,1::.3.:
. a ~y5

'0:
:Ir: . E.
c i 2:

 

 

 

.5 I}.
1.32:.3 2:...
. .. 1-1::

 

‘53:. (1.11....
4' ..

 

21.3... .
£11., 1..
A .t :5!

L:
...:..IJ.:.:
. . I 33...... p
. .It ‘-
(is...

u.
77.. I v.2:
T3: :. ..!A....:. u 1
2.4.2:... 72 at: .
.73.:1 I. 12 1 :
. z. .1532:
a if

. i

([2 :0

2.71:?!
I

:3 I
a: v:
x 3r:

L. x i; , , v1: 4
I A is? .3 . Vlizflttuf

in.
A.
ii... 1:: .e 4. 1| ..
2...: {3. .13.? Mi... 3
2. 2.. {(-39 1!.21 2
r 5:3... ..D....:?..Y.pl.u.

3.71.2

 

3.. fig”? ‘ ‘ . ,. it . ._, .1. ‘ ‘ .. ._ . . . . ‘ ,
.3 z E .. . “minivan“ c , . . ‘
a... ‘ , £35553 _.,.........

i x . :1..7I...5a

, A i. .151.

1

1.35:1... ’ i?
.15. a .c 1. .7:

 

 

LIBRARY
Michigan State
University

 

 

 

This is to certify that the

dissertation entitled

Fundamentai Study of the Ion Penreabiiity of Ultrathin,

Layered Poiyeiectr‘oiyte Fiims
presented by

Jerery James Harris

has been accepted towards fulfillment
of the requirements for

Ph.D degree in omsm

Mmm

‘5
Major professor /

Date 08/1 7/01

MS U is an Affirmative Action/Equal Opportunity Institution 0-12771

 

 

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.

 

9 WE WE

DATE DUE

DATE DUE

 

NOV 1 0 2002

 

 

U8 3 0 ZUUJ

 

W

 

 

 

 

 

 

 

 

 

 

6/01 c:/ClRC/DdoDue.p65-p.15

 

 

Fundamental Study of the Ion Permeability of Ultrathin, Layered Polyelectrolyte Films
By

Jeremy James Harris

AN ABSTRACT OF A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY

Department of Chemistry

2001

Professor Merlin L. Bruening

ABSTRACT

FUNDAMENTAL STUDY OF THE ION PERMEABILITY OF ULTRATHIN,
LAYERED POLYELECTROLYTE FILMS

By

Jeremy James Harris

Layered polyelectrolyte films are attractive as coatings and membranes because of
their ease of synthesis and versatility. The feasibility of utilizing these films, however,
will depend on their permeability and stability under relevant conditions. Cyclic
voltammetry and electrochemical impedance spectroscopy studies presented in this thesis
show that the permeability of poly(allylamine hydrochloride) (PAH)/poly(styrene
sulfonate) (PSS) coatings depends on the number of bilayers in the film, deposition
conditions, and solution pH. Unfortunately, even the most passivating films are
relatively permeable and unstable at high pH. To overcome this problem we employ
PAH/poly(acrylic acid) (PAA) films which, when heated, form a nylon-like coating that
dramatically decreases film permeability while increasing stability.

Ultrathin films such as layered polyelectrolytes are attractive membrane materials
because they potentially offer both high flux and high selectivity. Synthesis of such
membranes is challenging, however, as it requires defect-free deposition of a film at the
surface of a porous support. Scanning electron microscopy shows that adsorption of
layered polyelectrolyte films on porous alumina supports naturally results [in a defect-free

membrane. Ion transport through these membranes is primarily controlled by Donnan

interactions between the ions and the fixed charge present in the film, which results in
high selectivity between anions of different valence.

Layered polyelectrolyte films offer several means for controlling ion-flux through
these membranes. Since Donnan interactions at the film/solution interface greatly affect
ion—transport, capping the films with an additional anionic polyelectrolyte layer enhances
the monovalent/divalent anion selectivity of these membranes. While selectivity depends
on membrane surface charge, the flux through these membranes is affected by film
composition. PAH/FAA membranes show similar selectivities as PAH/PSS membranes
but with only 70% of the ion flux. To enhance selectivity while maintaining flux,
composite membranes were constructed that utilize the high flux of PAH/PSS
membranes along with the blocking effects of PAH/PAA membranes. Membranes
composed of 5 PSS/PAH bilayers capped with 2.5-bilayers of PAA/PAH show Cl'/SO42’
selectivity as high as 350. Utilization of charge, membrane composition, and cross-
linking provide for specific tailoring of membrane properties.

Utilization of polyelectrolyte membranes for water purification will require a
pressure gradient to achieve sufficient water flux. When used as nanofiltration elements,
PSS/PAH membranes show Ca2+ rejections of 90+% along with very high water flux.
Cross-linking and fluorination of polyelectrolyte membranes further increases NaJr and
3042' rejections, with respect to PSS/PAH membranes, while maintaining an appreciable
flux. Na+ and 3042' rejection values for composite membranes are >60 and >90%,
respectively, for both 50 and 1000 ppm solutions. Thus, polyelectrolyte membranes not

only offer easy means for membrane modification, but may provide solutions to current

problems in membrane separations.

To Mom, Dad, & Shannon

iv

ACKNOWLEDGEMENTS
“Even when on the right track, you ’11 get run over if you just sit there. ”

First, I would like to thank my advisor, Professor Merlin Bruening. You showed
me how to be an insightful scientist and one heck of a proofreader. Thank you for taking
a chance and allowing me to join your group. I have really enjoyed my four years at
MSU. Finally, how could I forget those “competitive” golf matches we engaged in
during the last four years and I do believe that I hold the edge.

Next, I would like to thank my family; Mom, Dad, Julie, Jessica, Grandma and
Grandpa, for all of their support and love in everything that I have pursued in my life.
Much thanks to my in-laws, Jim and Sheryl, for your support and evereflowing advice.
Mom & Dad, thanks for giving me the chance ($$) to further my education at Wooster
and thank you for the support while at MSU even though you didn’t quite understand
what I was doing. Did you ever think I would be where I am at today?

Shannon, my wife and my best friend, thank you for your love and support and
paying half the rent. It has been a blessing to have you by my side during this journey
and I look forward to many more. I have often told you that you are the reason I am
where I am today. Thanks.

Finally... The many friends that have been passengers with me on this four-year
trip whether you wanted to be or not. In no particular order; Dale, Wendy, Karen, Bob,
Kristi, Shannon, Art, Shawn, Pat, Kirk, Todd, Barb, Bill & Julie. Last but not least the
Bruening group, past and present; Dan, Wenxi, Keith, KP, Sandra, Anagi, Bo, Skanth,

Milind, and Jinhua. Thanks for the many “stimulating” conversations about chemistry

This work was partially supported by the
Department of Energy Office of Basic
Energy Sciences, the National Science
Foundation (CE-9816108), and the
Center for Fundamental Research at
Michigan State University. v

and life.

TABLE OF CONTENTS

List of Tables

List of Figures

List of Abbreviations

Chapter 1: Introduction and Background

I. Development of Thin Films

11. History and Development of Layered Polyelectrolyte Films

III. Ion-Permeability of Layered Polyelectrolyte Films

IV. Use of Layered Polyelectrolyte Films as Ion-Separation Membranes
V. Use of Layered Polyelectrolyte Films as Nanofiltration Membranes
VI. References

Chapter 2: Electrochemical and in Situ Ellipsometric Investigation of the
Permeability and Stability of Layered Polyelectrolyte Films

1. Introduction
II. Experimental
A. Chemicals and Solutions
B. Techniques
III. Results and Discussion
A. Characterization of Polyelectrolyte Films
B. Film Swelling
C. Investigation of Film Permeability to Fe(CN)63'/4'
D. Investigation of F ihn Permeability to Ru(NH3)62+/3+
E. Permeability as a Function of Time
F. Permeability as a Function of the Number of Bilayers

F. Effect of Supporting Electrolyte on the Permeability of Polyelectrolyte

Fihns

G. Comparing Permeabilities of PAH/PAA and PAH/PSS Multilayer Films

IV. Conclusions
V. References

Chapter 3: Synthesis of Passivating, Nylon-Like Coatings through
Cross-linking of Ultrathin Polyelectrolyte Films

1. Introduction

11. Experimental

IH. Results and Discussion
IV. References

vi

viii

ix

xiii

t—tO—l
UINNOth—I

21

21
22
22
24
27
27
30
33
41
45
46

48
50
51
53

56

56
57
58
68

Chapter 4: Layered Polyelectrolyte Films as Selective, Ultrathin Membranes

for Anion Separations 69
I. Introduction 69
11. Experimental 70
A. Materials 70
B. Transport Studies 71
III. Results and Discussion 73
IV. Conclusions 88
V. References 89
Chapter 5: Development of Polyelectrolyte Films as Nanofiltration
Membranes 92
I. Introduction 92
11. Experimental 93
A. Chemicals 93
B. Polyelectrolyte Deposition and Derivatization on Porous Alumina
Supports ‘ 94
C. Nanofiltration 95
III. Results and Discussion 98
A. Nanofiltration Performance of PSS/PAH Membranes 98
B. Modeling of Ion Transport Through Polyelectrolyte Membranes 102
C. Performance of Composite and Derivatizecl Polyelectrolyte Membranes 105
IV. Conclusions 105
V. References 108
Chapter 6: Conclusions and Future Work 110
1. Stability and Ion-Permeability of Layered Polyelectrolyte Films Deposited on
Solid Supports 110
II. Polyelectrolyte Films as Anion-Separation Membranes 11 1
III. Characterization of Nanofiltration Membranes Made from Layered
Polyelectrolyte Films 111
IV. Future Work 112

Appendix A: Derivazitation of Distribution Coefficient 114

vii

Table 2.1

Table 2.2

Table 2.3

Table 2.4

Table 4.1

Table 5.1

Table 5.2

Table 5.3

LIST OF TABLES

 

Average peak currents (uA/cmz) from cyclic voltammograms
(Fe(CN)6 ’) measured at PAH/PSS coated gold electrodes in different
buffered solutions. 34

Average R“ values (Q-cmz) from impedance plots (Fe(CN)63'/4’) of
PAH/PSS coated gold electrodes in different buffered solutions. 40

Average peak currents (uA/cmz) from cyclic voltammograms
(Ru(NH3)63+) at PAH/PSS coated gold electrodes in different buffered
solutions. 42

Average Ra values (Q-cmz) from impedance plots (Ru(NH3)62+/3+) of
PAH/PSS coated gold electrodes in different buffered solutions. 44

Anion flux values (moles cm'2 s") through bare alumina and alumina
substrates coated with between 1-5.5- or lO-bilayers of PAI-I/PSS. 80

Rejection and flux values (listed in parentheses) in m3 In2 (1'1 for various
polyelectrolyte fihns used as nanofiltration membranes. Rejection values
listed represent % decrease in permeate ion concentration relative to the
feed concentration in the collected permeate. The fluorinated film was
heated at 180 °C for 2 hrs. The salt solutions were NaCl, NaZSO4, and
CaClz. 99

Flux and rejection values for several commercially available nanofiltration
membranes and PSS/PAH membranes. Feed solutions were NaCl,
MgSO4, and CaClz. 101

Theoretical values for the concentration of fixed charge (M) in
polyelectrolyte membranes calculated using Equation 5.2 and rejection
values in Table 5.1. 104

viii

Figure 1.1

Figure 2.1

Figure 2.2

Figure 2.3

Figure 2.4

Figure 2.5

Figure 2.6

Figure 2.7

Figure 2.8

LIST OF FIGURES

 

Schematic diagram of the “Dip n’Rinse” procedure for the deposition

of a polyelectrolyte film on a gold electrode coated with a monolayer of
carboxylates. In the actual structure of these films, polycations and
polyanions are intermingled. 4

Ellipsometn'c thickness (squares) and FTIR-ERS absorbances at 1037 cm'1
(circles) for multilayered PAH/PSS films made in the presence of
supporting salt. The standard deviation for the absorbance values for each
bilayer is ~2%. The standard deviations for the ellipsometric thickness
values are <0.3 nm. 28

F TIR-ERS spectra of PAH/PSS films (made with supporting electrolyte)
composed of 1-5 bilayers. 29

FTIR-ERS spectrum of a 9-bilayer PAH/FAA film (made with no
supporting electrolyte) deposited on a gold electrode. 31

In situ ellipsometric thickness (circles) and cyclic voltammetry (5 mM

F e(CN)63 ') peak current values (squares) as a fimction of immersion

time for a lO-bilayer PAH/PSS film (made with supporting salt) in pH
3.2- (A), 6.3 (B), and 10-buffered solutions (C). 32

CVs (5 mM F e(CN)63 ') at a gold electrode coated with a five-bilayer
PAH/PSS film (made with supporting electrolyte). CVs were measured in
solutions buffered at pH 3.2 (Azdash), pH 6.3 (Assolid line), pH 10
(Bzcircles), and pH 3.2 after immersion in pH 10 (stquares). 36

Randles’ circuit used to fit impedance data where RC. = charge transfer
resistance, R8 = solution resistance, Cd; = double-layer capacitance, and Zw
= Warburg element. 38

Impedance data (Fe(CN)63”4‘) for a gold electrode coated with a five-
bilayer PAH/PSS film (made with supporting electrolyte). Plots were
obtained in solutions buffered at pH 3.2 (A), 6.3 (B), 10 (C), and 3.2 after
immersion in pH 10 (D). Solid circles represent experimental data and
Open circles show fits obtained with a Randles’ equivalent circuit. 39

(A) CVs (5 mM Ru(NH3)63+) of a gold electrode coated with a five-
bilayer PAH/PSS film (made with supporting electrolyte). CVs shown
were measured in solutions buffered at pH 10 (circles) and pH 3.2 after
immersion in pH 10 (squares). (B) Impedance data (5 mM Ru(NH3)62+/3+)
for a gold electrode coated with a five-bilayer PAH/PSS film (made with

ix

Figure 2.9

Figure 2.10

Figure 3.1

Figure 3.2

Figure 3.3

Figure 3.4

Figure 3.5

Figure 3.6

supporting electrolyte). Plots were obtained after immersion in pH 10-
buffered solutions (circles) and after reimmersion in pH 3.2-buffered
solutions (squares). 43

CVs (5 mM Fe(CN)63') of a gold electrode coated with a layer of MPA
(dashed line) or with a 1- (solid line), 2- (diamonds), 3- (circles), 4-
(squares), or 5-bilayer (diamonds) PAH/PSS fihn. CVs were measured in
solutions buffered at pH 6.3 and films were made from solutions
containing supporting electrolyte. 47

CVs (5 mM F e(CN).r,3 ', pH 6.3) of gold electrodes coated with a 20-
bilayer PAH/PSS film prepared with no supporting electrolyte (circles,
thickness of 16 nm), a 5-bilayer PAH/PAA fihn prepared with supporting
electrolyte (dashes, thickness of 44 nm), a 9-bilayer PAH/FAA prepared
with no supporting electrolyte (diamonds, thickness of 15 nm), or a 4-
bilayer PAH/PSS fihn prepared with supporting electrolyte (solid line with
x markers, thickness of 15 nm). 49

Schematic representation of cross-linking between FAA and PAH chains.
The actual structure of these films is likely highly intermingled not
discrete layers as shown in this representation. 59

FTIR-ERS spectra of 9-bilayer PAH/FAA fihns before heating (A), after
heating at 130 C (B), or after heating at 215 C (C). 60

XPS spectra of the N1, region of a 9-bilayer PAH/PAA film before (A)
and afier heating for 2 hours at 215 °C (B). Binding energy is referenced
to the C15 peak for hydrocarbons (284.6 eV). 62

Cyclic voltammograms (0.005 M Fe(CN)63' in 0.025 Na2HP04, 0.025
NaHzPO4, 1 M Na2SO4, pH 6.3) at gold electrodes coated with 9-bilayers
of PAH/PAA before (3A, dashed line), after heating a film at 130 °C for 2
hours (3A, solid line and 3B, dashed line), or after heating a film at 215 °C
(3B and C, solid lines). Scan rate of 0.1 V/s and electrode area of 0.1 cm2.
The electrode was immersed in the solution for 2 minutes prior to data
acquisition. 63

Impedance plots (0.005 M Fe(CN)63'/4' in 0.025 NazHPO4, 0.025
NaH2P04, 1 M NaZSO4, pH 6.3) for gold electrodes coated with 9-bilayer
PAH/PAA fihns before heating (A), after heating at 130 ”C (B), or after
heating at 215 °C (C). Scale bars apply to both x- and y-axes. Frequency
range 1-105 Hz. Sinusoidal voltage: 5 mV about E°'. The electrode was
immersed in solution for 5 minutes prior to data acquisition. 65

Modified Randles’ equivalent circuit used to model impedance data from
gold electrodes modified with cross-linked PAH/FAA films. 66

Figure 4.1

Figure 4.2

Figure 4.3

Figure 4.4

Figure 4.5

Figure 4.6

Figure 4.7

Figure 4.8

Figure 5.1

Dialysis cell used to measure ion-permeabilty of polyelectrolyte
membranes: conductivity meter (A), electric stirrer (B), conductivity
electrode (C), receiving phase (D), source phase (B), membrane (F), and
magnetic stir bar (G). 72

FESEM images of the filtrate side of a porous alumina substrate (A), and
alumina substrates coated with 2 (B), 4 (C), or 5 PAH/PS8 bilayers (D).
The circles in image C are likely defects in the film structure. Samples
were coated with ~5 nm Au for imaging. 74

F ESEM images of the permeate side of a bare alumina substrate before
film deposition (A), after restricting deposition to the filtrate side (B),

or following complete alumina immersion in polyelectrolyte

solutions (C). 76

Cross-sectional FESEM images of the filtrate side of a bare alumina
substrate (A) and an alumina substrate coated with 5-bilayers of
PAH/FAA (B). Film thickness 3 40 nm. 77

Cross-sectional FESEM image showing the lack of polyelectrolyte
deposition in the inner pores of an alumina substrate where the filtrate side
is coated with a lO-bilayer PAH/PSS film. 78

Plot of normalized receiving-phase conductivity (K2804) as a function of
time when the source phase was separated from the receiving phase by a
bare alumina substrate (open squares), or a 1- (diamonds), 2- (squares), 3-
(triangles), 4- (inverted triangles), 5- (circles), or 10-bilayer (X’s)
PAH/PS8 fihn. 81

Plot of normalized receiving-phase conductivity as a fimction of time
when the source phase was separated from the receiving phase by a 5-
bilayer PAH/PS8 film. Different symbols represent different salts: KCl
(triangles), KN 03 (squares), K2304 (diamonds), K2Ni(CN)4 (circles), and
K3Fe(CN)6 (inverted triangles). 82

FESEM images of the top of an alumina substrate coated with either a 4.5-
bilayer PAA/PAH fihn heated at 115 C (A) or a 5-bilayer PSS/PAH film
coated with 1.5-bilayers of PAA/PAH (B). Samples were coated with 5
nm Au for imaging. 87

Experimental setup for nanofiltration experiments: Ar tank (1), stainless
steel feed tank (2), centrifugal pump (3), prefilter (4), flowmeter (5),
pressure gauges (6), membrane cells (7), and graduated cylinders (8).
System pressure was 70 psi and flowrate was 18 mL/min. 96

xi

Figure 5.2

Figure 5.3

Schematic diagram of the nanofiltration membrane cell. (A) shows how
the cell is placed together: inlet/outlet ports (1&2), threads (3), rubber o-
ring (4), alumina membrane (5), porous stainless steel frit (6), cap that fn't
fits in and presses the membrane tight against O-ring (7), and threaded
bottom part of cell that top part screws into (8). The cell is assembled in
the direction of the arrow. (B) shows a bottom view of the upper portion
of cell: recessed inlet/outlet channels (1&2) and o-ring (3). 97

Reaction scheme for derivazitation of composite membranes with lH-lH-
perfluorooctylamine (PFOA). In step 1 the acid groups of PAA are
converted to a mixed anhydride. In step 2, the amine group of PF OA
reacts with the anhydride resulting in attachment of PFOA to the
composite film through an. amide linkage. 107

xii

F TIR-ERS

Hz

in
MPA
Q
PAH
PSS
FAA
Ra
Rs

5

T

V

ZA

23

ZX
Zw

LIST OF ABBREVIATIONS

double-layer capacitance

concentration of species A in membrane phase
concentration of species A in solution phase
cyclic voltammogram

iEpa _ EPCI in CV

equilibrium potential of couple

Fourier transform infrared — external reflectance
spectroscopy

Hertz

peak current

mercaptopropionic acid

ohms

poly(allylamine hydrochloride)
poly(styrene sulfonate)

poly(acrylic acid)

charge transfer resistance

solution resistance

second

temperature

volt

valence of ion A

valence of ion B .
valence of fixed charge in membrane
Warburg impedance

xiii

CHAPTER 1

Introduction and Background

1. Development of Thin Films

Controlled surface modification with ultrathin, organic films began with the work
of Langmuir and Blodgett in the 1930’s."2 Since then, there have been many advances
in this area of research with each discovery having its own unique benefits and
limitations?"6 One of the latest techniques for the modification of surfaces is the use of
layered polyelectrolyte films.6 Advantages of this deposition technique include an
increase in the types of substrates used for coating, increased film stability, and a wider
variety of film constituents. This thesis presents an investigation of the ion permeability
of layered polyelectrolyte films and their use as ion-separation membranes.

