaCl/CaCl2
solution. With feed solutions containing Ca2+, Mg”, and NaW, rejections are greater than
95% for both Ca2+ and MgZW, and the rejection of Na+ decreases substantially compared

to NF of solutions with only Na+ and Ca2+.

73

Table 2.6. Solution ﬂuxes and ion rejections from NF with Membrane K (ﬁve bilayers
of PSS/PAH deposited on porous alumina supports). NF experiments were performed at

4.8 bar using mixed feed solutions containing either Na+ and Ca2+ or Na+, Ca2+ and Mg2+

 

 

 

 

 

(chloride salts).
Feed Co;c:;1trat10n Rejection of Cations (%) Flux Selectivity
2 2 2 2 (m3/m2 (Na+ (Na+
+ + + + + +
Na Ca M N C M da
g a a g Y) / C$112+) /Mg2+)
l 1 O 28i5 98.5 i0.4 - 0.94i0.l4 501210 -
l 0.62a 0.3821 -5 i2 96.0 i0.2 95.2i0.4 0.79 i009 25 i2 21 i3

 

 

 

 

 

 

 

 

 

 

 

aCa2+ and Mg2+ concentrations were chosen to achieve 1 g/L of divalent cation with an

equimolar solution.

2.3.7. Streaming Potential Measurements

The above results suggest that the charge near the surface of the PEMs plays an
important role in ion separations, so we perform streaming potential measurements to
better correlate ion transport with surface charge. Figure 2.3 presents the C potentials
derived from streaming potential measurements in 1 mM KCl with both PSS/PAH and
PSS/PDADMAC ﬁlms deposited on 50 kDa PBS membranes. As expected, the 2;
potential is negative after deposition of a polyanion and positive after deposition of a
polycation.36’57‘58 The Q potential of the ﬁrst PSS layer is always more negative than the
others because the bare PES substrate has a I; potential of approximately -18 mV.
(Adsorption of PSS to the PBS membrane likely occurs via hydrophobic interactions.)
The positive l; potential induced by PDADMAC was 5-10 mV larger than that due to

PAH, presumably because of a greater overcompensation of the negative charge by

74

PDADMAC.59 Nevertheless, PEMs capped with PDADMAC show lower Mg2+

rejections than PEMs capped with PAH, presumably because of higher swelling, as noted

earlier.37

40; A
30: ‘
20$
105
o.‘
-10-

204‘ . ‘

'30 1 ﬂ

'40 I I I I I r I I I
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5

Number of Bilayers

I)
I}

Zeta Potential (mV)
bl

 

 

Figure 2.3. C potentials of [PSS/PAH]n (squares) and [PSS/PDADMAC]n (triangles)

ﬁlms on PBS substrates. Non-integer numbers of bilayers represent ﬁlms terminated
with PSS, whereas integer numbers of bilayers denote polycation-terminated ﬁlms.
Streaming potential measurements were performed in 1 mM KCl, and the standard

deviation of the measurements was about 1 mV.

Figure 2.4 demonstrates that I; potentials depend on both the ionic strength at
which the outer layer was deposited and the ionic strength of the probe solution. For
streaming potential measurements obtained with 1 mM KCl and membranes coated with

[PSS/PDADMAC]4+PSS/PAH ﬁlms, deposition of the outer PAH layer from 2.5 M NaCl

75

yields an 8 mV higher 2; potential than deposition of PAH from a solution containing 0.5
M NaCl. The ﬁlm with the top PAH layer deposited from 1.0 M NaCl shows an
intermediate Q potential. This is consistent with the supposition that deposition of the
terminal polycation layer from solutions of high ionic strength induces more charge near
the surface to increase Mg2+ rejection. Similar trends appear in the streaming potential
measurements with 5 and 10 mM KCl solutions, but for a given membrane the C potential
becomes less positive as KCl concentration increases.60 This is expected because the
diffuse double layer becomes more compact at higher ionic strengths, and more of the
surface-solution potential drop occurs between the ﬁlm surface and the plane of shear.
The t; potential represents only the potential drop between the plane of shear and the bulk

solution.

76

 

PO+

Zeta Potential (mV)

 

10-: \%

o 5 4 6 8 ' 1'0
Concentration of KCI (mM)

 

Figure 2.4. g potentials of [PSS/PDADMAC]4+[PSS/PAH] ﬁlms as a function of the

concentration of KCI employed in the streaming potential measurements. Polyelectrolyte
ﬁlms were deposited on PBS substrates using three different concentrations of NaCl
during deposition of the terminal PAH layer: 0.5 M NaCl, squares; 1.0 M NaCl, circles;

2.5 M NaCl, triangles.

To examine whether adsorption of Mg2+ may be responsible for the fact that Mg2+
rejections do not decrease with increasing MgC12 concentrations, we perform streaming
potential measurements using solutions containing three different concentrations of
Mng (Figure 2.5). In contrast to measurements in KCI, C potentials determined in
MgClz solutions increase with increasing concentrations of Mng. (The MgClz
concentrations in Figure 2.5 were chosen to duplicate the ionic strengths used for the

measurements in KCI in Figure 2.4.) Although the diffuse layer should still become

77

thinner at higher ionic strengths in MgClz solutions, adsorption of divalent Mg2+ on the
ﬁlm surface38 probably increases the amount of positive charge near the surface to make
the 6; potential more positive. This more positive l; potential likely overcomes any
decreases in rejection due to reduced Donnan exclusion at higher ionic strength, so
rejection is essentially independent of concentration. (The high standard deviations of C
potentials determined at 1.66 and 3.33 mM Mng are typical for streaming potential

measurements in these more concentrated solutions.)

35-

N (a)
01 O
1 . 4

Zeta Potential (mV)
to
O

15«

 

10 v I ' I ' U '
0 1 2 3 4
Concentration of MgCl2 (mM)

 

Figure 2.5. C potentials of [PSS/PDADMAC]4+[PSS/PAH] ﬁlms as a function of the

concentration of MgC12 employed in the streaming potential measurements.

Polyelectrolyte ﬁlms were deposited on PBS substrates using three different
concentrations of NaCl during deposition of the terminal PAH layer: 0.5 M NaCl,

squares; 1.0 M NaCl, circles; 2.5 M NaCl, triangles.

78

2.4. Conclusions

PEMs composed of PSS/PAH or hybrid PSS/PDADMAC + PSS/PAH ﬁlms are
effective in selectively removing Mg2+ from solutions containing NaCl and Mng, but
pure PSS/PDADMAC ﬁlms and commercial NF 270 membranes show only relatively low
rejections of Mg”. For PAH-terminated ﬁlms, Mg2+ rejection and Na+/Mg2+ selectivity
increase with the magnitude of the charge near the surface of the PEMs, and this can be
enhanced by increasing the ionic strength of the terminating PAH deposition solutions.
In single-salt experiments, MgC12 rejection did not vary signiﬁcantly with the MgClz
concentration in the feed solution. Streaming potential measurements suggest that as
Mng concentration increases, more Mg2+ adsorbs to the membrane to enhance charge
near the surface and overcome decreases in Donnan exclusion due to higher salt
concentrations. High rejections of both Ca2+ and Mg2+ in the presence of Na+ make these

membranes potentially useful for water softening.

79

 

2.5. References

(1) Ouyang, L.; Malaisamy, R.; Bruening, M. L. J. Membr. Sci. 2008, 310, 76-84.
(2) Das, G. J. Clin. Pharmacol. 1988, 28, 683-690.

(3) Ghizellaoui, S.; Taha, S.; Dorange, G.; Chibani, A.; Gabon, J. Desalination 2004,
171,133-138.

(4) Nanda, D.; Tung, K.-L.; Hsiung, C.-C.; Chuang, C.-J.; Ruaan, R.-C.; Chiang, Y.-
C.; Chen, C.-S.; Wu, T.-H. Desalination 2008, 234, 344-353.

(5) Wang, D.-X.; Su, M.; Yu, Z.-Y.; Wang, X.-L.; Ando, M.; Shintani, T.
Desalination 2005, 1 75, 219-225.

(6) Lu, X.; Bian, X.; Shi, L. J. Membr. Sci. 2002, 210, 3-11.

(7) Bruening, M. L.; Dotzauer, D. M.; Jain, P.; Ouyang, L.; Baker, G. L. Langmuir
2008, 24, 7663-7673.

(8) Eriksson, P.; Kyburz, M.; Pergande, W. Desalination 2005, 184, 281-294.
(9) Anim-Mensah, A. R.; Krantz, W. B.; Govind, R. Eur. Polym. J. 2008, 44.

(10) Tavares, C. R.; Vieira, M.; Petrus, J. C. C.; Bortoletto, E. C.; Ceravollo, F.
Desalination 2002, 144, 261-265.

(11) Wang, X.-L.; Tsuru, T.; Nakao, S.; Kimura, S. J. Membr. Sci. 1997, 135, 19-32.

(12) Wang, D.; Wu, L.; Liao, Z.; Wang, X.; Tomi, Y.; Ando, M.; Shintani, T. J.
Membr. Sci. 2006, 284, 384-392.

(13) Petersen, R. J. J. Membr. Sci. 1993, 83, 81-150.

(14) Du, R.; Zhao, J. J. Membr. Sci. 2004, 239, 183-188.

(15) Huang, R.; Chen, G.; Sun, M.; Gao, C. Carbohydr. Res. 2006, 341, 2777-2784.
(16) Li, L.; Wang, B.; Tan, H.; Chen, T.; Xu, J. J. Membr. Sci. 2006, 269, 84-93.
(17) Decher, G.; Hong, J. D.; Schmitt, J. Thin Solid Films 1992, 210-211, 831-835.
(18) Decher, G. Science 1997, 277, 1232-1237.

(19) Krasemann, L.; Toutianoush, A.; Tieke, B. J. Membr. Sci. 2001, 181, 221-228.

80

(20) Harris, J. J .; Stair, J. L.; Bruening, M. L. Chem. Mater. 2000, 12, 1941-1946.
(21) J in, W.; Toutianoush, A.; Tieke, B. Langmuir 2003, 19, 2550-2553.

(22) Chen, H.; Zeng, G.; Wang, Z.; Zhang, X. Macromolecules 2007, 40, 653-660.
(23) Zhang, G.; Yan, H.; Ji, S.; Liu, Z. J. Membr. Sci. 2007, 292, 1-8.

(24) Losche, M.; Schmitt, J .; Decher, G.; Bouwman, W. G.; Kjaer, K. Macromolecules
1998, 31, 8893-8906.

(25) Ladam, G.; Schaad, P.; Voegel, J. C.; Schaaf, P.; Decher, G.; Cuisinier, F. El:
Langmuir 2000, 16, 1249-1255.

(26) Steitz, R.; Leiner, V.; Siebrecht, R.; Klitzing, R. v. Colloids Surf, A 2000, 163,
63-70.

(27) Yoo, D.; Shiratori, S. S.; Rubner, M. F. Macromolecules 1998, 31 , 4309—4318.

 

(28) Shiratori, S. S.; Rubner, M. F. Macromolecules 2000, 33, 4213-4219.

(29) Mendelsohn, J. D.; Barrett, C. J .; Chan, V. V.; Pal, A. J .; Mayes, A. M.; Rubner,
M. F. Langmuir 2000, 16, 5017-5023.

(30) Krasemann, L.; Tieke, B. Langmuir 2000, 16, 287-290.
(31) Malaisamy, R.; Bruening, M. L. Langmuir 2005, 21, 10587-10592.
(32) Hong, S. U.; Malaisamy, R.; Bruening, M. L. Langmuir 2007, 23, 1716-1722.

(33) Tieke, B.; Toutianoush, A.; Jin, W. Adv. Colloid Interface Sci. 2005, 116, 121-
131.

(34) Farhat, T. R.; Schlenoff, J. B. J. Am. Chem. Soc. 2003, 125, 4627-4636.

(35) Stanton, B. W.; Harris, J. J .; Miller, M. D.; Bruening, M. L. Langmuir 2003, 19,
7038-7042.

(36) Adamczyk, Z.; Zembala, M.; Warszynski, P.; Jachimska, B. Langmuir 2004, 20,
10517-10525.

(37) Miller, M. D.; Bruening, M. L. Chem. Mater. 2005, 17, 5375-5381.
(38) Childress, A. E.; Elimelech, M. J. Membr. Sci. 1996, 119, 253-268.

(39) Chung, C. V.; Buu, N. Q.; Chau, N. H. Sci. Tech. Adv. Mater. 2005, 6, 246-250.

81

(40) Kotov, N. A. Nanostruct. Mater. 1999, 12, 789-796.

(41) Walker, S. L.; Bhattacharjee, S.; Hoek, E. M. V.; Elimelech, M. Langmuir 2002,
18, 2193-2198.

(42) Liu, X.; Bruening, M. L. Chem. Mater. 2004, 16, 351-357.

(43) J in, W.; Toutianoush, A.; Pyrasch, M.; Schnepf, J .; Gottschalk, H.; Rammensee,
W.; Tieke, B. J. Phys. Chem. B 2003, 107, 12062-12070.

(44) Carriere, D.; Krastev, R.; Schiinhoff, M. Langmuir 2004, 20, 11465-11472.

(45) Wong, J. E.; Rehfeldt, F.; Hanni, P.; Tanaka, M.; Klitzing, R. v. Macromolecules
2004, 37, 7285-7289.

(46) Gilron, J .; Gara, N.; Kedem, O. J. Membr.Sci. 2001, 185, 223-236.

(47) Levenstein, R.; Hasson, D.; Semiat, R. J. Membr. Sci. 1996, 116, 77-92.

(48) Harris, J. J .; Bruening, M. L. Langmuir 2000, 16, 2006-2013.

(49) Phuvanartnuruks, V.; McCarthy, T. J. Macromolecules 1998, 31, 1906-1914.
(50) Antipov, A. A.; Sukhorukov, G. B.; Mohwald, H. Langmuir 2003, 19, 2444-2448.
(51) Dubas, S. T.; Schlenoff, J. B. Macromolecules 2001, 34, 3736-3740.
(52) Lowack, K.; Helm, C. A. Macromolecules 1998, 31, 823-833.

(53) Schiinhoff, M. J. Phys. Condens. Matter 2003, I5, R1781-R1808.
(54) Choi, J .; Rubner, M. F. Macromolecules 2005, 38, 116-124.

(55) Peeters, J. M. M.; Boom, J. P.; Mulder, M. H. V.; Strathmann, H. J. Membr. Sci.
1998, 145, 199-209.

(56) Furusawa, K.; Iwamoto, M.; Matsumura, H. Bull. Chem. Soc. Jpn 1987, 60, 845-
849.

(57) Schwarz, B.; Schtinhoff, M. Langmuir 2002, 18, 2964-2966.

(58) Schwarz, S.; Eichhorn, K.-J.; Wischerhoff, E.; Laschewesky, A. Colloids Surf, A
1999, 159, 491-501.

(59) Adusumilli, M.; Bruening, M. L. Langmuir 2009, 25, 7478-7485.

82

(60) Sterthaus, R.; Wegner, G. Langmuir 2002, 18, 5414-5421.