Deposition of films using the Langmuir-Blodgett (LB) technique first showed that
organic molecules could be deposited onto a solid support with a high degree of order.'»2
The deposition of LB films begins with compression of a monolayer of amphiphilic
molecules on a liquid surface (typically water) in a Langmuir trough. The amphiphiles
orient themselves so that the hydrophilic end of the molecule is in contact with the
aqueous phase. Film deposition occurs as a substrate is lowered, perpendicular to the
liquid surface, into the trough. This results in deposition of one layer of molecules.
Subsequent layers are deposited by repeated immersions in the Langmuir trough.
Although LB films are highly ordered, they are limited to low thicknesses, and the weak
van der Waals forces holding the film together provide little stability.7 Another
disadvantage relates to the required amphiphilic nature of the absorbed molecules, which

restricts film composition severely.

In the early 1980’s, the development of self-assembled monolayers (SAMs)
provided a new and convenient method to construct modified surfacesg»9 The majority
of SAMs are alkanethiols, which rely on the formation of the M-S (M=Au, Cu, Fe, Ag)
bonds for adhesion between the substrate and adsorbate. SAMs offer greater
convenience for film deposition relative to LB films because the substrate is simply
immersed in a solution containing the self-assembling molecule. Although SAMs
provide a simple method for fihn deposition, they lack any means for stable formation of
multilayered films, and the M-S bond stability becomes an issue at high
temperatures. 10,1 1

The limitations imposed by SAMs and LB films elicited the development of new
procedures for thin film deposition that provide greater film stability while maintaining
control over film structure and composition. The introduction of layered polyelectrolyte
films by Decher and coworkers provided the ability to create thin films with well-defined
control over film thickness and composition.12 Multiple electrostatic interactions
between the individual polyelectrolyte layers provide the basis for film formation and
stability. The synthesis of these films, which involves alternating immersions of a
substrate in solutions containing polycations and polyanions, is simple and does not
require organic solvents or expensive instrumentation. An additional advantage of
layered polyelectrolyte films is their ability to deposit on a wide variety of substrates.
Both LB films and SAMs require smooth surfaces for defect-flee fihn deposition. With
polyelectrolyte films, one only needs a charged surface to begin deposition, because
adsorption of these films tends to cover defects.13 This furthers the possible applications

of layered polyelectrolyte fihns.

11. History and Development of Layered Polyelectrolyte Films

Iler was the first to show the premise for depositing films based on electrostatic
interactions with his work on multilayered, charged colloidal particles. ‘4 Decher and
Hong then extended this idea to the sequential adsorption of alternating layers of charged
polyrnersfi’12 Decher’s technique provided remarkable nanoscale control of film
thickness and composition. Additionally, the only prerequisite for inclusion of a material
in the film is that the moiety be charged, and this greatly increases the possibilities for
fihn structure and composition. The variety of films made by the alternating
polyelectrolyte deposition (APD) technique is illustrated by the use of DNA,15 redox-
active polymers,16"7 charged viruses,18 semiconductor particles,'9,20 conducting
polymers,21 and inorganic sheets,13’15’,22'25 in the formation of layered polyelectrolyte
films. The type of substrate used in the APD process also varies widely and includes
examples such as Si02,26’27 quartzflz,28 glass,29,30 Au modified with a charged
monolayer,“ charged colloids,”34 and polymers.”39 This wide range in film
composition has led to extensive research concerning the structure and application of
polyelectrolyte films.

Synthesis of layered polyelectrolyte films is a simple “Dip n’ Rinse” (Figure 1.1)
procedure, which begins with immersion of a charged substrate into a solution containing
an oppositely charged polyelectrolyte. A fihn forms due to electrostatic interactions
between the substrate and polyelectrolyte. This process is driven by entropy, which
increases due to the displacement of counter-ions from the surface by the adsorbing

polyelectrolyte chain. Rinsing of the substrate with water and immersion in a second

 

 

 

 

 

 

 

Figure 1.1: Schematic diagram of the “Dip n’ Rinse” procedure for the deposition of a
polyelectrolyte film on a gold electrode coated with a monolayer of carboxylates. In the
actual structure of these fihns polycations and polyanions are intermingled.

solution containing an oppositely charged polyelectrolyte yields another layer on the
surface. Repetition of this process produces multilayer films whose thickness depends on
the number of polyelectrolyte depositionsflzfio,41 pH of the deposition solutions,42 and
the supporting salt concentration in the polyelectrolyte solutions.3 1 41,43

Several groups examined the grth and structure of polyelectrolyte films using
techniques such as small angle X-ray scattering,44 X-ray and neutron reflectometry,“49
UV/V IS spectroscopy,27,30,50 scanning electron microscOpy,5l atomic force
microscopy,22,24’5' X-ray photoelectron spectroscOpy,26,36 time-of-flight secondary ion
mass spectrometry,”52 radiolabelling,53,54 quartz crystal microbalance gravimetry,55'57
and optical ellipsometryfi“,47 The prevailing theory on polyelectrolyte film structure is
that the individual layers are highly intermingled with surrounding layers.46»43,53
Schlenoff and coworkers used electrochemical techniques to investigate electron-transfer
through polyelectrolyte films containing redox-active layers. “5’53 These studies showed a
large degree of interrningling between individual polyelectrolyte layers, which resulted in
electron transfer between electro-active layer-pairs, even when separated by four non-
conducting layer-pairs.

Due to the high degree of entanglement between polyelectrolyte layers, there
appears to be one-to-one charge compensation between polyanion and polycation layers.
Several studies detected no counter ions in the bulk of polyelectrolyte films, supporting
the theory of one-to-one charge compensationfl‘i‘w’S3 Schlenoff and coworkers used
radiolabeled ions to show that charge compensation is >99% intrinsic within the bulk of
poly(diallyldimethylammonium)/PSS films, while at the surface there exists ~45% charge

excess with respect to the total layer charge.53 The excess charge at the surface is due to

the formation of loops and trains, which extend into the solution.59 The interface-induced
nonstoichometry caused by the formation of loops and trains at the film surface provides
the driving force for subsequent layer adsorption.59 Charge screening by supporting
electrolyte controls the degree of extrinsic charge; therefore, greater film thicknesses
result from increased salt concentrations in the deposition solutions.

After determining the nature of polyelectrolyte film structure, research began
focusing on possible applications of layered polyelectrolyte films. These films are
relevant to sensing,2’3,57,50»61 separations}849,52"S4 and electronics‘7’203"58 applications
because of their simple layer-by-layer synthesis and wide tunability. Additionally, these
films can naturally cover defects13 and coat substrates with unusual geometries. The
feasibility of applying these films will depend on their permeability and stability under
relevant conditions. Some applications, (e. g., sensing) require highly permeable films,
while others (e. g., anti—corrosion coatings) necessitate impermeable fihns. While there is
much information on the growth and structure of polyelectrolyte films in the “dry”
state,6,30,31494144494951,6536 there is less understanding of the structure of films exposed
to solvent.29»55,56»67 Our interest in using polyelectrolyte films for ion-separation

applications requires an understanding of solvated film properties.

III. Ion-Permeability of Layered Polyelectrolyte Films

Several groups performed fundamental studies of ion-transport through
polyelectrolyte films. Mfihwald and von Klitzing examined the transport of neutral
quenchers (to within 14 nm of the substrate) in polyelectrolyte films using time-

dependent fluorescence.29»68 These studies suggest two regions of differing

polyelectrolyte film structure as indicated by a change in the diffusion coefficients of the
quencher probes. The outer layers of the film are bound more loosely than the film bulk,
which probably has a higher degree of entanglement with neighboring layers. This gives
rise to decreases in the effective diffusion coefficients of the quencher probes in the film
interior. A second conclusion from these studies was that film structure depends strongly
on the concentration of supporting electrolyte in the polymer deposition solutions. Films
made without supporting electrolyte show drastic increases in fihn permeability as
compared to those made with supporting electrolyte due to the smaller degree of layer
interpenetration and smaller layer thicknesses.

Lindholm-Sethson and coworkers investigated the ion-permeability of layered
PAH/PSS fihns using electrochemical techniques.69 They examined the change in film
permeability to redox probes as a function of the number of bilayers in the film and the
individual bilayer thicknesses. Two types of films were used: thick and thin multilayer
films with bilayer thicknesses of 12 and 2 nm, respectively. Film thickness was
controlled by varying the amount of H2804 in both the PAH and PSS solutions. The
cyclic voltarmnetry of thin polyelectrolyte films in this recent study shows trends similar
to those in this thesis. F e(CN)64' permeability decreased with the addition of each
polyelectrolyte bilayer until after the addition of 8-bilayers, the CV response was
primarily due to capacitive current. When Os(bipy)33+ was used as a redox couple, the
film showed no inhibition of cation flux until the deposition of 5-bilayers. Deposition of
further layers greatly decreased flux. In the thick bilayer system, a thin film CV response

was observed indicating adsorption of the redox-active molecule in the film. As

expected, the amount of ion adsorption increased with the addition of each
polyelectrolyte bilayer.

In Chapter 2 of this thesis, we examine the permeability of two different types of
layered polyelectrolyte films as a function of solution pH, number of bilayers, and film
composition using electrochemical techniques. This study differs from Mohwald and von
Klitzing’s work as we describe the film permeability through the entire structure to the
substrate surface (They limited their tests to ~14 nm from the substrate surface). The
results presented here complement those of Lindholm-Sethson but, in addition, we
examine fihn permeability under various solution conditions and as a fimction of the film
structure. Our initial studies focus on layered poly(allylamine hydrochloride)/
poly(styrenesulfonate) (PAH/PSS) and PAH/poly(acrylic acid) (PAH/PAA) films in pH
3.2-, pH 6.3—, and pH lO-buffered solutions. Cyclic voltammetry indicates that solvent-
swollen film structure and permeability are highly dependent on solution pH.
Electrochemical impedance spectroscopy shows that multilayer PAH/PSS films block
gold electrodes and that passivation varies with the number of polyelectrolyte bilayers,
solution pH, and the concentration of salt used in film deposition. In contrast, PAH/FAA
films are highly permeable even after deposition of films as thick as 44 nm.

Although PAH/PAA films are highly permeable, the presence of carboxylate and
amine groups provides a convenient means to covalently cross-link these films and make
them highly blocking. Chapter 3 of this thesis discusses the permeability of such cross-
linked, nylon-like films and illustrates their increased stability over a wide pH range.
Fourier transform infrared external reflection spectroscopy (F TIR-ERS) and X-ray

photoelectron spectroscopy (XPS) confirm the conversion of carboxylate and ammonium

groups into amide bonds while electrochemical impedance spectroscopy shows that 11-

13 nm-thick cross-linked films have resistances as high as 105 Q-cmz.

IV. Use of Layered Polyelectrolyte Films as Ion-Separation Membranes

Anion separation membranes have along and rich history in many industrial
fields. Applications include salt production from seawater?“2 acid and dye recovery
from brine solutions,52,73'75 separation of organic acids during food processing,76
sensors-”’78 and separation of pollutants from groundwater.”81 The majority of anion
separation processes include polymer membranes used in conjunction with
electrodialysis. The properties (e.g. hydrophilicity70»72,32:84 and degree of cross-
linking7l’76) of the polymer membrane dictate flux and selectivity. Although polymer
membranes are extensively used, there are several drawbacks such as low porosity and
large membrane thicknesses73v75’3’ 34-36 (50-500 pm), which are due to manufacturing
limitations. In an attempt to enhance separations by these membranes, 3 polyelectrolyte
layer is deposited on anion-exchange membranes. In some cases, enhanced performance
is observed with this type of modification.63’87’88

The inherent disadvantages of thick membranes may be overcome with the
development of an ultrathin polymer skin on a highly porous support material. Layered
polyelectrolyte films deposited on porous alumina supports may provide a versatile
means for investigating such ion-separation membranes. Several groups recently
reported the use of polyelectrolyte fihns to modify porous polymer substrates for gas
separation.35,33’39a64 In some cases, these studies show selectivity enhancements with

polyelectrolyte films containing between 10 and 200 layers. A recent study showed the

applicability of polyelectrolyte films for the separation of monovalent and divalent
ions.89 However, none of these reports address the issue of whether film deposition
occurs at the membrane surface or within the pores. Determining the structure of
polyelectrolyte fihns deposited on porous substrates and the factors controlling transport
of mono-, di-, and trivalent anions of varying size through multilayered, ultrathin
polyelectrolyte films represents the second objective of this thesis.

Deposition of an ultrathin film on a porous support is a challenging endeavor, as
one must coat the substrate without depositing material in the pores of the support.
Porous alumina was chosen as the support material for these studies due to its high pore
density and well-defined porous structure. In addition, the regular support structure
allows for convenient field emission scanning electron microscope (F ESEM) studies to
determine film coverage and whether deposition occurs within the pores. Utilizing the
simple ADP procedure, polyelectrolyte films were deposited on porous alumina supports.
FESEM images show that as few as five-bilayers (<25 nm) of polyelectrolyte effectively
cover the surface without filling of underlying pores. '

Dialysis measurements with these membranes show that Donnan exclusion is the
primary selection mechanism in ion transport. Selectivity values for Cl'/SO42' and Cl'
/Fe(CN)63' transport were 6 and 500, respectively, for a 4.5-bilayer PSS/PAH membrane.
PAA/PAH membranes show similar selectivity values along with a 70% decrease in
overall ion flux. To increase selectivity, composite membranes were developed to utilize
the high Cl’ transport of PSS/PAH films and the high 8042' rejection and cross-linkability
of FAA/PAH membranes. These composite membranes contain a thin layer of

PAA/PAH deposited on a PSS/PAH precursor film and show a wide range of selectivity

10

values depending on the number of FAA/PAH layers and cross-linking temperatures.
Such films can form ultrathin barriers that allow high flux along with selective transport
of anions, where anion charge and size determine selectivity.

Krasemann and Tieke recently reported ion-permeation results for layered
polyelectrolyte films deposited on a polymeric support material.89 In this study, the
effect of bilayer thickness and charge density on ion transport was determined by varying
the pH and the supporting electrolyte concentration in the polymer deposition solutions.
With PAH/PSS films deposited from solutions containing supporting electrolyte, the Na+
flux steadily decreased with deposition of up to 30 bilayers but then was constant with
further bilayer depositions. Mg2+ flux decreased at a higher rate than Na+ flux and
continued decreases were observed for films with with up to 90 bilayers, resulting in
Nali/Mg2+ selectivity values of between 30 and 112. Anion selectivity values (Cl' over
SO42“) of between 17 and 45 were observed with similar films made in the presence of
supporting electrolyte. Films deposited with no supporting electrolyte showed similar
trends with respect to decreases in flux with increasing numbers of bilayers, but fluxes
were generally higher and selectivity lower for these fihns.

In addition to studying transport as a function of bilayer thickness and number of
bilayers, Krasemann and Tieke also investigated the effect of polyelectrolyte charge
density on ion transport through polyelectrolyte films. One way to alter charge density in
these films is through the incorporation of polyelectrolytes with differing charge
densities. As expected, the ion flux decreased and in some cases ion selectivity increased
as the charge density increased. A second method of changing the film charge density is

by varying the pH of deposition solutions. For films deposited at low pH values where

11

the majority of the amine groups of PAH are protonated, there is decreased cation flux
due to the increased charge repulsion between the film and cation.

Krasemann and Tieke attribute high selectivities to a bipolar membrane transport
mechanism where the ion encounters repulsion at each layer of oppositely charged
polyelectrolyte. They cite several reasons for their explanation. When supporting
electrolyte is present, the polyelectrolyte deposits in a globular form as opposed to a rod—
like structure. This globular form may not allow for the complete interpenetration of the
polyelectrolyte and therefore may result in a higher degree of free charge in the film. This
results in lower ion flux due to the increased Donnan repulsion produced by the
electrostatic interactions between the ion in solution and free charge in the fihn.

However, our studies see a maximum Cl'/SO42' selectivity of 6 with the deposition of 5-
bilayers of PSS/PAH with no further increase in selectivity with deposition of an
additional 5—bilayers. Moreover, simulations by Lebedev and coworkers show that a
multi-bipolar structure will not enhance selectivity.90 Our studies suggest that the outer
layer charge is the dominant factor in controlling flux as opposed to the multiple bilayer
structure suggested by Krasseman and Tieke. We suspect that in Krasemann and Tiekes
study they did not initially form an ultrathin film that fully covered the substrate.
Selectivity enhancement with increasing numbers of layers in the film is also suggestive
of a multi-bipolar membrane structure. Tieke suggests that selectivity would increase

with the number of bipolar interfaces encountered.

12

V. Use of Layered Polyelectrolyte Films as Nanofiltration Membranes

Although diffusion dialysis measurements provide an initial assessment of film
permeability, this technique is not a practical method for large-scale ion separations.
Such separations generally require electrodialysis or reverse osmosis (R0) membranes.
R0 processes utilize a pressure differential across the membrane along with partitioning
in the separation of solutes.” ’92 Traditionally, RO has been subdivided into 3 categories;
high-pressure RO, low-pressure R0, and nanofiltration, which is often referred to as
“soft” R0 (operating pressures of <200 psi). Size exclusion and charge in the membrane
both play a role in the rejection properties of nanofiltration membranes. Typical
applications include residential and industrial water softening along with filtration of
various organic constituents (e.g. humic substances, pesticides, and insecticides).7‘b93'96

There are a limited number of materials currently being used in membrane
separations. R0 and nanofiltration membranes are typically constructed from linear and
cross-linked polyamides and polyetherurea polymers.9'a92»97'100 The majority of these
membrane systems consist of a thick, highly porous support material coated with a thin
“skin”, which provides selectivity and determines membrane flux. Challenges currently
facing nanofiltration membranes are fouling, C12 resistance, and increasing salt rejections
while maintaining high fluxes at low Operating pressures. Present research on
membranes has focused on developing thinner composite layers, chemical derivatization
of commercial membranes, and developing new membrane materials related to
polyamide and cellulosic materials. Childs and coworkers have deviated from the notion
of “thinner is better” and have been modifying membranes with pore filling

polyelectrolytes to enhance membrane performancefilggdm Layered polyelectrolyte

13

films offer the possibility of creating ultrathin membranes as well as the ability to easily
derivatize membranes for increased salt rejection or introduction of anti-fouling
properties.

In the final chapter of this thesis, PSS/PAH films deposited on porous alumina
supports are tested under nanofiltration conditions. The high degree of charge in these
membranes provides moderate and high ion rejections for monovalent and divalent
species, respectively. In addition, high permeate fluxes are maintained with the high
level of ion rejections. Trends from the diffusion dialysis results, such as changes in
rejection with changes in the outer layer charge, are observed in the nanofiltration results.

The high rejections and permeate flux values compare very well with commercially

available nanofiltration membranes.

l4

VI. References

l) Langmuir, I. J. Am. Chem. Soc. 1917, 39, 1848-1852.

2) Blodgett, K. B. J. Am. Chem. Soc. 1935, 57, 1007-1010.

3) Swalen, J. D.; Allara, D. L.; Andrade, J. D.; Chandross, E. A.; Garodd, S.;
Israelachvili, J .; McCarthy, T. J .; Murray, R.; Pease, R. F .; Rabolt, J. F.; Wynne, K. J .;
Yu, H. Langmuir 1987, 3, 932-950.

4) Ulman, A. Introduction to Thin Organic Films: From Langmuir-Blodgett to Self-
Assembly; Academic Press: Boston, 1991.

5) Ulman, A. Chem. Rev. 1996, 96, 1533-1554.
6) Decher, G. Science 1997, 277, 1232-1237.

7) Embs, F.; Funhoff, D.; Laschewsky, A.; Licht, U.; Ohst, H.; Prass, W.; Ringsdorf, H.;
Wegner, G.; Wehrmann, R. Adv. Mater. 1991, 3, 25-30.

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

9) Nuzzo, R. G.; Dubois, L. H. Chem. Rev. 1996, 96, l533-1554.

10) Schlenoff, J. 8.; Li, M.; Ly, H. J. Am. Chem. Soc. 1995, 117, 12528-12536.

1 1) Nuzzo, R. G.; Fusco, F. A.; Allara, D. L. J. Am. Chem. Soc. 1987, 109, 2358-2368.
12) Decher, G.; Hong, J.-D.; Schmitt, J. Thin Solid Films 1992, 831-835.

13) Kleinfeld, E. R.; Ferguson, G. S. Chem. Mater. 1996, 8, 1575-1578.

14) Iler, R. K. J. Coll. Int. Sci 1966, 21 , 569-594.

15) Lvov, Y.; Decher, G.; Sukhorukov, G. Macromolecules 1993, 26, 5396-5399.

16) Laurent, D.; Schlenoff, J. B. Langmuir 1997, 13, 1552-1557.

17) Wu, A.; Yoo, D.; Lee, J .-K.; Rubner, M. F. J. Am. Chem. Soc. 1999, 121, 4883-4891.
18) Lvov, Y.; Haas, H.; Decher, G.; Mohwald, H. Langmuir 1994, 10, 4232-4236.

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

15

20) Kotov, N. A.; Déka'my, 1.; Fendler, J. H. J. Phys. Chem. 1995, 99, 13065-13069.
21) F on, A. C.; Rubner, M. F. Macromolecules 1995, 28, 7115-7120.

22) Kerimo, J.; Adams, D. M.; Barbara, P. F.; Kaschak, D. M.; Mallouk, T. E. J Phys.
Chem. B 1998, 102, 9451-9460.

23) Kim, H.-N.; Keller, S. W.; Mallouk, T. E. Chem. Mater. 1997, 9, 1414-1421.

24) Fang, M.; Kaschak, D. M.; Sutorik, A. C.; Mallouk, T. E. J Am. Chem. Soc. 1997,
119, 12184-12191.

25) Kleinfeld, E. R.; Ferguson, G. S. Science 1994, 265, 370-373.
26) Sukhishvili, S. A.; Granick, S. J. Chem. Phys. 1998, 109, 6861-6868.

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

28) Yang, X.; Johnson, S.; Shi, J.; Holesinger, T.; Swanson, B. Sensors and Actuators B
1997, 45, 87-92.

29) Klitzing, R. v.; Mohwald, H. Macromolecules 1996, 29, 6901-6906.
30) Ferreira, M.; Rubner, M. F. Macromolecules 1995, 28, 7107-7114.
31) Caruso, F .; Niikura, K.; Furlong, D. N .; Okahata, Y. Langmuir 1997, 13, 3422-3426.

32) Caruso, F .; Lichtenfeld, H.; Giersig, M.; Mohwald, H. J Am. Chem. Soc. 1998, 120,
8523-8524.

33) Caruso, F.; Caruso, R. A.; Mohwald, H. Science 1998, 282, 1111-1 1 14.

34) Donath, E.; Sukhorukov, G. B.; Caruso, F .; Davis, S. A.; Mfihwald, H. Angew.
Chem. Int. Ed. 1998, 37, 2202-2205.

35) Kotov, N. A.; Magonov, S.; Tropsha, E. Chem. Mater. 1998, 10, 886-895.
36) Chen, W.; McCarthy, T. J. Macromolecules 1997, 30, 78-86.

37) Delcorte, A.; Bertrand, P.; Wischeroff, E.; Laschewsky, A. Langmuir 1997, 13,
5125-5136.

38) Ackem, F. v.; Krasemann, L.; Tieke, B. Thin Solid Films 1998, 327-329, 762-766,

16

 

39) Levéisalmi, J .-M.; McCarthy, T. J. Macromolecules 1997, 30, 1752-1 75 7.

40) Hoogeveen, N. G.; Cohen Stuart, M. A.; F leer, G. J. Langmuir 1996, 12, 3675-3681.
41) Dubas, S. T.; Schlenoff, J. B. Macromolecules 1999, 32, 8153-8160. I

42) Yoo, D.; Shiratori, S. S.; Rubner, M. F. Macromolecules 1998, 31, 4309-4318.

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

44) Lvov, Y.; Decher, 6.; Has, H.; Mohwald, H.; Kalachev, A. Physica B 1994, 198,
89-91.

45) Sukhorukov, G. B.; Schmitt, J.; Decher, G. Ber. Bunsenges. Phys. Chem. 1996, 100,
948-953.

46) Schmitt, J .; Grunewald, T.; Decher, G.; Pershan, P. S.; Kjaer, K.; Losche, M.
Macromolecules 1993, 26, 7058-7063.

47) Tronin, A.; Lvov, Y.; Nicolini, C. Colloid Polym. Sci. 1994, 272, 1317-1321.

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

49) Decher, G.; Lvov, Y.; Schmitt, J. Thin Solid Films 1994, 244, 772-777. l

50) Sukhorukov, G. B.; Mohwald, H.; Decher, G.; Lvov, Y. M. Thin Solid Films 1996, ‘i'
220-223. E

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

52) Delcorte, A.; Bertrand, P.; Arys, X.; Jonas, A.; Wischeroff, E.; Mayer, B.;
Laschewsky, A. Surf Sci. 1996, 366, 149-165.

53) Schlenoff, J. B.; Ly, H.; Li, M. J Am. Chem. Soc. 1998, 120, 7626-7634.
54) Schlenoff, J. 8.; Li, M. Ber. Bunsenges. Phys. Chem 1996, 100, 943-947.

55) Lvov, Y.; Ariga, K.; Onda, M.; Ichinose, I.; Kunitake, T. Colloid Surf A 1999, 146,
337-346.

56) Xu, H.; Schlenoff, J. B. Langmuir 1994, 10, 241-245.

5 7) Caruso, F.; Niikura, K.; Furlong, D. N.; Okahata, Y. Langmuir 1997, 13, 3427-3433.

17

 

58) Schlenoff, J. B.; Laurent, D.; Ly, H.; Stepp, J. Adv. Mater. 1998, 10, 347-349.
59) Lowack, K.; Helm, C. A. Macromolecules 1998, 31, 823-833.

9

60) Montrel, M. M.; Sukhorukov, G. B.; Petrov, A. 1.; Shabarchina, L. 1.; Sukhorukov
B. I. Sensors and Actuators B 1997, 42, 225-231.

61) Lowy, D. A.; Finklea, H. O. Electrochimica Acta 1997, 42, 1325-1335.
62) Nam, S. Y.; Lee, Y. M. J. Membr. Sci. 1997, 161-171.

63) Mika, A. M.; Childs, R. F .; Dickson, J. M.; Mch, B. E.; Gagnon, D. R. J
Membr. Sci. 1997, 81-92.

64) Stroeve, P.; Vasquez, V.; Coelho, M. A. N.; Rabolt, J. F. Thin Solid Films 1996,
708-712.

65) Lvov, Y.; Decher, G.; Mohwald, H. Langmuir 1993, 9, 481-486.

66) Decher, G.; Eckle, M.; Schmitt, J .; Struth, B. Colloid & Interface Sci. 1998, 3, 32-39.
67) Shubin, V.; Linse, P. J Phys. Chem. 1995, 99, 1285-1291.

68) Klitzing, R. v.; Mohwald, H. Thin Solid Films 1996, 352-356.

69) Han, S.; Lindholm-Sethson, B. Electrochimica Acta 1999, 45, 845-853.

70) Saracco, G.; Zanetti, M. C. Ind. Eng. Chem. Res. 1994, 33, 96-101.

71) Sata, T. J Membr. Sci. 1994, 93, 117-135.

72) Sata, T.; Yamaguchi, T.; Kawamura, K.; Matsusaki, K. J Chem. Soc, Faraday
Trans. 1997, 93, 457-462.

73) Moon, P. J .; Parulekar, S. J.; Tsai, S. J Membr. Sci. 1998, 141, 75-89.
74) Xu, X.; Spencer, G. Desalination 1997, 114, 129-137.

75) Stachera, D. M.; Childs, R. F .; Mika, A. M.; Dickson, J. M. J Membr. Sci. 1998,
148, 199-127.

76) Gudematsch, W.; Krumbholz, C. H.; Strathmann, H. Desalination 1990, .79, 249-
260.

77) Toniolo, R.; Comisso, N.; Bontempelli, G.; Schiavon, G.; Sitran, S. Electroanalysis
1998, 10, 942-947.

18

 

78) Ugo, P.; Moretto, L. M.; Mazzocchin, G. A.; Guerriero, P.; Martin, C. R.
Electroanalysis 1998, 10, 1168-1173.

79) Hell, F.; Lahnsteiner, J .; Frischherz, H.; Baumgartner, G. Desalination 1998, 117,
173-180.

80) Fryxell, G. B; Lin, J .; Hauser, T. A.; Nie, Z.; Ferris, K. F.; Mattigod, S.; Gong, M.;
Hallen, R. T. Chem. Mater. 1999, 11, 2148-2154.

81) Sata, T.; Yamaguchi, T.; Matsusaki, K. J Chem. Soc, Chem Comm. 1995, 1153-
1154.

82) Sata, T.; Yamaguchi, T.; Matsusaki, K. J. Phys. Chem. 1995, 99, 12875-12882.
83) Sata, T.; Tagami, Y.; Matsusaki, K. J Phys. Chem. B 1998, 102, 8473-8479.
84) Sata, T.; Mine, K.; Higa, M. J Membr. Sci. 1998, 141, 137-144.

85) Zeng, X.; Ruckenstein, E. J Membr. Sci. 1998, 148, 195-205.

86) Narebska, A.; Warszawski, A. Sep. Sci. T echnol. 1992, 27, 703-715.

87) Mika, A. M.; Childs, R. F.; Dickson, J. M. J Membr. Sci. 1999, 153, 45-56.

88) Childs, R. F .; Mika, A. M.; Pandey, A. K.; McCrory, C.; Mouton, S.;'Dickson, J. M.
Sep. Pur. Tech. 2001, 22-23, 507-517.

89) Krasemann, L.; Tieke, B. Langmuir 2000, 16, 287-290.
90) Lebedev, K.; Ramirez, P.; Mafé, S.; Pellicer, J. Langmuir 2000, 16, 9941-9943.

91) Bhattacharyya, D.; Williams, M. E. Membrane Handbook, Bhattacharyya, D.;
Williams, M. E., Ed.; Van Nostrand Reinhold: New York, 1992, pp 263-280.

92) Brandt, D. C.; Leitner, G. F.; Leitner, W. E. Reverse Osmosis; Brandt, D. C.; Leitner,
G. F.; Leitner, W. E., Ed.; Van Nostrand Reinhold: New York, 1993, pp 1-8.

93) Bohdziewicz, J .; Bodzek, M.; Wasik, E. Desalination 1999, 121.

94) Berg, P.; Hagmeyer, G.; Gimbel, R. Desalination 1997, 113, 205-208.

95) Hofrnan, J. A. M. H.; Beerendonk, E. F.; Folrner, H. C.; Kruithof, J. C. Desalination
1997, 113, 209- 214.

19

96) Rautenbach, R.; Linn, T.; Eilers, L. J Membr. Sci. 2000, 174, 231-241.

97) Childress, A. E.; Elirnelech, M. Environ. Sci. T echnol. 2000, 34, 3710-3716.
98) Kim, K. J.; Chowdhury, G.; Matsuura, T. J Membr. Sci. 2000, 179, 43-52.

99) Kim, C. K.; Kim, J. H.;Roh,1. J.; Kim, J. J. J Membr. Sci. 2000, 165, 189-199.
100) Kosutic, K.; Kastelan-Kunst, L.; Kunst, B. J Membr. Sci. 2000, 168, 101-108.

101) Mika, A. M.; Childs, R. F .; Dickson, J. M. Desalination 1999, 121, 149-158.

20

CHAPTER 2

Electrochemical and in situ Ellipsometric Investigation of the Permeability and
Stability of Layered Polyelectrolyte Films
I . Introduction

Layered polyelectrolyte films1 are attractive for sensing}5 separations?“ anti-
corrosion,12 and electronics'345 applications because of their ease of synthesis and wide
variability. The feasibility of using these films will depend on their permeability and
stability under relevant conditions. Some applications, (e.g., sensing) require highly
permeable films, while others (e.g., anti-corrosion coatings) necessitate impermeable
films. While there is much information on the growth and structure of polyelectrolyte
films in the “dry” state,'»5"7'28 there is less understanding of the structure of films
exposed to solvent”:35 Our interest in using polyelectrolyte films for ion-separation
applications requires an understanding of solvated film properties.

Utilization of layered polyelectrolyte films as sensor or ion-separation materials
will depend critically on their stability and ion-permeability in aqueous solution. To
elucidate solvated film properties, electrochemical and in situ ellipsometric studies were
used to determine the permeability and stability of layered poly(allylamine
hydrochloride)/poly(styrenesulfonate) (PAH/PSS) and PAH/poly(acrylic acid) (PAA)
films. The permeability of these layered polyelectrolyte films to Fe(CN)63' and
Ru(NH3)(,3+ depends on solution pH, the number of bilayers in the film, whether
supporting electrolyte is present during film deposition, and the nature of constituent
polycations and polyanions. Cyclic voltammetry and impedance spectroscopy show that

film permeability is similar in pH 3.2- and pH 6.3-buffered solutions but increases

21

dramatically in alkaline solutions. In situ ellipsometry helps to explain these results.
Upon immersion in pH 3.2- and 6.3-buffered solutions, the thickness of PAH/PSS films
increases by 40%, but swelling is constant over time. At pH 10, these films initially
swell by 40% but then continue to swell for several minutes before delaminating. The
onset of increased swelling corresponds with dramatic increases in film permeability.
Both peak current (cyclic voltammetry) and charge transfer resistance (ac
impedance) depend on the number of polyelectrolyte bilayers and deposition conditions.
The structure of the first two bilayers is more porous than that of later bilayers. Adding
supporting electrolyte to deposition solutions results in thicker bilayers and changes film
permeability. For PAH/PSS fihns, the use of supporting electrolyte during film
formation results in a much less permeable fihn (comparing films of similar thickness).
In the case of PAH/PAA films, however, the use of supporting salt results in highly
permeable films (even with thicknesses as high as 44 nm). Thus, proper choice of
constituent polyelectrolytes and deposition conditions permits control over the

permeability of layered polyelectrolyte films.

11. Experimental
A. Chemicals and Solutions

Poly(allylamine) hydrochloride (Mw=50-65,000), sodium poly(styrenesulfonate)
(MW 2 70,000), and 3-mercaptopropionoic acid were used as received from Aldrich.
Hexaammineruthenium(III) chloride and hexaarnminerutheniumfll) chloride were used
as received from either Aldrich or Alfa Aesar. Poly(acrylic acid) (25% by wt. in water;

MW 3 90,000) was used as received from Acros. Solutions for electrochemical studies

22

contained 1.0 M NaZSO4 and were buffered with the following salts: 0.2 M CH3COOH +
0.0036 M CH3COONa (pH = 3.2), 0.025 M NaHzPO4 + 0.025 M Na2HP04 (pH = 6.3),
and 0.025 M N32CO3 + 0.025 M NaHC03 (pH = 10). Solutions for cyclic voltammetry
contained 0.005 M K3Fe(CN)6 or 0.005 M Ru(NH3)6Cl3, and solutions for impedance
measurements contained equimolar (0.005 M) K3Fe(CN)6/K4Fe(CN)6 or
Ru(NH3)6C13/Ru(NH3)6C12 in the above mentioned buffers. Ru(NH3)62+ undergoes
oxidation to Ru(NH3)(,3+ so quantitative electrolysis was performed until the open circuit
potential reached the formal redox potential of the couple (as determined from cyclic
voltammetry).

PSS deposition solutions (15 mL total volume, pH 2.3) contained 55 mg of
polymer and 1.5 mL of 0.01 M HCl. Solutions for preparing films in the presence of
supporting electrolyte also contained 0.5 M MnClz.17’21 PAH deposition solutions (15
mL total volume, pH 2.1) for PAH/PSS films contained 28 mg polymer and 1.5 mL of
0.01 M HCl. Deposition solutions were made following literature procedures.17 For
depositions in the presence of supporting electrolyte, PAH solutions contained 0.5 M
NaBr (PAH/PSS films) or 0.5 M NaCl (PAH/FAA films). The presence of Man causes
the precipitation of PAA at pH values of ~5 so we chose to use NaCl as the supporting
electrolyte for PAH/FAA deposition. PAA solutions were prepared by diluting 1.5 mL of
stock PAA solution (25% by wt.) to a final volume of 250 mL. Deposition solutions for
PAH/PAA films made with supporting salt (0.5 M NaCl) were adjusted to pH 4.5 with
NaOH or HCl. Deposition solutions for PAH/PAA films made without supporting salt
were adjusted to pH 5.6 (PAA) and 5.1 (PAH). The higher pH yields thinner films with

fewer loops and trains because of electrostatic repulsion between the monomer units.

23

Deposition times were five nrinutes for PAH and two minutes for PAA and PSS except
for PAH/PAA films made without supporting electrolyte. A five minute immersion for
both PAH and PAA was used for these films.

Adjusting the pH of polymer solutions is challenging, as polyelectrolytes tend to
adsorb to the pH electrode. Before pH adjustment the pH electrode was immersed in 0.1
M HCl followed by 0.1 M NaOH for ~1 min with water rinses following each immersion.
The polyelectrolyte solution is stirred constantly during the pH adj ustrnent and 0.1 M
HCl or 0.1 M NaOH was added dropwise. Following the addition of HCl or NaOH, the
pH electrode was immersed in the stirred solution until the reading stabilized. The
electrode was removed and stored in water between measurements. If the reading did not
stabilize or if adjustments took longer than ~15 min the electrode was rinsed again in HCl
and NaOH.

Gold slides (electron-beam evaporation of 200 nm of Au on 20 nm Ti on Si(100)
wafers) were UV/03 cleaned for 15 minutes and immediately immersed in an ethanolic
solution containing 0.002 M mercaptopropionoic acid (MPA) for 30 minutes. Slides
were then rinsed with ethanol followed by Millipore water (18 MQ-cm). To prepare one
polyelectrolyte bilayer, the MPA-coated slide was first immersed in PAH and then rinsed
under a continuous stream of water for one minute. Following deposition of PAH, the
slide was immersed in a solution of either PSS or PAA followed by a one minute rinse.
Additional bilayers were prepared similarly. Slides were dried with nitrogen only after

the desired number of bilayers were deposited.

24

B. Techniques

Electrochemical measurements were made using a standard three-electrode cell
containing a Ag/AgCl (3 M KCl) reference electrode and a platinum wire counter
electrode. The working electrode was a gold slide sealed in a plastic o-ring holder with
an exposed area of 0.1 cm2. A CH-Instruments 604 Electrochemical Analyzer was used
to perform the impedance and cyclic voltammetry measurements. A scan rate of 0.1
V/ sec was used for all cyclic voltammograms. Impedance measurements were taken at
frequencies ranging from 0.1 to 105 Hz using a 5 mV sinusoidal voltage centered around
E°' of Fe(CN)63'/4' or Ru(NH3)62+/3+. Solutions containing Ru(NH3)62+/3+ were purged
with nitrogen for five minutes prior to measurement, and the cell was kept under flowing
nitrogen for the duration of the experiment.

On each electrode, voltammetric and then impedance data were collected at several
pH values in the following order: pH 3.2, 6.3, 10, and pH 3.2 after immersion in pH 10.
Cyclic voltammetry (F e(CN)63 ' and Ru(NH3)63+) indicates that peak currents for
electrodes coated with 3-, 4-, and 5-bilayers of PAH/PSS are stable after a 20-minute
immersion in each of the buffered solutions. A 20-minute immersion was. thus used prior
to both voltammetry and impedance measurements. A two-minute immersion was
adequate for PAH/PAA films and 1-, 2-, and 20-bilayer PAH/PSS fihns as these films are
highly permeable. Between cyclic voltammetry and impedance experiments, the
electrode was rinsed with water. Charge transfer resistance (Rm) values (Table 2 & 4)
were obtained by fitting impedance data with a Randles’ circuit using complex non-linear

squares regression (LEVM 7.0 by J . Ross McDonald, distributed by Solartron).