83

Chapter 3
Multilayer Polyelectrolyte Nanoﬁltration Membranes for Selective

Phosphate Removal

This work was done in collaboration with Professor Seong Uk Hong of the
Department of Chemical Engineering at Hanbat National University, Daejeon, South

Korea, and the data were recently published in Journal of Membrane Science.1

3.1. Introduction

Chapter 2 discussed the use of polyelectrolyte multilayer ﬁlms (PEMs) on porous
alumina and polymeric supports as nanoﬁltration (NF) membranes for separation of
multivalent and monovalent cations. This chapter explores the formation of selective
PEM membrane skins for selective phosphate removal from chloride-containing solutions.
Variation of PEM composition, and especially whether the terminating layer is a
polycation or a polyanion, allows tailoring of membranes for speciﬁc separations.

The concentration of phosphorus in natural waters is critically important in a
variety of biological and chemical processes,2 and the discharge of excess phosphorus
into water bodies may cause eutrophication and subsequent deterioration of water
quality.3'4 Therefore, it is important to remove phosphate from wastewater before
discharge into natural water.5 (In this chapter, we refer to phosphate as P043” and all of

its protonated forms.) The US EPA recommends a regulatory limit of 0.05 mg

84

 

phosphorus/L for direct discharge into lakes and 0.1 mg phosphorus/L for indirect
discharge into lakes.”

Common phosphate removal techniques include chemical precipitation}9

biological remova1,'°"2 adsorption,”13

and two membrane-based processes, reverse
osmosis (R0) and NF.7’l4 Of these two membrane separation techniques, NF is
preferable for selective phosphate removal because it allows the passage of monovalent
ions to lower osmotic pressure. Hence, when compared with R0, NF allows selective
phosphate removal at relatively low pressures. 15"6

The goal of this project is to develop a highly permeable NF membrane that
rejects phosphate while passing common monovalent anions such as chloride. To
achieve high permeability, the membrane will most likely consist of a thin selective layer
deposited on a thicker porous support that provides the mechanical strength.”'9 This
study examines phosphate/chloride separations using membranes formed by layer-by-
layer adsorption of PEMs on porous alumina supports. PEMs are especially attractive for
these separations because the layer-by-layer method affords control over the ﬁlm
thickness, and the ﬁlm composition can be tuned for a wide range of permeation
properties.'7"8’2°'23 In general, ion transport through NF membranes is a function of both
the charge and effective pore size of the skin layer.“25 Layer-by-layer deposition of
polyelectrolytes naturally affords a charged surface, and the pore size can be controlled
through variation of the constituent polyelectrolytes and deposition conditions.”30

A recent study in our group reports the use of PEMs composed of

poly(styrenesulfonate) (PSS) and poly(diallyldimethylammonium chloride) (PDADMAC)

in selective removal of ﬂuoride from solutions containing chloride or bromide.3 '

85

Remarkably, (PSS/PDADMAC)4PSS ﬁlms on porous alumina supports exhibit
chloride/ﬂuoride or bromide/ﬂuoride selectivities >3 with minimum chloride or bromide
rejections, even though the Stokes’ radii of hydrated F' and Cl' differ by only 0.045 nm.

Table 3.1 lists the difﬁision coefficients and Stokes’ radii of some hydrated anions.

Table 3.1. Molecular weights (MW), Stokes’ radii (rs) and aqueous diffusion coefﬁcients

(D) of several anions.32

 

 

Anions Mw (g mol“) r, (nm) D (mz/s)

Hpo42' 96.0 0.560 0.4; x 10'9

H2P04' 97.0 0.279 0.88 x 10'9
F- 19.0 0.166 1.48 x 10'9
or 35.5 0.121 2.03 x 10'9
Br' 79.9 0.118 2.08 x 10'9

 

Given the fact that the Stokes’ radius of H2PO4' or HPO42' in water is at least 0.15
nm larger than that of or, size-based selective removal of H2P04' or HPO42’ in the
presence of Cl' should be possible. In addition, the form of phosphate ions varies with
pH, so phosphate rejections and chloride/phosphate selectivities can be tailored by
varying the pH of the feed solution. As Figure 3.1 shows, the fraction of phosphate that
is HPO42', which has both a larger size and a higher charge than HzPOa', increases with
solution pH for pH values between 5 and 10. Hence, the overall phosphate rejection

should increase with feed pH. This work explores the use of PSS/PDADMAC

86

membranes for selective phosphate removal as a function of both the feed pH and the
number of PSS/PDADMAC bilayers in the membrane. Studies of the separation of

phosphate and chloride with commercial NF membranes are included for comparison.

1.00

0.90 --
0.80 --
0.70 a,
0.60 -P
0.50 “-
0.40 --
0.30 --
0.20 +
0.10 ~-
0.00

Fraction

 

 

 

 

 

Figure 3.1. Fractional speciation of phosphate as a function of pH. (pKa values are 12.35,

7.20, and 2.15.”)

3.2. Experimental Section

3.2.1. Materials

Poly(sodium 4-styrenesulfonate) (PSS, MW = 70,000 Da) and

poly(diallyldimethylammonium chloride) (PDADMAC, MW = 100,000 — 200,000 Da, 20

wt% in water) were purchased from Sigma-Aldrich and used without ﬁuther puriﬁcation.

87

NaCl and NaH2P04 were obtained from CCI and also used as received. Porous alumina
supports (0.02 um Whatman Anodisk ﬁlters, average membrane thickness of 60 um)
were UV/O3 cleaned (Boekel UV-Clean Model 135500) with the feed side up for 15 min
before ﬁlm deposition. Deionized water (Milli-Q puriﬁcation system, 18.2 MQ-cm) was
used for membrane rinsing and preparation of the polyelectrolyte and feed solutions. The
two commercial NF membranes used in this study, NF 90 and NF 270, were kindly
provided by Dow-FILMTEC.
3.2.2. Film Formation on Porous Alumina

The pretreated alumina membrane was sandwiched in a home-made membrane
holder that exposes only the feed side of the membrane to the polyelectrolyte solutions.
Deposition started with immersion of the membrane surface in a 0.02 M PSS solution
(polyelectrolyte concentrations are always given with respect to the repeating unit) for 3
min. PSS was deposited ﬁrst due to the potential for electrostatic interactions with the
positively charged alumina support. Following a 1-min rinse with deionized water, the
support was exposed to a 0.02 M PDADMAC solution for 3 min and rinsed with
deionized water for l min. Additional layers were deposited similarly to give the desired
number of bilayers. Both polyelectrolyte deposition solutions contained 0.5 M NaCl as
supporting salt. After ﬁlm deposition, membranes were stored in deionized water until
use. Polyelectrolyte deposition time and salt concentrations were chosen based on
previous ion separation studies.18’3"34
3.2.3. Transport Studies

NF experiments were performed at room temperature (23 °C) using a cross-ﬂow

system described previously.19 The system was pressurized with Ar to 4.8 bar, and a

88

centrifugal pump circulated the feed solution through the system and across the
membrane surfaces. The ﬂow meter located between the pump and membrane cells was
set at 18 mL/min to minimize concentration polarization.19 The feed tank was ﬁlled with
a 2 L solution containing 1 mM NaH2P04 and 1 mM NaCl, and the pH of the feed
solution was adjusted with 1 M NaOH. After an initial 18 h of ﬁltration to condition the
membrane, permeate samples were collected every 15 min for 1 h. The feed solution was
sampled at the end of the experiment to take into account small concentration changes
during the ﬁltration. The reported ﬂuxes are usually the steady-state permeate ﬂuxes
after 18 h of ﬁltration, but in some cases, NF experiments at pH 5.6 were followed by
experiments at either pH 7 or 8.4, so only a ~6 h-ﬁltration was employed for conditioning
at pH 7 or 8.4. Ion chromatography (Dionex 600 Ion Chromatograph with an Ionpac
ASl6A column) with conductivity detection was used to determine the anion

concentrations in the feed and permeate samples.

3.3. Results and Discussion
Previous anion separations with PSS/PDADMAC-modiﬁed ultraﬁltration
membranes showed high Cl'/8042' and Cl'lF' selectivities along with a high solution

ﬂux,30,31,35

so we chose to examine this polyelectrolyte system for phosphate/chloride
separations. Additionally, PSS and PDADMAC are both strong polyelectrolytes whose
charge density should not vary signiﬁcantly at pH values >2. Thus, variations in
phosphate rejections with feed pH will stem primarily from the speciation of phosphate

but not changes in membrane properties. All the membranes in this study contain a PSS

89

BY-

 

terminating layer because the negative charge of the predominantly polyanionic surface
increases selectivities among monovalent and multivalent anionslg’3 ‘

In this section, we ﬁrst describe NF results with PSS/PDADMAC membranes and
then compare these results with the performance of commercial NF 90 and NF 270
membranes. Two parameters, solute rejection, R, and selectivity, S, of solute A over
solute B, characterize the selective removal of phosphate from chloride-containing
solutions. Equation (1) deﬁnes R, where Cp and Cf are the solute concentrations in the

permeate and feed, respectively, whereas equation (2) deﬁnes S, which can also be

expressed in terms of the rejections of the two solutes.

R = (1—cp /cf)x100% (1)

S- CAP .CBJ _100—RA(%)
CM C3,, 100—R,(%) (2)

 

 

3.3.1. Membrane Formation

To simultaneously achieve high permeability and selectivity, composite
membranes often contain a thin selective layer on a thicker porous support that provides
mechanical stability. In this work, the selective layer is a PEM ﬁlm prepared using layer-

36

by-layer deposition, and the ﬁlm thickness is in the range of 20-30 nm. An alumina

membrane with 0.02 pm surface pores and an approximate thickness of 60 um serves as
the porous support. Previous ion-transport studies showed that after deposition of 4-5

19.28

bilayers, the PEMs forms a defect-free ﬁlm on alumina supports, and SEM images

such as Figure 2.2 conﬁrm that complete surface coverage occurs without ﬁlling of the
28.37

underlying pores.

3.3.2. NF Experiments with PSS/PDADMAC Membranes

90

 

Table 3.2 shows solution ﬂuxes, solute rejections and chloride/phosphate
selectivities from NF with PSS/PDADMAC membranes. Without addition of NaOH, the
pH of the feed solution (1 mM NaHzPO4 and 1 mM NaCl) is 5.6, and under these
conditions most phosphate ions are present as H2P04' (Figure 3.1). In this case,
(PSS/PDADMAC)4PSS-coated alumina membranes exhibit 86% phosphate rejection, a
chloride/phosphate selectivity of 5.8, and a solution ﬂux of 2.2 m3/(m2 day). This result
is consistent with the size-based selectivity in a previous NF study with 1 mM NaF and 1
mM NaCl, in which ﬂuoride rejection was 73% and chloride rejection was 10% with the
same membrane composition.3 1 The higher rejection of H2P04' than F ' most likely stems

from the larger Stokes’ radius of H2PO4' (Table 3.1).

91

" ““ “'1

 

Table 3.2. Solution ﬂuxes, rejections and selectivities from NF with
(PSS/PDADMAC)XPSS-coated alumina membranes and feed solutions containing
NaHzPO4 (1 mM) and NaCl (1 mM). The transmembrane pressure was 4.8 bar.

a The extra 0.5 bilayer indicates that ﬁlms were terminated with a layer of PSS.

b Experiments were performed after NF experiments using a feed solution at pH 5.6

 

pH of feed solution = 5.6

 

Number of Flux Chloride Phosphate Chloride/Phosphate
bilayersa (m3/m2 day) Rejection (%) Rejection (%) Selectivity (-)

 

4.5 2.2i0.l 16.2i0.1 85.6il.6 58:06

5.5 2110.1 -2.1i4.7 74.8i4.2 4.1:tl.0

 

pH of feed solution = 7.0

 

Number of Flux Chloride Phosphate Chloride/Phosphate
bilayersa (m3/m2 day) Rejection (%) Rejection (%) Selectivity (-)

 

4.5b 2.2:01 33.1:1.3 918:07 82:05

5.5b 21:00 9.3:1.7 81.5:0.5 49:00

 

pH of feed solution = 8.4

 

Number of Flux Chloride Phosphate Chloride/Phosphate
bilayersa (m3/m2 day) Rejection (%) Rejection (%) Selectivity (-)

 

3.5 29:01 10:06 76.9:0.4 43:01
4.5b 24:01 209:6.7 98.3:02 48.3:58
5.5 24:00 12.4:10 97.7:0.1 37.4:1.4

 

92

The thicknesses of PEM ﬁlms usually increase with the number of bilayers,38‘40

and thicker ﬁlms are less likely to contain defects. Hence, we increased the number of
PSS/PDADMAC bilayers from 4.5 to 5.5 in an effort to improve chloride/phosphate
selectivity. However, the presence of the extra bilayer results in decreased phosphate
rejection (from 86% to 75%) and a drop in chloride/phosphate selectivity from 5.8 to 4.1.
Previous studies of chloride/ﬂuoride3 1 and chloride/sulfate separations38 also showed that
both rejections and selectivities decrease when the number of bilayers of
PSS/PDADMAC ﬁlms on alumina supports increases from 4.5 to 5.5. Streaming
potential measurements with PSS-terminated PSS/PDADMAC ﬁlms indicate that the
surface charge gradually changes from negative to positive as the number of adsorbed
bilayers increases, and this decrease in negative charge should result in less electrostatic
exclusion of the anions and lower rejections.38 However, because both Cl' and H2PO4'
are singly charged anions, their separation is mainly based on sieving of H2PO4', which
has a larger Stokes’ radius than Cl'. Recent studies suggest that the degree of hydration
of PEMs increases after deposition of the ﬁrst few layers.“43 Because the ﬁlms become
more swollen, the effective pore size of the PEMs might increase and causes a decline in
the selectivity.

After adjustment of the feed pH from 5.6 to 7.0, the phosphate ions should consist
of roughly 50% H2P04' and 50% HPO42' (Figure 3.1). Due to the presence of highly
rejected divalent anions on increasing the feed pH from 5.6 to 7.0, the overall phosphate
rejection increases from 86% to 92%, and the chloride/phosphate selectivity improves

from 5.8 to 8.2. Similar to what we observe in NF with a feed pH of 5.6, phosphate

93

 

rejection and chloride/phosphate selectivity decrease when the number of
PSS/PDADMAC bilayers increases from 4.5 to 5.5 at a feed pH of 7.0. The change in
pH does not affect the solution ﬂuxes, presumably because the membrane structure does
not vary signiﬁcantly with pH.

With a further increase in feed pH to 8.4, most phosphate ions exist as HPO42',
which has both a larger size and a higher charge than H2P04' (Table 3.1). As expected,
on increasing the feed pH from 7.0 to 8.4, phosphate rejection increases from 92% to
98%, and the chloride/phosphate selectivity reaches a value of 48 with
(PSS/PDADMAC)4PSS ﬁlms. For (PSS/PDADMAC)3PSS ﬁlms, the chloride/phosphate
selectivity is 4.3, and the phosphate rejection is only 77%. This result indicates that ﬁlms
of (PSS/PDADMAC)3PSS might not fully cover the support. Different from the results
obtained at lower feed pH values, phosphate rejection remains almost the same when the
number of PSS/PDADMAC bilayers increases from 4.5 to 5.5. This result suggests that
HPO42' has enough charge and size to be rejected by (PSS/PDADMAC)5PSS ﬁlms on
alumina supports.

3.3.3. NF Experiments with Commercial NF Membranes

For comparison, we investigated the separation performance of two commercially
available NF membranes, NF 90 and NF 270 from Dow-FILMTEC. These are both
polyamide thin-ﬁlm composite membranes with a negative surface charge, but NF 90 is a
denser membrane that typically gives higher salt rejection and lower ﬂuxes than NF 270.
Table 3.3 shows the solution ﬂuxes, solute rejections and chloride/phosphate selectivities

from NF experiments with these two membranes.