25

Ellipsometric measurements were obtained with a rotating analyzer ellipsometer
(Model M-44; J .A. Woollam) using WVASE32 software. Substrate parameters (n & k)
were measured after absorption of MPA. This insures that any changes in substrate
reflectivity due to Au-S bonds will not affect subsequent measurements. The angle of
incidence was 75° for all experiments. The thickness values for PAH/PSS films were
determined using 44 wavelengths between 414.0 and 736.1 nm. Refractive indices of
PAH/PSS films containing between 1- and 5-bilayers (made with supporting salt) were
assumed to be the same as those determined for a lO-bilayer fihn using ellipsometry.
Over this wavelength region, the refractive indices for a lO-bilayer PAH/PSS film ranged
from 1.53-1.66. A refractive index of 1.5 was assumed for films containing 20 PAH/PSS
bilayers (no supporting electrolyte) or 9 PAH/PAA bilayers (no supporting electrolyte),
as these films are not sufficiently thick for refractive index determination. Refractive
index and thickness values for PAH/PAA films (made with supporting salt) were
calculated simultaneously from the experimental L1’ and A values between the
wavelengths of 465.8 and 548.8 nm (11 values ranged from 1.56-1.60). This region
produced the best graphical fit, but calculation of the thickness over the entire wavelength
region changes the thickness by < 1.0 nm.

In situ ellipsometry was performed on lO-bilayer PAH/PSS films in each of the
buffered solutions using a home-built plastic cell with glass windows. The cell was first
placed on the sample stage without windows and the stage height was adjusted for
maximum intensity and flatness. The addition of the cell windows creates two reflections
(from the front and rear windows). To align the windows, they were adjusted so that both

reflections overlap the incident beam ensuring that the windows are perpendicular to the

26

be

ch

m

\Vl

all

1h;

[1]

HO?

beam (within 01°). The effect of the windows on measurements of dry films is <O.25°
for both ‘1’ and A. This possible error in ‘1’ and A, due to the windows, results in a ~l %
change in the calculated film thickness (~0.5 nm) as compared to measurements made
without the cell. Buffered solutions were then poured into the cell and measurements
were recorded as a function of time until a stable thickness was observed. In situ
ellipsometry was performed on four different lO-bilayer PAH/PSS films (dry thickness of
38 :1: 2 nm). The standard deviation of the swollen thickness in the four samples was less

than 5 nm in each buffered solution (pH 3.2, 6.3, and 3.2 after pH 10).

III. Results and Discussion
A. Characterization of Polyelectrolyte Films

Consistent with previous studies,'7a20»26,36’37 ellipsometry of "dry" films
demonstrates uniform, stepwise growth of PAH/PSS coatings as seen in Figure 2.1. The
thickness increase after the addition of each bilayer (with the exception of the initial
bilayer, which was 2.4 i: 0.3 nm thick) was 3.8 i 0.3 nm, which agrees with previous
reports.‘7’26,36’37 PAH/PSS films made with no supporting salt also exhibit linear growth
up to 20-bilayers. A 20-bilayer PAH/PSS film has a thickness of 16 :1: 2 nm. The
thickness of PAH/PAA films made in the presence of supporting salt increases linearly
with the number of bilayers,20 but growth of these films in the absence of salt yields
nonlinear growth. ‘2 The average thickness of a 5-bilayer PAH/PAA film
made with supporting electrolyte is 44 i 2 run, while the thickness of a 9-bilayer

PAH/PAA film made without salt is 15 fl: 1 nm.

27

0.05

 

A
O
—l

03
O

 

Thickness (nm)
'8
eoueqiosqv

_\
O
O

- 0.01

 

 

0 I; I #1 _I

0 2 4 6 8 100
Number of PAH/PSS Bilayers

Figure 2.1: Ellipsometric thickness (squares) and F TIR-ERS absorbances at 1037 cm'1
(circles) for multilayered PAH/PSS films made in the presence of supporting salt. The
standard deviation for the absorbance values for each bilayer is ~2%. The standard
deviations for the ellipsometric thickness values are <0.3 nm.

C0

de

p0}

COT.

in

FTIR external reflection spectroscopy (FTIR-ERS) also confirms the stepwise
growth of PAH/PSS films. Figure 2.2 shows the FTIR-ERS spectra of PAH/PSS films
containing up to 5-bilayers. The sulfonate absorbances, 1220, 1175, 1036, and 1005 cm'
', dominate the spectrum of PAH/PSS films, which agrees with previous spectra of
PAH/PSS/Immunoglobulin G fihns.l9 Symmetric and anti-symmetric -—NH3Jr
deformations are visible between 1625-1560 cm'1 and 1550-1505 cm", respectively.38
Absorbance values increase approximately linearly (Figure 2.1) with the number of
bilayers, which supports the ellipsometric data. The protonated and deprotonated acid
peaks, 1706 and 1575 cm", respectively, dominate the spectra of PAH/PAA films (Figure
2.3) and again increases in absorbance values occur with the addition of each bilayer of

polyelectrolyte.

B. Film Swelling

In situ ellipsometry allows one to investigate the swelling of polyelectrolyte films
in water. Upon immersion in pH 3.2- and 6.3-buffered solutions, the thickness of a 10-
bilayer PAH/PSS film increases ~40% due to absorption of water. Swelling remains
constant during the ten minutes of immersion (i 1 nm) as seen in Figure 2.4A&B. After
immersion in pH 3.2 and 6.3 buffers, refractive index values (450.5 nm) for the PAH/PSS
film decrease by ~4% (1.60 to 1.54). This decrease is consistent with a final volume
fraction of between 30 and 40% water (n=1.33). The percent swelling of films at pH 3.2
and 6.3 is reasonable as a previous study indicates that the volume of PAH/PSS fihns at

100% humidity is 2 40% water.27 Upon immersion in a pH lO-buffered solution, the film

29

Fig

 

I 0.005 ‘ ‘

Absorbance

 

 

r-‘NOD h

M

1600 1400 1200 1000
Wavenumbers (cm'l)

 

Figure 2.2: FTIR—ERS spectra of PAH/PSS films (made with supporting salt) composed
of 1-5 bilayers.

30

 

Absorbance

 

 

I I I I

1800 1700 1600 1500 1400
Wavenumbers (cm'1)

 

Figure 2.3: FTIR-ERS spectrum of a 9-bilayer PAH/PAA film (made with no
SUpporting electrolyte) deposited on a gold electrode.

31

 

 

 

Current (uA)
(wu) sseuxogql

 

 

 

 

 

 

 

Time (min)

Figure 2.4: In situ ellipsometric thickness (circles) and cyclic voltammetry (5 mM
Fe(CN)63') peak current values (squares) as a function of immersion time for a lO-bilayer
PAH/PSS film (made with supporting salt) in pH 3.2- (A), 6.3- (B), and lO-buffered
solutions (C).

32

swells as in lower pH buffers, but continues to swell for approximately seven additional
minutes (Figure 2.4C). Over the following ten minutes, the film thickness returns to
approximately its original value. A subsequent immersion in a pH 3.2-buffered solution
results in a reduced amount of stable swelling (20% increase from dry thickness).

The additional swelling at pH 10 is probably due to deprotonation of the amine
groups of PAH. At pH 10, this deprotonation causes a charge imbalance to develop inthe
film. To compensate the remaining negative charges, the film must absorb Na+ from
solution30 and probably some additional water. The subsequent decrease in thickness at
high pH is likely due to delamination of the film and this is suggested by ex situ
experiments. Immersion of a PAH/PSS film (made with supporting salt) into a pH 10-
buffered solution, rinsing, and measurrnent of a “dry” thickness indicate that up to 30%
of the film is lost due to delamination during the rinsing and drying process.
Re-immersion in pH 3.2-buffered solutions and subsequent rinsing results in the same
“dry” thickness as that after immersion in pH lO-buffered solutions. Although in situ and
ex situ experiments show some delamination, both studies indicate that a significant
portion of the film remains on the substrate after immersion in a pH lO-buffered solution.
This is probably due to the high degree of overlap and interrningling of the polymer
layers.25’27 The fihn instability shown by in situ ellipsometry suggests, however, that
PAH/PSS fihns will probably not be useful for sensing or separation applications in

alkaline solutions.

33

de

1‘“

EU

\‘0

bu

it

“‘1
(LI

C. Investigation of Film Permeability to Fe(CN)63'/"‘

The effectiveness of polyelectrolyte films as ion-separation membranes will
depend on their ability to selectivity inhibit transport of specific ions. Electrochemistry
provides an initial assessment of film permeability and serves as a sensitive probe of
structural changes in PAH/PSS fihns. Cyclic voltammetry affords qualitative
comparisons of the passivating abilities of different films based on peak currents and
voltammogram shape. Peak currents (ip) in cyclic voltammograms (CVs) of F e(CN)63‘
vary with pH as given in Table 2.1 There is little difference between CVs in pH 3.2 and
pH 6.3 solutions. However, in the case of 4- and 5-bilayer films immersed in pH 10-
buffered solutions, ip increases ~50-fold with respect to values at pH 3.2 and 6.3. The
increase in peak current values at high pH may be due to an increase in the number of
pinholes to the substrate (increased swelling) as well as film delamination. A second
immersion in pH 3.2 buffer yields nearly the same value of ip as in basic solution.

The shape of CVs (Figure 2.5) also helps in understanding the processes occurring at the
electrode-fihn interface. At pH 3.2 and 6.3, CVs of a 5-bilayer film-coated

electrode are broad and plateau-shaped (Figure 2.5A), suggesting current originating
from an array of microelectrodes (pinholes)39,40 or slow diffusion through‘the
polyelectrolyte film.“ At pH 10, CVs show larger peak currents along with a quasi-
reversible voltammogram (AEp = 180 i 25 mV) (Figure 2.5B). This change in CV shape
at pH 10 points to a transition to semi-infinite linear diffusion.40 At pH 3.2 after
immersion in pH 10, the CVs show an even more reversible response (AEp = 100 :t 8
mV). Although ip values at pH 10 and pH 3.2 after immersion in pH 10 are similar in

magnitude (Table 2.1), the decrease in peak splitting (AEP) suggests variation in film

34

lat
me:

Table 2.1: Average peak currents (uA/cmz) from cyclic voltammograms (Fe(CN)(,3')
measured at PAH/PSS coated gold electrodes in different buffered solutions

 

 

 

 

 

 

 

PAH/PSS pH 3.2 pH 6.3 pH 10 pH 3.2 after
Bilayers (11) pH 10
5 91:1 71:1 400i60 4001:60
4 10:1 9::1 550140 4701:130
3 201:1 201:1 600110 660150
2 480 :t 200 440 i 170 500 :t 70 720 1: 50
1 700 1: 40 670 i 50 670 1: 70 830 :t 140
MPA monolayer 890 i 30 770 :t 30 848 i 30 890 i 40

 

 

 

 

 

 

 

35

 

12 . -. .—

 

0.6 - 0'

 

oofi' -

 

 

 

Current (uA)

 

 

 

 

0% dz do
Potential (V vs. Ag/AgCl, 3 M KCI)

Figure 2.5: CVs (5 mM Fe(CN)63‘) at a gold electrode coated with a 5-bilayer PAH/PSS
film (made with supporting electrolyte). CVs were measured in solutions buffered at pH
3.2 (Azdash), pH 6.3 (Azsolid line), pH 10 (B:circles), and pH 3.2 after immersion in pH
10 (B:squares).

36

 

at:

1011

run

indi

$311

the:

01:

structure or charge. The comparable ip values with the increased reversibility at pH 3.2
after immersion in pH 10 is somewhat surprising. Typically, a more reversible
voltammogram indicates better access to the electrode, which should result in higher peak
currents. It is possible that after a second immersion in pH 3.2 buffer there are large
individual defect areas of semi-infinite diffusion leading to decreased AEp values. At the
same time, however, there may be fewer defects, and hence ip values could be similar to
those observed in voltammograms at pH 10. Impedance data (below) suggest that the
total active area of the electrode is greater at pH 10 than at pH 3.2 after immersion in pH
10 buffer.

Impedance spectroscopy is a complementary technique to cyclic voltammetry that
often allows quantitative determination of kinetic and diffusion parameters.42”43
Impedance methods are also attractive because of the small sinusoidal potentials that are
employed, as opposed to the wide potential window used in cyclic voltammetry. To
interpret impedance data, we used the Randles’ circuit shown in Figure 2.6. For highly
blocking films (Ra ~ 104 Q-cmz) we also used a modified Randles’ circuit that contains
film capacitance and resistance elements.44 Both models yield similar (within 20%) RC,
values for the highly resistive systems. The circuit element of most interest to this work
is R“ because it often relates directly to the accessibility of the underlying gold
electrode.“-45 Figure 2.7 illustrates the fits obtained with a Randles’ equivalent circuit to
the experimental impedance data in each of the buffered solutions

Impedance data for a 5-bilayer PAH/PSS film (Table 2.2) confirm the trends shown
by cyclic voltammetry. Increases in peak current values should be accompanied by

decreases in charge transfer resistance values. The difference between RC, values for a

37

Fio

tit:

 

 

Figure 2.6: Randles’ circuit used to fit impedance data where Ra = charge transfer
resistance, & = solution resistance, Car = double-layer capacitance, and Zw = Warburg
element.

38

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

10000 A ' 12000 r
A B
o 0 O. O
0 0 . <9 0.0 O
O p 0
5000 - O 2 ' 6000 i o 9
o o
o 0 o o
0.. 0..
000 00‘
o
{N
0 - E 0 g
0 5000 10000 c.) 0 6000 12000
Ci
1 80 f g 260 u
C N D
90 ~ - 130 - -
(5% ‘3 9 onav 00 0.2 gooooév
O
0 L 0 l
0 90 1 80 0 1 30 260
zR (Q-cmz)

Figure 2.7: Impedance data (F e(CN)63'/4') for a gold electrode coated with a 5-bilayer
PAH/PSS film (made with supporting electrolyte). Plots were obtained in solutions
buffered at pH 3.2 (A), 6.3 (B), 10 (C), and 3.2 after immersion in pH 10 (D). Solid
circles represent experimental data and open circles show fits obtained with a Randles’

equivalent circuit.

39

Table l.
coated gr

Table 2.2: Average Rm values (Q—cmz) from impedance plots (Fe(CN)63'/4') of PAH/PSS

coated gold electrodes in different buffered solutions.

 

 

 

 

 

 

 

 

 

 

 

3111;511:185) pH 3.2 pH 6.3 pH 10 1.1113110
5 12000 1 3000 13000 1 2000 100 1 20 170 1 70
4 10000 1 1000 9000 1 1000 50 1 10 80 1 50
3 5000 1 350 4000 1 300 40 1 2 70 1 30
2 1013 20110 20110 1014
1 211 712 613 211

 

40

 

S-bilayer film (Fe(CN)63'/4‘) measured in pH 3.2- and 6.3-buffered solutions is less than
10%. Cyclic voltammetry indicates a dramatic increase in fihn permeability at pH 10 and
pH 3.2 afler immersion in pH 10, which is supported by the impedance data. For
electrodes coated with S-bilayers, Rct decreases by 120-fold in pH 10 solutions and by
70-fold in pH 3.2 solutions afier immersion in pH 10 buffered solutions (both with

respect to initial measurements at pH 3.2). Lower Rct values at pH 10 than at pH 3.2 after
immersion in pH 10 suggest a larger active electrode area at pH 10.

Impedance plots for 5-bilayer films in pH 3.2- and pH 6.3-buffered solutions
(Figure 2.7A&B) show only a quarter circle with no visible Warburg line, indicating
complete kinetic control of the electron transfer process.42 The complex impedance plots
for 5-bilayer films in pH 10 and pH 3.2 after immersion in pH 10 (Figure 2.7C&D) are
characterized by the appearance of small, nearly complete semicircles in the high
frequency domain and a straight line at low frequencies. The response at low frequencies
is indicative of facile kinetics43 and greater access to the underlying electrode.
D. Investigation of Film Permeability to Ru(NH3)¢52+/3+
As with F e(CN)63', the permeability of PAH/PSS films to Ru(NH3)63+ increases in

alkaline solutions (Table 2.3). However, the trend in the shape of CVs at pH 10 and pH
3.2 after immersion in pH 10 is different for Ru(NH3)(,3+ than for Fe(CN)63'. For
R11(NH3)¢53+ the voltammogram is more reversible at pH 10 as seen in Figure 2.8A. The
difference between the two couples is likely due to their charge. At pH 3.2, Ru(NH3)63+
may experience some repulsion from ammonium groups in the film. Impedance data

acquired with Ru(NH3)62+’3+ (Table 2.4) are consistent with the trends in the cyclic

41

Table 2.3: Average peak currents (uA/cmz) from cyclic voltammograms (Ru(NH3)63+) at

PAH/PSS coated gold electrodes in different buffered solutions.

 

 

 

 

 

 

 

 

 

 

 

 

pH 3.2
Blila 658:1) pH 3.2 pH 6.3 pH 10 after
y pH 10
5 1701100 80140 730170 4701150
4 240 1 30 180 1 20 790 1 20 640 1 50
3 280 1 40 180 1 30 700 1 40 600 1 100
2 640 1 40 600 1 90 590 1 50 580 1 160
1 700 1 20 720 1 40 650 1 40 730 1 160
MPA monolayer 880 1 20 860 1 20 790 1 40 800 1 40

 

42

 

 

Current (11A)

 

 

 

 

01) 412 414
Potential (V vs Ag/AgCl, 3M KCI)

 

 

 

 

250 '
B
"E
‘3
S3 125' ‘
é
NI .l-. I ' I.
.[Z . M
0 125 250
ZR (Q—sz)

Figure 2.8: (A) CVs (5 mM RU(NH3)63+) of a gold electrode coated with a 5-bilayer

PAH/PSS film (made with supporting electrolyte). CVs shown were measured in

solutions buffered at pH 10 (circles) and pH 3.2 after immersion in pH 10 (squares). (B)

Impedance data (5 mM RU(NH3)62+I3+) for a gold electrode coated with a 5-bilayer

PAH/PSS film (made with supporting electrolyte). Plots were obtained after immersion

in pH lO-buffered solutions (circles) and after reimmersion in pH 3.2-buffered solutions
Squares).

43

Table 2.4: Average Ru values (Q-cmz) from impedance plots (Ru(NH3)62+/3+) of
PAH/PSS coated gold electrodes in different buffered solutions.

 

 

 

 

 

 

131;}; is pH 3.2 pH 6.3 pH 10 afigilpilzl 0
5 740 1 400 800 1 400 20 :i: 10 170 :1: 80
4 300130 360120 1011 80150
3 180:1: 100 220140 1012 120130
2 916 20125 51:2 50:1:10
1 No discernible Rm

 

 

 

 

 

 

 

44

 

1‘01

501

1'31

161

:11

111

in

voltammetry data, which indicate the highest permeability (lowest Rct values) at pH 10.
The domination of the Warburg line at pH 10 (Figure 2.8B), which is indicative of facile
kinetics (mass transfer control), is also consistent with higher penneabilities in basic
solutions. The increase in Rd at pH 3.2 after immersion in pH 10 compared to Ra

values at pH 10 shows that a portion of the film remains on the electrode surface. Both

F e(CN)63 '/ 4' and Ru(NH3)62+/3+ impedance data suggest a decrease in the active area of the

electrode after reimmersion in acidic solutions.

E. Permeability as a Function of Time

Fihn swelling observed with in situ ellipsometry indicates that permeability will
likely depend on immersion time in alkaline solutions. Thus, we decided to investigate
how cyclic voltammetry varies with time. Figure 2.4 shows how peak currents vary with
time when an electrode coated with a lO-bilayer PAH/PSS film is immersed in pH 3.2-,
6.3—, and lO-buffered solutions. Immersions in pH 3.2- and 6.3-buffered Solutions, show
a minimal increase in ip with time, as seen in Figure 2.4A&B and over this period there is
no change from the plateau-shaped voltammogram.

Immersions in a pH 10-buffered solution do not show steady ip values. A 20-
minute immersion in pH 10-buffered solution results in a 30-fold increase in ip for an
electrode coated with a lO-bilayer PAH/PSS fihn. For the first four rrrinutes, the CVs are
stable with only an ~20% increase in i,. Subsequently, ip rapidly increases and begins to
stabilize after 10 minutes. Swelling of PAH/PSS films, as determined by in situ
ellipsometry, occurs on the same time scale as increases in ip as seen in Figure 2.4C.

Voltammograms for a 10 bilayer-coated electrode in pH lO-buffered solution change

45

in

10

111

1'1

611

DE

11

from those characteristic of a microelectrode system or slow diffirsion through the film
(plateau peaks) to those of a quasireversible system during a 20-minute immersion. The
changes in voltammogram shape (plateau to quasireversible) and the increases in ip
indicate that the fihn is undergoing pH-induced structural changes creating greater access

to the underlying electrode.

F. Permeability as a Function of the Number of Bilayers

Increasing the number of bilayers in the film should increase charge transfer
resistance and decrease peak current values because pathways to the electrode surface
will become fewer and longer. This is indeed the trend we generally observe using cyclic
voltammetry. The most drastic changes in electrode blocking occur upon introduction of
the third bilayer as seen in Figure 2.9. For one- and two-bilayer films, CVs contain a
quasireversible response while CVs of electrodes coated with a three-bilayer film show
plateau-shaped voltammograms at low pH values. Peak currents (Fe(CN)(,3') at
electrodes covered with a one- or two-bilayer film are 85% and 55%, respectively, of ip
values for a MPA-modified gold electrode at low pH. The presence of a third bilayer,
however, results in an ~40-fold reduction in peak current at low and moderate pH values
(Table l).

Impedance spectroscopy also shows the dependence of permeability-on the number
of polyelectrolyte bilayers on the surface. For electrodes coated with one- and two-
bilayer films, the Warburg diffirsion element dominates the complex impedance plots.
Addition of a third bilayer results in complete kinetic control as seen by the appearance

of only a semicircle in the complex impedance plot at low pH values. There is a >200

46

 

 

Current (uA/cmz)

 

 

0.4 0.2

0.0
Potential (V vs Ag/AgCI, 3M KCI)

 

 

F igurc 2.9: CVs (5 mM Fe(CN)63') of a gold electrode coated with a layer of MPA
(dashed line) or with a 1- (solid line), 2- (triangles), 3- (circles), 4- (squares), or 5-bilayer
(diamonds) PSS/PAH film. CVs were measured in solutions buffered at pH 6.3 and films
were made from solutions containing supporting electrolyte.

47

fold increase in charge transfer resistance between a two and three bilayer fihn as
opposed to the small increase in RC, values between one and two bilayer films (Table 2.2:
low and moderate pH). Electrochemical data with Ru(NH,),2+’3+ (Table 2.3&2.4) also
show similar changes in permeability with an increasing number of polyelectrolyte
layers.

The nonlinear decrease in permeability with the number of bilayers suggests a
different structure in the first two bilayers than in the rest of the film. Previous studies
indicate that the first few polyelectrolyte bilayers have a different thickness than later
bilayers”,28 With each subsequent bilayer, the charge is multiplied until stepwise
growth is achieved. This agrees with the electrochemical data, which suggest that there is
greater access to the underlying gold electrode within the first two bilayers. These data
show that separations applications will likely require several polyelectrolyte bilayers to

be effective.

F. Effect of Supporting Electrolyte on the Permeability of Polyelectrolyte Films
The addition of supporting salt to deposition solutions changes the structure and

thickness of polyelectrolyte films by screening the ionic charges and allowing for the
formation of loops and trains};46 Polyelectrolytes in solutions without supporting salt
assume a linear structure, as there is a large amount of charge-charge repulsion between
monomer units.46»47 For separation and anti-corrosion applications it is important to
understand how bilayer structure affects transport properties of the film. Voltarnmetric
and impedance measurements (F e(CN)(,3 '/ 4') indicate that PAH/PSS films made from

solutions containing no supporting salt are much more permeable than fihns made with

48

supporting salt as seen in Figure 2.10. A 20-bilayer PAH/PSS film prepared without salt
(thickness = 16 nm) is much more permeable than a four-bilayer PAH/PSS fihn prepared
with salt (thickness = 15 nm). An electrode coated with a 20-bilayer PAH/PSS fihn
shows an 80-fold higher ip value and a 1500-fold lower Rct value than an electrode coated
with a four-bilayer film made with salt (pH 6.3). Voltarnmetric and impedance
measurements do not indicate that solution pH changes the permeability of 20-bilayer
PAH/PSS films. However, small changes in ip are difficult to detect when values are
nearly as high as those of a bare gold electrode. The thin, flat PAH/PSS bilayers

prepared without salt offer a vastly more open film structure than films containing a large

number of loops and trains.

G. Comparing Permeabilities of PAH/PAA and PAH/PSS Multilayer Films

PAH/PAA films made with no supporting electrolyte are similar in permeability to
PAH/PSS fihns prepared without salt. Peak current values (Fe(CN)(,3' in pH 6.3-buffered
solutions) for electrodes coated with PAH/PAA films (nine-bilayers; ~15 nm) are 470 1:
110 tiA/cm2 compared to 710 :1: 7O jiAlcm2 for an electrode coated with a 20-bilayer
PAH/PSS film (thickness = 16 nm). CVs of a PAH/PAA-coated electrode show a nearly
reversible response (AEp~80mV) similar to that observed for PAH/PSS fihns made with
no supporting salt (Figure 2.10). The somewhat lower ip value for PAH/PAA films may
be due to the larger bilayer thickness12 and more interrningling of the polymer layers.

Interestingly, fabrication of PAH/PAA fihns in the presence of supporting salt does
not yield the same decrease in permeability seen with PAH/PSS films. The permeability

of PAH/PAA films increases with the formation of loops and trains in the polymer

49

00
O

 
 
 

A
O

 

Current mm

0.4 0.2 0.0
Potential (V vs Ag/Cl, 3M KCI)

Figure 2.10: CVs (5 mM Fe(CN)63', pH 6.3) of gold electrodes coated with a 20-bilayer
PAH/PSS film prepared with no supporting electrolyte (circles, thickness of 16 nm), a 5-
bilayer PAH/PAA film prepared with supporting electrolyte (dashes, thickness of 44 nm),
a 9-bilayer PAH/PAA prepared with no supporting electrolyte (diamonds, thickness of 15
nm), or a 4-bilayer PAH/PSS film prepared with supporting electrolyte (solid line with x
markers, thickness of 15 nm).

50

coating. This may be due to the fact that PAA does not contain bulky pendant groups.

An electrode coated with a 5-bilayer PAH/PAA fihn made with supporting salt (thickness
of 44 nm) has an average ip value of 690 :1: 80 j.rA/cm2 at pH 6.3. This is a 50% increase
from the ip value (pH 6.3) for a nine-bilayer PAH/PAA film made with no supporting
electrolyte (thickness of 15 nm). Impedance measurements of PAH/PAA films (made
with or without the addition of supporting salt) and PAH/PSS films (made without
supporting salt) also indicate high penneabilities as the impedance plots only contain the
Warburg element. Apparently, the addition of supporting salt to PAH/PAA films makes

for a more Open and loose structure as seen by the increase in ip values.

IV. Conclusions
The permeability Of polyelectrolyte films varies widely with the number of bilayers

in the film, ambient pH, and deposition conditions. Permeability of PAH/PSS films made
with supporting salt decreases nonlinearly with the number Of bilayers because the
structure of the initial layers is different from that of outer layers. Film permeability is
relatively independent of solution pH at low and moderate values but increases drastically
in basic solutions. In situ ellipsometric studies indicate that PAH/PSS films swell in
water by ~40% at low and moderate pH values. At high pH, the film swells further and
then begins to delaminate. This additional swelling accounts for higher penneabilities in
basic solution. Addition of supporting salt during film deposition dramatically alters the
structure of PAH/PSS films. The formation of loops and trains in PAH/PSS films due to

the presence of supporting electrolyte probably yields the large decrease in film

permeability.

51

The permeability of PAH/PAA films prepared in the presence of supporting salt is
much greater than that of PAH/PSS films. Apparently, the formation of loops and trains
in S-bilayer PAH/PAA films does not have the same effect as in PAH/PSS films. This
observation is likely due to differences in size and hydrophobicity of the pendant groups

of PAA and PSS. Chapter 3 examines the ion-separation properties of polyelectrolyte

films deposited on permeable substrates.

52

V. References

l) Decher, G. Science 1997, 27 7, 1232-1237.
2) Lowy, D. A.; F inklea, H. O. Electrochimica Acta 1997, 42, 1325-1335.

3) Montrel, M. M.; Sukhorukov, G. B.; Petrov, A. I.; Shabarchina, L. 1.; Sukhorukov, B.
I. Sensors and Actuators B 1997, 42, 225-231.

4) Yang, X.; Johnson, 8.; Shi, J .; Holesinger, T.; Swanson, B. Sensors and Actuators B
1997, 45, 87-92.

5) Caruso, F .; Niikura, K.; Furlong, D. N.; Okahata, Y. Langmuir 1997, 13, 3427-3433.

6) Mika, A. M.; Childs, R. F.; Dickson, J. M.; McCarry, B. E.; Gagnon, D. R. J Membr.
Sci. 1997, 81-92.

7) Nam, S. Y.; Lee, Y. M. J. Membr. Sci. 1997, 161-171.
8) Leviisalmi, J .-M.; McCarthy, T. J. Macromolecules 1997, 30, 1752-1757.
9) Ackem, F. v.; Krasemann, L.; Tieke, B. Thin Solid Films 1998, 327-329, 762-766.

10) Stroeve, P.; Vasquez, V.; Coelho, M. A. N.; Rabolt, J. F. Thin Solid Films 1996,
708-712.

11) Krasemann, L.; Tieke, B. Langmuir 2000, 16, 287-290.

12) Harris, J. J.; DeRose, P. M.; Bruening, M. L. J Am. Chem. Soc. 1999, 121, 1978-
1979.

13) Kotov, N. A.; Dékany,1.; Fendler, J. H. J Phys. Chem. 1995, 99, 13065-13069.

14) Fou, A. C.; Rubner, M. F. Macromolecules 1995, 28, 7115-7120.

15) Schlenoff, J. B.; Laurent, D.; Ly, H.; Stepp, J. Adv. Mater. 1998, 10, 347-349.

16) Wu, A.; Yoo, D.; Lee, J .-K.; Rubner, M. F. J Am. Chem. Soc. 1999, 121, 4883-4891.
17) Caruso, F.; Niikura, K.; Furlong, D. N.; Okahata, Y. Langmuir 1997, 13, 3422-3426.

18) Ferreira, M.; Rubner, M. F. Macromolecules 1995, 28, 7107-7114.

53

l9) Caruso, F .; Furlong, D. N.; Ariga, K.; Ichinose, 1.; Kunitake, T. Langmuir 1998, 14,
4559-4565.

20) Yoo, D.; Shiratori, S. S.; Rubner, M. F. Macromolecules 1998, 31, 4309-4318.

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

22) Decher, 6.; Hong, J.-D.; Schmitt, J. Thin Solid Films 1992, 831-835.
23) Decher, G.; Lvov, Y.; Schmitt, J. Thin Solid Films 1994, 244, 772-777.
24) Lvov, Y.; Decher, G.; MOhwald, H. Langmuir 1993, 9, 481-486.

25) Schmitt, J.; Grunewald, T.; Decher, G.; Pershan, P. S.; Kjaer, K.; Losche, M.
Macromolecules 1993, 26, 7058-7063.

26) Tronin, A.; Lvov, Y.; Nicolini, C. Colloid Polym. Sci. 1994, 2 72, 1317-1321.

27) LOsche, M.; Schmitt, J.; Decher, G.; Bouwman, W. G.; Kjaer, K. Macromolecules
1998, 31 , 8893-8906.

28) Hoogeveen, N. G.; Cohen Stuart, M. A.; Fleer, G. J. Langmuir 1996, 12, 3675-3681.
29) Schlenoff, J. 8.; Li, M. Ber. Bunsenges. Phys. Chem 1996, 100, 943-947.
30) Schlenoff, J. B.; Ly, H.; Li, M. J Am. Chem. Soc. 1998, 120, 7626-7634.
31) Xu, H.; Schlenoff, J. B. Langmuir 1994, 10, 241-245.

32) Shubin, V.; Linse, P. J Phys. Chem. 1995, 99, 1285-1291.

33) Lvov, Y.; Ariga, K.; Onda, M.; Ichinose, 1.; Kunitake, T. Colloid Surf A 1999, 146,
337-346.

34) Klitzing, R. v.; MOhwald, H. Thin Solid Films 1996, 352-356.
35) Klitzing, R. v.; MOhwald, H. Macromolecules 1996, 29, 6901-6906. .

36) Lvov, Y.; Decher, G.; Haas, H.; MOhwald, H.; Kalachev, A. Physica B 1994, 198,
89-91.

37) Lvov, Y.; Decher, G.; Sukhorukov, G. Macromolecules 1993, 26, 5396-5399.

54

38) Lin-Vien, D.; Colthup, N. B.; F ateley, W. G.; Grasselli, J. G. The Handbook of
Infrared and Raman Characteristic Frequencies of Organic Molecules; Academic Press:

New York, 1991.

39) Menon, V. P.; Martin, C. R. Anal. Chem. 1995, 67, 1920-1928.
40) Chailapakul, 0.; Crooks, R. M. Langmuir 1995, 11, 1329-1340.
41) Savéant, J.-M. J Electroanal. Chem. 1991, 302, 91-101.

42) Sabatani, E.; Rubinstein, I. J Phys. Chem. 1987, 91, 6663-6669.

43) Bard, A. J .; Faulkner, L. R. Electrochemical Methods: Fundamentals and
Applications; John Wiley & Sons: New York, 1980.

44) Zhao, M.; Bruening, M. L.; Zhou, Y. F.; Bergbreiter, D. E.; Crooks, R. M. Isr. J
Chem. 1997, 3 7, 277-286.

45) Sabatani, E.; Cohen-Boulakia, J .; Bruening, M. L.; Rubinstein, I. Langmuir 1993, 9.

46) Steeg, H. G. M. v. d.; Stuart, M. A. C.; Keizer, A. d.; Bijsterbosch, B. H. Langmuir
1992, 8, 2538-2546.

47) Dobrynin, A. V.; Rubinstein, M. Macromolecules 1999, 32, 915-922.

55

 

d:

CHAPTER 3

Synthesis of Passivating, Nylon-Like Coatings through Cross-linking of Ultrathin
Polyelectrolyte Films
1. Introduction

Although deposition of PAH/PSS films provides a convenient means for
passivating gold electrodes, these films are still too permeable to be used for practical
substrate protection. One reason for the permeability of PAH/PSS films is the high
degree of swelling observed in these films as seen in Chapter 2. Another drawback Of the
PAH/PSS system is its instability in alkaline solutions, which results from the
deprotonation of amine groups of PAH. One possible means for improving the stability
and passivation characteristics of polyelectrolyte films is to replace the ionic interactions
holding the film together with stronger covalent linkages. .

This chapter describes the heat-induced formation of covalent cross-links from the
formation of amide bonds between the carboxylate and ammonium groups of PAH/PAA
films.l Unlike many layered polyelectrolyte fihns, these cross-linked, nylon-like films
are stable over a wide pH range and highly impermeable. This stabilization strategy
resembles that used by other groups when forming layered polymer films with covalent
linkages.2'4 However, post-deposition cross-linking of polyelectrolytes preserves the
advantages inherent in the APD process (e.g., no need for organic solvents and a

convenient deposition procedure).

11. Experimental

56

To synthesize PAH/PAA films on gold, we first form a monolayer of
mercaptopropionic acid (MPA) by immersing a gold-coated silicon wafer (Si(100) wafers
coated with 20 nm Ti and 200 nm of Au by electron beam evaporation) in 1.5 mM MPA
in ethanol for 30 minutes. (Use of MPA, rather than a longer molecule such as
mercaptoundecanoic acid, avoids possible blocking of the electrode by the monolayer.)
Upon deprotonation, the monolayer provides a negatively charged surface. Immersion of
the MPA-modified gold slide in a PAH solution (0.02 M with respect to the repeating
unit, pH adjusted to 5.1 with NaOH) for 5 minutes yields a layer of positively charged
polymer on the substrate. Subsequent immersion in a PAA solution (0.01 M with respect
to the repeating unit, pH adjusted to between 5.6 and 5.7 with NaOH) for 5 minutes
results in deposition of the first negatively charged layer. Supporting electrolyte was not
used in the deposition of these ultrathin PAH/PAA films. Repetition of this procedure
allows synthesis Of films with a large number of polyelectrolyte layers. We rinsed
samples with Milli-Q water (18 MQ-cm) for one minute after the deposition of each
layer, and we dried with N2 only after deposition of the desired number of layers.

Procedures for the electrochemical and ellipsometric characterizations of cross-
linked PAH/PAA were the same as described in Chapter 1 for PAH/PSS films. The peak
current values from the cyclic voltammograms was determined as the current value at the
standard reduction potential of Fe(CN)63' (~260 mV). Baselines were sometimes difficult
to draw for CV’s with films heated at 215 °C due to the low faradaic current. XPS data
was collected on a PHI 540 Small Spot X—ray Photoelectron Spectrometer by Dr. Per
Askeland. The N15 region of a 9-bilayer PAH/PAA film before and after heating for 2

hours at 215 °C was used to determine presence of amide cross-links. Binding energy is

57

referenced to the Cl, peak for hydrocarbons (284.6 eV). XPS (survey spectra) was unable

to detect the inclusion of either Na+ or Cl' ions into the polyelectrolyte film.

111. Results and Discussion

The first step in the formation of cross-linked layered polyelectrolytes is the
desposition of PAH/PAA fihns. Optical ellipsometry indicates nonlinear bilayer growth
for PAH/PAA films deposited in the absence of supporting electrolyte. The thickness
increase for the first 1-5 bilayers ranges from 4 to 17 nm/bilayer, while for bilayers 6-9
the increase is 5.5 1 1.1 nm/bilayer. The absence of supporting electrolyte contributes to
the nonlinear thickness increases and the low individual layer thicknesses. The change in
bilayer thickness is consistent with previous studies},6 which indicate that a precursor
film is needed prior to regular film deposition.

After deposition of PAH/PAA films, cross-linking can be effected simply by
heating the sample under N; as shown in Figure 3.1. F TIR-ERS spectra of 9-bilayer
PAH/PAA fihns verify that cross-linking between ammonium groups of PAH and
carboxylate groups of PAA occurs by formation of an amide bond (Figure 3.2). Before
heating (Figure 3.2 A), the most notable peaks are due to the -COOH carbonyl and
-COO' asymmetric and symmetric stretches (1709, 1572, and 1399 cm], respectively).7v8
Both carboxylate groups and a small fraction of protonated acid are present because the
deposition was performed at pH 5.6. After heating the slide at 130 °C, the carboxylate
peaks decrease, and amide peaks appear at ~1540 and 1670 cm’1 as seen in Figure 3.2 B,

Some protonated acid remains, but its absorbance peak shifts to higher wavenumbers

58

12518218 25 a 13211 8 203 828% 6: 8.158125 Ewe 183 a 18.1
821:0 passage 338 a; .8120 51 use <5 8238 1511-188 .8 1101981212 01%me "mm 8:”...—

 

59

 

 
   
     
  
  

Amide I

(D
0
C
(0
2
8 / B
.0
<1:
Amide ll
\

   

   

 

 

 

1600 1 500 1400

Wavenumbers (cm'1)

1 800 1700

Figure 3.2: F TIR-ERS spectra of 9-bilayer PAH/PAA films before heating (A), after
heating at 130 C (B), or after heating at 215 C (C).

60

indicating a change in hydrogen bonding or chemical environment. Similar spectral
changes occur when a PAH/PAA fihn is heated at 215 °C (spectrum C), but the amide to
acid peak-height ratio is larger suggesting more extensive cross-linking. Heating
individual solvent-cast PAH and PAA films at 130 °C does not produce the amide peaks
seen in Figure 3.2.

XPS also demonstrates the formation of amide bonds (N 1, spectrum) in heated
PAH/PAA fihns. Figure 3.3A exhibits two peaks that we attribute to amino (398.8 eV)
and ammonium (400.8 eV) nitrogens in PAH/PAA films9. After heating (215 °C for 2
hours), the XPS spectrum (Figure 3.38) shows a single peak at 399.4 eV, which we
assign to an amide nitrogen.9 The peak in Figure 3.3B probably also contains a small
contribution from residual ammonium and amino nitrogens. XPS can also provide
insight into whether counterions are included in the film structure. Consistent with
previous studies,‘Oa” XPS (survey spectra) was unable to detect the inclusion Of either
Na+ or Cl' ions in the polyelectrolyte film.

Cyclic voltammetry demonstrates the dramatic decrease in film permeability that
occurs upon cross-linking. Figure 3.4 shows cyclic voltammograms at a 9_-bilayer
PAH/PAA film-coated gold electrode before and after heating at 130 or 215 °C for 2
hours. Cross-linking at 130 °C reduces the peak current to about 1% of its initial value
(Figure 3.4B). Heating at 215 °C has an even more dramatic effect on permeability. The
voltammogram for this system shows an additional ~85% decrease in peak current from
that of a fihn heated at 130 °C (Figure 3.4C). The decreased permeability shown by

PAH/PAA films heated at 215 °C is probably due to the higher degree of cross-linking

61

 

10F.fi f. r . .1 1. . .11.