94

 

Table 3.3. Solution ﬂuxes, rejections, and selectivities from NF with commercially

available membranes and feed solutions containing NaHzP04 (1 mM) and NaCl (1 mM).

The transmembrane pressure was 4.8 bar.

 

pH of feed solution = 5.6

 

 

 

 

 

 

 

 

 

Flux Chloride Phosphate Chloride/Phosphate P
Membrane (m3/m? day) Rejection (%) Rejection (%) Selectivity ('l
NF 90 1.4-1.0.1 79.3:3.7 98.7i0.7 22.0i-12.5 :
NF 270 1.1:0.2 55.5:11.7 87.6:05 36:08
pH of feed solution = 7.0 l”
Flux Chloride Phosphate Chloride/Phosphate
Membrane (m3/m2 day) Rejection (%) Rejection (%) Selectivity (-)
NF 90 131-0.3 97.4104 99.3:05 5.4i2.8
NF 270 15:01 61.8:98 91.2:4.3 4.9:1.3
pH of feed solution = 8.4
Flux Chloride Phosphate Chloride/Phosphate
Membrane (mil/m2 day) Rejection (%) Rejection (%) Selectivity (-)
NF 90 12:01 97.9i0.6 99.7:0.1 7.4:|:1.1
NF 270 20:01 71 2:71 94.8:12 55:02

 

95

At a feed pH of 5.6 where most of the phosphate ions are present as H2PO4', NF
90 shows 99% phosphate rejection, 79% chloride rejection and a chloride/phosphate
selectivity of 22. In the case of NF 270, phosphate and chloride rejections are 88% and
56%, respectively, and the selectivity is only 4. Although phosphate rejections are higher
with both of these membranes than with the PSS/PDAMAC-modiﬁed alumina
membranes, NF 90 and NF270 membranes are less permeable than porous alumina coated
with PSS/PDADMAC ﬁlms, perhaps because the polyamide skin layers of the
commercial membranes are denser than PSS/PDADMAC ﬁlms. (At the relatively low 1
mM solute concentrations in the present study, osmotic pressure is less than 0.1 bar at
100% rejection. This value is negligible compared to the transmembrane pressure of 4.8
bar.)

On going from a feed pH of 5.6 to 7.0, phosphate rejection increases with both NF
90 and NF 270 membranes (Table 3.3) due to the fact that half of the phosphate ions exist
as HP042' (Figure 3.1). Chloride rejection also increases in both cases, probably because
the membrane surface became more negatively charged at the higher pH. Although more
than 99% of phosphate is rejected by NF 90 membranes, the selectivity is only 5 because
the chloride rejection is 97%. A similar selectivity occurs with NF 270.

After increasing the feed pH to 8.4, phosphate rejection with NF 90 and NF 270
increases even more because most of the phosphate ions are present as HP042'.
Speciﬁcally, NF 90 membranes reject more than 99.5% of phosphate with a relatively
low ﬂux of 1.2 m3/(m2 day), and NF 270 exhibits a phosphate rejection of 95% with a
ﬂux of 2 m3/(m2 day). However, chloride/phosphate selectivity is less than 8 in both

cases due to high chloride rejection. In summary, these two commercial NF membranes

96

 

present high rejections for both phosphate and chloride, which makes them attractive in
complete salt removal applications. In contrast, the combination of high phosphate
rejection, low chloride rejection, and high ﬂux make PSS/PDADMAC ﬁlms particularly

promising in selective phosphate removal in the presence of monovalent anions.

3.4. Conclusions

Layer-by-layer adsorption of PSS/PDADMAC ﬁlms on porous alumina supports
provides a simple way to fabricate high-ﬂux NF membranes for selective removal of
phosphate in the presence of chloride. Because a higher fraction of the phosphate ions
are present as HPOaz' when the pH of the feed solution increases from 5.6 to 7.0 or 8.4,
phosphate rejections and chloride/phosphate selectivities of PSS/PDADMAC membranes
increase with the feed pH. Although the two commercial NF membranes we examined
(NF 90 and NF 270) exhibit high phosphate and chloride rejections, the higher ﬂuxes and
chloride/phosphate selectivities of PSS/PDADMAC membranes make them potentially

attractive in selective phosphate removal applications.

97

Pin

 

3.5. References

(1) Hong, S. U.; Ouyang, L.; Bruening, M. L. J. Membr. Sci. 2009, 32 7, 2-5.

(2) Chitrakar, R.; Tezuka, S.; Sonoda, A.; Sakane, K.; Ooi, K.; Hirotsu, T. J. Colloid

Interface Sci. 2006, 297, 426-433.

(3) Zhang, G.; Liu, H.; Liu, R.; Qu, J. J. Colloid Interface Sci. 2009, 335, 168-174.
(4) Rodrigues, L. A.; da Silva, M. L. C. P. Colloids Surf, A 2009, 334, 191-196.
(5) Saha, B.; Chakraborty, S.; Das, G. J. Colloid Interface Sci. 2009, 331, 21-26.
(6) Mueller, D. K.; Helsel, D. R. In US. Geological Survey Circular, 1136, 1999.

(7) Kumar, M.; Badruzzaman, M.; Adham, S.; Oppenheimer, J. Water Res. 2007, 41,
221 1-2219.

(8) Hosni, K.; Ben Moussa, S.; Chachi, A.; Ben Amor, M. Desalination 2008, 223,
337-343.

(9) Katz, 1.; Dosoretz, C. G. Desalination 2008, 222, 230-242.

(10) Clark, T.; Stephenson, T.; Pearce, P. A. Water Res. 1997, 31 , 2557-2563.
(11) de-Bashan, L. E.; Bashan, Y. Water Res. 2004, 38, 4222-4246.

(12) Barat, R.; van Loosdrecht, M. C. M. Water Res. 2006, 40, 3507-3516.

(13) Xu, X.; Gao, B.; Wang, W.; Yue, Q.; Wang, Y.; Ni, S. Colloids Surf, B 2009, 70,
46-52.

(14) Balannec, B.; Vourch, M.; Rabiller-Baudry, M.; Chaufer, B. Sep. Purif T echnol.
2005, 42, 195-200.

(15) Petersen, R. J. J. Membr. Sci. 1993, 83, 81-150.

(16) Eriksson, P.; Kyburz, M.; Pergande, W. Desalination 2005, 184, 281-294.
(17) Liu, X.; Bruening, M. L. Chem. Mater. 2004, 16, 351-357.

(18) Miller, M. D.; Bruening, M. L. Langmuir 2004, 20, 11545-11551.

(19) Stanton, B. W.; Harris, J. J.; Miller, M. D.; Bruening, M. L. Langmuir 2003, 19,
7038-7042.

98

1.“

 

F”—

(20) Leviisalmi, J.-M.; McCarthy, T. J. Macromolecules 1997, 30, 1752-1757.
(21) Krasemann, L.; Tieke, B. Langmuir 2000, 16, 287-290.
(22) Hong, S. U.; Bruening, M. L. J. Membr. Sci. 2006, 280, 1-5.

(23) Tieke, B.; Toutianoush, A.; Jin, W. Adv. Colloid Interface Sci. 2005, 116, 121-
131.

(24) Wang, D.-X.; Su, M.; Yu, Z.-Y.; Wang, X.-L.; Ando, M.; Shintani, T.
Desalination 2005, I 75, 219-225.

 

:-
(25) Wang, X.-L.; Tsuru, T.; Nakao, S.; Kimura, S. J. Membr. Sci. 1997, 135, 19-32. 5
(26) Dubas, S. T.; Schlenoff, J. B. Macromolecules 1999, 32, 8153-8160.

(27) Antipov, A. A.; Sukhorukov, G. B.; Mohwald, H. Langmuir 2003, 19, 2444-2448.

(28) Harris, J. J.; Stair, J. L.; Bruening, M. L. Chem. Mater. 2000, 12, 1941-1946. L

(29) Choi, J .; Rubner, M. F. Macromolecules 2004, 38, 116-124.

(30) Hong, S. U.; Malaisamy, R.; Bruening, M. L. J. Membr. Sci. 2006, 283, 366-372.
(31) Hong, S. U.; Malaisamy, R.; Bruening, M. L. Langmuir 2007, 23, 1716-1722.
(32) Handbook of Chemistry and Physics, CRC Press, Boca Raton, 2000-2001.

(33) Smith, R. M.; Martel], A. E. Critical stability constants; Plenum Press: New York,
1976; Vol. 4.

(34) Ouyang, L.; Malaisamy, R.; Bruening, M. L. J. Membr. Sci. 2008, 310, 76-84.
(35) Malaisamy, R.; Bruening, M. L. Langmuir 2005, 21 , 10587-10592.
(36) Miller, M. D.; Bruening, M. L. Chem. Mater. 2005, I 7, 5375-5381.

(37) Hong, S. U.; Miller, M. D.; Bruening, M. L. Ind. Eng. Chem. Res. 2006, 45, 6284-
6288.

(3 8) Adusumilli, M.; Bruening, M. L. Langmuir 2009, 25, 7478-7485.

(39) Hubsch, E.; Fleith, G.; Fatisson, J .; Labbe, P.; Voegel, J. C.; Schaaf, P.; Ball, V.
Langmuir 2005, 21, 3664-3669.

99

(40) Kujawa, P.; Moraille, P.; Sanchez, J.; Badia, A.; Winnik, F. M. J. Am. Chem. Soc.
2005, 12 7, 9224-9234.

(41) Séilomaki, M.; Laiho, T.; Kankare, J. Macromolecules 2004, 37, 9585-9590.

(42) Ladarn, G.; Schaad, P.; Voegel, J. C.; Schaaf, P.; Decher, G.; Cuisinier, F.
Langmuir 2000, I 6, 1249-1255.

(43) Tanchak, O. M.; Yager, K. G.; Fritzsche, H.; Harroun, T.; Katsaras, J .; Barrett, C.
J. Langmuir 2006, 22, 5137-5143.

100

 

Chapter 4
Catalytic Hollow-Fiber Membranes Prepared Using
Layer-by-Layer Adsorption of Polyelectrolytes and

Metal Nanoparticles

This work has recently been submitted for publication in Catalysis Today.

4.1. Introduction

In the 1990’s, Decher and coworkers developed layer-by-layer (LBL) adsorption
of polyelectrolytes for the formation of functional thin ﬁlms. This technique allows
incorporation of a wide range of materials including polymers, lipids, nanoparticles,
proteins and dye molecules, into thin ﬁlms."4 Chapters 2 and 3 described the use of LBL
adsorption to create nanoﬁltration membranes for selective ion removal. In a further
demonstration of the versatility of the LBL technique, Chapters 4 and 5 explore layer-by-
layer adsorption of polyelectrolytes and metal nanoparticles to create catalytic
membranes.

Hollow ﬁber membrane reactors, which initially appeared in the 19808,5 are
attractive because their geometry provides a high surface area to volume ratio‘"'8 and

9-11

should facilitate gas-liquid reactions. Moreover, the hollow ﬁber module is the

membrane conﬁguration with the highest packing density. Remarkably, the ratio of

membrane surface area to module volume can reach values as high as 30 000 mz/m3.l2

101

 

Thus, a number of recent studies examine hollow ﬁbers as catalyst supports in membrane
reactors.9’l3'15
Metal nanoparticles are especially effective catalysts because of their relatively

16.17

large percentage of surface atoms. Additionally, these materials often possess

unusual electronic properties due to their unique size, which is between the bulk and

molecular regimes.”18

For most practical catalytic applications, however, the
nanoparticles must be immobilized on solid supports to prevent aggregation and facilitate
catalyst recovery.'6"9'25 Common methods for making supported noble metal catalysts
include impregnation, coprecipitation (CP), and deposition-precipitation (DP).26 In the
case of impregnation, the support is typically immersed in a metal salt solution, heated in
air, and the absorbed metal species are ﬁnally reduced in hydrogen.”29 For the CP 30’3 l
or DP methods,”32 the noble metal precursor is either coprecipitated with the support
precursor or directly deposited on the support surface prior to precipitation with another
reagent. These methods usually yield metal particles of nonuniform size and shape, and
the degree of catalyst dispersion depends on various factors such as the type of support,
the deposition pH, and the concentration of precursors in the solutions.2(”27’33 To make
heterogeneous catalysts with more control over particle size and shape, metal
nanoparticles can be synthesized in solution and then loaded on the support .2034

LBL deposition of polycations and polyanions provides a particularly simple way
to deposit a high density of well-deﬁned metal nanoparticles on a wide range of

substrates.”39

Alternating adsorption of charged nanoparticles and an oppositely
charged polyelectrolyte affords control over the amount of nanoparticles deposited

through variation of the number of adsorbed layers, and a wide range of particle sizes and

102

 

compositions are available through tuning of the nanoparticle synthesis.”44 The LBL
technique is effective for membrane modiﬁcation because it can occur on a wide variety
of substrate geometries,45 including ﬂat surfaces, cylindrical particles, and the pores of
ﬂat sheet membranes.35’4°‘4 1’44

This chapter describes modiﬁcation of hollow ﬁber microﬁltration (MF)
membranes through LBL adsorption of polyelectrolytes and citrate-stabilized metal
nanoparticles to create catalytic reactors (Figure 4.1). The use of MP substrates with
sufﬁciently large pores allows ﬁlm formation without pore blockage. SEM images show
that adsorption of poly(styrenesulfonate)/poly(allylamine hydrochloride)/Au-nanoparticle
(PSS/PAH/AuNP) ﬁlms results in unifome distributed nanoparticles on the exterior and
pore walls of both polyethersulfone (PBS) and polysulfone (PS) MF hollow ﬁber
membranes. These nanoparticles are highly active in catalyzing the reduction of 4-

nitrophenol with NaBHa to give 4-aminophenol, but conversion decreases gradually with

time probably because of catalyst fouling.

103

 

   

8H;

- N02 T02
4' OH 3H4- OH

 
   

Hollow Fiber Membrane

Polyelectrolyte-
AuNP ﬁlm

Figure 4.1. Schematic diagram of the cross section of a porous hollow ﬁber membrane
modiﬁed with a polyelectrolyte/AuNP ﬁlm. In this study, the ﬁbers serve as catalytic

reactors for the reduction of 4-nitrophenol by NaBH4.

4.2. Experimental Section
4.2.1. Materials

Poly(sodium 4-styrenesulfonate) (PSS, Mw = 70,000 Da), poly(allylamine
hydrochloride) (PAH, Mw= 56,000 Da), HAuCl4 - 3H20, sodium citrate, NaBH4 and 4-
nitrophenol were used as received from Sigma-Aldrich. Deionized water (Milli-Q
puriﬁcation system, 18.2 MQ-cm) was employed for membrane rinsing and preparation
of polyelectrolyte and reaction solutions.

PES MF hollow ﬁber membranes (MicroPES Capillary membrane Type TF10)
were manufactured by MEMBRANA (Wuppertal, Germany). These membranes have a

speciﬁed maximum pore size in the membrane skin of 0.5 i 0.1 um, an inner diameter of

104

 

300:40 um, and a wall thickness of 100i25 urn. Figure 4.2 shows the SEM images of

the ﬁber cross section, the shell and lumen surfaces, and the interior of the ﬁber.

 

Figure 4.2. SEM images of PBS hollow ﬁber membranes: (a) cross-sectional surface, (b)

shell surface, (c) lumen surface, and (d) the interior of the membrane wall.