 

Arbitrary Units
8
l
l

4 :

2 .
i.

 

 

AAAAAAAAAA

o 403‘ ‘ 401 399 397
Binding Energy (eV)

 

Figure 3.3: XPS spectra of the N1, region of a 9-bilayer PAH/PAA fihn before (A) and
after heating for 2 hours at 215 °C (B). Binding energy is referenced to the C1, peak for
hydrocarbons (284.6 eV).

62

 

 

 

 

 

 

Current (11A)

 

Potential (V vs Ag/AgCl, 3M KCI)

Figure 3.4: Cyclic voltammograms (0.005 M Fe(CN)63' in 0.025 NazHPOa, 0.025
NaH2P04, l M Na2804, pH 6.3) at gold electrodes coated with 9 bilayers of PAH/PAA
before (3A, dashed line), after heating a fihn at 130 °C for 2 hours (3A, solid line and 3B,
dashed line), or after heating a film at 215 °C (38 and C, solid lines). Scan rate of 0.1 V/s

and electrode area of 0.1 cm2. The electrode was immersed in the solution for 2 minutes
prior to data acquisition.

63

suggested by FTIR-ERS, but there could also be some decrease due to annealing of the
polyelectrolyte fihn or the underlying gold surface.

Impedance spectroscopy allows quantitative comparison of the permeability of
fihns before and after cross-linking. Figure 3.5 shows typical plots of impedance for
electrodes coated with 9-bilayers of PAH/PAA before and after heating at 130 °C or 215
°C. In the case of the unheated film (Figure 3.5A), we see a small semicircle at high
frequencies and the beginnings of a straight line at low frequencies that is characteristic
of semi-infiinite linear diffusion.” The diameter of the semicircle is approximately equal
to the charge transfer resistance, R“, which is indicative of the accessibility of the
electrode. The impedance plot for a film after cross-linking at 215 °C (Figure 3.5C)
contains only 1/8 of a circle showing that the reaction is completely kinetically
controlled.” Using the modified Randle’s equivalent circuit13 shown in Figure 3.6, we
modeled the electrode-film interface and calculated values for R“, film resistance (Rf),
film capacitance (Cr), double-layer capacitance (Cd) and the diffusion impedance (Zw).
The equivalent circuit provides an excellent fit to the experimental data. Ra increases by
a factor of 200 upon cross-linking at 130 °C (8X103 Q-cmz), and Rf values are 6.5><103 9-
cm2. Heating of films at 215 °C has an even more dramatic effect on the electrode
system. Rf and RC, values for these systems (8><104 Q-cm2 and 1X 105 Q—cmz,
respectively) are 1-2 orders of magnitude higher than for films heated at 130 °C. These
results are in good agreement with cyclic voltammetry.

Cross-linking of PAH/PAA films dramatically increases their stability at high pH
values. Prior to cross-linking, immersion of a PAH/PAA fihrl in a pH 10-buffered

solution (0.025 M Na2C03/0.025 M NaHCO3 in 1 M NaZSOa) for 20 minutes results in

64

 

 

 

 

 

 

 

 

'Zlm (Q'szl

 

 

 

ZR (Q—cmz)

Figure 3.5: Impedance plots (0.005 M Fe(CN)e3"4' in 0.025 NazHPO4, 0.025 NaHzPO4, l
M Na2S04, pH 6. 3) for gold electrodes coated with 9- bilayer PAH/PAA films before
heating (A), after heating at 130 °C (B), or after heating at 215 °C (C). Scale bars apply
to both x- and y-axes. Frequency range 1-105 Hz. Sinusoidal voltage: 5 mV about E°'.
The electrode was immersed in solution for 5 nrinutes prior to data acquisition. Scale bar
is representative for both x- and y-axes.

65

 

 

 

 

Figure 3.6: Modified Randles’ equivalent circuit used to model impedance data from
gold electrodes modified with cross-linked PAH/PAA fihns.

66

an ~30% decrease in ellipsometric thickness. We observe similar instability at high pH
When working with PAH/PSS films. At pH 10, a large fraction of the amine groups in
these films are deprotonated and many of the electrostatic bonds linking the layers are no
longer present. In contrast, cross-linked PAH/PAA fihns show no decrease in fihn
thickness after immersion in pH 10 buffer. As a control, we immersed PAH/PSS films,
which were heated at 130 °C for 2 hours in pH 10 buffer for 20 minutes. These films still
show an ~30% decrease in ellipsometric thickness.

In conclusion, heat-induced cross-linking between PAH and PAA occurs via
formation of amide bonds and stabilizes these films over a wide pH range. Cyclic
voltammetry and impedance spectroscopy show that film permeability decreases
dramatically after cross-linking and depends upon heating conditions. we are
investigating using these nylon-like, cross-linked films as ultrathin membranes or

corrosion resistant coatings. '4

67

IV. References

1) March, J. Advanced Organic Chemistry; 4th ed.; John Wiley: New York, 1992.
2) Beyer, D.; Bohanon, T. M.; Knoll, W.; Ringsdorf, H. Langmuir 1996, 12, 2514-2518.

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

4) Lee, B.-J.; Kunitake, T. Langmuir 1994, 10, 557-562.

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

6) Caruso, F .; Niikura, K.; Furlong, D. N.; Okahata, Y. Langmuir 1997, 13, 3422-3426.
7) Yoo, D.; Shiratori, S. S.; Rubner, M. F. Macromolecules 1998, 31 , 4309-4318.

8) Socrates, G. Infrared Characteristic Group Frequencies; John Wiley & Sons: New
York, 1980.

9) Wagner, C. D.; Riggs, W. M.; Davis, L. E.; Moulder, J. F .; Muilenberg, G. E.
Handbook of X-Ray Photoelectron Spectroscopy; Wagner, C. D.; Riggs, W. M.; Davis, L.
E.; Moulder, J. F .; Muilenberg, G. E., Ed; Perkin Elmer: Eden Prairie, MN, 1979, pp 40-
41.

10) Schlenoff, J. B.; Ly, H.; Li, M. J Am. Chem. Soc. 1998, 120, 7626-7634.

11) Sukhorukov, G. B.; Schmitt, J .; Decher, G. Ber. Bunsenges. Phys. Chem. 1996, 100,
948-953.

12) Bard, A. J .; Faulkner, L. R. Electrochemical Methods: Fundamentals and
Applications; John Wiley & Sons: New York, 1980.

13) Rubinstein, 1.; Sabatani, E.; Rishpon, J. J Electrochem. Soc. 1987, 134, 3078.

14) Dai, J .; Jensen, A. W.; Mohanty, D. K.; Emdt, J .; Bruening, M. L. Langmuir 2001,
I 7, 931-937.

68

CHAPTER 4

Layered Polyelectrolyte Films as Selective, Ultrathin Membranes for Anion
Separations

I. Introduction

Ultrathin membranes are attractive for separation processes because they can
simultaneously allow both high flux and high perlnselectiv’ty.l'3 Typically, such
membranes consist of polymer “skins” (<50 nm thick) on a highly permeable support.2
The support provides mechanical stability, but its high permeability presents minimal
mass-transfer resistance. Martin and coworkers used ultrathin polymer membranes
deposited on porous alumina substrates in the development of anion-permselective
sensors as well as gas-separation membranes."5 Regen showed that even Langmuir-
Blodgett films offer permselectivity in some gas separations.“9 Deposition of highly
selective, stable, ultrathin membranes remains a challenge, however, because of the
difficulty of firlly covering a porous substrate without filling the pores.

Synthesis of high-flux composite membranes requires methods for deposition of
ultrathin, defect-free films on highly permeable supports. Layer-by-layer deposition of
polyelectrolytes on porous alumina (0.02 pm pore diameter) provides a convenient and
versatile way to produce such membranes. Electron microscopy shows that S-bilayers
(~25 nm) of poly(allylamine hydrochloride) (PAH)/poly(styrene sulfonate) (PSS) are
sufficient to cover porous alumina and that underlying pores are not clogged during the

deposition process. The selectivity of anion transport through these membranes increases

69

 

with the number of bilayers until the substrate is fully covered. 5-bilayer PAH/PSS
membranes have Cl'/SO42' and Cl'/Fe(CN)63' selectivity values of 5 and 1000,
respectively. PAH/poly(acrylic acid) (PAA) membranes show selectivity values similar
to those of PAH/PSS membranes but with a 3-fold decrease in anion flux. Selectivity in
both of these systems likely results from Donnan exclusion. By combining the properties
of both PAH/PAA and PAH/PSS membranes it is possible to achieve increased Cl'/SO42‘

values while maintaining high flux.

[1. Experimental
A. Materials

Poly(allylamine hydrochloride) O’AH) (Aldrich, Mw=70,000), poly(styrene
sulfonate) (PSS) (Aldrich, Mw=70,000), poly(acrylic acid) (PAA) (Mw=90,000, 25 wt%
solution, Alfa Aesar), MnClz (Mallinckrodt), NaBr (Aldrich), NaCl (Mallinckrodt),
KzNi(CN)4 (Aldrich), KCl (Baker), KNO3 (CCI), K2804 (CCI), K3Fe(CN)6
(Mallinckrodt), and salicylic acid (Spectrum) were used as received. Prior to film
deposition, the porous alumina support (Whatman® AnodiscTM 0.02 pm membrane
filters) was UV/03 cleaned for 15 minutes (Boekel UV-CleanTM model 135500).
Polyelectrolyte deposition then began by immersing the alumina support in a solution of
PAH (0.02 M with respect to the monomer unit, pH 2.3 for PAH/PSS films or pH 4.5 for
PAH/PAA films) for S-minutes. PAH solutions also contained either 0.5 M NaBr
(PAH/PSS deposition) or 0.5 M NaCl (PAH/PAA deposition). Supporting electrolytes
were chosen based on literature preparations of polyelectrolyte films.31 37,48 The alumina

membrane has two distinct sides: the filtrate side has a skin layer of 0.02 mm diameter

70

pores and the permeate side has 0.2 pm diameter pores. We used a holder to limit
polyelectrolyte deposition to the filtrate side. A l-minute rinse with water (Milli-Q, l8
MQ-cm) followed PAH deposition and all other polyelectrolyte deposition steps. To
deposit the polyanion, the PAH-coated alumina support was immersed in either a solution
of PSS or PAA. After the deposition of the desired number of polyelectrolyte layers, the
membrane was dried with N2. Polyelectroltye deposition began with the polycation but
the alumina substrate is also positively charged. Ellipsometiy experiments on Al coated
Si substrates shows that a layer of polyelectrolyte is deposited regardless of charge.

A Hitachi S—47OO field-emission scanning electron microscope (F ESEM) operating with
accelerating voltages of between 6 and 8 kV was used to image polyelectrolyte films
deposited on alumina membranes. The use of low voltages with the FESEM allows
imaging of polymer films while reducing charging artifacts and specimen damagefios”

Samples were coated with 5 nm of Au for FESEM imaging.

B. Transport Studies

The dialysis apparatus consists of two glass cells (volumes of 100 mL) connected
by a 2.5 cm-long neck in which the membrane separates the feed and permeate sides
(Figure 4.1). The exposed area of the membrane was 2.0 cm2. The permeate cell
contained Milli-Q water and the feed cell contained the appropriate 0.1 F salt solution
(KCl, KN O3, K2SO4, KzNi(CN)4, or K3Fe(CN)5). The measured pH values of the
unbuffered 0.1 F salt solutions were 4.8 (KCl), 5.3 (KNO3), 5.5 (K2804), 6.4
(KzNi(CN)4), and 6.2 (K3FC(CN)6). These pH values are well below the pK value for

ammonium groups and well above the pK value of styrene sulfonic acid. Thus, pH

71

 

I 152 its

DUE!
DUE]
DUE]

 

m]

 

 

 

 

 

 

 

 

 

allo'

 

 

®

 

 

 

Figure 4.1: Dialysis cell used to measure ion-permeabilty of polyelectrolyte membranes:
conductivity meter (A), electric stirrer (B), conductivity electrode (C), receiving phase
(D), source phase (B), membrane (F), and magnetic stir bar (G).

72

should have little effect on transport in this system. However, in the case of PAH/PAA
membranes, there is a possibility that pH may affect transport. As the selectivities in the
PAH/PAA system were similar to those of PAH/PSS, we assume that the pH differences
in these unbuffered solutions have little effect on transport. Salicylic acid (SA) was also
used as a probe anion. Solutions of salicylic acid were adjusted to a pH of 6 with KOH to
deprotonate the acid. The concentration of salt in the permeate cell was monitored at 10
minuteintervals for 90 minutes using conductivity measurements (Orion Model 115
conductivity meter). To normalize receiving phase conductivity values, we divide by the
conductivity of the source phase. Both feed and permeate solutions were stirred
vigorously to minimize concentration polarization at the membrane surface. Permeability
measurements made with different stirring speeds showed no significant change in flux.
The order in which the anions were investigated was Cl', NO3', 8042‘, Ni(CN)42', and

F e(CN)(,3 '. After determining the flux of each anion, the membrane was immersed in
Milli-Q water for 15 minutes.

The integrity of the film during the permeability experiments was tested with a
second Cl' run performed after the experiment with F e(CN)(,3 '. The C1' flux changed less
than 10% for PAH/PSS membranes, but PAH/PAA membranes showed an ~50%
decrease. Three or more different membranes were tested for each type of

polyelectrolyte film.

III. Results and Discussion

To synthesize ultrathin polyelectrolyte membranes, the layered film must cover

the substrate without filling underlying pores. Although a few reports describe the layer-

73

Ilvvrlllllt
687nm

O.
1‘1

7 a v 16 5mm “5 on SE “)1213/13‘3910 43'

 

Figure 4.2: FESEM images of the filtrate side of a porous alumina substrate (A), and
alumina substrates coated with 2 (B), 4 (C), or 5 PAH/PSS bilayers (D). The circles in
image C are likely defects in the film structure. Samples were coated with ~5 nm Au for
1maging.

74

by-layer deposition of polyelectrolyte films on highly porous substrates)?"20 they do not
emphasize surface coverage or the likely possibility of polyelectrolyte deposition in
substrate pores. F ESEM images (Figure 4.2) of PAH/PSS films on porous alumina show
that coverage of the filtrate side of the membrane gradually increases with the deposition
of each subsequent bilayer. Complete pore coverage occurs after the addition of about 5
PAH/PSS bilayers. Micrographs of the alumina permeate side (Figure 4.3) show that this
surface remains polyelectrolyte free even after deposition of lO-bilayers on the filtrate
side. This demonstrates that deposition of the polyelectrolyte does not occur throughout
the membrane. FESEM micrographs of 5-bilayer PAH/PAA membranes also show a
defect free film, but complete pore coverage occurs with fewer depositions as the bilayer
thickness of PAH/PAA films is greater than that of PAH/PSS films.

Cross—sectional F ESEM images show that the pores of the membrane are
unblocked after film deposition. Figure 4.4 shows a lO-bilayer PAH/PSS film deposited
on the membrane surface (filtrate side), with little polyelectrolyte in the alumina pores.
There appears to be some polyelectrolyte in the upper part of the pores as the appearance
of small cracks is likely due to a small amount of polyelectrolyte deposition. However,
the majority of film deposition occurs at the support surface and the pores remain open,
even if they contain a bilayer or two of polyelectrolyte. Figure 4.5 shows the interior
pore structure of an alumina support that was coated with a lO-bilayer PAH/PSS film.
The absence of polyelectrolyte in this image shows that polyelectrolytes are unable to
penetrate much past the skin layer on the filtrate side of the support. The cross-sectional

micrograph of the lO-bilayer PAH/PSS membrane (Figure 4.4) indicates a film thickness

75

 

 

Figure 4.3: F ESEM images of the permeate side of a bare alumina substrate
before film deposition (A), after restricting deposition to the filtrate side (B), or
following complete alumina immersion in polyelectrolyte solutions (C).

76

7 (”4": 12 7mm 17‘? l)‘ 9:“ 1| 1.’1(U?WJO 18 47

 

Figure 4.4: Cross-sectional FESEM images of the filtrate side of a bare alumina -
substrate (A) and an alumina substrate coated with 5-bilayers of PAH/PAA (B). Film

thickness 2 40 nm.

77

 

Figure 4.5: Cross-sectional F ESEM image showing the lack of polyelectrolyte
deposition in the inner pores of an alumina substrate where the filtrate side is coated with
a lO-bilayer PAH/PSS film.

78

of ~35 nm, which is comparable to ellipsometric thicknesses for lO—bilayer films
deposited on solid substrates.21 Images in Figures 4.2, 4.3, 4.4, and 4.5 show collectively
that deposition of layered polyelectrolyte films is a simple means of creating ultrathin
membranes on porous supports.

Because pore coverage depends on the number of polyelectrolyte bilayers,
selectivity and flux through these membranes also vary with the number of PAH/PSS
bilayers. Chloride flux shows little change with the number of PAH/PSS bilayers, while
5042‘ transport decreases 5-fold with the addition of 10 bilayers (Table 4.1). Under the
same conditions, F e(CN)(,3' flux decreases 330-fold. Figure 4.6, which shows receiving
phase conductivity as a function of time, illustrates the effect of each additional PAH/PSS
bilayer on S042' flux. The selectivity ratio for C1" to SO42} which is determined by
dividing anion flux values, reaches its maximmn value of 5 after the deposition of 5-
bilayers.22 This is consistent with the FESEM images that show complete surface
coverage with few defects after the deposition of 5-bilayers (Figure 4.2). Prior to
complete pore coverage, most sulfate transport likely occurs through membrane defects.
Upon full coverage, however, sulfate must enter and diffuse through the membrane
causing a decrease in flux. Resistance to mass transfer of Cl' through the polyelectrolyte
film must be minimal because flux is hardly affected by full coverage of the support.

Figure 4.7 compares the fluxes of various anions through a 5-bilayer PAH/PSS
membrane. The smallest measured fluxes for PAH/PSS membranes occur with
F e(CN)¢53 , because its high charge produces the largest electrostatic repulsion between
anion and membrane. The linearity of the conductivity values with time, which occurs

with each membrane tested, demonstrates steady-state anion transport where receiving-

79

Table 4.1: Anion flux values (moles cm'2 s") through bare alumina and alumina
substrates coated with between 1-5.5- or lO-bilayers of PAH/PSS.

 

 

 

 

 

 

 

 

 

 

Number
of cr Flux NO3’ Flux so} Flux Ni(CN)42' Flux Fe(CN)63' Flux
Bilayers
alumina 6.7x 10'8 3: 3% 6.7><10'8 a 6% 5.1><10’8 i 4% 4.8x10'8 s 1% 5.0><10‘8 s 7%
1 6.6x 10'8 a 4% 4.8x 10'8 a 6% 4.4x10'8d: 10%
2 5.7x10‘8 a 7% 4.5x10'“ :t 3% 4.1><10‘8 a 7%
3 5.9x10'8 :2 5% 2.9x10'8 at 8% 1.3x10'84 40%
4 5.1x10‘8a13% l.5><10'8i 14% 2.7><10'°i 10%
5 6.5><1O'8:t 25% 6.1><10'8:1: 20% 9.9x10'9 a 8% 4.7x10'9s 10% ' 2.1><10"°i 30%
5.5 4.6><10'8 a 6% 5.0x10'8 s 4% 2.2x10'8a 20% 8.6XI0'9i 30% 1.9x10"°a 20%
10 4.5x10'8n 1% 5.1x10'8n4% 8.9X10'9zl: 1% 3.6XI0'9t3% 1.4x10"°n 20%

 

 

 

 

 

 

80

 

0.06 t :—

 

0.05 r

2

0.03 i

 

0.02 -

Normalized Conductivity

0.01 '

 

 

 

Time (min)

Figure 4.6: Plot of normalized receiving-phase conductivity (K2804) as a function of
time when the source phase was separated from the receiving phase by a bare alumina
substrate (open squares), or a 1- (diamonds), 2- (squares), 3- (triangles), 4- (inverted
triangles), 5- (circles), or lO-bilayer (X’s) PAH/PSS film.

81

 

Normalized Conductivity

 

 

 

Time (min)

Figure 4.7: Plot of normalized receiving—phase conductivity as a function of time when
the source phase was separated from the receiving phase by a 5-bilayer PAH/PSS film.
Different symbols represent different salts: KC] (triangles), KNO3 (squares), K2804
(diamonds), KzNi(CN)4 (circles), and K3Fe(CN)6 (inverted triangles).