A module containing PS MF hollow ﬁber membranes was purchased from GE
Healthcare (Model #: CFP-6-D-6A), and individual ﬁbers were removed from the module
and repotted. These ﬁbers have an average skin pore size of 0.65 pm, an inner diameter
of 750 um, and a wall thickness of 225 um. Figure 4.3 contains SEM images of these

ﬁbers.

105

 

Figure 4.3. SEM images of PS hollow ﬁber membranes: (a) cross-sectional surface, (b)

shell surface, (c) lumen surface, and (d) the interior of the membrane wall.

4.2.2. Preparation of Citrate-Stabilized Au Nanoparticles

Au nanoparticles were prepared by the citrate reduction method,18’46’47 where
citrate serves as both the reducing agent and the nanoparticle stabilizer. In a 250 mL
round-bottom ﬂask equipped with a condenser, 100 mL of 1 mM HAuCl4 was heated to a
rolling boil while stirring. Ten mL of 38.8 mM sodium citrate solution was also heated to
a rolling boil and then rapidly added to the Au solution. Aﬁer about 15 s, the solution
changed from colorless to blue and then burgundy. The mixture was subsequently stirred
with heating for another 10 min, and then without heating for 15 min. The nanoparticle

solution was kept in an amber glass bottle and stored in a reﬁigerator until use. Prior to

106

particle deposition, the nanoparticle solution was diluted 9:1 with deionized water. TEM

images show that the particle diameter is 12 i1 nm.“

4.2.3. Modification of the Hollow Fiber Membranes

Following a literature procedure, the hollow ﬁbers were potted in a poly(vinyl
chloride) (PVC) tube with epoxy glue purchased from Axson Technologies (Cat. Number:
Adekit A 130).48 All the modules in this study contained one ﬁber, with an effective
ﬁber length (distance between the potting glue on the two ends) of about 22 cm for PBS
ﬁbers and 15 cm for PS ﬁbers. Prior to modiﬁcation, the potted PES ﬁber was rinsed
with water at 0.7 bar for 30 min. In the modiﬁcation steps, all the solutions were passed
through the membrane in the lumen to shell ﬂow conﬁguration with cross ﬂow (see
Figure 4.4) using a pressure of 0.7 bar. Cross ﬂow was used during modiﬁcation in an

effort to minimize any pore blockage caused by the polyelectrolytes and nanoparticles.

 

——>_.‘ - -

Solution L

Inlet ’
1 Permeate

‘ ___) Retentate
Outlet

 

 

Outlet

Figure 4.4. Schematic diagram of the ﬂow conﬁguration for modifying a hollow ﬁber
membrane with a PSS/PAH/AuNP ﬁlm. Flow occurs both axially along the membrane

and from the inside to the outside of the ﬁber.

A preﬁlter (2.5 pm Ahlstrom qualitative ﬁlter paper) was used to remove
particles that can potentially block the ﬁber, and this ﬁlter paper was replaced after

deposition of each layer. Polyelectrolyte adsorption solutions contained 2 mM PSS or

107

 

PAH and 50 mM NaCl (polyelectrolyte concentrations are always given with respect to
the repeating unit). The modiﬁcation started with passage of 100 mL of PSS solution
through the ﬁber (volume includes both transmembrane and cross ﬂow). This initial
adsorption of PSS is likely driven by hydrophobic interactions between PSS and the PES
ﬁber.49’50 The ﬁber was then rinsed with 300 mL of water to remove any extra
polyelectrolytes. Subsequently, 100 mL of PAH solution was passed through the hollow
ﬁber membrane followed by another 300-mL water rinse. Nanoparticles were loaded by
passing 100 mL of citrate-stabilized Au colloid solution (~0.1 mM in Au atoms) through
the membrane prior to rinsing with 300 mL of water. The permeate of the nanoparticle
solution was initially colorless and then became light pink when the ﬁber was nearly
saturated with Au nanoparticles.

Modiﬁcation of the PS ﬁbers occurred similarly with slight modiﬁcations due to
the relatively high ﬁber permeabilities. Prior to polyelectrolyte deposition, the potted PS
ﬁber was rinsed with water at 0.34 bar for 5-10 min. All the solutions were passed
through the membrane in the same ﬂow conﬁguration (Figure 4.4) as for the PES ﬁber,
but at a lower pressure of 0.34 bar, and a back pressure regulator (valve) was installed in
the lumen ﬂow after the membrane to maintain comparable permeate and retentate ﬂow
rates. The polyelectrolyte adsorption solutions contained 0.8 mM PSS or PAH and 20
mM NaCl. The volumes of polyelectrolyte solutions (250 mL) were higher for
modiﬁcation of PS ﬁbers than for modiﬁcation of PBS ﬁbers because of a higher ﬂow
rate through the PS ﬁbers. Due to this higher volume, we also decreased the
concentration of PSS and PAH for PS modiﬁcation. 500 mL of water was used for

rinsing between each layer deposition, and 400 mL of citrate-stabilized Au colloid

108

 

solution (~0.1 mM in Au atoms) was passed through the PS ﬁber to load Au
nanoparticles.
4.2.4. Characterization of the Hollow Fiber Membranes

SEM images of bare and coated hollow ﬁber membranes were taken using a
Hitachi S-4700 II ﬁeld-emission scanning electron microsc0pe (F ESEM). For shell and
lumen surface characterization, 5 nm of Au was sputter-coated (Pelco SC-7 auto sputter
coater) on the surface before imaging. For cross-sectional images, the ﬁber was dipped
in ethanol for 2 min, fractured in liquid nitrogen, and the cross section was sputter coated
with 5 nm of Au.

To determine the amount of Au loaded in the ﬁber, the membrane was immersed
in aqua regia (3 parts HCl and 1 part HNO3) for 5 min, and this solution was then diluted
and analyzed for Au by atomic absorption spectroscopy (AAS) (V arian Spectra Atomic
Absorption-200 Spectrophotometer).

The water permeability of the PES ﬁber was tested in a shell to lumen ﬂow
conﬁguration (no cross ﬂow in the shell) at room temperature with pressures ranging
from 0.7 to 2.1 bars. All reported ﬂuxes are calculated with respect to the surface area of
the outside of the ﬁber. Hydraulic permeability values, which were obtained from the
slopes of plots of ﬂux versus pressure, were determined before and after each
modiﬁcation step. For the PS ﬁber, water permeability was determined only at 0.7 bar
because the membrane manufacturer does not recommend using the ﬁbers under
pressures higher than 1 bar.

4.2.5. Catalytic Reactions and Leaching Studies

109

 

Reduction of aqueous 4-nitrophenol with sodium borohydride served as a model
reaction for examining the catalytic activity of the modiﬁed membranes. A feed tank was
ﬁlled with 0.5 or 1 L of a freshly prepared solution containing 0.5 mM 4-nitrophenol and
25 mM NaBHa, and this solution was forced through the hollow ﬁber membrane under
pressure. Permeate samples were collected every 5 or 10 min, and the concentration of 4-
nitrophenol in the feed and permeate was determined from the solution absorbance at 400
rim.“5 1 (UV-Vis spectra were acquired with a Perkin-Elmer Lambda 40
spectrophotomer.) To investigate the extent of catalyst leaching during the reaction, the
amount of Au in the feed and permeate samples was determined using inductively
coupled plasma optical emission spectroscopy (ICP-OES) (V arian 710-ES ICP Optical

Emission Spectrometer).

4.3. Results and Discussion
4.3.1. Membrane Characterization

The ﬁrst sign of successful modiﬁcation of the hollow ﬁber membranes with An
nanoparticles is a change in the ﬁber color from white to red. SEM images, elemental
analysis by AAS, and water permeability measurements afford a more quantitative
picture of the ﬁber modiﬁcation.
4.3.1.1. SEM Characterization of Immobilized Nanoparticles

Figure 4.2 shows SEM images of bare PES hollow ﬁber membranes. These ﬁbers
have a layered, porous structure on both lumen and shell sides, but pore sizes are smallest
at the shell surface and become larger in the interior of the ﬁber wall, and small again at

the lumen surface. The extensive internal surface area of the membrane pores should

110

 

allow a high loading of Au nanoparticles, and the relatively large pores (maximum pore
size around 2 pm) facilitates ﬁlm adsorption without pore blockage. However, pore
blockage sometimes occurs during ﬁlm adsorption, especially in ﬁber sections that
contain somewhat smaller pores (see below). Figure 4.5 shows SEM images of the shell,
lumen, and interior surfaces of PES hollow ﬁber membranes modiﬁed with

PSS/PAH/AuNP ﬁlms.

 

Figure 4.5. SEM images of PES hollow ﬁber membranes coated with PSS/PAH/AuNP

ﬁlms: (a) shell surface, (b) lumen surface, and (c) the interior of the membrane wall.

Adsorption of the PSS/PAH/AuNP ﬁlms occurs while ﬂowing solution along the

lumen side of the ﬁber and also through the ﬁber wall (Figure 4.4). Consequently,

nanoparticles appear on all surfaces of the membranes. The nanoparticles are well
separated with no visible aggregation, and the particle size is consistent with TEM
images of the as-prepared particles.4°‘4' Figure 4.6 shows that modiﬁcation of PS ﬁbers

with PSS/PAH/AuNP ﬁlms also results in a uniform Au nanoparticle distribution.

 

Figure 4.6. SEM images of PS hollow ﬁber membranes coated with PSS/PAH/AuNP

ﬁlms: (a) shell surface, (b) lumen surface, and (c) the interior of the membrane wall.

4.3.1.2. Catalyst Loading
SEM provides information on nanoparticle deposition in small spots with surface
areas of 10’s of umz, but these images may not give a good indication of Au loading on a

macroscopic scale. To determine the overall Au content in a membrane and whether the

112

nanoparticles are evenly distributed along the length of the ﬁber, we cut coated PES
ﬁbers into several pieces and dissolved the nanoparticles in each piece in aqua regia for
subsequent Au analysis by AAS. Table 4.1 summarizes these analyses and shows that

the Au deposition in the PES ﬁber is very uniform.

Table 4.1. Au loading along PES ﬁbers coated with PSS/PAH/AuNP ﬁlms. The length

of each segment is approximately 5 cm.

    

 

 

 

 

 

Segment A Segment C
-' I. - ~12 .s..; ,, .,_ '~;-n--.«t...- :1 .7 :5..- ,_ -,_-._. 3.14:4, .n-s .-
" amt-ﬁﬁ'vrangr ~“"--,..-.:...;-' «.... ' - .33! ... an. .-.
Segment B Segment D
Segment A Segment B Segment C Segment D Average
(nmol) (nmol) (nmol) (nmol) (nmol)
*Fiberl 357 371 347 375 363i13
*Fiber 2 345 339 326 370 345 :1: 18

 

 

 

 

 

 

 

 

* The reported values are normalized to a ﬁber length of 5 cm.

Remarkably, the amount of Au in each piece of the ﬁber differs by less than 8%
from the average loading. The average amount of Au in a 5 cm-long PES ﬁber is 350
nmol, which corresponds to 0.3 mg of Au in a ﬁber with a length of 22 cm. Gold loading
in a PS ﬁber (determined with the same method) is 1.2 mol in a 5 cm segment or about
1 mg in a ﬁber with a length of 22 cm. The higher loading in the PS membranes likely
occurs because the cross-sectional area of the walls of these ﬁbers is 5.5 times that of
PES ﬁbers. The loading in both types of membranes is comparable to that in a chitosan

ﬁber that was modiﬁed by the in-situ reduction method.14 In that case, palladium

113

chloride was absorbed on the chitosan and subsequently reduced by hydrogen gas to give
a Pd loading of 0.25 to 2.5 mg in a 50 cm long ﬁber.l4

Adsorption of additional PAH/Au nanoparticle bilayers should further increase
the Au loading in the ﬁlm.“ However, due to challenges with blocking of some
membranes by the polyelectrolyte ﬁlms (see below), the catalytic studies described here
focus on modiﬁcation with only one PAH/AuNP layer.
4.3.1.3. Water Permeability

Although a previous study showed no detectable plugging of pores during
deposition of poly(acrylic acid)/PAH/AuNP ﬁlms in ﬂat-sheet porous alumina
membranes (pore size of 0.2 um),41 the hollow ﬁber ﬂow conﬁguration and the spongy
membrane structure in the present case may lead to pore blockage. Figure 4.7 shows the
water permeabilities of PES ﬁbers after each step in the deposition of PSS/PAH/AuNP
ﬁlms. (Permeabilities are determined by measuring the pressure dependence of pure

water ﬂux in the shell to lumen ﬂow mode.)

114

 

Water Permeability
(m3/m2 day bar)
N 0) A 01
O O O O

_L
O

       

Bare PSS PSS/PAH PSS/PAH/Au
Figure 4.7. Water permeabilities of PES hollow ﬁber membranes before (bare) and after

adsorption of different polyelectrolyte/AuNP ﬁlms.

The bare PES ﬁber exhibits a water permeability of 41 m3/(m2 day bar), which is
in the high range of ﬂuxes through MF membranes.12 After PSS deposition, water
permeability decreases by less than 10%, but adsorption of one bilayer of PSS/PAH
reduces permeability by 30% compared to a bare ﬁber. Moreover, adsorption of Au
nanoparticles on the PSS/PAH ﬁlm results in a water permeability of only 22 m3/(m2 day
bar), about half of the value for the bare ﬁber. This decrease in water permeability is
somewhat surprising because typical PSS/PAH ﬁlms are only 1 nm thick,52’53 and the
pores sizes in these membranes are in the range of several hundred nanometers.
Comparison of Figure 4.2-(c) and Figure 4.5-(b) (SEM images of lumen surfaces of the
PES ﬁbers before and aﬁer modiﬁcation) suggests that ﬁlm deposition partially occludes
the lumen surface, which may account for the 50% drop in permeability after deposition

of the PSS/PAH/AuNP ﬁlm.

115

 

We should note that in some cases adsorption of PSS/PAH/AuNP ﬁlms in PES
ﬁbers decreases ﬂow to around 7% of that before modiﬁcation. SEM images of different
bare fibers indicate that the pore size varies greatly from roll to roll or perhaps changes
with time (Figure 4.8). As expected, smaller pores are more prone to plugging during
ﬁlm adsorption. To overcome this problem, we employ PS ﬁbers as substrates because
these ﬁbers have a permeability of 360 m3/(m2 day bar), which is about 9 times the
permeability of the PES ﬁbers. Adsorption of PSS/PAH/AuNP ﬁlms in the PS ﬁbers

decreases permeability by less than 10%.

   

Figure 4.8. SEM images of the shell surfaces of bare PES hollow ﬁber membranes taken

ﬁ'om different rolls.

4.3.2. Catalysis with Nanoparticle-Coated Fibers

116

We employed reduction of 4-nitrophenol with NaBI-I4 (Scheme 4.1) as a test
reaction for examining the catalytic activity of hollow ﬁbers loaded with Au
nanoparticles. This reaction is attractive because it does not proceed in the absence of
catalyst, and yet it rapidly occurs in the presence of Au nanoparticles and can be easily
monitored by UV-Vis spectrophotometryf’1’54’55 In a typical reaction, the feed solution
contains 0.5 mM 4-nitrophenol and 25 mM NaBH4 so the reducing agent is in large

excess throughout the reaction.

N02 NI—I2

A
+ NaBH4 ” >

 

OH OH

Scheme 4.1. Au—catalyzed reduction of 4-nitrophenol with NaBHa.

4.3.2.1. Catalytic Reduction Using PES Fibers Coated With PSS/PAH/Au NP Films
We initially performed reduction of 4-nitrophenol in the lumen-shell ﬂow
conﬁguration with the lumen outlet plugged (Figure 4.9). Because most of the catalyst is
in the ﬁber wall, the reaction should primarily occur in the interior of the PES ﬁber.
Figure 4.9 shows the 4-nitrophenol conversion over a 2-h period. Initially, more than
99% of the 4-nitrophenol passing through the membrane is converted to 4-aminophenol

at a permeate ﬂux of 0.24: 0.07 mL/(cm2 min), which corresponds to a residence time of

600i 175 ms in the membrane (assuming 30% porosity). Thus, the nanoparticles are

117

 

highly catalytically active. After 1 h, more than 98% of the 4-nitrophenol passing
through the membrane is still reduced, but the conversion slowly decreases to 80% after
two hours of permeation through the membrane. One possible reason for the decline in
conversion is that the amount of reducing agent in the feed tank decreases with time due
to continuous NaBH4 degradation. A previous study reports that NaBH4 decomposes
completely to NaBOz and hydrogen gas in 20 min at 75°C.56 Although the membrane
reactions occur at room temperature in our studies, an equation developed by Kreevoy
suggests that 50% of the NaBH.; should decompose in approximately one hour.57
Additionally, hydrogen is produced continuously during the reaction, and this gas might
be trapped in the limited space in the ﬁber to induce a drop in the effective membrane
surface area. The permeate ﬂux ﬂuctuated by 40% dming the experiment, perhaps

because of entrapment and release of hydrogen bubbles.

118

 

 

 

 

 

 

 

Inlet

Permeate
Outlet
100 4609000 9 o .
O o
A O
e\°, 80 I ° .
8
3 60
U 1
31’
E 40
8
(“f 20
0 1 _._-____:_ A,

 

0 20 40 60 80 100 120
Reaction Time (min)

Figure 4.9. Percent reduction of 4-nitrophenol during passage of a 0.5 mM 4-nitrophenol,
25 mM NaBH4 aqueous solution through a PBS hollow ﬁber membrane coated with a
PSS/PAH/AuN P ﬁlm. The reaction occurs in the lumen-shell ﬂow conﬁguration without
cross ﬂow as shown in the diagram above the graph. (The two red cylinders represent

outlet plugs.) The effective ﬁber length is 22 cm, and the applied pressure is 1.4 bar.

To prevent trapping of hydrogen gas in the lumen of the ﬁber, the reduction
reaction was carried out in the shell to lumen conﬁguration without cross ﬂow as shown
in Figure 4.10. In this case the ﬁber is perpendicular to the ground, and any trapped
hydrogen should pass through the lumen outlet along with the permeate. The red plug on

the shell (Figure 4.10) can be removed periodically to release H2 that accumulates in the

119

 

shell. With this ﬂow mode, the permeate ﬂux is 0.7i0.1 mL/(cm2 min) with the same
applied pressure of 1.4 bar as in the lumen-shell ﬂow where the ﬂux is only 0.24i0.07
mL/(cm2 min). The residence time in the membrane is only 210:1:30 ms in the shell to

lumen ﬂow mode. Additionally, to minimize the effect of NaBH4 degradation on

conversion, we replaced feed solution every 60 min.

Solution
Outlet

 

Solution
Inlet

—_

[-.-

Figure 4.10. Schematic diagram of the shell to lumen solution ﬂow conﬁguration used in

 

 

 

 

most catalytic reactions in this study. (The two red cylinders represent outlet plugs.)

As Figure 4.11 shows, in the shell to lumen ﬂow conﬁguration, 4-nitrophenol
reduction is more than 99% for 30 min and then decreases to 83% in an hour. After
replacing the feed solution, the 4-nitrophenol conversion recovers to values as high as
96%, but then begins decreasing again, so conversion is only 70% aﬁer two hours. When
the feed is replaced for a second time after 2h, the conversion only recovers to 77% and

ﬁnally drops to 60% after three hours. The incomplete recovery of conversion after

120

 

replacing the feed shows that degradation of NaBH4 is not the sole cause for the

conversion decline.

100 0909909..

I .... I
A O O
°\o l o.;. 6.. :
r“ 80 ~ 0
'0 I .. I .009...
8 l . ....
' I I O
s 60 ~ . .
g l I
-I-0 40 ‘ . .
g I I
I I
I. l. I I
m 20 . . .
0- . .
o _ __ - ___.___1_. .-l -:___-_ F I -0: :. ..__

0 20 40 60 80 100 120 '140 160 180
Reaction Time (min)

Figure 4.11. Percent reduction of 4-nitrophenol during passage of a 0.5 mM 4-
nitrophenol, 25 mM NaBH4 aqueous solution through a PES hollow ﬁber membrane
coated with a PSS/PAH/AuNP ﬁlm. Flow through the membrane occurs in the shell to
lumen conﬁguration without cross ﬂow (Figure 4.10). The effective ﬁber length is 22 cm,
and the applied pressure is 1.4 bar. The dashed, vertical lines represent the addition of a

fresh 4-nitrophenol/NaBH4 solution.

A second possible reason for the conversion decline is that reaction byproducts
block the catalytic sites of the nanoparticles.58 Comparison of SEM images of the lumen
surfaces of modiﬁed PES ﬁbers before (Figure 4.5-(b)) and after the NaBH4/4-
nitrophenol reaction (Figure 4.12) indicates that deposits form during the reaction.

Interestingly, no deposits appear in the ﬁber wall where most of the catalyst is located.

121

 

Another possible cause of the conversion decline, catalyst leaching, is discussed in more

detail below.

 

Figure 4.12. SEM image of the lumen surface of the catalytic hollow ﬁber after use in 4-

nitrophenol reduction.

4.3.2.2. Catalytic Reduction Using PS Fibers Coated With PSS/PAH/Au NP Films
Reduction of 4-nitrophenol also occurs rapidly when using PSS/PAH/AuNP-
modiﬁed PS hollow ﬁber membranes. Figure 4.13 shows 4-nitrophenol conversion over
a 4-h period in which the feed is replaced after 80 and 180 min. Because of the high
permeability of modiﬁed PS membranes, the permeate ﬂux is 1.4-:02 mL/(cm2 min)
under a pressure <0.2 bar. (This ﬂux corresponds to a membrane residence time of
240i30 ms). Conversion is more than 99.9% for the ﬁrst 30 min and decreases to 92%
after 80 min. Once the feed solution is replaced, however, conversion recovers to around
99.9% and remains more than 99% for another 40 min. Conversion drops to 90% after
180 min, and replacing the feed solution a second time does not result in recovery of the

catalytic activity.

122

 

 

100 F9090. o o . p000... o . o . I
O ' o I
s 1 I 1..
g 90 H I | .9000 9 ° 0
o I I
s 1 - .
a: 80 it I I
E I I
5 I I
1 I I
g 70 j I |
O. I I
I I
60 1 J,:::: ,_:_1. _|,___ - 3
0 40 80 120 160 200 240
Reaction Time (min)

Figure 4.13. Percent reduction of 4-nitrophenol during passage of a 0.5 mM 4-
nitrophenol, 25 mM NaBI-la aqueous solution through a PS hollow ﬁber membrane
coated with a PSS/PAH/AuNP ﬁlm. The reaction takes place in the shell to lumen ﬂow
conﬁguration without cross ﬂow using an effective ﬁber length of 15.5 cm, and the ﬂux
through the membrane is 1.4 mL/(cm2 min). The dashed, vertical lines represent the

addition of a fresh 4-nitrophenol/NaBI-I4 solution.

To determine whether the conversion decline stems from catalyst leaching, the Au
concentration in the permeate samples was analyzed by ICP-OES. The permeate
contains less than 5 ppb Au throughout the reaction period, which corresponds to less
than 3% leaching of the Au in the membrane. Hence, Au leaching is not a likely cause
for conversion decline in this experiment.

To minimize the effect of NaBH4 degradation on conversion decline, a 4-h 4-

nitrophenol reduction was performed where the feed solution (500 mL of 0.5 mM 4-

123

 

nitrophenol and 25 mM NaBH4) was replaced every 30 min. Figure 4.14 shows the plot
of conversion vs. reaction time at a ﬂux of 1.4 mL/(cm2 min). 4-Nitrophenol conversion
is at least 99.9% for 90 min, and >99.7% for another 65 min. Even after 3 h, more than
99% of the 4-nitrophenol is reduced. However, during the 4th hour of reaction,
conversion drops to between 95 and 98%. The Au concentration in the permeate samples
is again < 5 ppb, showing that conversion declines are not due to leaching of Au. The
maximum 5% conversion decline is signiﬁcantly less than the conversion drop of 12% in
the experiment where the feed solution was only replaced after 80 and 180 min (Figure
4.13). Conversion declines are less severe when the feed is replaced more often, and
when the feed is not replaced, the NaBHa solution becomes less reducing. Thus, the
fouling mechanism may include oxidation of aminophenol in the less reducing solution

and formation of polymer products that foul the catalyst.

124

A 100 roooorooooroooor .ooo".oooof¢....* :
é . I I I I I I.
8 98 1 I I I I I I5 I . .
. I I I 1 l I I o
S 96 L I I I I I I. I. n
'3 j I I 1 I I I ’.¢ol
m l I I I I 1 I I
'E 94 f I 1 l I I I I
8 I I I I I I I
._ I l I I I I I
d: 92 I I I I I I I
I I 1 I I I 1
90 L: ..I. H ”1-, --l. owns... :1- ___.. I ___ _1--
0 30 60 90 120 150 180 210 240

Reaction Time (min)

Figure 4.14. Percent reduction of 4-nitrophenol during passage of a 0.5 mM 4-
nitrophenol, 25 mM NaBH4 aqueous solution through a PS hollow ﬁber membrane
coated with a PSS/PAH/AuNP ﬁlm. The reaction takes place in the shell to lumen ﬂow
conﬁguration without cross ﬂow. The effective ﬁber length is 15.5 cm, and the ﬂux
through the membrane is 1.4 mL/(cm2 min). The dashed, vertical lines represent the

addition of a fresh 4-nitrophenol/NaBHa solution.

4.4. Conclusions

LBL adsorption of polyelectrolytes and Au nanoparticles provides a simple and
effective way to modify the pores of hollow ﬁbers and prepare catalytic membrane
reactors. This procedure results in a dense layer of unaggregated nanoparticles in the
membrane pores, although as pore size decreases, plugging of membranes can be a
problem. Catalytic reduction of 4-nitrophenol with NaBH4 shows that the immobilized
nanoparticles are highly active, but their activity declines over time, possibly because of

membrane fouling from oxidation products of aminophenol. The shell to lumen ﬂow

125

 

conﬁguration is probably better suited for this type of reaction than lumen to shell ﬂow

because in the former case, gas is less likely to build up in the ﬁber.

126

4.5. References
(1) Wang, Y.; Angelatos, A. S.; Caruso, F. Chem. Mater. 2008, 20, 848-858.
(2) Peyratout, C. S.; Dahne, L. Angew. Chem. Int. Ed. 2004, 43, 3762-3783.

(3) Quinn, J. F .; Johnston, A. P. R.; Such, G. K.; Zelikin, A. N.; Caruso, F. Chem.
Soc. Rev. 2007, 36, 707-718.

(4) Ariga, K.; Hill, J. P.; Ji, Q. Phys. Chem. Chem. Phys. 2007, 9, 2319-2340.
(5) Butler, J. S.; Gordon, 3.; Harrison, I. R. J. Appl. Polym. Sci. 1988, 35, 1183-1190.
(6) Kleinert, A.; Feldhoff, A.; Schiestel, T.; Caro, J. Catal. Today 2006, 118, 44-51.

(7) Peters, T. A.; Fontalvo, J.; Vorstrnan, M. A. G.; Keurentjes, J. T. F. Chem. Eng.
Res. Des. 2004, 82, 220-228.

(8) Wang, H.; Tablet, C.; Schiestel, T.; Werth, S.; Caro, J. Catal. Commun. 2006, 7,
907-912.

(9) van der Vaart, R.; Lebedeva, V. 1.; Petrova, I. V.; Plyasova, L. M.; Rudina, N. A.;
Kochubey, D. 1.; Tereshchenko, G. F.; Volkov, V. V.; van Erkel, J. J. Membr. Sci. 2007,
299, 38-44.

(10) Jiang, H.; Wang, H.; Werth, S.; Schiestel, T.; Caro, J. Angew. Chem. Int. Ed. 2008,
37, 9341-9344.

(1 l) Lebedeva, V. 1.; Gryaznov, V. M.; Petrova, I. V.; Volkov, V. V.; Tereshchenko, G.
F.; Shkol’nikov, E. I.; Plyasova, L. M.; Kochubey, D. 1.; van der Vaart, R.; van Soest—
Verecammen, E. L. J. Kinet. Catal. 2006, 47, 867-872.

(12) Mulder, M. Basic Principles of Membrane technology; Kluwer Academic
Publishers: Dordrecht, 1996.

(13) Israni, S. H.; Nair, B. K. R.; Harold, M. P. Catal. Today 2009, 139, 299-311.

(14) Peirano, F .; Vincent, T.; Quignard, F.; Robitzer, M.; Guibal, E. J. Membr. Sci.
2009, 329, 30-45.

(15) Caro, J .; Caspary, K. J .; Hamel, C.; Hoting, B.; Kolsch, P.; Langanke, B.;
Nassauer, K.; Schiestel, T.; Schmidt, A.; Schomacker, R.; Seidel-Morgenstern, A.;
Tsotsas, E.; Voigt, 1.; Wang, H.; Warsitz, R.; Werth, S.; Wolf, A. Ind. Eng. Chem. Res
2007, 46, 2286-2294.

(16) Astruc, D.; Lu, F.; Aranzaes, J. R. Angew. Chem. Int. Ed. 2005, 44, 7852-7872.

127

(17) Pool, R. Science 1990, 248, 1186-1188.
(18) Pachon, L. D.; Rothenberg, G. Appl. Organomet. Chem. 2008, 22, 288-299.

(19) Mei, Y.; Sharma, G.; Lu, Y.; Ballauff, M.; Drechsler, M.; Irrgang, T.; Kempe, R.
Langmuir 2005, 21, 12229-12234.

(20) Comotti, M.; Li, W.-C.; Spliethoff, B.; Schﬁth, F. J. Am. Chem. Soc. 2006, 128,
917-924.

(21) Park, K. H.; Jang, K.; Kim, H. J.; Son, 3. U. Angew. Chem. Int. Ed. 2007,46, “’1.
1152-1155. *

(22) Valden, M.; Lai, X.; Goodman, D. W. Science 1998, 281, 1647-1650.

(23) Jiang, Y.; Gao, Q. J. Am. Chem. Soc. 2006, 128, 716-717.

 

(24) Bengtson, G.; Fritsch, D. Desalination 2006, 200, 666-667. w
(25) Xu, J .; Bhattacharyya, D. J. Phys. Chem. C 2008, 112, 9133-9144.

(26) Zhong, L.-S.; Hu, J.-S.; Cui, Z.-M.; Wan, L.-J.; Song, W.-G. Chem. Mater. 2007,
19, 4557-4562.

(27) Gates, B. C. Chem. Rev. 1995, 95, 511-522.

(28) Sarkar, J .; John, V. T.; He, J .; Brooks, C.; Gandhi, D.; Nunes, A.; Ramanath, G.;
Bose, A. Chem. Mater. 2008, 20, 5301-5306.