82

phase concentration is negligible compared to that in the source phase. Five-bilayer
PAH/PSS membranes show selectivity ratios of 300, 50, and 20 for Cl‘lFe(CN)63' , 8042'
/Fe(CN)63', and Ni(CN)42'/Fe(CN)63', respectively (Table 4.1 and Figure 4.7). The ability
of PAH/PSS membranes to nearly eliminate F e(CN).g,3 ' flux suggests that these
membranes could be useful for the dialytic purification of large, highly charged
molecules.23'7-5

Once complete pore coverage occurs, the presence of additional polyelectrolyte
bilayers decreases anion flux, but has little impact on selectivity. The Cl'/SOaz', Cl'
/Fe(CN)63', Cl'/Ni(CN)42', and SO42'/Ni(CN)42‘ selectivity are nearly the same for 5 and
lO-bilayer PAH/PSS membranes. On going from 5 to 10 PAH/PSS bilayers, anion flux
decreases only between 10 and 30% (Table 4.1). Kraseman and Tieke report that
transport selectivity does depend on the number of PAH/PSS bilayers.26 However, the
films in that study were deposited under different conditions and on different substrates.

The minimal decrease in the flux of highly charged anions on going from 5— to
lO-bilayer membranes suggests that the major selectivity factor is Donnan exclusion near
the surface.26 Hindered diffusion would yield a significant decrease in anion flux with
increasing membrane thickness, while Donnan exclusion due to uncompensated charge at
the membrane surface should be relatively constant once full coverage is achieved.?-7~28
The large decreases in 8042‘, Ni(CN)42', and F e(CN)(,3 ' fluxes upon addition of the first 5
PAH/PSS bilayers likely result from increased Donnan exclusion as surface coverage
increases. Also consistent with a Donnan exclusion model is the minimal reduction in Cl‘
and N03" flux (Table 4.1) due to the much smaller repulsive forces generated by the

monovalent anions.27 One advantage of an electrostatic exclusion mechanism in

83

PAH/PSS membranes is that it provides monovalent/divalent and monovalent/trivalent
selectivity with little inhibition of the monovalent anion flux (Table 4.1).

The majority of uncompensated charge in layered polyelectrolyte films lies near
the film/solution interface”,30 so one should expect a change in anion flux if the outer
layer is a polycation rather than a polyanion. We tested this possibility by depositing an
additional PAH layer on a 5-bilayer film. This yields a membrane with a positive surface
charge, while providing a minimal change in membrane thickness. The additional PAH
layer has little effect on C1' and N03' transport (Table 4.1) as the flux of these anions is
already close to that of bare alumina. However, there is a 2-fold increase in 8042' and
Ni(CN)42' flux when the outermost charge is positive (Table 4.1). This suggests that the
outer layer charge does have a major effect on flux. In contrast, however, there is little
change in F e(CN)(,3 ' flux when the outer layer charge is positive. Apparently, large
repulsive forces between internal regions of the membrane and trivalent F e(CN)63 ' are
dominant for decreasing Fe(CN)63' flux. This is somewhat surprising as Schlenoff
showed that strongly dissociated polyelectrolytes provide complete intrinsic charge
compensation.29 However, charge compensation appears to depend on deposition
conditions and the polyelectrolyte used.“ For PAH/PSS films Lowack and Helm suggest
significant extrinsic charge compensation, which could result in Donnan exclusion within
the polyelectrolyte membrane.28

Flux of the potassium salt of salicylic acid through PAH/PSS films supports the
hypothesis that Donnan exclusion (rather than hindered diffusion) is the dominant factor
governing transport selectivity. We chose to work with salicylate because it is a large

monovalent anion with a diffusion coefficient 15% smaller than that of Fe(CN)(,3'.32 The

84

large size of salicylate is manifested by the fact that its flux through bare alumina (3.2 x
10'8 moles cm'2 s") is the lowest of any anion tested. The presence of 5 PAH/PSS
bilayers on alumina reduces salicylate flux two-fold while reducing Cl' flux by <30°/o.
This suggests some size selectivity between monovalent anions. However, most
selectivity among anions occurs as a result of Donnan exclusion as shown by the fact that
5 PAH/PSS bilayers reduce F e(CN)63' flux 220-fold and 8042‘ flux 5-fold. This occurs in
spite of the fact that salicylate, SOaz', and F e(CN)63 ' have similar diffusion
coefficients.32,33 These results show that Donnan exclusion is the major factor affecting
anion transport, but there may be some effect of size for larger anions.

Transport selectivity is similar in PAH/PAA and PAH/PSS membranes, although
fluxes are considerably lower through PAH/PAA membranes. F ive-bilayer PAH/PAA
membranes show between a 2- and 10—fold reduction in anion flux compared to 10-
bilayer PAH/PSS membranes (Table 4.1), which have similar thicknesses as seen with
FESEM. The cwsofi and Cl'/Fe(CN)6-3' selectivity values for PAH/PAA films, 4 and
230, respectively, are similar to 10-bilayer PAH/PSS selectivity values. The decreases in
anion flux are likely due to a tighter structure in PAH/PAA membranes than in PAH/PSS
membranes. If the effect was due to higher charge densities, we would expect to see
higher Cl'/SO.12' selectivities but this is not the case.

Decreased flux through PAH/PAA membranes is in contrast to our previous work
with PAH/PAA films deposited on gold electrodes (Chapter 1). In those studies, cyclic
voltammetry of Fe(CN)63' was hardly affected by the presence of 5-bilayers of
PAH/PAA, suggesting rapid Fe(CN)63' flux through these films. We speculate that in the

present case, PAH/PAA swelling is constricted at the surface of the alumina substrate.

85

Swelling of films on solid gold electrodes is uninhibited and may result in an open and
highly permeable film. Fe(CN)63’ absorption may also play a role in the contrasting
results. In the electrochemistry, absorption would result in higher peak currents
indicating a higher permeability due to the higher concentration at the electrode surface.
But, in the membrane studies ion absorption would have the effect of increasing the
effective charge of the membrane therefore increasing the repulsive effects and
decreasing the observed permeability.

In an attempt to utilize properties of both PAH/PSS and PAH/PAA membranes,
composite membranes consisting of a thin PAH/PAA film deposited on a PAH/PSS
precursor film show increased Cl'lSOaz' selectivity.34 The PAH/PSS film allows for high
flux while the PAH/PAA fihn dramatically decreases 8042' transport. Composite
membranes consisting of 2.5 bilayers of PAA/PAH deposited on S-bilayers of PSS/PAH
show Cl'/SOaz' selectivity of 150 and cross-linking of this membrane at 115 °C (as
described in Chapter 3) increases selectivity 4-fold. F ESEM images show changes in
surface morphology when PAA/PAH films are deposited on the precursor PSS/PAH film
(Figure 4.8). As observed in previous experiments the outer layer charge of the film
dramatically affects membrane transport. In one case, Cl'/SOaz' selectivity decreases
from 360 to 2 upon deposition of one layer of PAH.34 Construction of composite
membranes provides for additional means to easily tailor the transport of layered

polyelectrolyte membranes.

86

'1"
" . . 667 nm
.. dew16,6mmx45.0kSE(U)'5726/9903;17 ' w

 

8.0kV152mm x450k sewn/11120001012
Figure 4.8: FESEM images of the top of an alumina substrate coated with either a 4.5-

bilayer PAA/PAH film heated at 115 C (A) or a 5-bilayer PSS/PAH film coated with 1.5-
bllayers of PAA/PAH (B). Samples were coated with 5 nm Au for imaging.

87

 

IV. Conclusions

Ultrathin polyelectrolyte films on porous supports show selectivities typical of
anion exchange membranes. Using 5-bilayer PAH/PSS membranes on CelgardTM
supports Krasemann and Tieke report a C1'/so.2' selectivity value of 17, which is
somewhat higher than our value of 7.26 This higher value may be a result of a different
film structure due to a difference in preparation conditions or substrate topography. Our
Cl'/SOaz' selectivity value of 7 is two-fold higher than a previous dialysis study using
interpolymer type carboxylic ion-exchange membranes.35 Electrodialysis studies with
copolymer membranes show selectivity values between one and 100, where higher values
usually result from cross-linked surfaces.36“"O Composite membranes show higher
selectivity values without the need for cross-linking but upon cross-linking selectivity can
be greatly enhanced. Ultrathin, polyelectrolyte membranes show a wide range of
selectivity values depending on fihn composition and charge. These properties should
provide easy means for the tailoring of specific properties in anion separation

membranes.

88

V. References

1) Liu, C.; Espenscheid, M. W.; Chen, W.-J.; Martin, C. R. J. Am. Chem. Soc. 1990, 112.
2) Liu, C.; Martin, C. R. Nature 1991, 352, 50—52.

3) Liang, W.; Martin, C. R. Chem. Mater. 1991, 3, 390-391.

4) Ugo, P.; Moretto, L. M.; Mazzocchin, G. A.; Guerriero, P.; Martin, C. R.
Electroanalysis 1998, 10, 1168-1 173.

5) Liu, C.; Chen, W. J.; Martin, C. R. J. Membr. Sci. 1992, 65, 113-128.

6) Conner, M. D.; Janout, V.; Kudelka, 1.; Dedek, P.; Zhu, J .; Regen, S. L. Langmuir
1993, 9, 2389-2397.

7) Albrecht, 0.; Laschewsky, A.; Ringsdorf, H. Macromolecules 1984, 17, 937-940.

8) Hendel, R. A.; Nomura, E.; Janout, V.; Regen, S. L. J. Am. Chem. Soc. 1997, 119,
6909-6918.

9) Dedek, P.; Webber, A. S.; Janout, V.; Hendel, R. A.; Regen, S. L. Langmuir 1994, 10,
3943-3945.

10) Kim, K. Y.; Dickson, M. R.; Chen, V.; Fane, A. G. Micron Microscopia Acta 1992,
23, 259-271.

11) Kim, K. J .; Fane, A. G. J. Membr. Sci. 1994, 88, 103-114.

12) Nam, S. Y.; Lee, Y. M. J. Membr. Sci. 1997, 161-171.

13) Ackem, F. v.; Krasemann, L.; Tieke, B. Thin Solid Films 1998, 327-329, 762-766.
14) Kotov, N. A.; Magonov, S.; Tropsha, E. Chem. Mater. 1998, 10, 886-895.

15) Stroeve, P.; Vasquez, V.; Coelho, M. A. N.; Rabolt, J. F. Thin Solid Films 1996,
708-712.

16) Levéisalmi, J .-M.; McCarthy, T. J. Macromolecules 1997, 30, 1752-1757.

17) Dauginet, L.; Duwez, A.-S.; Legras, R.; Demoustier-Charnpagne, S. Langmuir 2001
17, 3952-3957. ’

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

19) Tieke, B.; Ackem, F. v.; Krasemann, L. Eur. Phys. E 2001, 5, 29-39.

89

20) Meier-Haack, J .; Lenk, W.; Lehmann, D.; Lunkwitz, K. J. Membr. Sci. 2001, 184,
233-243. -

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

22) Selectivity values can be calculated directly from flux ratios because source- phase
concentrations were always 0.1 F.

23) Chen, D.-H.; Wang, S.-S.; Huang, T.-C. J. Chem. Tech. Biotechnol. 1995, 64, 284-
292.

24) Akerman, S.; Viinikka, P.; Svarfvar, B.; J arvinen, K.; Kontturi, K.; Nasman, J.; Urtti,
A.; Paronen, P. J. Controlled Release 1998, 50, 153-166.

25) Saimoto, H.; Shigemasa, Y. Polym. Adv. Technol. 1999, 10, 39-42.

26) Krasemann, L.; Tieke, B. Langmuir 2000, 16, 287-290.

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

28) Lowack, K.; Helm, C. A. Macromolecules 1998, 31, 823-833.

29) Schlenoff, J. B.; Ly, H.; Li, M. J. Am. Chem. Soc. 1998, 120, 7626-7634.
30) Dubas, S. T.; Schlenoff, J. B. Macromolecules 1999, 32, 8153-8160.

31) Hoogeveen, N. G.; Cohen Stuart, M. A.; Fleer, G. J. Langmuir 1996, 12, 3675-3681.
32) Stojek, Z.; Ciszkowska, M.; Osteryoung, J. G. Anal. Chem. 1994, 66, 1507-1512.

33) Bard, A. J .; Faulkner, L. R. Electrochemical Methods: Fundamentals and
Applications; John Wiley & Sons: New York, 1980.

34) Stair, J. L.; Harris, J. J .; Bruening, M. L. Chem. Mater. in press.
35) Wycisk, R.; Trochimczuk, W. M. J. Membr. Sci. 1992, 65, 141-146.
36) Sata, T. J. Membr. Sci. 1994, 93, 117-135.

37) Sata, T.; Yamaguchi, T.; Kawamura, K.; Matsusaki, K. J. Chem. Soc., Faraday
Trans. 1997, 93, 457-462.

38) Sata, T.; Yamaguchi, T.; Matsusaki, K. J. Phys. Chem. 1995, 99, 12875-12882.

90

 

39) Sata, T.; Tagami, Y.; Matsusaki, K. J. Phys. Chem. B 1998, 102, 8473-8479.

40) Sata, T.; Mine, K.; Higa, M. J. Membr. Sci. 1998, 141, 137-144.

91

CHAPTER 5

Development of Polyelectrolyte Films as N anofiltration Membranes

I. Introduction

Nanofiltration (NF) membranes are used widely in the areas of water softening
and purification, wastewater reclamation, and dye-salt separations. ' ‘6 The majority of NF
membranes consist of a dense, thin active layer deposited on a thick support material, e. g.
an ultra or microfiltration membrane. The composition of the thin active layer
determines the separation characteristics of the membrane while the support material
provides mechanical rigidity to the system. Typically, solute transport through a NF
membrane is a firnction of the membrane charge and pore size.7'IO

Nanofiltration represents a relatively new area of research in the membrane field
and has separation characteristics between those of microfiltration and reverse osmosis.
The technique is often referred to as “soft” reverse osmosis, and typical operating
pressures are < 200 psi},10 The goal of nanofiltration is high water flux with moderate to
high rejection of ions at low operating pressures. This differs from reverse osmosis,
which typically requires NaCl rejections in the range of 99-100°/o. The separation
mechanism in nanofiltration relies on both the size and charge of the analyte. For
separating ions, the thin active layer in these membranes is highly charged and the
effective pore size is generally between 1 and 5 nm.7'9rl "'3

Deposition of layered polyelectrolyte films provides one means of creating highly

charged, ultrathin membranes and provides some control over both charge density and

92

membrane composition. ‘4 Chapters 2 and 4 illustrate the high degree of multivalent ion
rejection that is possible with polyelectrolyte films. The cross-linking strategy described
in Chapter 3 should provide a means for decreasing the effective pore size of
polyelectrolyte membranes, which should firrther increase ion rejections. Another
possible means for modifying the separation properties of polyelectrolyte membranes is
the derivatization of PAA layers in composite membranes to incorporate various moieties
into the fihn structure. As described in Chapter 4 there are numerous means to tailor the
ion-penneability of polyelectrolyte membranes.

This chapter discusses the use of layered polyelectrolyte films as nanofiltration
membranes. The high degree of multivalent ion rejection shown by these films (Chapter
4) supports the proposition of using polyelectrolyte fihns for nanofiltration applications.
Rejection and flux values should depend on the outer layer charge of the membrane and
the concentration of the feed solutions. Along with high rejection values these ultrathin
membranes should allow for high water flux. Performance of composite membranes and

membranes derivatized with hydrophobic groups will also be discussed.

11. Experimental
A. Chemicals

Polyelectrolytes and salts were used as described in previous chapters. Ethyl
chloroformate (Aldrich), N-methylrnorpholine (Aldrich), dimethyl fonnamide (Aldrich),
lH-lH-perfluorooctylamine (Lancaster), ethanol (Aldrich), and ethyl acetate (Aldrich)

were used as received.

93

B. Polyelectrolyte Deposition and Derivatization on Porous Alumina Supports

Polyelectrolyte membrane deposition followed the procedure described in Chapter
4 except that polyelectrolyte deposition began with the polyanion. In Chapter 4
polyelectrolyte deposition began with the polycation even though the alumina membrane
has a positive charge below pH values of 8.15 Ellipsometric studies on Al coated Si
wafers suggest that a layer of polyelectrolyte absorbs irrespective of the polyelectrolyte
charge. Regardless, the initial layer should have little if any effect on the transport
characteristics of the membrane. Composite membranes used for subsequent
derivatization consist of a 1.5-bilayer PAA/PAH fihn deposited on a precursor 5-bilayer
PSS/PAH film. For the PAA/PAH fihns, the deposition solutions contained 0.02 M
polyelectrolyte and 0.5 M NaCl and were adjusted to a pH of 4.5.

The presence of —COOH groups in composite membranes allows for the
derivatization of these films with perfluorooctylarnine (PF OA).l6 The derivatization of
procedure begins with activation of the COOH groups in the PAA layer via a 10 min
immersion in an ethyl chloroformate solution (100 11L in 10 mL DMF/ 100 11L N-
methylmorpholine). After rinsing with ethyl acetate, the membrane was immersed in a
solution of 0.01 N PF 0A for one hour. Upon removal from the PFOA solution, the
membrane was rinsed with ethanol and dried with N2. The extent of derivatization was
monitored using films of similar composition deposited on Au substrates. In that case,
Au substrates were first coated with MPA followed by polyelectrolyte deposition, which

followed the procedure described in Chapters 2 and 3.

94

C. Nanofiltration

Nanofiltration experiments were conducted in the home built cross-flow system
described schematically in Figure 5.1. The system was pressurized to 70 psi with Ar.
The feed tank was used to store the various brine solutions (~3 L), which consisted of 50
or 1000 ppm solutions of NaCl, NaZSOa, or CaClz. The concentrations were calculated
with respect to the ion of interest (i.e. Na+, S042} or Ca2+). 2000 ppm MgSO4 solutions
were also used. An AcetatePlusTM (0.45 pm pore size) membrane (Osmonics) was used
as a prefilter to remove any rust or particulate matter from the feed stream. The flow
meter was set to 18 mL/min, which is at least lOO-fold greater than the membrane flux.
This flowrate is high enough to minimize concentration polarization and low enough to
minimize delamination of the polyelectrolyte film. Increasing the flowrate to 40 mL/min
has little if any effect on ion-rej ection or permeate flux. The membranes were housed in
a home built stainless steel cell as shown in Figure 5.2.17 The active area of the
membrane contained within the o-ring was 1.6 cm2. Pressure gauges were used to
monitor possible pressure gradients that might arise from a fouled membrane cell.
Graduated cylinders were used to collect the permeate stream from the membrane cells.

Once the system was pressurized, the membranes were allowed to equilibrate with
the flowing solution for between 30 and 60 min. Upon equilibration permeate aliquots
were collected every 30 min. The run was terminated upon the collection of 3 aliqouts,
that showed similar (within 20%) flux and rejection values. The membranes were tested
first with the 50 ppm solution and then with the 1000 ppm solution. New membranes
were used when salt solutions were changed and 3 membranes were tested for each salt

solution.

95

 

 

 

 

 

 

 

 

 

'w

Figure 5.1: Experimental setup for nanofiltration experiments: Ar tank (1), stainless
steel feed tank (2), centrifiigal pump (3), prefilter (4), flowmeter (5), pressure gauges (6),
membrane cells (7), and graduated cylinders (8). System pressure was 70 psi and
flowrate was 18 mL/min.17

96

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 5.2: Schematic diagram of the nanofiltration membrane cell. (A) shows how the
cell is placed together: inlet/outlet ports ( 1&2), threads (3), rubber o-ring (4), alumina
membrane (5), porous stainless steel frit (6), cap that frit fits in and presses the membrane
tight against O-ring (7), and threaded bottom part of cell that top part screws into (8).

The cell is assembled in the direction of the arrow. (B) shows a bottom view of the upper
portion of cell: recessed inlet/outlet channels (1&2) and o-ring (3). l 7

97

III. Results and Discussion

A. Nanofiltration Performance of PSS/PAH Membranes

As seen in Chapter 4, only 4.5-bilayers of PSS/PAH are needed to completely
coat the porous structure of the alumina support. Nanofiltration experiments with
PSS/PAH membranes focused on films consisting of 4.5 or 5 bilayers of PSS/PAH.
Under 70 psi of pressure, the bare alumina membrane shows little if any resistance to
pure water transport (flowrate through the membrane of >10 mL/min). Upon deposition
of 4.5 bilayers of PSS/PAH, however, pure water flux decreases to 11 mL/hr. This
dramatic decrease in flux occurs with a film that is only 20 nm thick.