(29) Nikolla, E.; Schwank, J .; Linic, S. J. Am. Chem. Soc. 2009, 131, 2747-2754.
(30) Akande, A. J .; Idem, R. 0.; Dalai, A. K. Appl. Catal., A 2005, 287, 159-175.

(31) Kozlov, A. 1.; Kozlova, A. P.; Asakura, K.; Matsui, Y.; Kogure, T.; Shido, T.;
Iwasawa, Y. J. Catal. 2000, 196, 56-65.

(32) Claus, P.; Bruckner, A.; Mohr, C.; Hofmeister, H. J. Am. Chem. Soc. 2000, 122,
11430-11439.

(33) Wolf, A.; Schﬁth, F. Appl. Catal., A 2002, 226, 1-13.
(34) Zheng, N.; Stucky, G. D. J. Am. Chem. Soc. 2006, 128, 14278-14280.

(35) Cho, J .; Caruso, F. Chem. Mater. 2005, 17, 4547-4553.

128

(36) Ostrander, J. W.; Mamedov, A. A.; Kotov, N. A. J. Am. Chem. Soc. 2001, 123,
1101-1110.

(37) Caruso, F.; Caruso, R. A.; Mohwald, H. Science 1998, 282, 1111-1114.
(38) Jiang, C.; Markutsya, S.; Tsukruk, V. V. Langmuir 2004, 20, 882-890.

(39) Kidambi, S.; Dai, J.; Li, J.; Bruening, M. L. J. Am. Chem. Soc. 2004, 126, 2658-
2659.

(40) Dotzauer, D. M.; Bhattacharjee, S.; Wen, Y.; Bruening, M. L. Langmuir 2009, 25 ,
1865-1871.

(41) Dotzauer, D. M.; Dai, J .; Sun, L.; Bruening, M. L. Nano Lett. 2006, 6, 2268-2272.
(42) Bhattacharjee, S.; Bruening, M. L. Langmuir 2008, 24, 2916-2920.

(43) Bhattacharjee, S.; Dotzauer, D. M.; Bruening, M. L. J. Am. Chem. Soc. 2009, 131,
3601-3610.

(44) Kidambi, S.; Bruening, M. L. Chem. Mater. 2004, 1 7, 301-307.

(45) Schdnhoﬁ‘, M. J. Phys. .' Condens. Matter 2003, 15.

(46) Enﬁstﬁn, B. V.; Turkevich, J. J. Am. Chem. Soc. 1963, 85, 3317-3328.

(47) Kumar, S.; Gandhi, K. S.; Kumar, R. Ind. Eng. Chem. Res 2007, 46, 3128-3136.
(48) Rouaix, S.; Causserand, C.; Aimar, P. J. Membr. Sci. 2006, 277, 137-147.

(49) Malaisamy, R.; Bruening, M. L. Langmuir 2005, 21, 10587-10592.

(50) Kotov, N. A. Nanostruct. Mater. 1999, 12, 789-796.

(51) Pradhan, N.; Pal, A.; Pal, T. Colloids Surf, A 2002, 196, 247-257.

(52) Ouyang, L.; Malaisamy, R.; Bruening, M. L. J. Membr. Sci. 2008, 310, 76-84.
(53) Miller, M. D.; Bruening, M. L. Chem. Mater. 2005, 17, 5375-5381.

(54) Esumi, K.; Miyamoto, K.; Yoshimura, T. J. Colloid Interface Sci. 2002, 254, 402-
405.

(55) Ghosh, S. K.; Manda], M.; Kundu, S.; Nath, S.; Pal, T. Appl. Catal., A 2004, 268,
61-66.

129

(56) Soler, L.; Macanas, J .; Muﬁoz, M.; Casado, J. Int. J. Hydrogen Energy 2007, 32,
4702-4710.

(57) Kreevoy, M. M.; Jacobson, R. W. Ventron Alembic 1979, 15, 2-3.

(58) Vincent, T.; Guibal, E. Langmuir 2003, 19, 8475-8483.

 

130

Chapter 5
Reduction of 4-Nitrophenol with Sodium F oérmate Using Catalytic

Flat-Sheet Membranes Containing Pd Nanoparticles

5.1. Introduction

Alternating adsorption of polyelectrolyte/metal nanoparticle ﬁlms in porous
supports is a convenient method to prepare membrane reactors with high catalytic activity,
and this technique also affords control over the amount of catalyst deposited. Another
prominent advantage of this membrane-modiﬁcation method is its applicability to
deposition of various catalysts including Au, Pd, Pt, and Ag nanoparticles."7 The
previous chapter demonstrated both the preparation of hollow ﬁber membrane reactors by
alternating adsorption of polyelectrolytes and Au nanoparticles and the high catalytic
activity of these nanoparticles in the reduction of 4-nitrophenol by NaBH4. Although
NaBH4 is a very effective reducing agent, it is highly corrosive and not safe for use in
closed systems due to the continuous release of hydrogen gas. A previous study reported
that aqueous NaBH4 decomposes completely to NaBOz and hydrogen gas in 20 min at
75°C,8 hence, it is necessary to add fresh NaBH4 periodically for reductions that take
place over extended periods (>30 min in some cases).

Catalytic transfer hydrogenation with the aid of hydrogen donors is a useful
alternative to NaBH4 reduction or classical hydrogenation by H19” The use of a
hydrogen donor rather than H2 avoids the risks and constraints associated with hydrogen

gas.11 Among various hydrogen donors, formic acid and its salts are uniquely reactive,

131

 

due to the C02 release and subsequent bicarbonate formation that accompany hydrogen
donation.12 Alkali metal formate salts are of particular interest because they are mild in
nature and readily available.13 In addition, formate salts are recyclable hydrogen donors
because their dehydrogenated product, bicarbonate salts, can be separated from the
reaction mixture and subsequently rehydrogenated to formate under mild reaction
conditions.14 Equation (1) shows the typical catalytic transfer hydrogenation process for
reaction of nitroarene and formate, whereas equation (2) represents the regeneration of

formate by hydrogenation of the bicarbonate product from equation (1).15

ArN02+ 3Hco,-+ H20 3911i ArNH2+ 3Hco,- (1)

HC03'+ H2 :2 HC02‘+ H20 (2)
In this chapter we use HCOONa, a safer, more cost-effective, and more
environmental friendly compound than NaBI-I4, for reduction of 4-nitrophenol catalyzed
by Pd nanoparticles in membranes. F lat-sheet membranes serve as the nanoparticle

supports in these preliminary studies.

5.2. Experimental Section
5.2.1. Materials

Poly(sodium 4-styrenesulfonate) (PSS, MW = 70,000 Da), poly(allylamine
hydrochloride) (PAH, Mw = 56,000 Da), poly(ethyleneimine) (PEI, branched, Mw =
25,000 Da), KdeCIG, sodium citrate, sodium formate, and 4-nitrophenol were used as
received ﬁ'om Sigma-Aldrich. Poly(acrylic acid) (PAA, Mw=5,000 Da) was purchased

from Polysciences and was used without any further puriﬁcation. Deionized water

132

(Milli-Q puriﬁcation system, 18.2 MQ-cm) was employed for membrane rinsing and
preparation of polyelectrolyte or reaction solutions. The pH of the polyelectrolyte
solutions was adjusted with dilute HCl or NaOH.

Anodisc alumina membranes with a pore size of 0.2 um and a membrane
thickness of 60 pm were purchased from Whatman. For pretreatment, these membranes
were UV/O3 cleaned (Boekel UV-Clean Model 1135500) with the feed side up for 15
min. Poly(ethersulfone) (PES) microﬁltration (MF) ﬂat-sheet membranes were
purchased from Sterlitech Corporation (cat. number PE50225100). These membranes
have a skin pore size of 0.2 pm and a membrane thickness of 110-150 um. Prior to
modiﬁcation, about 30 mL of deionized water was passed through these membranes
using the setup depicted in Figure 5.3. Figure 5.1 shows the SEM images of the skins

and cross-sectional surfaces of the alumina and PBS membranes.

133

 

a,
E
%

 

Figure 5.1. SEM images of alumina (a and b) and PBS (c and d) ﬂat-sheet membranes.

Images (a) and (c) show skin layers, and images (b) and (d) are cross sectional views of

the membranes.

5.2.2. Preparation of Citrate-Stabilized Pd Nanoparticles

The Pd nanoparticles were prepared using the well-established citrate reduction
method.3 In a 100 mL round-bottom ﬂask equipped with a condenser, 50 mL of an
aqueous 0.393 mM solution of potassium hexachloropalladate(IV) was heated to reﬂux.
A 1% sodium citrate solution (6 mL) was added to the Pd solution, and the mixture was
reﬂuxed continuously for 4 h. The resulting dark brown nanoparticle solution was kept

in an amber glass bottle and stored in a refrigerator until use. Prior to particle deposition,

134

 

the nanoparticle solution was diluted 4:1 with deionized water. TEM studies show that

Pd nanoparticles prepared by this method have a diameter of 7 :1: 1 nm (Figure 5.2)”.

b)

Number

 

456789101112
Nanoparticle Diameter (nm)

 

Figure 5.2. a) TEM image of citrate-stabilized Pd nanoparticles. b) Histogram of Pd

nanoparticle diameters

5.2.3. Modification of Flat-Sheet Membranes

Both alumina and PBS membranes were placed in an ultraﬁltration cell (Amicon
stirred cell 8010, Millipore, exposed membrane area of 3.1 cm?) for modiﬁcation and
testing. A peristaltic pump (Cole—Farmer Masterﬂex C/L) attached to the permeate side
of the cell pulled all the solutions through the membranes, and the ﬂow rate was set
between 3 and 4 mL/min (Figure 5.3). Deposition started with passage of 10 mL of 0.02
M PAA solution (pH adjusted to 4.5) for alumina membranes and 10 mL of 0.02 M PSS
solution (no pH adjustment, pH 6.2) for PBS membranes (polyelectrolyte concentrations
are always given with respect to the repeating unit). Thirty mL of deionized water was
then pumped through the membrane to remove any excess polyelectrolyte. Subsequently,
10 mL of 0.02 M PAH (pH adjusted to 4.5) or 0.02 M PEI (pH adjusted to 9) was passed

through the membrane followed by another water rinse. Pd nanoparticles were then

135

loaded by passage of a citrate-stabilized Pd colloid solution (0.1 mM in Pd atoms)
through the membrane until the permeate color changed from colorless to pale brown.
(The volume of Pd colloid loading solution was about 50 and 100 mL for PAH-
terminated and PEI-terminated ﬁlms, respectively.) Finally, the coated membrane was
rinsed with water and stored until use. All the polyelectrolyte solutions contained 0.5 M

NaCl.

     

Membrane

Peristaltic pump

Figure 5.3. Apparatus employed for membrane modiﬁcation and catalytic reactions.

5.2.4. Characterization of the Flat-Sheet Membranes

SEM images of bare and coated ﬂat-sheet membranes were taken using a Hitachi
S-4700 II ﬁeld-emission scanning electron microscope (FESEM). For skin surface
characterization, 5 nm of Au was sputter-coated (Pelco SC-7 auto sputter coater) on the
surface before imaging, whereas for cross-sectional images, both the cross section and the
skin surface were sputter coated with 5 nm of Au. Membrane cross-sections were
prepared as described in chapter 2.

To determine the amount of Pd loading, the membrane was immersed in 1 mL of

aqua regia (3 parts HCl and 1 part I-IN03) for 5 min, and this solution was then diluted

136

 

and analyzed for Pd by inductively coupled plasma optical emission spectroscopy (ICP-
OES) (Varian 710-ES ICP Optical Emission Spectrometer).
5.2.5. Catalytic Reactions and Leaching Studies

The catalytic activities of the modiﬁed membranes were examined by reduction
of aqueous 4-nitrophenol with HCOONa at room temperature. A solution containing 4-
nitrophenol and HCOONa was passed through the membrane using a peristaltic pump as
shown in Figure 5.3. Permeate samples were collected at speciﬁc time intervals, and the
concentration of 4-nitrophenol in the feed and permeate solutions was determined from

the solution absorbance at 400 nm."17

(UV -Vis spectra were acquired with a Perkin-
Elmer Lambda 4O spectrophotomer.) A typical reaction mixture contained 0.5 mM 4-
nitrophenol and 0.05 M HCOONa, with a pH adjusted to 4.5 with dilute formic acid.

To determine the extent of catalyst leaching during the reaction, the amount of Pd
in the feed and permeate samples was analyzed using ICP-OES. To qualitatively
examine catalyst leaching, I examined the change in 4-nitrophenol concentration in the
permeate solution (after exiting the membrane) over time. If leached catalyst is present

in the permeate, the reaction should continue to take place after permeation to decrease

the 4-nitrophenol concentration in the sample.

5.3. Results and Discussion
Figure 5.4 shows the structures of the polyelectrolytes investigated in this study.
PSS is usually deposited ﬁrst on polymeric membranes due to its hydrophobic

18.19

interactions with polymer surfaces, and because PAA does not adsorb as well on such

substrates. In contrast, PAA binds strongly to alumina and serves as the initial layer in

137

 

porous alumina membranes. Comparing the polycations branched PEI and PAH, the
branched PEI forms thicker ﬁlms that have more binding sites for noble metal

nanoparticles. Thus, ﬁlms containing PEI have a higher catalyst loading than ﬁlms

 

containing PAH.
n
F“
n '1
COOH NHJCI’

$03“ Na“ Poly(acrylic acid) Poly(allylamine l

Poly(styrene sulfonate) PAA hydrogmande) :

PSS E
. i:-

RH ﬁ~ﬁn3
”2 n

H3ﬁ/\/"\/\ﬁﬂa

Branched Poly(ethylenimine)
PEI

Figure 5.4. Structures of the polyelectrolytes used in this study.

5.3.1. SEM Characterization of Immobilized N anoparticles

Figure 5.1 shows SEM images of bare alumina and PBS membranes. Alumina
membranes have standard cylindrical pore structures with a relatively uniform skin pore
size distribution, whereas PES membranes have a layered structure and a relatively wide
range of skin pore sizes (approximately 0.2-0.5 um). Moreover, the PES membranes

have a smaller porosity than alumina membranes. Although these two substrates differ in

138

structure, SEM images of the cross-sections of membranes coated with polyanion/PEI/Pd
nanoparticle (PdNP) ﬁlms show that Pd nanoparticles are loaded in the membrane pores
of both materials (Figure 5.5). Due to the small size of Pd nanoparticles (7 nm diameter),
some aggregation may occur in alumina membranes to give particles that are larger than
expected. However, the low resolution of SEM and the sputter coating process could also
cause particles to appear larger than their actual size. The density of Pd nanoparticles in
PBS membranes is not as high as that in alumina membranes, and it is difﬁcult to
characterize the small Pd particles in the PES membranes. However, the color of the PES
membrane changes from white to dark brown aﬁer ﬁlm deposition, conﬁrming the
loading of Pd nanoparticles. ICP-OES analysis also shows that the Pd loading in alumina
and PBS membranes is 320 pg and 390 ug, respectively. Taking into account the larger
PES membrane thickness (110-150 um) compared to alumina (60 mm), the Pd loading per

cm3 of membrane is lower in PBS membranes than in alumina.

 

Figure 5.5. SEM images of cross-sections of ﬂat-sheet membranes coated with

polyanion/PEl/PdNP ﬁlms. (a) alumina and (b) PES.

5.3.2. Catalysis with Pd Nanoparticle-Loaded Alumina Membranes

139

Guibal and coworkers extensively studied reduction of aqueous nitrophenol with
HCOONa as the hydrogen donor.”22 According to their work, when using a chitosan-
supported Pd catalyst, the most rapid nitrophenol reduction occurs at a pH between 3 and
4.20 Thus, we began our studies by investigating the effect of feed pH on the rate of 4-
nitrophenol reduction in modiﬁed membranes.
5.3.2.1. Effect of Feed pH on 4-Nitrophenol Conversion

pH is an important variable in catalytic reactions because it affects both the
charge of the substrate and the dissociation of the hydrogen donor (HCOONa in this
study). Without any adjustment, the pH of the reaction solution (0.5 mM 4-nitrophenol
and 0.05 M HCOONa) is 6.2. Passage of this solution through a PAA/PEI/PdNP-coated
alumina membrane at a ﬂux of 0.027 mL cm'2 s'1 (residence time of 68 ms, assuming a
30% porosity) results in the reduction of only ~1% of the 4-nitrophenol. When the pH is
adjusted to 4.5, however, passage through the membrane at the same ﬂux yields >99.8%
conversion of 4-nitrophenol, and the conversion remains essentially constant over a 5-h
reaction period (Figure 5.6). By lowering the feed pH to 4.5, less charge on the citrate-
stabilized nanoparticles may facilitate the adsorption of formate ions and 4-nitrophenol

onto catalytic active sites and allow for a faster reaction.

140

 

.100.0;°°°99999000000090
e

 

 

23 I
1: 99.6 ~
0

g
1, 99.2 ~
.2"

,, 98.8 [L-
5

2 98.4

g .

98.0 _ _ _ r __-.t__ _ __,;-V-,-_. wwwh- .L L- _

O 50 100 150 200 250 300

Reaction Time (min)

Figure 5.6. Percent reduction of 4-nitrophenol during passage of a 0.5 mM 4-nitrophenol,
50 mM HCOONa aqueous solution (pH adjusted to 4.5) through an alumina membrane
coated with a PAA/PEI/PdNP ﬁlm (ﬂux = 0.027 mL cm'2 s", residence time in the

membrane of 68 ms).

In preliminary experiments with PBS membranes, feed solutions at pH 4.5 and 3.8
gave similar conversions, but about 4.6% of the Pd was leached from a PSS/PEI/PdNP
membrane at pH 3.8 over an 80-min permeation (0.5 mM 4-nitrophenol, 50 mM
HCOONa, ﬂux of 0.035 mL cm'2 5"), whereas only 2.2% of the Pd was leached ﬁom a
similar membrane at a pH of 4.5 under the same conditions. Because the citrate ions
become less charged at low pH, the Pd nanoparticles may be less stable and more subject
to leaching at pH 3.8 than at pH 4.5. Thus, we employed a pH of 4.5 for all subsequent
studies.

5.3.2.2. Dependence of 4-Nitrophenol Conversion on HCOONa Concentration

141

 

At a constant 4-nitrophenol concentration and pH, the percent reduction of 4-
nitrophenol upon passage through a catalytic alumina membrane varies signiﬁcantly with
the concentration of HCOONa in the feed solution. Figure 5.7 shows that more than
99.5% of 4-nitrophenol is reduced when 20 mM HCOONa (40-fold excess) is present in
the feed. In contrast, conversion is only ~70% when using 10 mM HCOONa (20-fold
excess). To insure high rates of reduction, subsequent experiments were performed using

50 mM HCOONa.

1oo ‘
80 '
o L; -4; . , i . i ,. . .;
50 40 30 20

17.5 15 12.5 10
Concentration of HCOONa (mM)

#0)
OD

Percent Reduced (%)
N
0

Figure 5.7. Effect of HCOONa concentration on the percent reduction of 4-nitrophenol
during passage of a 0.5 mM 4-nitrophenol, HCOONa aqueous solution (pH adjusted to
4.5) through an alumina membrane coated with a PAA/PEI/PdNP ﬁlm. (ﬂux = 0.027 mL
cm"2 s", residence time in the membrane of 68 ms, conversion was measured after 5 min

of permeation)

5.3.2.3. SEM Characterization of Deposit Formation on Alumina Membranes

142

 

As noted in the study of Au-catalyzed 4-nitrophenol reduction by NaBHe (chapter
4), deposits that form on the lumen surface of catalytic hollow ﬁber membranes might
stem ﬁ‘om oxidation of aminophenol.20 Similar deposit formation occurs on the alumina
membrane surface during the reduction of 4-nitrophenol with HCOONa. Figure 5.8
shows that stirring the solution above the membrane during the reaction may reduce the
amount of material deposited. However, these deposits do not appear in SEM images of
the membrane pores, where most of the catalyst is located. Nevertheless, plugging of

membrane pores could decrease the available catalyst surface area.

 

Figure 5.8. SEM images of the skin surfaces of PAA/PEl/PdNP-coated alumina
membranes alter their use in catalyzing the reduction of 4-nitrophenol by HCOONa (a)
with stirring and (b) without stirring of the solution above the membrane. The reaction
solution contained 0.5 mM 4-nitrophenol and 50 mM HCOONa, and permeation

occurred at a ﬂux of 0.021 mL cm'2 s'1 for 1 h.

5.3.3. Catalysis with Pd Nanoparticle-Loaded PES Membranes

As mentioned earlier, one advantage of the layer-by-layer (LBL) method is its

applicability to different materials including ceramics and polymers. By using PSS as the

143

initial layer, polyelectrolyte/Pd nanoparticle ﬁlm formation can take place in PES

membranes as well as in porous alumina.
5.3.3.1. 4-Nitrophenol Reduction in PES Membranes Coated with PSS/PEllPdNP
Films

Figure 5.9 shows the percent conversion of 4-nitrophenol during passage through
a PBS membrane coated with a PSS/PEI/PdNP ﬁlm. At the beginning of the experiment,
99% of the 4-nitrophenol is reduced, but the conversion decreases gradually to 97% after

1 h. This conversion decline could be due to either catalyst leaching or catalyst fouling.

 

100 ‘t‘
A . O 0
2% 98 l ° 9 e .
.3 as r ’
g a
> 94 :
I: l
8 92 »
90 I,__,_,_-__,_-_-_ _____s--, e. __,s_- we»--- __,H_.__
O 10 20 30 40 50 60
' Time (min)

Figure 5.9. Percent reduction of 4-nitrophenol during passage of a 0.5 mM 4-nitrophenol,
50 mM HCOONa aqueous solution (pH adjusted to 4.5) through a PES membrane coated
with a PSS/PEI/PdNP ﬁlm (ﬂux = 0.022 mL cm'2 s", residence time = 173 ms, assuming

30% porosity).

5.3.3.2. Catalyst Leaching

144

 

 

The above results show that conversion declines over time and that at least some
leaching of Pd occurs. Below we ﬁrst discuss capping of ﬁlms to minimize leaching
along with qualitative and quantitative studies of leaching as a function of ﬂow rate and
ﬁlm composition.
5.3.3.2.]. Use of a Capping Layer to Minimize Catalyst Leaching

One way to minimize catalyst leaching is to add a capping layer on the
polyelectrolyte/PdNP ﬁlms. Because the citrate-stabilized Pd nanoparticles are
negatively charged, an additional polycation layer can be deposited to protect the
nanoparticles. However, Figure 5.10 shows that conversion still declines dramatically
with time during passage of 4-nitrophenol, HCOONa solutions through a PBS membrane
coated with a PEI-capped (PSS/PEI/PdNP/PEI) ﬁlm. Almost complete conversion of 4-
nitrophenol occurs in the ﬁrst 5 minutes of the experiment, but a dramatic drop to 85%
conversion follows in the next 10 min. We also examined the concentration of 4-
nitrophenol in several permeate samples as a function of time elapsed since permeation
through the membrane. For the permeate sample taken at 6 min of ﬁltration, conversion
increased from 92% to almost complete conversion alter the sample sat for one day,
which suggests that some Pd is present in the permeate to catalyze continued conversion
of the 4-nitrophenol. Smaller increases in conversion with time after permeating through
the membrane occurred for subsequent permeate samples. These results suggest that Pd
leaching takes place even in the presence of a PEI capping layer, but leaching may be
most extensive in earlier stages of permeation. We should note that in a control
experiment, the concentration of 4-nitrophenol was constant with time (for at least 1 day)

in formate-containing feed solutions.

145

E".
!

 

 

 

 

100 -.. ...- o

l .

l
2:: 96 I ' ____.- ‘
c i loAnalyzedwithin15 min
° 92 l" i l A 1’: 1h i
'5 g o :I nalyzeda er
5 , .
g 88 L ‘ o “Analyzed afterZh 1
g l . g V°Analeedaﬁer1£9v r

l

r

l

t .

80 l __ ___.._.;,__L..._.____’
O 5 10 15
Time (min)

Figure 5.10. Conversion of 4-nitrophenol during passage of a 0.5 mM 4-nitrophenol, 50
mM HCOONa aqueous solution (pH adjusted to 4.5) through a PES membrane coated
with a PSS/PEI/PdNP/PEI ﬁlm (ﬂux = 0.06 mL cm'2 s], residence time = 67 ms).
Samples were analyzed 15 min, diamonds; 1 h, squares; 2 h, triangles; and 1 day, open

circles,,after passing through the membrane.

Figure 5.11 shows the conversion as a function of ﬁltration time during passage of
a 4-nitrophenol, HCOONa solution through a PBS membrane coated with a
PSS/PEI/PdNP/PAH ﬁlm. The conversion, which is lower for this membrane than
typical PSS/PEI/PdNP/PAH-coated PES membranes, perhaps because of defects, is
essentially constant with permeate volume in this case. For permeate samples taken in
the ﬁrst 5 min, increases in conversion with the time elapsed prior to analysis suggest that
there is some Pd in the permeate. However, after the ﬁrst 6 min of permeation, no

detectable changes in 4-nitrophenol concentration appear as a function of time elapsed

146

 

prior to analysis. This suggests minimal leaching occurs after 6 min of permeation.
Comparison of Figures 5.10 and 5.11 suggests that the PAH capping layer more
effectively prevents catalyst leaching than a PEI capping layer, but this conclusion needs

further veriﬁcation. In the following experiments, all polyelectrolyte/PdNP ﬁlms are

 

 

 

 

capped with a PAH layer.
100 l
A 96 L: 9 :
o\"’ l 1 0;
Z 92 l’ - E 8 6 : OAnalyzedwithin15min§
'3 g IAnalyzed after1 h
g 88 l AAI'IBIYZBd after2h
8 ° AMY?! 51192.an
o
34
l
I
80 ' — “.--“..-"n-..
0 5 10 15 20

Time (min)

Figure 5.11. Conversion of 4-nitrophenol during passage of a 0.5 mM 4-nitrophenol, 50
mM HCOONa aqueous solution (pH adjusted to 4.5) through a PBS membrane coated
with a PSS/PEI/PdNP/PAH ﬁlm (ﬂux = 0.06 mL cm'2 s", residence time = 67 ms).
Samples were analyzed 15 min, diamonds; 1 h, squares; 2 h, triangles; and 1 day, open

circles, after passing through the membrane.

5.3.3.2.2. Effect of Solution Flux and Feed Concentration on Catalyst Leaching
To conﬁrm the ability of the PAH capping layer to minimize catalyst leaching, we

carried out the reduction reaction at a lower ﬂow rate using a PES membrane coated with

147

a PSS/PEI/PdNP/PAH ﬁlm. Figure 5.12 shows that at the lower ﬂux value, more than
99.5% of the 4-nitrophenol is reduced throughout a 1-h permeation, although there is a
small conversion decline between 50 and 80 min of permeation. In comparison, a Pd-
loaded PES membrane without a capping layer showed a 2% conversion decline in 1 h at
an even lower ﬂow rate (Figure 5.9). The comparison between Figure 5.9 and 5.12
suggests that use of PAH is capping layer is helpful to minimize Pd leaching and
maintain a high catalytic activity. However, the differences in initial conversions for the

two membranes complicate the comparison.

 

 

100.0 g..-n.. . _ . . .
$ .
99.6 ‘- I
g
.5 99.2 i
2 1
g ,
c 98.8 '—
o
o 3
98.4 r
98.0 l———-—----—-—-——-— . - ,-_._.-.-_..- ..
0 10 20 30 40 50 60 70 80 90
Time (min)

Figure 5.12. Percent reduction of 4-nitrophenol during passage of a 0.5 mM 4-
nitrophenol, 50 mM HCOONa aqueous solution (pH adjusted to 4.5) through a PBS

membrane coated with a PSS/PEI/PdNP/PAH ﬁlm (ﬂux = 0.035 mL cm'2 s", residence

time =110 ms).

148

 

High ﬂuxes might induce more catalyst leaching from the ﬁlm. To better
examine this effect, we increased the ﬂux more than 3-fold relative to the experiment in
Figure 5.12. Figure 5.13 shows the conversion change using the PSS/PEI/PdNP/PAH
ﬁlm-coated PES membrane at this high solution ﬂux. Conversion remains more than
99.5% in the ﬁrst 20 min, but then it drops drastically to as low as 30% in 50 min. ICP-
OES analysis of permeate samples and the total loading of the membranes show that
about 6.3% of Pd leaches out in this experiment (Figure 5.13) while only 2% Pd leaching

is detected when the reduction is carried out at lower ﬂux (Figure 5.12).

100.0 Jvﬁ- —|»- J- I.-- ... 1.. .. .. -. -- .. _- -- -- -- . .- -- .- -.. -_ .. - .-. -

l
l
I
80.0 { _
1
g 60.0 l
r . -
g 40.0 ,L .
o z .
0 v
20.0 i
i
0.0 L--————--—--~-——-——'~r--—--.-.r.._.._...-___r----_._-..--____J
0 10 20 30 40 50 60

Time (min)

Figure 5.13. Percent reduction of 4-nitrophenol during passage of a 0.5 mM 4-
nitrophenol, 50 mM HCOONa aqueous solution (pH adjusted to 4.5) through a PBS
membrane coated with a PSS/PEI/PdNP/PAH ﬁlm (ﬂux = 0.12 mL cm’2 s", residence

time =33 ms).

149

The high solution ﬂux could result in greater shear forces that enhance catalyst
leaching, but another possible explanation for decreases in catalytic activity is that at
higher ﬂux, more 4-nitrophenol is reduced and more catalyst fouling may occur. (In fact,
leaching of 2 to 6% of the Pd may have only a minor effect on catalytic activity.) To
differentiate between the effects of ﬂux and fouling, we performed 4-nitrophenol
reduction at a the ﬂow rate used to obtain Figure 5.13 but with a feed that contains only
0.1 mM 4-nitrophenol with 100 fold excess HCOONa. Under these conditions, more
than 99.6% of the 4-nitrophenol is reduced without any signiﬁcant conversion decline
over 45 min (Figure 5.14). This suggests that the primary reason for conversion decline
is catalyst fouling from byproducts of 4-aminophenol and that fouling is a function of the
amount of material reacted at the catalyst. However, comparison of the amount of Pd in
the permeate solution and in the membrane aﬁer this experiment did indicate 8.7% Pd

leaching with most leaching happened in the early stage of the experiment.

150

loo 0 [- ». .—~. .— .--— .—.-.- ...—— vs 4— —- ~v~ M .w" ~—- ...~ .- --.-. ...... .. .... '~ «'4- u~ -.. --.. -... .... 8... a, --.. .... -.. .. —_ «w r- ..-_ M.