PSS/PAH membranes show impressive ion-rejection values that vary with the
outer layer charge of the membrane. As seen in Table 5.1, ion-rejection values increase
when the outer layer charge matches the charge of the ion of interest. Ca2+ rejection
values for 50 and 1000 ppm solutions, for example, increase from 92 to >95 % upon
deposition of a polycation layer. The same trend is also seen with S042} whose rejection
increases fiom 47 and 55% to 54 and 65%, for 50 and 1000 ppm solutions, respectively,
when the outer layer of the membrane is terminated with PSS as opposed to PAH. Na+
rejection values increase dramatically (2-to 3-fold) upon deposition of a polycation layer
for both 50 and 1000 ppm brine solutions. This trend in rejection values is consistent
with a Donnan repulsion model, which is discussed further in Section B. The
performance of polyelectrolyte membranes also depends on the concentration of brine in

solution. If rejection in polyelectrolyte membranes is represented by a Donnan

98

 

 

 

 

EEEAA

 

 

 

 

 

5.8 Eé A . . .
32 cm 8 a; 8 cm 8
3 Woo 3 cc 8 +:<<.$2&
+nm<fimme
3.8 $08 $08 $08 fine :2: m. 2:4.ng
co 3 co R mm a. + man/Emma
a 4 m. c a 4 46.8 8 4 m. s s 4 a: 8 4 was 8 4 8.8
o 4 no a 4 mo 0 4 mm o 4 t. o 4 me o 4 g 32.5de
$3 4 n: so 4 9: £3 4 N.; ad 4 n: so 4 4.: god 4 m: .
48 «48 548 3% 3.2 4:. 415$de
+~aU +~mU -vam -~¢Om +mZ +aZ 055502
can 82 eon cm. and 82 can an and 82 can on

 

 

 

 

 

 

 

 

$0.3 can .VOmSZ .534 263 £8528 :3 ofi. .mE N 8a Do o2 Hm vogue:

83 SE 333846 2:. 98252. 366:8 65 E countenance Bow 2: 8 6338 countenance
:2 8888a S 888% .x. Ecmoaou was: mun—9 :38 FM .8523an couch—moan: mm pom: mEE
owboboo—oboa 3223 com 76 N.E MS 5 $865553 5 835 329 5.: Ba 538.3% "fin «Bah

99

repulsion model, as suggested earlier, then rejection should decrease at higher
concentrations due to increased charge screening by the electrolyte solution. N 3"
rejection does follow this trend (Table 5.1) as rejection decreases by 66 and 43% for 4.5-
and 5-bilayer PSS/PAH films, respectively, as the feed concentration increases from 50 to
1000 ppm. However upon increasing salt concentration, SOaz' rejection values actually
increases by ~20% for both 4.5- and 5-bilayer PSS/PAH membranes while Ca2+
rejections show little if any change due to feed concentration.

One possible explanation for increased SOaz' rejections at higher feed
concentrations is that ion adsorption increases the amount of charge at the surface of
polyelectroltye films.18 The increased rejection values at high feed concentrations would
seem to indicate an increase in the effective charge of the polyelectrolyte membrane,
which would be consistent with a higher degree of ion adsorption at higher ion
concentrations. The high Ca2+ rejection values (>90%) for 4.5-bilayer PSS/PAH
membranes also suggest ion absorption, as one would expect rejection to be low for this
negatively capped membrane.

The rejection values for PSS/PAH membranes are similar to values obtained with
commercially available membranes as shown in Table 5.2. '9 One advantage of these
membranes is their high water flux, which is on average 1.3 m3m'zd" (Table 5.1), a value
greater than the majority of industrial nanofiltration membranes. The combination of a
high degree of surface charge and the minimal thickness of polyelectrolyte membranes
leads to high water fluxes while maintaining high ion rejections. In most cases the flux is
unaffected by the concentration of salt in the feed stream. This is unexpected, as the flux

should decrease at higher feed concentrations due to increased osmotic pressure. The

100

 

Table 5.2: Flux and rejection values for several commercially available nanofiltration
membranes and PSS/PAH membranes. Feed solutions were NaCl, MgSOr, and CaClz.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Pressure 3 -2 -1 Na+ S047”. Caz+
Membrane (psi) Flux (m m d ) Rejection Rejection Rejection
NF—200' 70 0.82 >95%4 35.50%5 I
NF-90‘ 70 0.91 85-95%' >95%4 7
Desal 51+ 50 0.94 41%2 81%6 I
Hydro 7540s 50 0.34 33%2 68%6 I
(PSS/PAH)4.5 70 >12 16%3 94%4 92%3 I
(PSS/PAH)5 70 >0.82 48%3 95%4 98%3 I
Dow F ilmTec Membranes‘ Osmonics+ HydranauticsS
‘ 2000 ppm NaCl
2 60 ppm NaCl
3 1000 ppm solutions
4 2000 ppm MgSO4
5 500 ppm CaCl2

6 34 ppm CaClz

101

increase in osmotic pressure decreases the effective pressure applied to the membrane

and should affect the results of both 1000 ppm Na)r and Ca2+ solutions as their osmotic
pressures are 31 and 27 psi, respectively. As seen in Table 5.1 there is little change in
permeate flux at the higher feed concentrations. In addition to high permeate flux, the
ease of tailoring the properties of the membrane by changing the outer layer charge is
highly advantagous. This provides one with the ability to tune the quality of the permeate

stream depending on the final application of the purified water.

B. Modeling of Ion Transport Through Polyelectrolyte Membranes

Modeling of the experimental data should provide a better understanding of the
possible mechanisms involved in ion transport through polyelectrolyte membranes.
Since charge interactions between the membrane and ions in solution appear to have a
major effect on ion transport, Donnan theory may apply. This theory arises from the
interfacial potential (Donnan Potential) that is created when a boundary separates two
solutions of varying concentration. The potential is described by equation 5.1, where (I) is
the Donnan potential, R is the gas constant, T is temperature, ZA is the charge of the ion,
and CASOI and CAmcm are the concentrations of ion A in the solution and membrane,

respectively.

 

RT CAmem
don _ mam _ 301 = 1n 5 1
(D - ¢ ¢ ZAF CASOI ‘

102

The distribution coefficient (Equation 5.2) can be derived from Equation 5.1 (Appendix 1
contains a detailed derivation) and this value correlates the concentration of ions in
solution with that of ions in the membrane.20 In this equation, Zx and Cx are the charge
and concentration of fixed sites in the membrane. If flux through a membrane is

sufficiently slow, equation 5.2 can be used to calculate ion rejection.

CAmem IZAlCASOI llAI/lZBI 5 .2
CASOI ( IZAICAmcm + IZX|meem )

 

To compare the performance of PSS/PAH membranes with this model, the
concentration of fixed charge within the membrane, Cx, was calculated frdm rejection
values using equation 5.2. These values (Table 5.3) were then compared to previously
measured values of charge in polyelectrolyte films. Several studies determined the
amount of charge present in polyelectrolyte films using various techniqueng’Z"22 Using
radiolabeled ions, Schlenoff determined that the amount of uncompensated charge at the
film surface is 4.3>< 1 0-10 mo] cm‘z.18 Using Schlenoff’s value of uncompensated
charge,18 and assuming 1:1 charge compensation in the membrane bulk and a 2-4 mm
thick charged outer layer, the concentration of fixed charge at the membrane surface
would be between 1.1 and 2.2 M. This value of uncompensated charge is much greater
than the charge required (Table 5.3) for the observed rejection values of PSS/PAH
membranes. Theoretically, the rejection values for ions should be even higher in
polyelectrolyte membranes, but flux may be high enough that ion concentrations in the

membrane are greater than equilibrium values.

103

Table 5.3: Theoretical values for the concentration of fixed charge (M) in
polyelectrolyte membranes calculated using Equation 5.2 and rejection values in Table
5.1.

 

50 ppm 1000 ppm 50 ppm 1000 ppm 50 ppm 1000 ppm
Na+ Na+ S042' 8042' Ca2+ Ca2+

 

 

3.2 x 10'3 2.0 x 10'2 1.1x10'3 2.8 x 10'2 1.1x10'2 2.2 x 10‘2

 

 

 

 

 

 

104

 

C. Performance of Composite and Derivatized Polyelectrolyte Membranes

In addition to control of ion transport through variation of surfacecharge,
composite membranes can be derivatized or cross-linked to utilize size exclusion and
hydrophobicity effects to alter ion transport. Previous studies show that ion-transport
selectivity through polyelectrolyte membranes can be enhanced by using
PSS/PAH+PAA/PAH composite membranes.23 As seen in Table 5.1 Na+ rejections for
such membranes remain relatively low, but S042' rejections increase for these composite
membranes compared to PSS/PAH membranes. Ca2+ rejections decrease for both 50 and
1000 ppm solutions with respect to PSS/PAH membranes. The trend of increasing 8042'
and Ca2+ rejection at higher feed concentrations, observed with PSS/PAH membranes, is
also seen with the composite membranes.

Composite membranes that are derivatized with PFOA (Figure 5.3) and then
cross—linked show even higher S042' rejections (99%) while maintaining moderate Ca2+
and Na+ rejections. Increased SOaz' rejections are reasonable due to the increased
hydrophobicity of the derivatized membrane. The SO42' anion is highly hydrated and
very hydrophilic.24’25 Post-deposition cross-linking and/or derivatization shows the
versatility of polyelectrolyte membranes for ion separations. However, these

modifications do decrease flux by a factor of about 3.

IV. Conclusions
Layered polyelectrolyte films provide a convenient means for the construction of
nanofiltration membranes. The ion-rejection values of these membranes are competitive

with commercially available membranes, while their water permeability is higher than

105

commercial systems. By changing the outer layer charge in PSS/PAH membranes or by
using composite membranes, rejection and flux can be easily tailored to fit the needs of a
desired applications. These attributes along with ease of synthesis makes layered

polyelectrolyte films attractive for constructing ion-separation membranes.

106

5-Bilayer PSS/PAH
Film + 1.5-Bilayer
PAA/PAH

 

 

  
   

Ethyl Chloroformate
N-Methylmorpholine
(1) DMF

10 min

 

 

0.01 M PFOA
DMF (2)
1 h

 

 

slflimlifilfllw’filnllar

 

Figure 5.3: Reaction scheme for derivazitation of composite membranes with lH-lH-
perfluorooctylamine (PFOA). In step 1 the acid groups of PAA are converted to a rrnxed
anhydride. In step 2, the amine group of PF OA reacts with the anhydrlde resultmg in
attachment of PF 0A to the composite film through an amide linkage.

107

V. References

1) Xu, X.; Spencer, G. Desalination 1997, 114, 129-137.
2) Yaroshchuk, A.; Staude, E. Desalination 1992, 86, 115-134.

3) Hofrnan, J. A. M. H.; Beerendonk, E. F.; Folmer, H. C.; Kruithof, J. C. Desalination
1997, 113, 209-214.

4) Bohdziewicz, J .; Bodzek, M.; Wasik, E. Desalination 1999, 121.

5) Berg, P.; Hagmeyer, G.; Gimbel, R. Desalination 1997, 113, 205-208.

6) Rautenbach, R.; Linn, T.; Eilers, L. J. Membr. Sci. 2000, 174, 231-241.

7) Childress, A. E.; Elimelech, M. Environ. Sci. T echnol. 2000, 34, 3710—3716.

8) Afonso, M. D.; Hagmeyer, G.; Gimbel, R. Separation Purif T echnol. 2001, 22-23,
529-541.

9) Bowen, W. R.; Mohammad, A. W.; Hilal, N. J. Membr. Sci. 1997, 126, 91-105.

10) Bhattacharyya, D.; Williams, M. E. Membrane Handbook; Bhattacharyya, D.;
Williams, M. E., Ed.; Van Nostrand Reinhold: New York, 1992, pp 263-280.

11) Kim, C. K.; Kim, J. H.; Rob, 1. J.; Kim, J. J. J. Membr. Sci. 2000, 165, 189-199.
12) Ko§utic, K.; Kastelan-Kunst, L.; Kunst, B. J. Membr. Sci. 2000, 168, 101-108.
13) Shibata, M.; Kobayashi, T.; Fujii, N. J. Appl. Polym. Sci. 2000, 75, 1546-1553.
14) Decher, G. Science 1997, 277, 1232-1237.

15) Parks, G. A. Chem. Rev. 1965, 65, 177-198.

16) Bruening, M. L.; Zhou, Y.; Aguilar, G.; Agee, R.; Bergbreiter, D. E.; Crooks, R. M.
Langmuir 1997, 13, 770-778.

17) The author would like to thank Keith Harris for the original design of the
nanofiltration cell.

18) Schlenoff, J. B.; Ly, H.; Li, M. J. Am. Chem. Soc. 1998, 120, 7626-7634.

108

19) Performance of commercially available membranes can be found at
www.osmonics.com/products or www.dow.com/liquidseps/pc/product.htm.

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

21) Hoogeveen, N. (3.; Cohen Stuart, M. A.; Fleer, G. J. Langmuir 1996, 12, 3675-3681.
22) Ruths, J .; Essler, F .; Decher, G.; Riegler, H. Langmuir 2000, 16, 8871-8878.
23) Stair, J. L.; Harris, J. J .; Bruening, M. L. Chem. Mater. in press.

24) Collins, K. D.; Washabaugh, M. W. Quar. Rev. Biophys. 1985, 18, 323-422.

25) Marcus, Y. J. Chem. Soc. Faraday Trans. 1991, 87, 2995-2999.

109

CHAPTER 6

Conclusions and Future Work
1. Stability and Ion-Permeability of Layered Polyelectrolyte Films Deposited on
Solid Supports

Chapter 2 shows that layered polyelectrolyte films offer a convenient means for
surface modification. The deposition of PAH/PSS and PAH/PAA fihns follows a regular
and controlled pattern, which allows for nanoscale control of film thickness and
composition. Once deposited these films are stable in solutions at low and moderate pH
values but delaminate and undergo irreversible structural changes in alkaline solutions.
However, upon cross-linking, PAH/PAA films show increased stability at high pH due to
the formation of covalent linkages between the individual polyelectrolyte layers as
discussed in Chapter 3.

The permeability of polyelectrolyte films is related to the film stability and
composition. At pH 3.2 and 6.3, the permeability of PAH/PSS films to F e(CN)r,3 ’ is low
and stable with time. Upon immersion in pH 10 solutions, the permeability of PAH/PSS
films drastically increases at rates similar to those for film delamination. Even though a
significant portion of polyelectrolyte film remains on the electrode, the structure is
altered, allowing for a continued high ion-flux even upon reimmersion in a pH 3.2
solution. The permeability of PAH/PSS films also depends on film thickness. Decreases
in permeability occur with increases in the number of bilayers up to lO-bilayers.
Covalent cross-linking of PAH/PAA films increases film stability and dramatically
decreases film permeability. By varying cross-linking temperature, the permeability of

PAH/PAA films can be easily controlled due to the extent of amidation. Cross-linking

110

and varying the number of film bilayers provide two means for tailoring the permeability

of polyelectrolyte films.

11. Polyelectrolyte Films as Anion-Separation Membranes

Polyelectrolyte fihns show promise as ion-separation membranes due to their high
charge and wide tunability. Chapter 4 discusses the ion—permeability, as determined from
dialysis measurements, of these films deposited on porous alumina supports. Cl'/SO.12'
selectivity values for both PSS/PAH and PAA/PAH films are ~5, but with PAA/PAH
membranes there is a 70% decrease in ion flux. The use of composite membranes, which
consist of a thin PAA/PAH film deposited on a precursor PSS/PAH fihn, results in CI'
/SOaz' selectivity values as high as 350. In these membranes the outer layer charge is the
major factor in controlling the ion-permeability characteristics of the membrane. In most
cases, anion selectivity values decrease upon deposition of a cationic layer at the

membrane solution interface.

III. Characterization of Nanofiltration Membranes Made from Layered
Polyelectrolyte Films

Chapter 5 shows that polyelectrolyte fihns are feasible candidates for
nanofiltration membranes due to their high degree of charge and high multivalent ion
rejection illustrated in Chapters 2 and 4. Polyelectrolyte membranes provide rejection
values of 98 and 70% for Ca2+ and S042} respectively, while maintaining moderate Na+
rejection. As in Chapter 4, the ion rejection values are highly dependent on the charge of
the outer layer of the polyelectrolyte film. Ion-rej ections increase when the

polyelectrolyte membrane is terminated with the same charge as the ion in solution. In

111

addition to high rejection values, these membranes show high water fluxes, which is
critical for possible industrial applications. Further derivatization with hydrophobic
groups and/or cross-linking increases certain rejection values, but fluxes decrease.
Overall, polyelectrolyte films provide an easy means for membrane construction while

providing a high level of membrane performance.

IV. Future Work

The primary drawback with polyelectrolyte membranes as described in this thesis
is the porous alumina support material. While providing well-defined pores and a high
pore density, the material is too fragile for practical applications. Work must be done to
find a better support such as a polymeric membrane, which will provide mechanical
stability but at the same time increased durability. Unfortunately, commercially available
polymeric membranes suffer from low pore density, and therefore reduced fluxes.
Another approach would be to use polyelectrolyte films to modify the separation
properties of commercially available anion-exchange or nanofiltration membranes. This
should provide an easy and practical means to tailor the properties of the membrane.

The derivatizable nature of PAH/PAA films provides several possible means for
modifying separation characteristics of the membrane. Since the major separation
mechanism in nanofiltration membranes is charge exclusion, the templating of metal ions
in the film may provide a means for increasing the effective charge of the membrane
upon removal of the ions. Also templating of molecules may provide for a means of
creating well-defined pores in the membrane that would provide structure selective

transport of molecules. The derivatization of polyelectrolyte membranes touched on in

112

 

 

Chapter 5 opens many opportunities for introducing new separation mechanisms (i.e.
hydrophobicity) to the field of nanofiltration. In conclusion, this thesis presents the
advantages and possibilities of using layered polyelectrolyte films as ion-separation

membranes.

113

APPENDIX 1

Appendix 1 represents the derivazitation (page 115) of the distribution coefficient

so') and within the membrane

of anion (a) with respect to its concentration in solution (Ca
(Camcm). Equation 1 and 2 equate the electrochemical and standard chemical potentials,
respectively, between species (a) in the solution and membrane phase. This can also be
done for the cation (equation 2a). The chemical potential between the membrane and
solution phases (equation 4 and 4a) is solved by rearrangement of equation 2 and this
represents the Donnan potential for species (a) and (c), respectively. Equation 5
represents the electroneutrality condition between species (a) and (c) in solution, while
equation 5a describes the electroneultrality conditions between the species in solution and
the fixed charge of the membrane. Since we are concerned with the anionic species
equations 5 and 5a are solved for CCSOI and Ccmem resulting in equations 6 and 6a,
respectively. Substituting equations 6 and 6a into equation 4 produces equation 7.
Raising equation 7 to the -za power produces equation 8. Inverting the left side of

equation 8 (equation 9) followed by inserting absolute values for 2a and 2c results in

equation 10.

114

(1)

(2)

(26)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

APPENDIX 1

firm = 1'14“”
m” + RT InC "‘°"‘ + z Firm: n, °'+ RT lnC, 5"'+ z ,F¢’°‘
"“m + RT lnC '"°"‘ + z F=¢'“°"' tlc‘°'+ RT lnC so'+ z F¢5°'

mam sol_ RT C_a 30' mem sol RT C sol
(I) ' (I) =-Z—Fn In Ca —mem (3a) (1) ' ¢

(0:;0111'1/23:(Cccmom)'1/Zc

zaCasol + ZcCcsol = 0 (5a) zaCamem + ZCCcmem + Z"Cumem = 0

ll
5

 

 

.zaCa (6a) Ccmem = 'szx ' 28C,

C sol =
c zc 2.

 

_ -1/zc

~(ZXC‘ mom + 28 Ca mom) - ( Camem )‘1/2.
z(-z -z C __:,_') C“

Z Zc mam
zaCamem+ chmem 9/ : C——a-—
2803301 To

_

 

 

 

 

 

l'fi-i'l

 

'2 Zc mem
280880. a, = L
2.03”“ + 2.03““ C:

2. / m...
|z,|C,’°' I I IZ‘L C.
_' sol
12.10.”°"‘+ 12.10.”: C-

115

l—I

 

 

NT“

 

  

 

  

MICHIGAN STATE llB
Iiiwi‘iI HI‘I‘I Iii All: 214' iI

1293 021770

 

 

 

      

- . -_—‘_—‘ ‘A