 

)- I I I I . . - .
i I
i I
99.6 :—
i
E I
c 99.2 i
.2 ,
2 I
g r
c 98.8 ;
o
0 i
98.4 i
i
98.0 1.... ﬂ_ _. -Jtﬁir— -... ..-... MWWME .4
0 5 10 15 20 25 30 35 40 45 50
Time (min)

Figure 5.14. Percent reduction of 4-nitrophenol during passage of a 0.1 mM 4-
nitrophenol, 10 mM HCOONa aqueous solution (pH adjusted to 4.5) through a PBS
membrane coated with a PSS/PEI/PdNP/PAH ﬁlm (ﬂux = 0.12 mL cm'2 s", residence

time =33 ms).

Figure 5.15 shows the conversion of 4-nitrophenol using a feed solution
containing 0.1 mM 4-nitrophenol with 100 fold excess HCOONa at a high solution ﬂux
over a 3-h reaction period. Conversion remains essentially constant at >99.5% for 150
min and deceases to 85% in the last 30 min. Pd leaching mainly happens during the ﬁrst
60 min and less than 2 ppb Pd is detected in the perrnate during the rest of the experiment.
This result conﬁrms that the conversion decline is not due to catalyst leaching. As

mentioned above, we speculate that the conversion decline after 150 min is the result of

151

catalyst fouling by oxidation products of 4-aminophenol. The precipitous drop in

conversion may occur as pores become fully occluded.

100'0 _.-' "I" "I" 'I" '“ﬂii'ifaII-‘fi' ~ * .. ‘~ v.
I

I I

 

A 95.0

E5 .

c

.2 I

2 90.0 t I

Q I

> I

C I

0 I
° 85.0 I -

I
80.0 “A— - - - --—----—-~--— — -—~'-— —---m....- .. "_--..t - ...... - ....-
0 40 80 120 1 60 200

Time (min)

Figure 5.15. Percent reduction of 4-nitrophenol during passage of a 0.1 mM 4-
nitrophenol, 10 mM HCOONa aqueous solution (pH, adjusted to 4.5) through a PBS

membrane coated with a PSS/PEI/PdNP/PAH ﬁlm (ﬂux = 0.12 mL cm'2 s", residence

time =33 ms).

5.4. Conclusions

Layer-by-layer adsorption of polyelectrolytes and Pd nanoparticles in porous
membranes yields catalytic reactors that are active in the reduction of 4-nitrophenol with
HCOONa, which is a more cost-effective and environment friendly reagent than NaBH4.
This LBL method is applicable to both ceramic and polymeric materials, and the resulting

membranes allow a high conversion of 4-nitrophenol to 4-aminophenol, but the catalytic

152

activity declines over time. When low concentrations of reactants pass through the
membrane at relatively low ﬂow rates, high catalytic activities can be maintained for
several hours, but higher ﬂow rates and 4-nitrophenol concentrations lead to more rapid
conversion declines. The decline in conversion is probably due to catalyst fouling, which
may stem from the oxidation aminophenol, as we noted earlier in the study of Au-

catalyzed reduction of 4-nitrophenol with NaBH4.

153

5.5. References
(1) Dotzauer, D. M.; Dai, J .; Sun, L.; Bruening, M. L. Nano Lett. 2006, 6, 2268-2272.

(2) Dotzauer, D. M.; Bhattacharjee, S.; Wen, Y.; Bruening, M. L. Langmuir 2009, 25,
1865-1871.

(3) Dokoutchaev, A.; James, J. T.; Koene, S. C.; Pathak, S.; Prakash, G. K. S.;
Thompson, M. E. Chem. Mater. 1999, I I, 2389-2399.

(4) Dotzauer, D. M.; Abusaloua, A.; Miachon, S.; Dalmon, J .-A.; Bruening, M. L.
Appl. Catal., B 2009, 91, 180-188.

(5) Feng, Z.; Yan, F. Surf Interface Anal. 2008, 40, 1523-1528.

(6) Goulet, P. J. G.; dos Santos, D. S.; Alvarez-Puebla, R. A.; Oliveira, O. N.; Aroca,
R. F. Langmuir 2005, 21 , 5576-5581.

(7) Puniredd, S. R.; Yin, C. M.; Hooi, Y. S.; Lee, P. S.; Srinivasan, M. P. J. Colloid
Interface Sci. 2009, 332, 505-510.

(8) Soler, L.; Macanas, J .; Munoz, M.; Casado, J. Int. J. Hydrogen Energy 2007, 32,
4702-4710.

(9) Brieger, G.; Nestrick, T. J. Chem. Rev. 1974, 74, 567-580.

(10) Johnstone, R. A. W.; Wilby, A. H.; Entwistle, I. D. Chem. Rev. 1985, 85, 129-170.
(11) Basu, B.; Das, P.; Das, S. Mal. Diversin 2005, 9, 259-262.

(12) Rajagopal, S.; Spatola, A. F. J. Org. Chem. 1995, 60, 1347-1355.

(13) Zoran, A.; Sasson, Y.; Blum, J. J. Mol. Catal. 1984, 26, 321-326.

(14) Wiener, H.; Blum, J .; Feilchenfeld, H.; Sasson, Y.; Zalmanov, N. J. Catal. 1988,
110, 184-190.

(15) Wiener, H.; B1um,J.; Sasson, Y. J. Org. Chem. 1991, 56, 4481-4486.
(16) Dotzauer, D. M., Ph. D Dissertation, Michigan State University, 2009.
(17) Pradhan, N.; Pal, A.; Pal, T. Colloids Surf, A 2002, 196, 247-257.
(18) Malaisamy, R.; Bruening, M. L. Langmuir 2005, 21 , 10587-10592.

(19) Ouyang, L.; Malaisamy, R.; Bruening, M. L. J. Membr. Sci. 2008, 310, 76-84.

154

 

(20) Vincent, T.; Guibal, E. Langmuir 2003, 19, 8475-8483.
(21) Guibal, B.; Vincent, T. J. Environ. Manage. 2004, 71, 15-23.

(22) Vincent, T.; Guibal, E. Environ. Sci. Technol. 2004, 38, 4233-4240.

155

 

Chapter 6

Summary and Future Work

This dissertation investigates two applications of functional membranes prepared
with layer-by-layer (LBL) adsorption: ion separations and catalysis. Chapters 2 and 3
describe high-ﬂux nanoﬁltration (NF) membranes that selectively remove multivalent
ions in the presence of monovalent ions, and the formation of these membranes through
LBL adsorption of polycations and polyanions allows tuning of separations through
control over the terminating layer. Deposition parameters such as number of bilayers and
the pH and ionic strength of the polyelectrolyte deposition solutions can be varied to
Optimize cation or anion separations. The best membranes prepared in these studies
exhibit both higher ﬂuxes and higher selectivities than commercially available NF
membranes. Chapters 4 and 5 discuss catalytic membrane reactors that contain a high
density of well separated metal nanoparticles deposited. by LBL adsorption of
polyelectrolytes and metal nanoparticles. This adsorption method occurs with substrates
of different geometries (hollow ﬁber and ﬂat sheet membranes) and various noble metal
nanoparticles including Au and Pd. Hollow ﬁber membrane reactors loaded with Au
nanoparticles show high catalytic activity in the reduction of 4-nitrophenol with NaBI-I4,
and >99% initial conversion of 4-nitrophenol to 4-aminophenol by HCOONa also takes
place in ﬂat sheet membranes containing Pd nanoparticles. However, a slight conversion
decline over time appears in both hollow ﬁber and ﬂat sheet membranes. This
conversion decline probably stems from catalyst fouling by byproducts of 4-aminophenol

oxidation.

156

 

Although polyelectrolyte multilayer ﬁlms (PEMs) on ultraﬁltration supports show
promising NF properties, there are still several challenges to overcome in this area.
Because electrostatic interactions hold the layers together, the ﬁlm structure is sensitive
to changes in solution pH or ionic strength."3 Release of polymer material and even hole
formation occur upon dramatic changes in pH or ionic strength.3’4 One way to overcome
this instability is to incorporate covalent bonding between different layers in the ﬁlm.
Several studies reported formation of covalently linked PEMs through UV- or heat-
induced cross-linking.5'7 Perhaps the biggest challenge in utilizing PEMs is that their
formation involves many adsorption and rinsing steps. The use of spin- and spray-
coating methods could potentially speed up the ﬁlm formation processs'm and facilitate
the commercial application of PEMs.

Because Donnan exclusion plays an important role in ion separations with NF
membranes, increasing the charge density at the surface of PEMs improves their ion-
transport selectivities. One way to increase surface charge is to prepare the terminating
polyelectrolyte layer in a deposition solution of high ionic strength.”'13 Another
approach is to introduce net, ﬁxed charge into the PEMs by partial derivatization of
polyelectrolytes with photolabile ﬁmctional groups that photolyze to charged species.l4
While the underivatized ionic functional groups allow for formation of PEMs through
electrostatic interactions, postdeposition UV irradiation of partially derivatized ﬁlms
results in the formation of ﬁxed-charge sites that are not charge compensated by a
neighboring polyelectrolyte. Such sites may greatly increase cation or anion selectivities.

This thesis describes preliminary work on developing catalytic membrane reactors

through LBL adsorption of polyelectrolytes and metal nanoparticles. One important

157

 

advantage of the LBL method is its ability to control the catalyst location in a tubular
membrane.15 In a gas-liquid reaction, localization of the catalyst at the interior of a
tubular membrane (Figure 6.1) can increase catalytic activity. The combination of
hollow ﬁber membranes, whose geometry provides a high surface area to volume ratio,
with catalyst localization in the ﬁber may further enhance the region of gas-liquid-
catalyst contact to facilitate gas-liquid reactions.l6"3. Such a membrane reactor system

might be useful in applications including hydrogenation and oxidation reactions.""'21

    
 

Catalytic
layer

Porous
membrane

Figure 6.1. Schematic diagram of a gas-liquid contactor in a hollow ﬁber membrane.

Several recent studies utilized star polymers with globular architecture and

multiple arms in LBL assembly.22’23

Due to their unique mechanical properties and
signiﬁcantly higher degree of functionality compared to linear analogues, star polymers
can serve as an inexpensive alternative to dendrimers for stabilizing catalytic metal

nanoparticles.”27 Because the thickness of LBL ﬁlms of star polymers is comparable to

158

 

 

that of conventional PEMs,23 their high charge density make these materials promising in
forming highly selective membranes in ion separations.

Overall, the versatility of LBL assembly should allow for formation of a wide
range of membrane compositions to further improve both catalytic and NF membranes.

However, issues of stability and the economic feasibility of the LBL procedure must be

addressed.

 

159

References
(1) Itano, K.; Choi, J .; Rubner, M. F. Macromolecules 2005, 38, 3450-3460.

(2) Ladhari, N.; Hemmer], J .; Haikel, Y.; Voegel, J .-C.; Schaaf, P.; Ball, V. Appl.
Surf Sci. 2008, 255, 1988-1995.

(3) Mjahed, H.; Voegel, J .-C.; Senger, B.; Chassepot, A.; Rameau, A.; Ball, V.;
Schaaf, P.; Boulmedais, F. Soﬁ Matter 2009, 5, 2269-2276.

(4) Dejeu, J .; Diziain, S.; Dange, C.; Membrey, F.; Charraut, D.; Foissy, A. Langmuir
2008, 24, 3090-3098.

(5) Ma, N.; Wang, Y.; Wang, Z.; Zhang, X. Langmuir 2006, 22, 3906-3909.

(6) Dai, J .; Jensen, A. W.; Mohanty, D. K.; Emdt, J .; Bruening, M. L. Langmuir 2001,
1 7, 931-937.

(7) Wu, G.; Shi, F .; Wang, Z.; Liu, Z.; Zhang, X. Langmuir 2009, 25, 2949-2955.
(8) Schlenoff, J. B.; Dubas, S. T.; Farhat, T. Langmuir 2000, 16, 9968-9969.

(9) Lee, S.-S.; Hong, J.-D.; Kim, C. H.; Kim, K.; Koo, J. F.; Lee, K.-B.
Macromolecules 2001, 34, 5358-5360.

(10) Bruening, M.; Dotzauer, D. Nat. Mater. 2009, 8, 449-450.

(11) Stanton, B. W.; Harris, J. J .; Miller, M. D.; Bruening, M. L. Langmuir 2003, 19,
7038-7042.

(12) Malaisamy, R.; Bruening, M. L. Langmuir 2005, 21, 10587-10592.
(13) Ouyang, L.; Malaisamy, R.; Bruening, M. L. J. Membr. Sci. 2008, 310, 76-84.

(14) Dai, J .; Balachandra, A. M.; Lee, J. 1.; Bruening, M. L. Macromolecules 2002, 35,
3164-3170.

(15) Dotzauer, D. M.; Abusaloua, A.; Miachon, S.; Dalmon, J .-A.; Bruening, M. L.
Appl. Catal., B 2009, 91 , 180-188.

(16) Israni, S. H.; Nair, B. K. R.; Harold, M. P. Catal. Today 2009, 139, 299-311.

(17) Peirano, F.; Vincent, T.; Quignard, F.; Robitzer, M.; Guibal, E. J. Membr. Sci.
2009, 329, 30-45.

160

 

(18) Lebedeva, V. 1.; Gryaznov, V. M.; Petrova, I. V.; Volkov, V. V.; Tereshchenko, G.
F.; Shkol’nikov, E. 1.; Plyasova, L. M.; Kochubey, D. 1.; van der Vaart, R.; van Soest-
Verecammen, E. L. J. Kinet. Catal. 2006, 47, 867-872.

(19) van der Vaart, R.; Lebedeva, V. I.; Petrova, I. V.; Plyasova, L. M.; Rudina, N. A.;
Kochubey, D. I.; Tereshchenko, G. F .; Volkov, V. V.; van Erkel, J. J. Membr. Sci. 2007,
299, 38-44.

(20) Caro, J .; Caspary, K. J .; Hamel, C.; Hoting, B.; Kolsch, P.; Langanke, B.;
Nassauer, K.; Schiestel, T.; Schmidt, A.; Schomacker, R.; Seidel-Morgenstern, A.;
Tsotsas, B.; Voigt, 1.; Wang, H.; Warsitz, R.; Werth, S.; Wolf, A. Ind. Eng. Chem. Res.
2006, 46, 2286-2294.

(21) Sirkar, K. K.; Shanbhag, P. V.; Kovvali, A. S. Ind. Eng. Chem. Res. 1999, 38,
3715-3737.

(22) Connal, L. A.; Li, Q.; Quinn, J. F .; Tjipto, E.; Caruso, F.; Qiao, G. G.
Macromolecules 2008, 41 , 2620-2626.

(23) Kim, B.-S.; Gao, H.; Argun, A. A.; Matyjaszewski, K.; Hammond, P. T.
Macromolecules 2009, 42, 368-375.

(24) Yoo, H.-S.; Watanabe, T.; Hirao, A. Macromolecules 2009, 42, 4558-4570.

(25) Hirao, A.; Watanabe, T.; Ishizu, K.; Ree, M.; Jin, S.; Jin, K. S.; Defﬁeux, A.;
Schappacher, M.; Carlotti, S. Macromolecules 2009, 42, 682-693.

(26) Hirao, A.; Matsuo, A.; Watanabe, T. Macromolecules 2005, 38, 8701-8711.

(27) Zheng, Q.; Pan, C.-Y. Macromolecules 2005, 38, 6841-6848.

161

 

 

   

"IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII“