Biomedical Applications of Nanostructured Polymer Films
by
ARC&WES
Jonathan Brian Gilbert
MASSACHURi8
OF TECHNOLOGY
M.S. Chemical Engineering Practice
Massachusetts Institute of Technology
IE
JUN 3 0 2014
L IB RARI EE
B.S. Chemical Engineering
Johns Hopkins University
SUBMITTED TO THE DEPARTMENT OF CHEMICAL ENGINEERING IN PARTIAL
FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY IN CHEMICAL ENGINEERING
at the
MASSACHUSETTS INSTITUTE OF TECHNOLOGY
June 2014
©2014 Massachusetts Institute of Technology. All rights reserved.
Signature of Author: ....
Signature redacted,.......................................
ment of Chemical Engineering
May 5, 2014
-/7
Signature redacted
Certified by:..................................................
Robert E. Cohen
St. Laurent Professor of Chemical Engineering
Thesis Co-Supervisor
Certified by: ..................
S ignature redacted
...........................
Michael F. Rubner
TDK Professor of Materials Science and Engineering
Thesis Co-Supervisor
Signature redacted
Accepted by: ..........................
..............................
Patrick Doyle
Professor of Chemical Engineering
Chairman, Committee for Graduate Students
1
Biomedical Applications of Nanostructured Polymer Films
by
Jonathan Brian Gilbert
Submitted to the Department of Chemical Engineering on May 5 th, 2014 in partial fulfillment of
the requirements for the degree of Doctor of Philosophy in Chemical Engineering
ABSTRACT
Functional polymeric thin films are often stratified with nanometer level structure and
distinct purposes for each layer. These nanostructured polymeric materials are useful in a wide
variety of applications including drug delivery, tissue engineering, controlling condensation and
polymeric batteries; all of which will be discussed in this work.
The first area of my thesis will detail the use of C60 cluster-ion depth profiling X-ray
Photoelectron Spectroscopy (XPS) to fundamentally understand how thin film structure and
function relate. This method has the unique capability to determine the atomic composition and
chemical state of polymeric thin films with <10nm nanometer depth resolution without any
chemical labeling or modification. Using this technique, I probed the nanostructure of functional
thin films to quantify the interlayer diffusion of the biopolymer chitosan as well as demonstrate
methods to stop this diffusion. I also explored the role of interlayer diffusion in the design of
hydrophobic yet antifogging 'zwitter-wettable' surfaces. Additionally, I probed the lithium
triflate salt distribution in solid block copolymer battery electrolytes (PS-b-POEM) to understand
the lithium-ion distribution within the POEM block.
In the second area of my thesis, I show how the nanostructure of materials control the
function of polymeric particles in vitro and in vivo. One example is a 'Cellular Backpack' which
is a flat, anisotropic, stratified polymeric particle that is hundreds of nanometers thick and
microns wide. In partnership with the Mitragotri group at UCSB, we show that cellular
backpacks are phagocytosis resistant, and when attached to a cell, the cell maintains native
functions. These capabilities uniquely position backpacks for cell-mediated therapeutic delivery
and we show in vivo that immune cells attached to backpacks maintain their ability to home to
sites of inflammation. In addition, we have designed polymeric microtubes that can control their
orientation on the surface of living cells. Inspired by chemically non-uniform Janus particles, we
designed tube-shaped, chemically non-uniform microparticles with cell-adhesive ligands on the
ends of the tubes and a cell-resistant surface on the sides. Our results show that by altering the
surface chemistry on the end versus the side, we can control the orientation of tubes on living
cells. This advance opens the capability to control phagocytosis and design cellular materials
from the bottom up.
Thesis Supervisors: Robert E. Cohen and Michael F. Rubner
Respective Titles: St. Laurent Professor of Chemical Engineering and TDK Professor of
Materials Science and Engineering
2
Acknowledgements
Many people have contributed to my education and to the work presented in this thesis.
Professor Rubner and Professor Cohen have been essential in developing my capabilities as a
scientist and I thank them immensely for their time and effort. Their unique contributions to my
research have always pushed me to ask fundamental questions so my research can have the
largest impact possible. My thesis committee members, Professor Irvine and Professor
Hammond, also served as valuable resources in my research. I am particularly grateful for
Professor Irvine's generosity with his resources as I have enjoyed the benefits of his tissue
culture space and excellent microscopes. I also have appreciated the help and support of his
friendly group members. I am also grateful to Professor Hammond who provided constant
positive support and input.
Many students in the Rubner and Cohen groups have also given essential scientific and
personal advice. I am particularly indebted to Albert Swiston, who pioneered the cell backpack
work and was instrumental in developing applications. His leadership and advice in the lab as I
started my research career was very helpful. I also would like to thank Hyomin Lee and Siddarth
Srinivasan who also entered the Cohen lab in January 2010 and have been friends and excellent
collaborators. It has been a joy growing as scientists with them and I am excited to see what their
careers will hold. Other past members of the lab including Gary Chia, Jonathan DeRocher and
Shree Chhatre have been essential in my growth. I also would like to acknowledge my other
current labmates, Rosanna Lim and Justin Kleingartner, for their input and support. I am
especially thankful for Rosanna and her constant positive attitude. It has been a pleasure working
with her to further develop the backpacks.
During my thesis research, I have also have enjoyed working with wonderful collaborators
across the US and the world. Aaron Anselmo, Nishit Doshi and Professor Mitragotri of UCSB
have been very helpful as we have pushed toward the in vivo application of cell backpacks. I am
also very thankful for Ming Luo and Professor Epps at the University of Delaware as we have
worked towards the depth profiling analysis of block copolymer thin films. Grinia Nogueira,
Mariana de Moraes and Roberta Polak also have shaped my graduate career and I am
appreciative of Mariana and Roberta's hospitality in my short research trip to UNICAMP in
Campinas, Brazil. Maria Ochoa, Janice O'Brien and Harini Suresh's work as UROPs also added
essential parts to my thesis. I also thank Girma Endale for his work in immune cell bilayers. I
would also like to acknowledge the help of the CMSE staff Libby Shaw and Tim McClure.
Libby's help and support was essential in developing the best procedures for depth profiling XPS
of polymeric thin films.
Furthermore, the constant support of my friends and family has been essential. In particular, I
would like to thank Tim Palmer for his constant support and advice in my research and my
personal life. Mitch Wang was also an invaluable ally in the first year of classes, and was always
willing to help refine an idea or a presentation throughout graduate school. Other classmates and
housemates including Jonathan Harding, David Borelli, Katie Quinn, Kamil Khan, Shamel
Merchant, Matt Glassman and many others provided a wonderful social network of friends. I
would also like to thank the network of friends at Hope Fellowship Church for their constant
support and accountability in my faith. My family's support and love has been non-stop and I
3
cannot thank them enough. Most importantly, I would like to thank my wonderful, beautiful, and
intelligent wife Stephanie for her constant support and love. She has listened to my worries and
troubles and has always supported me in love. I could not have done this without her.
4
Table of Contents
1.
List of Figures .........................................................................................................................
2.
List of Tables ........................................................................................................................
18
3.
Introduction and Background ............................................................................................
19
7
3.1
Introduction ....................................................................................................................
19
3.2
N anostructured Thin Film s ..........................................................................................
19
4.
3.2.1
Fundam ental Studies of Nanostructured Film s....................................................
21
3.2.2
Biomedical Applications of N anostructured Film s.............................................
23
Depth Profiling XPS Analysis of Interlayer Diffusion in Polyelectrolyte Multilayers ....... 25
4.1
Introduction....................................................................................................................
25
4.2
Experim ental Set-up ...................................................................................................
29
4.3
Diffusion of Chitosan in Hydrogen Bonded M ultilayers ...........................................
30
4.4
D isplacem ent of PEO from Film by Chitosan ...............................................................
34
4.5
Electrostatic Blocking Layer Stops Chitosan Diffusion ............................................
36
4.6
pH Sensitivity of the Hydrogen Bonded Region.........................................................
40
4.7
Conclusions ....................................................................................................................
41
4.8
M aterials and Methods...............................................................................................
42
4.9
Supplem ental Inform ation..........................................................................................
45
5.
Designing Zw itter-wettable Surfaces.................................................................................
51
5.1
Introduction ....................................................................................................................
52
5.2
Results and Discussion...............................................................................................
54
5.3
M ethods..........................................................................................................................
64
5.4
Supporting Inform ation...............................................................................................
69
Ion D istribution in Block Copolym er Electrolyte Thin Films...........................................
71
6.
6.1
Introduction ....................................................................................................................
72
6.2
Results and Discussion...............................................................................................
75
6.3
Supporting Inform ation...............................................................................................
86
6.4
M aterials and Methods ...............................................................................................
88
7.
Cellular Backpacks ...............................................................................................................
91
7.1
Introduction and Background......................................................................................
91
7.2
Phagocytosis Resistance of Cell Backpacks ..............................................................
96
5
7.2.1
Introduction.............................................................................................................
7.2.2
Results.............................................................................................................
7.2.3
D iscussion.............................................................................................................
96
-.... 97
105
107
Experim ental Section............................................................................................
Monocyte Hitchhiking: Delivery of Phagocyte Resistant Backpacks to Lungs .......... 113
7.2.4
7.3
7.3.1
Supporting Figures................................................................................................
121
7.3.2
Experim ental Section............................................................................................
124
Future Applications ......................................................................................................
129
Tissue Engineering Applications .................................................................................
Controlled A ggregate Size using Cell Backpacks ................................................
7.5.1
130
7.4
7.5
130
Immune Cell Bilayers Assembled Using Layer-by-Layer Processing ................. 131
137
Supplem ental Inform ation............................................................................................
7.5.2
7.6
8.
Orientation-Specific Attachment of Polymeric Microtubes on Cell Surfaces....................
139
8.1
Introduction..................................................................................................................
139
8.2
141
8.3
Fabrication of Cellular Tubes.......................................................................................
Effect of Chem istry on the Cell Surface Orientation...................................................
8.4
Conclusion and Future Outlook ...................................................................................
147
8.5
Experim ental D etails ....................................................................................................
148
8.6
Supplem ental Figures...................................................................................................
150
143
Future W ork ........................................................................................................................
151
10. Citations ...................................................................................................................----....
153
11. Appendix......................................................................................................................---..
165
9.
11.1
Chem istry of 'Released' Side of Free Floating Film s..................................................
6
165
1. LIST OF FIGURES
Figure 3-1: Procedure to create electrostatic Polyelectrolyte Multilayers (PEMs) A) The surface
is alternately dipped in solutions of oppositely charged polymers and rinsed in between.
B) The polymers are adsorbed and sequentially build a film. - Adapted from (13) ........ 20
Figure 4-1: Schematic of systems used to test (A) chitosan diffusion into the hydrogen bonded
region and (B) electrostatic blocking layer effectiveness. The number after the polymer
abbreviation is the deposition solution pH...................................................................
30
Figure 4-2: Diffusion of chitosan into hydrogen bonded films. Spectra of hydrogen bonded
(PAA3/PEO3) films exposed to chitosan solution for different amounts of time. (A) 1 min
exposure (B) 3 min exposure (C) 10 min exposure (D) 60 min exposure to chitosan. The
color scheme is the same as that of Figure IA. Red spectra represent chitosan infused
areas, yellow spectra represent the hydrogen bonded (PAA3/PEO3) area and black
spectra represent the (PDAC4/SPS4) adhesion layer. (E) Quantification of A-D to
determine the atomic concentration of nitrogen with depth in the film. Data points are
individual dots and line shown is the result of a Savitzky-Golay five point quadratic
alg orithm ...........................................................................................................................
31
Figure 4-3: High resolution Cls XPS depth profiling of a hydrogen bonded film exposed to
chitosan solution. (A) 1 min (B) 3 min exposure to chitosan. The color scheme is the same
as that of Figure 4-lA. Red spectra represent chitosan infused areas, yellow spectra
represent the hydrogen bonded (PAA3/PEO3) areas and black spectra represent the
(PDAC4/SPS4) adhesion layer. Comparing the red chitosan diffused areas to the yellow
hydrogen bonded areas the chitosan diffused areas have a lower signal at 286.5eV since
PEO has diffused out. (C) Comparison of Cls spectra with different chitosan exposure
times to the initial yellow hydrogen bonded area. The longer exposure to chitosan
solution, the more the PEO signal at 286.5eV decreases. All chitosan exposed spectra
were from ~450 nm above the glass surface.................................................................
35
Figure 4-4: Effect of a blocking layer on interlayer diffusion of chitosan (A) C Is and Ni s regions
from depth profiling XPS of a hydrogen bonded sample with a (PAH3/SPS3) 3 5 blocking
layer topped with (HA3/CHI3) 35 . The color scheme is the same as that of Figure 4-lB.
7
Red
spectra represent (HA3/CHI3),
green spectra represent
the (PAH3/SPS3)
electrostatic blocking layer, yellow spectra represent the (PAA3/PEO3) hydrogen bonded
region and black spectra represent the (PDAC4/SPS4) adhesion layer. (B) Quantification
of the nitrogen signal for different blocking layer systems tested. Data points are
individual dots and the line is the result of a Savitzky-Golay five point smoothing
algorithm. All films had (HA3/CHI3) 3 5 deposited on top of the blocking layer.......... 37
Figure 4-5: Chemical structure of the polymers in this study...................................................
42
Figure 4-6: Increasing X-ray Exposure Time Decreases Carboxyl Atomic Concentration: To
study the effect of X-ray exposure and sputtering conditions identical (PAA3/PEO3) 30.5
(PAH3/SPS3) 9 .5 samples were exposed to a variety of X-ray exposure levels while depth
profiling. Data was analyzed with Multipak software (Physical Electronics, USA) and
then the carboxyl peak at 289eV was fitted using CasaXPS software. Different XPS data
acquisition parameters were tested along with different C6 0 * cluster ion sputtering
conditions to determine conditions with minimal decrease in carboxyl signal, high signal
to noise ratio and many data points while depth profiling. The carboxyl signal was chosen
to determine X-ray damage since it is highly sensitive to degradation. It was determined
that the condition 5 shown above provided a good balance of these demands and the final
acquisition parameters are found in Table 4-3. The degradation of the film due to
extended X-ray exposure or sputtering is a limitation that should be carefully considered
by others before using this technique. C6 0 deposition was not observed since the C-C
bond at 285eV did not systematically increase............................................................
46
Figure 4-7: XPS Depth profile in absence of chitosan. (A) Nis signal from a (PAA3/PEO3) film
before exposure to chitosan. The only Nis signal is from the (PDAC4/SPS4) adhesion
layer. (B) PDAC from (PDAC4/SPS4) adhesion layer does not diffuse into the hydrogen
bonded region (PAA3/PEO3) on top of it even after 15 hours in pH 3 water as shown by
47
the absence of nitrogen signal above the adhesion layer. .............................................
Figure 4-8: GATR-FTIR results showing ionization of PAA upon chitosan diffusion. Two distinct
1
peaks from the carboxylic acid functional group of PAA are shown: one at -1550 cm is
associated with the asymmetric stretching of the ionized carboxylate (COO-) and the
other at -1710 cm' which is associated with the C=O stretching of the carboxylic acid
8
(COOH) groups. In comparison to the pure hydrogen bonded sample (PAA3/PEO3), the
sample exposed to chitosan for 10 minutes has a higher amount of ionized carboxylate
gro up s................................................................................................................................
47
Figure 4-9: High resolution Cls XPS depth profiling of a hydrogen bonded film exposed to
chitosan solution. (A) 10 min and (B) 60 min exposure to chitosan. The color scheme is
the same as that of Figure 4-A. Red spectra represent chitosan infused areas, yellow
spectra represent the hydrogen bonded (PAA3/PEO3) areas and black spectra represent
the (PDAC4/SPS4) adhesion layer...............................................................................
48
Figure 4-10: Pure component Cls XPS spectra. (A) Poly(ethylene oxide) (B) Chitosan (C)
P oly(acrylic acid)..............................................................................................................48
Figure 4-11: N Is spectra from depth profiling XPS of hydrogen bonded films with different
blocking layers. All samples were topped with (HA3/CHI3) 3.5. (A) No blocking Layer (B)
(PAH3/SPS3) 9.5 layer (C) (PAH3/SPS3) 3 5 layer (D) Single (PAH3) layer ..................
49
Figure 4-12: (CHI3/HA3) 3 .5 on top of a hydrogen bonded region. Independent of whether CHI or
HA is deposited first onto the hydrogen bonded region, the chitosan diffuses fully
through the film as seen by the nitrogen signal throughout the film. ............................
50
Figure 4-13: Comparison of CHIlO nitrogen signal before and after PBS exposure. CHIlO
sample did not dissolve after a 30 minute PBS soak. It was stabilized due to the diffusion
and electrostatic cross-linking of chitosan as seen by high levels of nitrogen throughout
th e film ..............................................................................................................................
50
Figure 5-1: (a) Structures of the polymers described in this paper. Polycations and polyanions are
labeled with parenthesis (+), (-), respectively. (b) Water advancing contact angle of the
samples tested. (c) Photographs taken immediately after transfer to ambient lab
conditions (22 ± 1 'C, 40 ± 10% RH) from a -1 C refrigerator (1 hr). Only the
hydrophilic surface and zwitter-wettable surface resisted fog formation. (d) Schematic
representation of hydrophilic surface with sessile drop and during condensation. (e)
Schematic representation of zwitter-wettable surface with sessile drop and during
conden sation .....................................................................................................................
9
54
Figure 5-2: (a) Schematic diagram of the polymer patch fabrication process. (b) Scanning
electron microscopy (SEM) image of the polymer patch. Scale bar is 50 pm. Outlet image
shows the enlarged SEM image and the average carbon and fluorine signal obtained from
energy dispersive X-ray spectrometry (EDS). (c) Optical microscopy image of the
polymer patch after exposure to humid air. (d) ESEM images of the condensation of
water vapor on a polymer patch. Water droplets start to form on the off-patch areas and at
the edges of the patches. ...............................................................................................
57
Figure 5-3: (a) Schematic of depth-profiling x-ray photoelectron spectroscopy (XPS) used to
acquire atomic concentration profile and molar ratio profile of various samples tested.
Here, atomic concentration of (CHI/Nafion) 3 coated hydrophilic surface is shown as an
example. (b) Molar ratio of hydrophilic surface ((CHI/CMC) 30 ) with depth in the film. (c)
Molar ratio of (CHI/Nafion) 3 coated hydrophilic surface with depth in the film. (d) Molar
ratio of PPFDA coated hydrophilic surface with depth in the film...............................
59
Figure 5-4: (a) Diffusivity (D) of water vapor in hydrophilic reservoir and (CHI/Nafion) 3 coated
hydrophilic reservoir (b) Solubility of water vapor in hydrophilic reservoir and
(CHI/Nafion) 3 coated hydrophilic reservoir. (c) Schematic representation of the effect of
the ability of thin hydrophobic capping layer to transport water molecules: For low
diffusivity capping layer, water molecules nucleate and grow on the surface while for
high diffusivity capping layer, water molecules in the vapor phase prefer to directly
imbibe into the underlying hydrophilic reservoir. .........................................................
61
Figure 5-5: (a) Water contact angle evolution over time for (CHI/CMC) 30 subsequently coated
with 1-, 2-, and 3- bilayers of (CHI/Nafion) by LbL technique. (b) Water contact angle
evolution over time for various samples tested in this work. ......................................
69
Figure 5-6: Advancing and receding contact angle of water, diiodomethane, and hexadecane on
(CHI/Nafion) 3 coated hydrophilic reservoir. ...............................................................
Figure 6-1: Fabrication of lamellar PS-POEM lithium containing films. .................
70
76
Figure 6-2: (A) Optical images of gradient thickness PS-POEM films (neat, 12:1 salt and 6:1
salt) annealed at 135 'C for 6 h; (B) AFM height images and corresponding sections
10
show that the difference between the high and low regions is Lo, which is consistent with
SAXS m easurement of the bulk system ........................................................................
77
Figure 6-3: X-ray reflectivity profile for PS-POEM films with salt ratio 12:1 and 6:1. The solid
line denotes measured profile and the "o" symbol denotes fitted profile.....................
79
Figure 6-4: Depth Profiling XPS of 36 kDa PS-POEM without lithium salt. (A) Schematic of
depth profiling XPS analysis. (B) Atomic concentration versus thickness above the
substrate. (C) and (D) Ols and Cl s photoelectron spectra respectively, showing the
alternating intensity of a lamellar block copolymer film. The red and blue spectra are
spectra primarily in the POEM and PS region respectively. .........................................
81
Figure 6-5: XPS depth profiling of 36kDa PS-b-POEM with 6:1 Li salt added. (A) Atomic
concentration profile with depth. (B) Zoom-in of low concentration and overlay of
fluorine atomic ratio comparison. (C) Distribution of PEO containing side block within
the film analyzed via Ols signal minus Ols signal from the salt (dashed green), fitting the
Cls peak (dashed blue) or Fis ratio to POEM (dashed red) (D), (E), (F) and (G) 3D
spectra of the Cls, Lils, Fis and Ols regions respectively. The Lils figure is rotated for
clarity .................................................................................................................................
Figure 6-6: Bulk TEM image of salt-doped PS-POEM with EO:Li
=
84
6:1. Dark domains
correspond to RuO4 -stained POEM; B. AFM height image of salt-doped PS-POEM thin
film with EO:Li = 4:1, indicating a mixed cylinder and lamella structure.................... 86
Figure 6-7: Peak fitting to the different regions of the C Is spectra. A) POEM region displaying
peaks from the POEM backbone (C-C) PEO side chain (PEO), ester linkage (Ester
Carbon) and lithium Salt (CF3). B) PS region displaying Carbon bonds (C-C) a very
small peak from PEO side chain (PEO) and C-C bond shake up peak due to pi stacking.
...........................................................................................................................................
86
Figure 6-8: Effect of Inelastic Mean Free Path (IMFP) on Depth Profiling Interface Analysis.
Comparing experimental data 'o' to the modeled system we see that though the IMFP
would affect the shape of the profile, but not to the extent seen in our data. ...............
II
87
Figure 7-1: Fabrication of Backpacks on a Surface. Reprinted with permission from Swiston,
A.J., Cheng, C., Urn, S.H., Irvine, D.J., Cohen, R.E., and Rubner, M.F. Nano Lett. 8,
4446-4453 (2008). Copyright 2008 American Chemical Society. ................................
92
Figure 7-2: Potential Methods of Cell Attachment A) Attachment of red labeled B-cells onto
FITC labeled (hyaluronic acid/chitosan) patches (reproduced from Al Swiston). scale bar
1 Opm B) Attachment of unlabeled Monocytes onto (PAA4/PAH4) patches activated with
EDC-NHS. scale bar l0pm C) Attachment of unlabeled Monocytes onto green labeled
IgG presenting patches. scale bar 50pm . .......................................................................
93
Figure 7-3: Attachment Scheme of Cell Backpacks a) PEM posts with three regions remains on
the surface after the removal of the photoresist. b) The posts are incubated with cells to
attach them to the posts c) The hydrogen bonding region is then dissolved, resulting in
cells with backpacks attached. Reprinted with permission from Swiston, A.J., Cheng, C.,
Urn, S.H., Irvine, D.J., Cohen, R.E., and Rubner, M.F. Nano Lett. 8, 4446-4453 (2008).
Copyright 2008 American Chemical Society. .............................................................
95
Figure 7-4: Backpacks for cell-based drug delivery devices: (a) schematic of the different
backpack layers, all of which deposited on top of a photoresist-patterned glass substrate
and (b) scanning electron micrograph of a backpack with a PMAA/PVPON release
region fabricated by the above strategy. The average diameter is ~ 6 pm and the
backpacks have a flat disk shape. (c) Time lapse microscopy image sequence of
macrophages interacting with HA/CHI coated backpacks. The arrow indicates an
internalized backpack. All other backpacks (n=7) are attached to macrophage surfaces for
the entire time lapse sequence of 50 minutes, exhibit strong resistance to macrophage
phagocytosis. (d) Quantitative analysis of 6 pm diameter particles internalization by
macrophages. Data is presented in terms of internalization percentage (percentage of
particles observed that were internalized by phagocytosis). At least 100 particles were
observed for each condition. Amine modified spheres act as a positive control and show a
high internalization percentage. HA coated backpacks show significantly lower
internalization percentage compared to both HA coated spheres and amine modified
spheres. (e) Scanning electron micrographs showing the interaction of backpacks and
spheres with macrophages after 3 hours of incubation in standard cell culture conditions:
12
(i) HA coated backpack attached to the surface of a macrophage, and (ii) three 6 pm
diameter HA coated spheres internalized within a macrophage....................................
99
Figure 7-5: Backpack attachment to macrophages does not affect cellular health and functions.
(a) Time lapse video microscopy showing that a backpack-laden macrophage can still
efficiently internalize 3 pm PS spheres (MP: Macrophage, BP: Backpack, PS-S: 3 pm
polystyrene sphere). The backpack is not internalized during the entire duration of the
time lapse sequence. (b) Differences in migration distance of macrophages with and
without backpacks 3 hours were found to be insignificant. (c) MTT assay results in terms
of proliferation potential, where 1 indicates healthy proliferation capacity and 0 indicates
com plete cell toxicity......................................................................................................
102
Figure 7-6: Controlled release of FITC-BSA from therapeutic backpacks. (a) Fabrication
schematic (b) Controlled release profile of FITC-BSA in-vitro. The normalized release of
FITC-BSA occurred over multiple hours and was not dependent on the loading
percentage of the spray emulsion used. Red and blue lines correspond to 7.5 wt% and 25
w t% , respectively............................................................................................................105
Figure 7-7: Design of cellular backpacks and confirmation of antibody presentation. (a)
Schematic of a cellular backpack. (b) TRITC-Streptavidin and mouse-IgG-biotin are
localized to the backpack and activity is shown by secondary antibody FITC-anti-mouseIgG labeling. (c) If mouse-IgG-biotin is not added to the backpack, no specific FITCanti-m ouse-IgG labeling is seen......................................................................................
116
Figure 7-8: Attachment of cellular backpacks to WEHI-265.1 monocytes and effect on monocyte
ability to differentiate into macrophages. (a) Confocal and brightfield overlay images of:
(i) a BP (green) wrapping around a stained monocyte (red) and (ii) monocytes (unstained)
attaching to an array of patterned BPs (green). (b) Brightfield images of: (i) WEHI-265.1
monocytes
in normal culture media at 48 hours, (ii)
WEHI-265.1
monocytes
differentiated into macrophages via incubation with phorbol 12-myristate 13-acetate
(PMA) treated culture media at 48 hours, (iii) FACS sorted monocytes with BPs (green)
attached prior to incubation for 48 hours with 50nM PMA, and (iv) J774 macrophages.
Scale bar = 10 pm . ..........................................................................................................
13
117
Figure 7-9: Transmigration of WEHI-265.1 monocytes and in vivo targeting. (a) Analysis of
monocyte migration (MCP- 1 initiated) through HUVEC layer for monocytes alone (black
squares), monocytes with backpacks attached (black circles) and negative control of
media alone without monocytes (black triangles). Values represent mean ± SD (n = 3).
(b) Analysis of BP migration through HUVEC layer when attached to monocytes (black)
and in free solution without monocytes present (hatched) at 24 and 48 hours. Values
represent mean ± SD (n = 3). (c) Percent of injectable dose per gram (%ID/g) 3Hbackpacks attached to monocytes in normal mice (black bars) and LPS induced lung
inflamed mice (hatched bars) at 6 hours. Values represent mean ± SEM (n = 4).
Statistical difference (P < 0.01) is seen only in lungs.....................................................
119
Figure 7-10: Preparation of Cellular Backpacks i) incubation of Cell Backpacks with streptavidin
(can be TRITC labeled) ii) Exposure to mouse IgG-biotin to dock onto streptavidin iii)
Backpacks washed and ready for attachment. ................................................................
121
Figure 7-11: Secondary Labeling to Check Activity of the Antibody on Backpack Array (no pH
sensitive release region). i) and ii) show backpacks exposed to IgG-b or not respectively.
iii) After exposure to anti-mouse IgG-FITC, if the antibody is active and present the
sample will fluoresce (iii), if not there will be no signal (iv). ........................................
121
Figure 7-12: FACS plots and approximate gating for backpack attachment to WEHI-265.1
monocytes. (i) Representative FACS output for WEHI-265.1 monocytes alone with
approximate gate. (ii) Representative FACS output for backpacks alone with approximate
gate. (iii) Representative FACS output for solution containing conjugates, unmodified
monocytes and unattached backpacks with approximate gates. (iv) Representative FACS
data of monocyte-bound BPs via FACS, showing unmodified monocytes (red),
unattached BPs (purple), and conjugates (green) showing both >50% BPs attached to
monocytes and >50% of monocytes with at least 1 BP attached at a 1:1 (cell:BP)
incubation ratio. Inset shows representative histogram as a function of fluorescent
intensity from the sam e experim ent................................................................................122
Figure 7-13: Toxicity assay for backpacks and WEHI-265.1 monocytes. 24 hour MTT assay
showing limited toxicity to monocytes incubated with BPs functionalized with IgG-b.
V alues represent m ean + SD (n = 3)...............................................................................
14
123
Figure 7-14: Schematic of Monocyte Targeting in vivo Experiment. Acute inflammation caused
by LPS will cause the monocytes to localize to the lung (left), while in the non-inflamed
case (right) m inimal accumulation will occur. ...............................................................
123
Figure 7-15: Expression of ICAM-1 following LPS induction in BALB/c mice. ICAM-1
expression in lungs 24 hours after intranasal administration of 0 pig, 5 jig, and 1 Opg LPS.
12 4
.........................................................................................................................................
Figure 7-16: Cluster Size Dependence on Patch Diameter A) Using a 7pm backpack B) Using a
1 5pm backpack. R= (# cells)/(# Backpacks) (167) ........................................................
131
Figure 7-17: Bilayers of B-cells stabilized by HA and CHI on a (HA3/CHI3) substrate A) First
layer of the cell structure and the structure is primarily green. B) The second layer of the
cellular bilayer and the structure is primarily red. Scale bar is 20 pm. ..........................
132
Figure 7-18: Inclusion of synthetic RNA into a film: A) Film containing (Poly(I:C)5/LPEI5) was
built and is stable through acetone sonication. Thickness of (Poly(I:C)5/LPEI5) 20 was
3 5 0 nm . ............................................................................................................................
13 7
Figure 7-19: 7pm Quantum Dot full BPs as excited by 543nm HeNe Laser A) Backpack
Structure (PMAA2/PVPON2)
30.5
Structure
30.5
(PMAA2/PVPON2)
Structure (PMAA2/PVPON2)
(PAH3/MNP4)10.5
(PAA4/PAH-b4)8 B) Backpack
(PAH5/QD5) 30.5 (PAA4/PAH-b4)8
30 .5 (PAH3/MNP4) 10.5
(QD5/PAH5)
30
C)
Backpack
(PAA4/PAH-b4) 8 D)
Emission of backpacks shown in A-C. Backpacks were released into PBS and excited at
543nm on a spectrofluorom eter......................................................................................
Figure 8-1: Fabrication Scheme of Orientation Controlled Microtubes.....................................
138
142
Figure 8-2: Scanning electron micrographs of cell tubes with (PAA/PAH-Rhod) 40 outside and
(HA/FITC-CHI)40 inside on the surface of a syringe filter.............................................
143
Figure 8-3: Confocal microscopy images of heterostructured tubes dried on a slide along with
cartoons of the structure. Red represents the cell-resistant rhodamine conjugated
(PAA/PAH)
multilayers and green represents the cell-adhesive FITC conjugated
(HA/CHI) multilayers. a) (PAA/PAH-Rhod)40 b) (PAA/PAH-Rhod)40 outside and
15
(HA/FITC-CHI) 40 inside c) (HA/FITC-CHI) 4 outside and (PAA/PAH-Rhod)40 inside. Full
image scale bars are 10 pm and inset scale bars are 5pm................................................
144
Figure 8-4: Controlled Orientation of Microtubes on the Surface of B-cells: a) Samples were
analyzed using confocal microscopy and microtubes with cell-adhesive (HA/CHI) on the
inside caused more end-on attachment. b) Confocal images of cell tube interactions i)
(PAA/PAH-Rhod) 40 ii) (PAA/PAH-Rhod)40 outside and (HA/FITC-CHI)40 inside iii)
(HA/FITC-CHI) 4 outside and (PAA/PAH-Rhod)40 inside. Scale bar 10 pm. c) Ratio of
'End-on' to 'Side-on' showing effect of tube structure d) Cellular viability of the B-cells
after 24 hr incubation with tubes. Viability was not altered by the addition of (HA/CHI)
outside tubes added at a 1:1 cell to tube ratio. ................................................................
146
Figure 8-5: Representative confocal images of cell-tube orientation study after 1-2 hours a)
(PAA/PAH-Rhod) 40 b) (PAA/PAH-Rhod)40 outside and (HA/FITC-CHI)40 inside c)
(HA/FITC-CHI)4 outside and (PAA/PAH-Rhod)40 inside. .............................................
147
Figure 8-6: Confocal microscopy images of heterostructured tubes dried on a slide along with
cartoons of the structure. Red represents the cell-resistant rhodamine conjugated
(PAA/PAH) multilayers and green represents the cell-adhesive FITC conjugated
(HA/CHI) multilayers. a) (PAA/PAH-Rhod)40 b) (PAA/PAH-Rhod)40 outside and
(HA/FITC-CHI)40 inside c) (HA/FITC-CHI)4 outside and (PAA/PAH-Rhod)40 inside.
Image scale bars are 1 Opm ..............................................................................................
150
Figure 11-1: PEM System studied for analysis of released free floating film chemistry ........... 165
Figure 11-2: A) Method for flipping of the film prior to XPS, pH 7.4 PBS dissolves the hydrogen
bonding region, allowing the film to be flipped. B) Cls XPS spectra of an as-deposited
(PAA3.0/PEG3.0)20.5,
(PAA3.0/PAH3.0)lo
and
flipped
(PAA3.0/PEG3.0)20.5
(PAH3.0/MNP4.0) 10 film. PAA's carboxyl Cls peak in the flipped film shifts to lower
binding energy as would be expected when paired with an amine. PEG's Cls peak also
significantly shrinks in the flipped sample, suggesting the ejection of PEG upon PAH
disrupting the PAA-PEG hydrogen bonds......................................................................
16
166
Figure 11-3: XPS spectra and atomic percentages for a flipped (PAA3.0/PEG-SH3.0)20.5
(PAH3.O/MNP4.0)
10
film. The peaks for C, N, 0, and S are as indicated - all other peaks
are due to residual salt from PBS. Inset shows the S peak at 168eV..............................
17
169
2. LIST OF TABLES
Table 4-1: Diffusion of Chitosan in a Swollen Hydrogen Bonded Film ..................................
33
Table 4-2: pH Sensitivity of the Hydrogen Bonded Region......................................................
40
Table 4-3: Detailed XPS Acquisition Parameters......................................................................
44
Table 4-4: Swelling of (PAA3/PEO3) in pH3 water as measured by wet ellipsometry ...........
45
Table 5-1: Thickness of the multilayer films used in this work. .............................................
56
Table 6-1: Atom ic Composition Ratios.....................................................................................
87
Table 6-2: IMFP information used for modeling from Tanuma et al.(1 10). Also shown is the
relationship between the Intensity (I) of each electron from different depths (d) within the
film depending on its IM FP (k).....................................................................................
88
Table 6-3: Detailed XPS Acquisition Parameters for BCP Depth Profiling ............................
90
Table 7-1: Explored Payloads in a Cellular Backpack ............................................................
94
18
3. INTRODUCTION AND BACKGROUND
3.1 Introduction
My thesis is focused around the versatile potential of structured polyelectrolyte
multilayer thin films. These easy-to-fabricate polymeric thin films can have an exceptionally
wide range of applications depending on the materials and the chosen nanostructure. The Rubner
and Cohen groups alone have researched functional films that are antibacterial(l), antifogging(2), optically active(3), drug delivery systems(4), stimuli-responsive(5) and many other
properties.
My thesis focuses on two areas. First I study the nanostructure of stratified
polyelectrolyte multilayer films primarily through the use of depth profiling X-ray Photoelectron
Spectroscopy (XPS) to fundamentally understand how thin film structure and function relate.
This method has the unique capability to determine the atomic and chemical state of a thin film
with depth with nanometer resolution without any chemical labeling or modification. Exploring
the application of this technique I probed the nanostructure of functional thin films for
applications in antifogging surfaces and solid block copolymer batteries. In the second area of
my thesis I apply insights derived from this powerful technique to effectively create functional
biomedical materials including flat anisotropic polymeric particles we call 'cellular backpacks'
and also orientation controlled microtubes. I show the how the film polymeric film nanostructure
controls the function in vitro and in vivo.
3.2 Nanostructured Thin Films
One method to fabricate nanostructured films is through the use of polyelectrolyte
multilayers (PEMs). These films are formed by dipping a surface of any shape into a solution of
19
species that adsorbs and forms a conformal layer. The advantages of this method include that it is
simple to perform, aqueous based, and can contain a diversity of materials. Previous work has
used proteins(6), small molecules(7), nanoparticles(8), natural and synthetic polymers(9, 10),
DNA(1 1), inorganic platelets(12) and many other materials as material to build functional thin
films.
The film growth can be driven by a variety of interactions including hydrogen bonding,
biological recognition, hydrophobicity and covalent interactions, but the most common method
is electrostatic interactions as first described by Decher in the 1990s. The figure below shows the
process of using electrostatic interactions to create a bilayer of polymers on a surface (13).
A
2
L
3
4II
B
1 Poyanion
3. Polycafion
2. Wash
4. WaSh
Figure 3-1: Procedure to create electrostatic Polyelectrolyte Multilayers (PEMs) A) The
surface is alternately dipped in solutions of oppositely charged polymers and rinsed in
between. B) The polymers are adsorbed and sequentially build a film. - Adapted from (13)
Depending on the choice of materials, PEMs can have a wide range of applications. For
applications in biology, the PEM technique is especially relevant since the active materials can
be deposited under physiological conditions, allowing for the incorporation of biomolecules. As
a result, PEMs are important to biomedical problems such as drug delivery, tissue engineering
and biomedical materials (14). Another application of PEMs explored is the control of water
20
condensation for applications in antifogging surfaces. Enabled by the capability to easily deposit
films using the layer-by-layer assembly method, we studied the antifogging properties of a
variety of materials ranging from hydrophilic biopolymers and hydrophobic fluorine containing
polymers.
Furthermore, by altering the materials included, the films can become stimuli responsive.
One type of stimuli responsive PEM of particular interest are hydrogen bonding films. If a film
of hydrogen bonding polymers is raised above a critical pH value, then the ionization of the
polymers will result in the film dissolving. This can be used to create free floating films which
have a variety of biomedical applications (15, 16). The nomenclature for PEM films for the rest
of this thesis follows the convention ("poly 1 "X/"poly2"Y)z, where X and Y represent the pH of
the polymer solutions used during assembly, and Z is the total number of bilayers deposited.
3.2.1
Fundamental Studies of Nanostructured Films
Functional thin films often require stratified structures with a distinct purpose for each
layer and nanometer level control over the structure(1 7). However, due to dynamic diffusion of
the species between layers, the composition and function of these layers may not be as desired.
Analysis of the diffusion of species between layers has proven difficult with other methods
(FRET, neutron reflectivity, confocal microscopy, FTIR, SIMS), but high-resolution depth
profiling X-ray Photoelectron Spectroscopy (XPS) using polymer-friendly cluster-ion C60
sputtering can detect new details. This technique has never before been applied to functional
nanostructured films, and allows for the direct determination of the composition and chemical
bonding properties of polymeric films.
21
Applying this technique to a model nanostructured film in Chapter 4, we explored
common issues in the field such as the dynamic competition between hydrogen bonding and
electrostatic interactions during fabrication, blocking interlayer diffusion, the exchange of film
components with a surrounding solution, and the kinetics of interlayer diffusion. Using high
resolution depth profiling XPS we analyzed the diffusion of chitosan, a commonly used
biopolymer, within the PEM. Furthermore, we quantified the effect of this diffusion on the film
structure and function and finally showed a method to stop this interlayer diffusion. We propose
that our results address some common issues of interlayer diffusion and could be applied to
multiple fields utilizing organic thin films.
I further applied this technique to two very distinct problems: one in surface science and
one in polymer batteries. In Chapter 5, in collaboration with Hyomin Lee of the Rubner/Cohen
Group, we analyzed the molecular profile and nanostructure of zwitter-wettable surfaces to
understand the role of polymer interlayer diffusion on antifogging and antifrosting surfaces.
Zwitter-wettable surfaces are unique surfaces with the capability to absorb molecular water
directly into the film, yet still present a hydrophobic character to macroscopic water droplets.
This uncommon combination provides the ability to overcome the limitations of current
hydrophilic anti-fogging surfaces.
In Chapter 6, I demonstrate the resolution limits of the depth profiling technique to
resolve the structure of lamellar block copolymer systems. This work is in collaboration with
Ming Luo of the Epps Group at the University of Delaware. We analyze aligned lamellar films
of polystyrene-b-poly(oxyethylene methacrylate) that have been swollen with lithium salts for
applications in solid state lithium-ion batteries. This block copolymer architecture is an exciting
type of solvent-free polymer electrolyte, since the PEO based region can solubilize the lithium
22
ions while the polystyrene can provide the mechanical rigidity needed to maintain the separation
of the anode and cathode(1 8). For this application, we are interested in whether XPS depth
profiling can resolve the distribution of lithium within the nanostructured PEO region. Previous
research has seen significant increases in conductivity with larger (higher molecular weight)
domains and it has been hypothesized that the lithium ions are concentrated in the center of the
POEM. However, the experimental proof is limited(19). Using depth profiling XPS, we believe
that better analysis capabilities may enable a thorough understanding of the phenomenon and
potentially allow for more effective polymer electrolyte design.
3.2.2
Biomedical Applications of Nanostructured Films
The second area of my thesis is focused on the design of nanostructured materials for
biological applications. In particular, I have chosen to focus on drug delivery since effective
material delivery to desired regions in the body has the potential to greatly improve human
health. Drug delivery has been enabled by designing the surface chemistry and, more recently,
the shape of particles for therapeutic targeting (20). In particular, recent research from the
Mitragotri group at UCSB clarified the effect of shape on cellular response and applied these
insights for enhanced drug delivery(21). Furthermore, recent studies have started using cellmediated drug delivery which uses the cell's natural function, such as homing to inflammation or
disease sites, to deliver the desired therapeutic materials(22). This synergistic combination has
great potential to increase the efficiency of drug delivery in the future.
In Chapter 7, I discuss the development of a uniquely shaped polymeric particle termed a
cellular backpack that conjugates strongly to the surface of immune cells through specific
interactions and has many applications in cell-mediated delivery. The technology was first
developed by Albert Swiston in the Rubner/Cohen group in 2008 and he initiated our work with
23
immune cells and UCSB. In general, backpacks are anisotropic, stratified polymeric particles
that are hundreds of nanometers thick and microns wide(23). Since a backpack leaves most of
the cell surface unaltered, the biological functions and characteristics of cells have not shown to
be greatly changed upon its conjugation with the cell surface. Additionally, Swiston noticed that
the backpacks were not being phagocytosed, and in partnership with the Mitragotri group at
UCSB, we further explored this potential. Our continuing collaboration with the Mitragotri group
has shown that the backpack is uniquely positioned for cell-mediated therapeutic delivery in
areas such as inflammation or cancer.
Chapter 8 discusses cellular microtubes as another application of nanostructured
materials for biomedical applications. The cellular microtube is built on the idea that shape,
surface chemistry and size can be designed to increase the function of a material. In particular, it
has been found that the local shape of an anisotropic particle in contact with the cell determines
the internalization rate of the particle(24). This insight drives the desire to design anisotropic
particles that orient themselves on the surface of living cells to either promote or resist
internalization depending on the desired purpose. For example, one can imagine using the
controlled orientation interactions of anisotropic particles to form stable cell-biomaterial hybrids
for cell-mediated drug delivery. Inspired by the use of chemically non-uniform Janus or patchy
particles to control the local orientation of synthetic particles in colloid systems(25), we designed
a tube-shaped, chemically non-uniform microparticle with the capability to control its orientation
on cell surfaces. This advancement opens the possibility to design new cell-biomaterial hybrids
for a variety of biomedical applications.
24
4. DEPTH
PROFILING
XPS
ANALYSIS
OF
INTERLAYER
DIFFUSION
IN
POLYELECTROLYTE MULTILAYERS
Reproduced with permission from Gilbert, J.B, Rubner, M.F, and Cohen, R.E. PNAS, 2013 110
6651-6656. Copyright 2013 National Academy of Sciences, USA.
4.1 Introduction
Layer-by-Layer assembly of polyelectrolyte multilayers (PEMs) allows for the precise
deposition of ultrathin organic films that can conformally coat features of any shape and size.
These films can incorporate a variety of species, leading to a wide range of applications
including antifogging(26), anti-reflection(27), drug delivery(14, 23, 28-30), fuel cells(12) and
responsive materials(31). Since the multilayer films are assembled through a sequential selflimiting adsorption process onto a substrate(13), a major advantage of the technique is the ability
to constrain the location of certain materials within the film at the nanoscale by simply
controlling the order of material deposition. Such nanoscale spatial control has allowed the
creation of complex periodic heterostructures not easily realized by other deposition
techniques(32). In some cases however, diffusion of the constituent macromolecular species in
and out of the film may occur, changing the desired film stratification, composition, growth
profile and properties(17). This type of diffusion, known as interlayer diffusion, is often
detrimental to the desired properties but also could be an opportunity to impart new functions.
Yet, interlayer diffusion has proven difficult to fully characterize and control and a more
thorough understanding is needed.
25
Interlayer diffusion in PEM films can be detrimental if the desired stratified
heterostructure is lost during the assembly process, a post-assembly treatment or in use. For
example, in the cases of the sequential release of therapeutics(30), structural color(33), organic
LED devices(34), solar cells(35) and on-demand release of PEM films(23, 36), the loss of
stratification due to interlayer diffusion results in the loss of the desired function. However in
other cases, such as surface planarization for the creation of higher efficiency dye-sensitized
solar cells(35) or ordering of the internal or surface arrangement of PEM films(37, 38), interlayer
diffusion can be harnessed to provide functional benefits. Interlayer diffusion can also be used to
tune material properties such as the critical dissolution pH in hydrogen bonding systems through
the addition of small amounts of electrostatic cross-links(36) or viscoelasticity through diffusion
of stiffer polymer components(39). In all of these cases, interlayer diffusion must be understood
and controlled.
Interlayer diffusion also affects the mechanism of PEM film growth. Some polymer
systems show linear growth where the bilayer thickness is invariant with deposition cycles, while
other systems show exponential growth with progressively increasing bilayer thicknesses(40). It
is widely believed that exponential growth arises from rapid interlayer diffusion of polymers
throughout the film during the fabrication steps(40, 41), however some disagree with this
conclusion(42). Improved analysis techniques that provide spatial information about the location
of specific molecules within a multilayer thin film are therefore clearly needed.
Due to the importance of understanding interlayer diffusion, a variety of techniques have
been employed to analyze it with varying degrees of success. These techniques include confocal
microscopy(43-45), fluorescence resonance energy transfer (FRET)(46, 47), FTIR(48), neutron
reflectivity(49-5 1) and X-ray reflectometry(52). Confocal microscopy is limited in spatial
26
sensitivity since films much thicker than the typical PEM thickness (500 nanometers or less) are
required due to a relatively low z-resolution(43-45). FRET is more sensitive but relies on
fluorescent modification of polymers for indirect measurements of diffusion(46, 47). FTIR can
provide valuable information on the exchange of polymers in solution with PEM film
components, but commonly probes the full thickness of the film, limiting the ability to spatially
resolve the effects of diffusion(48). Neutron reflectivity and x-ray reflectometry(49-52), require
nuclear contrast and electron density contrast respectively in the film, and commonly require the
use of special deuterated polymers. In comparison, x-ray photoelectron spectroscopy (XPS) is a
highly sensitive surface analysis method that probes the top 10 nm of a film. When combined
with sputtering or etching sources to slowly remove material between analysis cycles without
damaging underlying material, depth profiling XPS enables high resolution chemical analysis of
polymer films. The information provided by this technique could expand the understanding of
how to control PEM structure, what compositional/structural changes occur with interlayer
diffusion and when polymers in the film exchange with deposition/post-assembly solutions.
Development of less destructive sputtering or etching sources has been the enabling step
in advancing polymer depth profiling capabilities. Many depth profiling techniques use singleatom sputtering sources like argon, applicable to inorganic materials but severely damaging to
polymers(53-55). Only recently have cluster ion sources like C60 + been used in conjunction with
XPS for analysis of polymer films with depth(53, 56, 57). Cluster ion C6o0 sputtering is much
less damaging since the energy transfer from the ion to the material occurs primarily at the
surface, minimizing the chemical damage deep into the film(58). Therefore most of the damaged
material is removed from the surface, minimizing its interference with the proper analysis of the
exposed surface (59).
27
One strategy in the fabrication of functional PEM films is the use of blocking layers to
minimize interlayer diffusion. Earlier studies(30, 44, 60-62) have shown that the properties of a
successful blocking layer depend on the diffusing species under consideration and the conditions
of diffusion. Some have found that covalent cross-linking is the only way to stop interlayer
diffusion of polymers(30, 63) while others have noted that electrostatic interactions can be
utilized to stop interlayer diffusion (44, 64, 65). Of interest to this study is how blocking layers
enable the controlled production of free-floating PEM films by maintaining the desired
dissolution properties of a sacrificial region that anchors a pH stable PEM film to a substrate
surface. Once released, these free-floating assemblies have been used for tissue engineering(66)
and drug delivery(23, 29).
In this paper we designed a model system that enables the study of common attributes of
interlayer diffusion found in many PEM systems including blocking layer effectiveness. The
sacrificial component of this model PEM system is a hydrogen bonded region (poly(acrylic
acid)/poly(ethylene oxide)) that is insoluble at low pH but becomes soluble at a critical higher
pH. A pH stable PEM system based on chitosan and hyaluronic acid (HA) is assembled on top of
this sacrificial region with the goal of creating an on-demand pH triggered release of the
chitosan/HA multilayer films. A related approach has previously been used by us to create
cellular backpacks that attach to immune system cells via specific interactions between HA and
CD-44 receptors on the cell surface(23). From these earlier studies, it became apparent that the
assembly of chitosan/HA onto a hydrogen bonded region rendered the entire multilayer system
insoluble under pH conditions that should dissolve the sacrificial region. Thus, this work seeks to
determine if interlayer diffusion of chitosan/HA causes the changes in solubility and how
suitable blocking layers can prevent these changes. From a fundamental perspective, this model
28
system allows the exploration of elements such as the competition between hydrogen bonding
and electrostatic interactions in multilayers, the design of an effective blocking layer, the
exchange of film components with a surrounding solution and the extent and kinetics of
interlayer diffusion. Utilizing XPS depth profiling data acquired with C60+ cluster ion sputtering,
we find that chitosan diffuses effectively into the hydrogen bonded region of the multilayer film
and displaces the hydrogen bonded component poly(ethylene oxide). In addition, we show that
this interlayer diffusion process can be blocked completely with only a single adsorbed layer of a
polycation.
4.2 Experimental Set-up
The model PEM films examined in this work were assembled on glass slides coated with
an 80nm poly(diallyldimethylammonium chloride) (PDAC) and poly(styrene sulfonate) (SPS)
adhesion promoting multilayer. Above this adhesion layer is a hydrogen bonded film comprised
of poly(acrylic acid) (PAA) and poly(ethylene oxide) (PEO). The solubility of this hydrogen
bonded PEM system is pH sensitive and above a pH of 3.6, it will dissolve in water(67). Figure
4-1 shows a cartoon representation of the multilayer heterostructures examined in this work
along with the experimentally determined thicknesses of the various regions of the multilayer. In
the top case (Figure 4-1A), the experiments involved immersing a hydrogen bonded multilayer
film in a chitosan solution for varying amounts of time. In the bottom case (Figure 4-1B),
blocking layers containing varying numbers of poly(allylamine hydrochloride) (PAH) and SPS
layers were deposited on top of the hydrogen bonded region followed by the assembly of a
hyaluronic acid (HA) and chitosan (CHI) multilayer.
29
A
Chitosan Diffusion Test
Thickness
Exposed to pH 3 CHI solution forX min-+
Hydrogen Bonded Region: (PAA3/PE03)
Adhesion Layer: (PDAC4/SPS4)
B
-475nm
-+
>
~80nm
Thickness
Blocking Layer Test
(HA3/CHI3)-+
Electrostatic Blocking Layer (PAH3/SPS3)->
Hydrogen Bonded Region: (PAA3/PE3)-+
~20nm
0-1 Onm
-300nm
Adhesion Layer (PDAC4/SPS4)->
~80nm
Figure 4-1: Schematic of systems used to test (A) chitosan diffusion into the hydrogen
bonded region and (B) electrostatic blocking layer effectiveness. The number after the
polymer abbreviation is the deposition solution pH.
These stratified films were dried and then analyzed using depth profiling XPS paired with
C6o sputtering to collect CIs, Ols, N Is and Si2p high resolution spectra. It is important to note
that prolonged X-ray exposure and C6o0 sputtering can alter the chemical composition of PEMs
and decrease the interface resolution (56, 68). As described in Fig. S1, the choice of XPS data
acquisition parameters and sputtering conditions is very important since long periods of X-ray
exposure reduced the 0 to C ratio and in particular the signal of the carboxyl peak at
-289eV(69). As a result we chose acquisition parameters and C60 + sputtering conditions to
minimize the total X-ray exposure time while still obtaining an acceptable resolution and signal
to noise ratio at each point in the depth profile.
4.3 Diffusion of Chitosan in Hydrogen Bonded Multilayers
To explore the question of whether the adsorbed chitosan diffuses into the swollen
hydrogen bonded region and by how much, the nitrogen signal from the amine on chitosan was
analyzed as a function of depth for hydrogen bonded multilayers exposed to a 0.1% (w/v)
chitosan solution at pH 3 for a specified amount of time. The chitosan solution acted as an
infinite supply for diffusion of chitosan into the hydrogen bonded region. All samples were
30
rinsed with water for four minutes and dried with nitrogen gas prior to analysis. The compiled
spectra for chitosan exposure times of 1, 3, 10 and 60 minutes (CHI 1, CH13, CHI 10 and CH160)
are plotted in Figure 4-2A-D. Color was added to highlight the approximate locations of the
distinct regions of the PEM film, using the same color scheme shown in Figure 4-IA. The depth
of the (red) chitosan region was determined by analyzing the intensity of the N Is signal with
depth. When the Ni s signal dropped to background levels the spectrum was colored yellow to
denote the hydrogen bonded region. Finally, the (black) adhesion layer starts when the Nis
signal increases at the base of the film due to the presence of nitrogen-containing PDAC.
A
3000
1minCHI
B
3ooo
3min CHI
E
m
20002000
4
3.5
_
-
-
3mm~ CHI
S31
800
400
C 3000 C~~~~
404 402 400 3W6
Binding Energy (*V)
10min
lommnCHI
-
400
Thickness
D
20W0
4
above glass (nm)
402
Binding Energy
CHI
-lmin
-00
Thickness above gas (nm)
no CHI
1in CH
60mmn CHI
25
(eV)
2
oi ~
8OinH
CHIao
3oo
00i
20001
8
ThFig
ness ab
.
0
gless (nm)
.
820
1
na e
y
(*v)
0
Thickness above glass (nm)
402 400
Binding Enery (eV)
00
400
200
0
Thickness above glass (nm)
Figure 4-2: Diffusion of chitosan into hydrogen bonded films. Spectra of hydrogen bonded
(PAA3/PEO3) films exposed to chitosan solution for different amounts of time. (A) 1 min
exposure (B) 3 min exposure (C) 10 min exposure (D) 60 min exposure to chitosan. The
color scheme is the same as that of Figure IA. Red spectra represent chitosan infused areas,
yellow spectra represent the hydrogen bonded (PAA3/PEO3) area and black spectra
represent the (PDAC4/SPS4) adhesion layer. (E) Quantification of A-D to determine the
atomic concentration of nitrogen with depth in the film. Data points are individual dots and
line shown is the result of a Savitzky-Golay five point quadratic algorithm.
The spectra from Figure 4-2 were analyzed to determine the atomic percentage of
nitrogen with depth as seen in Figure 4-2E. A film not exposed to chitosan is shown in Figure
4-7. The concentration of nitrogen in the multilayer film increased systematically with time of
exposure to the chitosan solution. Also, the maximum depth at which an appreciable nitrogen
31
signal was observed increased with time. Separate experiments showed that the PDAC from the
adhesion layers does not enter the hydrogen bonded region during the assembly process even
after many hours at pH 3 (Figure 4-7). Thus, the only source of nitrogen in the film, above the
80nm adhesion layer, is from the chitosan that diffused from the top of the hydrogen bonded
region. For the samples CHI 1, CHI3, and CHI 10 the location of the diffusion front, (where NI s
concentration is 50% of the maximum value), advanced 181, 238 and 299 nm respectively as
measured in dry films. Since the final dry thickness remained relatively constant, independent of
chitosan diffusion depth, the diffusion of chitosan does not greatly expand or collapse the film.
As a result, the dry diffusion distance of chitosan directly correlates with the thickness of the
portion of the (PAA3/PEO3) film that was altered by chitosan diffusion. Therefore, to estimate
the diffusion coefficient, the chitosan penetration distances were multiplied by a factor of 2.5 to
account for the 250% swelling of the (PAA3/PEO3) film in pH 3 water (Table 4-1 and Table
4-4). After 60 minutes in the chitosan solution (CHI60), the chitosan has diffused through the
entire hydrogen bonded region as seen by the uniformly high nitrogen content throughout the
film in Figure 4-2E. These results clearly show that the adsorbed chitosan diffuses into the
hydrogen bonded region. Chitosan is known to be highly diffusive due to a lower charge density
than typical polyamines like PAH and the presence of multiple hydrogen bonding acceptors(45,
46).
32
Table 4-1: Diffusion of Chitosan in a Swollen Hydrogen Bonded Film
Sample
CHI 1 min
CHI 3 min
CHI 10 min
t (sec)
360
720
900
Dry film
distance (nm)
181
238
299
Swollen film
distance (nm)
452
594
748
2
D (cm /sec)
1.42E-12
1.23E-12
1.56E-12
Dry distance multiplied by 2.5 to account for 250% film swelling at pH 3
To estimate the diffusion coefficient of chitosan into the hydrogen bonded region we
used the data in Table 4-1 and the characteristic diffusion length, L= 4L
(70). As seen in
Table 4-1 the calculated diffusion coefficient is consistent for the three time points sampled and
is -1.4* 1012 cm 2 /sec. Recent reports on interlayer diffusion coefficients in polyelectrolyte
multilayers range from 1020 cm 2 /sec for SPS in linearly growing (PAH/SPS) films(49) to 10-7
cm 2 /sec for poly(L-lysine) in exponentially growing poly(L-lysine)/hyaluronic acid films(40).
This wide range of reported interlayer diffusion coefficients is due to a fundamental difference in
the film growth mechanism between linearly and exponentially growing films. In linear growth
conditions, the deposited polymers generally interact only with the top surface and thus generally
have interlayer diffusion coefficients below 10-' 7 cm 2/sec(49-5 1). In comparison, exponentially
growing systems require some amount of interlayer diffusion to occur during the dipping
cycle(40) and as a result have higher reported interlayer diffusion coefficients in the range of 10
16 Cm2/sec
to 10-7
Cm
2
/sec depending on the conditions and polyelectrolytes used(40, 46). A
recent paper by Lundin et al. (46) used fluorescence resonance energy transfer (FRET) and
showed that the interlayer diffusion of chitosan in exponentially growing films made of chitosan
and heparin was ~10~"
cm 2 /sec for 150kDa chitosan. Our reported interlayer diffusion
coefficient of about 10-2 cm 2 /sec for chitosan of roughly the same molecular weight is larger,
33
but well within the range of other exponentially growing polymer systems previously studied. In
addition, the diffusion coefficient we report would be higher than the diffusion coefficient of
chitosan in a pure film of chitosan/heparin since Xu et al. showed that weaker matrix interactions
enable a higher diffusion coefficient(50). Given that in our study chitosan is diffusing in a
hydrogen bonded (PAA3/PEO3) matrix with a low interaction strength as measured by
dissolution pH and is without internal structure as measured by neutron reflectivity(71) it would
thus allow for a higher diffusion coefficient than the more strongly interacting electrostatic
matrix of chitosan and heparin.
4.4 Displacement of PEO from Film by Chitosan
One possible effect of interlayer diffusion in PEMs is the displacement of materials from
the film into solution(48, 60, 72). In our model system, interlayer diffusion of chitosan into the
hydrogen bonded region changes the dominant interaction from hydrogen bonding between PAA
and PEO to electrostatic interactions between PAA and chitosan. FTIR confirmed these new
electrostatic interactions. In films with large amounts of chitosan diffusion, the PAA ionization
level increased due to the titration of carboxylic acid groups to carboxylate groups by cationic
chitosan (Figure 4-8)(9). Therefore upon chitosan diffusion, electrostatic interactions between
chitosan and PAA displace the weaker hydrogen bonding interactions between PEO and PAA.
As a result, the hydrogen bonding acceptor PEO is no longer associated with the film and can
diffuse out.
Enabled by the high sensitivity of XPS, the displacement of PEO can be directly
explored. Figure 4-3 shows the high resolution CIs data from the CHI 1 and CHI3 samples using
the color scheme from Figure 4-1A (CHI10 and CHI60 data in Figure 4-9). The red region is
infiltrated with chitosan and its location in this region was determined by analyzing the intensity
34
of the Ni s signal with depth as in Figure 4-2. In the remaining depth of the film, the Ni s signal
is very low (<0.5% atomic conc. N), signifying that in this yellow region, the film contains little
to no chitosan. To analyze the displacement of PEO from the film, the Cl s spectra from the (red)
chitosan infiltrated regions were compared to the spectra in the (yellow) hydrogen bonded
regions of the film. In particular we focus on the change in Cl s signal intensity at 286.5eV. Both
PEO and chitosan have a signal at this point, but since the extent of chitosan diffusion can be
independently determined by the nitrogen signal, the changes in Cls spectra can be used to
analyze the displacement of PEO from the film. The Cl s spectra of pure PEO, chitosan and PAA
can be seen in Figure 4-10.
A
C
1min CHI
4000
3500
S3000
PAA3/PEO3
2000-
3000
S1000
-
CHM1
-
CHI3
CH110
2500 4a
Thickness above glass (nm)
2ce
B
282
290 288 286
Binding Energy (eV)
3min CHI
-
CHI60
2000
1500
4000
1000
3000
2000-
500
1000
0
200 400
Thickness above glass (nm)
290288288 284282
Binding Energy (eV)
290 288 286 284
Binding Energy (eV)
Figure 4-3: High resolution Cis XPS depth profiling of a hydrogen bonded film exposed to
chitosan solution. (A) 1 min (B) 3 min exposure to chitosan. The color scheme is the same as
that of Figure 4-1A. Red spectra represent chitosan infused areas, yellow spectra represent
the hydrogen bonded (PAA3/PEO3) areas and black spectra represent the (PDAC4/SPS4)
adhesion layer. Comparing the red chitosan diffused areas to the yellow hydrogen bonded
areas the chitosan diffused areas have a lower signal at 286.5eV since PEO has diffused out.
(C) Comparison of Cis spectra with different chitosan exposure times to the initial yellow
35
hydrogen bonded area. The longer exposure to chitosan solution, the more the PEO signal
at 286.5eV decreases. All chitosan exposed spectra were from ~450 nm above the glass
surface.
As shown in Figure 4-3 A and B, the red regions where chitosan has diffused into the
film, have a markedly lower signal at 286.5eV than the yellow hydrogen bonded region. The
change in the signal intensity at 286.5eV is highlighted in Figure 4-3 C which compares the
chitosan infiltrated regions from Figure 4-3 A and B to the (PAA3/PEO3) hydrogen bonded
region. The spectra of all chitosan exposed samples were obtained from ~450nm above the glass
surface to minimize differences due to X-ray exposure time or C6o0 sputtering time. The decrease
in signal intensity at 286.5eV is a result of chitosan diffusion displacing PEO and allowing it to
diffuse out of the film. Since chitosan also has a peak at 286.5eV, if PEO was not diffusing out
of the film, this signal would increase. Figure 4-3 C also shows that the decrease in the signal at
286.5eV correlates with the holding time in chitosan solution, which is consistent with the
diffusion of PEO out of the film.
4.5 Electrostatic Blocking Layer Stops Chitosan Diffusion
In many cases it is desirable to stop interlayer diffusion to maintain distinct functional
regions of a multilayer heterostructure. The effect of electrostatic blocking layers on the
diffusion of chitosan into the hydrogen bonded region was investigated using the film
architecture shown in Figure 4-1 B. Unlike the previous study, above the hydrogen bonded
region there is an electrostatic blocking layer that varies from a single layer of PAH (<mnm) to
9.5 bilayers of (PAH3/SPS3) (10nm). On top of the blocking region, the final region is a 20nm
(HA3/CHI3) multilayer film. To determine the location of the distinct regions of the PEM film
and apply the color scheme shown in Fig. 1 B, information from the C Is and N Is spectra was
combined. For example, in Figure 4-4A the transition from the red HA/CHI region at the surface
to the green (PAH3/SPS3) blocking region was determined by the change in shape of the Cis
36
spectrum. The shape change is a result of more carbon-carbon bonds at 285eV and fewer carbonoxygen bonds at 286.5eV in the green region compared to the red region. Since XPS analyzes
approximately the top 1Onm and the blocking layers are less than 1Onm, the green spectrum
representing the blocking layers likely contains signal from an adjacent region as well. Yet the
C Is spectra of the blocking layers remain distinct from the red and yellow C 1s spectra. The end
of the green region was determined by the drop in N Is signal and thus the yellow hydrogen
bonded region begins. Finally, the black adhesion layer starts when the NIs signal increases at
the base of the film due to the presence of nitrogen-containing PDAC. These data reveal that
each of the various regions of the multilayer heterostructure illustrated in Figure 4-1B can be
identified in XPS depth profile spectra.
A
B
Cls
Blocking Layer
-
4000.
--
2000
Thickness
(PAH3)
(PAH3/SPS3)3.5
(PAH3/SPS3)9.5
none
<1nm
4nm
lonm
4.5
0
4
100 200
300
Thickness above glass (nm)
Nis
2
284 22
290288
Binding Energy (eV)
z 3.53.
e2.5
Nis
3000,
2
2000
2 1.5
E
1000
0.5
0
100200
Thickness above glass (nm)
A
400 300
Binding Energy (eV)
400 300 200 100
0
Thickness above glass (nm)
Figure 4-4: Effect of a blocking layer on interlayer diffusion of chitosan (A) Cis and Nis
regions from depth profiling XPS of a hydrogen bonded sample with a (PAH3/SPS3) 3 5
blocking layer topped with (HA3/CH13) 3 5 . The color scheme is the same as that of Figure
4-1B. Red spectra represent (HA3/CHI3), green spectra represent the (PAH3/SPS3)
37
electrostatic blocking layer, yellow spectra represent the (PAA3/PEO3) hydrogen bonded
region and black spectra represent the (PDAC4/SPS4) adhesion layer. (B) Quantification of
the nitrogen signal for different blocking layer systems tested. Data points are individual
dots and the line is the result of a Savitzky-Golay five point smoothing algorithm. All films
had (HA3/CHI3) 3 5 deposited on top of the blocking layer.
The Nis spectra from depth profiling samples with blocking layers were analyzed to
determine the atomic percentage of nitrogen with depth as seen in Figure 4-4B (Ni s spectra from
all samples with blocking layers seen in Figure 4-11). From Figure 4-4B it is clear that all three
electrostatic blocking layers tested (PAH3/SPS3)z (z=0.5, 3.5, 9.5) with approximate thicknesses
of <1nm, 4nm and I0nm respectively, effectively stop the diffusion of chitosan into the hydrogen
bonded region as seen by the absence of any detectable nitrogen signal in the bulk of the film.
Remarkably, even though the films spent over 30 minutes in chitosan solution during the
(HA/CHI) film fabrication, even a single adsorbed layer of PAH was sufficient to block its
diffusion into the hydrogen bonded region. At pH 3, PAH is a fully charged polycation (pKa
~8)(9) so it has a strong electrostatic interaction with the ionizable polyanion PAA found at the
top of the hydrogen bonded region(73). Due to the strong electrostatic interaction, the PAH
blocking layer is kinetically trapped at the top of the film (48) and effectively stops the diffusion
of chitosan into the hydrogen bonded region under the conditions employed in this study. Recent
literature (30, 44, 60, 63, 64) has shown that different blocking layers work well for different
polymer systems and annealing conditions. In some cases covalently cross-linked blocking layers
are needed(30, 63), but in other cases electrostatic blocking layers can stop interlayer diffusion as
well(44, 64). In the conditions tested in this work, a single electrostatic blocking layer is all that
was needed to block the interlayer diffusion of chitosan.
In the absence of a blocking layer, chitosan from the (HA3/CHI3) multilayer film diffuses
into the entire hydrogen bonded film during fabrication, producing a high nitrogen signal
38
throughout as seen in Figure 4-4B. Even though the film has fewer total layers deposited than the
multilayer films with blocking layers, the interlayer diffusion of chitosan allows for a large
increase in the final film thickness as shown by the leftmost data point in Figure 4-4B. This large
increase in thickness is characteristic of exponential growth which is caused by the diffusion of
polymers and polymer pairs in and out of the film during deposition(40). As a result we expect
that chitosan from not only the first bilayer deposition, but from the subsequent depositions to be
present throughout the film. Similar chitosan diffusion into the hydrogen bonded region is also
observed when the order of polymer deposition is switched from hyaluronic acid first
(HA3/CHI3) to chitosan first (CHI3/HA3) as revealed in Figure 4-12. As a result of
macromolecules like chitosan diffusing throughout a film, hydrogen bonded films could be used
as scaffolds for easily loading of drugs or other macromolecules of interest(14, 74).
An interesting question to ask concerning the use of depth profiling XPS with C60+
sputtering is what level of vertical resolution is possible and is it sufficient to probe PEM
heterostructure interfaces? The interface used to analyze this resolution was the sharp boundary
between the (PAH3/SPS3) blocking layers and the hydrogen bonded region seen in Figure 4-4B.
The precipitous drop of nitrogen signal from the blocking layer to the hydrogen bonded region
occurs between two data points or approximately 15nm. Reducing the sputtered thickness
between successive XPS spectra would allow for the interface resolution to increase to
approximately 1 Onm since this is commonly the depth of analysis during an XPS cycle.
However, near this limit of resolution, extended sputtering time may cause radiation induced
diffusion and surface roughening which need to be considered during experimental design(68).
39
4.6 pH Sensitivity of the Hydrogen Bonded Region
The purpose of a blocking layer is to maintain the distinct properties of each region of a
multilayer film. In our model system the as-assembled (PAA3/PEO3) hydrogen bonded region
dissolves above pH 3.6(67). If the film is altered by sufficient chitosan interlayer diffusion, the
multilayer becomes insoluble at neutral pH. Therefore to test whether the desired properties of
the hydrogen bonded region can be maintained through the use of blocking layers, all films
characterized previously were exposed to a buffered salt solution of PBS pH 7.4 for 30 minutes.
After 30 minutes the residual dry film thickness was compared to the initial dry film thickness to
determine if the hydrogen bonded film dissolved. As see in Table 4-2, all multilayer films with a
blocking layer dissolved in PBS. Therefore even a single layer of PAH is able to effectively
block chitosan diffusion and maintain the pH sensitive solubility of the hydrogen bonded region.
Table 4-2: pH Sensitivity of the Hydrogen Bonded Region
Figure 4-4
Figure 4-2
Top layer
Blocking Layer
Dissolve in PBS pH 7.4?
(HA3/CHI3) 3 5
(PAH3)
(HA3/CHI3) 3 5
(HA3/CHI3) 3 5
(HA3/CHI3) 3 5
(PAH3/SPS3)
(PAH3/SPS3) 9 5
none
Yes
Yes
Yes
No
CHI 60
CHI 10
CHI 3
none
none
none
No
No
No
1
none
No
CHI
After exposure to pH 7.4 PBS for 30 min, the remaining thickness was compared to the initial
thickness to determine if the hydrogen bonded region dissolved
In the absence of a blocking layer, the multilayer films no longer dissolved due to pH
stable electrostatic crosslinks formed between the diffused chitosan and PAA. In some cases like
CHI1, CHI3 and CHI10 this result was unexpected since chitosan had not diffused all the way
through the film (Figure 4-2E) before PBS exposure. However, depth profiling of CHI10 after
40
PBS exposure shows that chitosan has diffused throughout the film and stabilized it to pH
changes (Figure 4-13). Therefore, in the absence of a blocking layer, chitosan interlayer diffusion
was stopped by drying for analysis but continues after exposure to PBS solutions. Since
hydrogen bonded PEMs require a minute or two to dissolve(75), this brief time allows for further
chitosan diffusion, rendering the film insoluble in PBS.
4.7 Conclusions
XPS with C6o0 cluster ion sputtering is a powerful technique for analyzing the atomic
composition and chemical state of organic nanostructured films. With the correct choice of
conditions to minimize sample damage it can directly determine to within 15 nm the location of
polymers through the thickness of a film, allowing for analysis of interlayer diffusion; in addition
to testing the efficacy of various blocking layers. Using our model system we have shown that
chitosan is highly diffusive with an interlayer diffusion coefficient around 1.4* 10-12 cm 2 /sec in
hydrated hydrogen bonded (PAA3/PE03) films. Also, the high resolution capabilities of XPS
show the displacement of hydrogen bonded PEO in favor of electrostatic interactions between
chitosan and PAA. Finally, various thicknesses of PAH-containing blocking layers were
explored, including a single layer of PAH that can stop the diffusion of chitosan into the
hydrogen bonded region.
We believe that the ability to control and measure the interlayer diffusion in PEMs will
have applications in a variety of areas. For example, exchanging hydrogen bonding interactions
for electrostatic may be important for loading of biological molecules through post assembly
modification of films. Furthermore, the design of films capable of blocking macromolecule
diffusion is relevant to structured films, sequential drug release and production of free-floating
PEM films.
41
4.8 Materials and Methods
Materials: poly(acrylic acid) (PAA, Aldrich, M=450kDa), poly(allylamine hydrochloride)
(PAH,
Aldrich,
M=15kDa),
poly(ethylene
glycol)
(PEO,
PolySciences,
M=20kDa),
poly(diallyldimethylammonium chloride) (PDAC, Aldrich, M=200-35OkDa in 20% aqueous
solution), poly(styrene sulfonate) (SPS, Aldrich, M=70kDa), hyaluronic acid (HA, from
Streptococcus equip, Fluka, M~1580kDa), acetic acid (Sigma) and low MW chitosan (CHI,
deacetylation 0.9, Sigma, M= 50-190kDa) were used as received. The nomenclature for PEMs
follows (polylX/poly2X)z, where X is the pH of the polymer solutions and z is the number of
bilayers deposited (1 bilayer
=
polyl+poly2). A non-integer value of z indicates the assembly
was terminated with polyl.
PDAC
SPS
PAH
PEO
PAA
HCI
NH 2
SO3'
CHI
HA
OH
[_
Na
0
Hc~3
N H 4H
H
0O
NH2 "
L
H
H
-
x deacetylated
Figure 4-5: Chemical structure of the polymers in this study
Multilayer Film Deposition: Polymers solutions were made from Milli-Q 18.2MQ water.
Solutions of PAA, PDAC and SPS were 0.01M and solutions of PEO, HA and CHI were 0.1%
(w/v). CHI solutions included 0.1M of acetic acid to aid dissolution. PDAC and SPS solutions
for the adhesion layer had 0.1M NaCl at pH 4.0. All other solution pHs were adjusted to pH 3.0
with 1M HCl and no added salt. Glass substrates were sequentially dipped in the polymer
solutions using an automated Zeiss programmable slide stainer or nanoStrata dipping unit.
42
Substrates were held in polymer solutions for 10 minutes and then rinsed for a total of 3 minutes
in water with mild agitation. The time in chitosan solution was altered for diffusion studies, but
the rinse cycle employed the same time profiles. Dry film thickness was measured with a P-16
profiler (KLA Tencor Corp.). Layer-by-Layer assembly was executed using the Zeiss
programmable slide stainer apparatus for all of the regions except for the HA/CHI regions, which
were built in the nanoStrata unit. The dipping time for polymers in the Zeiss apparatus was 10
minutes, followed by two pH 3 Milli-Q water rinses for 2 and 1 minutes with mild agitation. In
the nanoStrata unit, polymer depositions were done for 10 minutes, followed by three pH3 Milli-
Q water rinses:
one for 2 minutes and two for 1 minute, while the substrate was rotating at about
100rpm. To minimize the changes that could arise from drying we dried all samples in Figure
4-2 and Figure 4-3 with nitrogen immediately after the aqueous assembly was complete. All
samples in Figure 4-4 were dried passively at ambient conditions. Time from Table 4-1 started
when the sample entered solution and ended when the sample was dried with N2 .
X-ray Photoelectron Spectroscopy (XPS):
Chemical composition of the surface was
characterized using a PHI Versaprobe II X-ray photoelectron spectrometer with a scanning
monochromated Al source (1486.6 eV, 50 W, spot size 200 um). The takeoff angle between the
sample surface and analyzer was 450 and the X-ray beam collected Cls, Nls, Ols and Si2p
elemental information while rastering over a 200 pm x 700 pm area. Detailed XPS acquisition
parameters are found in Table 4-3. Depth profiling was accomplished using the instrument's C60+
ion source operated at 10 kV, 10 nA and rastered over a 3x3 mm area at an angle 700 to the
surface normal. Sputtering occurred in 1 minute intervals while the sample was moved using
concentric Zalar rotation at 1 rpm. Atomic composition was determined based on photoelectron
peak areas and the relative sensitivity factors provided in PHI's Multipak processing software.
43
All data were background subtracted, smoothed using a five point quadratic Savitzky-Golay (SG) algorithm and charge-corrected so that the carbon-carbon bond has a binding energy of 285.0
eV. The surface of the glass was defined as the point at which the atomic concentration of silicon
reached 5% in the depth profiling data. The thickness as measured by profilometry was
compared to the number of sputter cycles that occurred before reaching the surface of the glass.
Data was plotted using Matlab.
Table 4-3: Detailed XPS Acquisition Parameters
Element
Binding Energy Range
Pass Energy
Data spacing (eV/step)
Sweeps
CIs
Nis
Ols
Si2p
280-290
393-404
526-537
95-107
11.75
117.4
117.4
117.4
0.200
0.125
0.500
1.000
3
6
1
1
(eV)
(eV)
pH Sensitivity of Hydrogen Bonded Region: Films were tested by a 30 minute immersion in
PBS (pH 7.4) with 100rpm agitation. After drying with N2 , the thickness was measured using
profilometry and was compared to the initial thickness to determine if the hydrogen bonded
region dissolved.
Film details listed in order of fabrication (substrate to exposed surface).
Fig. lA Chitosan Diffusion test: (PDAC4/SPS4) 15.5 (PAA3/PEO3)
30.5
(CHI3 exposed for
designated time)
Fig. lB Electrostatic Blocking Layers test: (PDAC4/SPS4) 15.5 (PAA3/PEO3) 30 .s (PAH3/SPS3)z
(z=0,0.5,3.5,9.5) (HA3/CHI3) 3.5
As others have found, the thickness/bilayer of (PAA3/PEO3) region can vary between batches
(67, 76). As expected in the hydrogen bonded regions, the molar ratio of PAA to PEO in
44
(PAA3/PEO3) based on the qualitative peak intensity is approximately 1:1 matching literature
(75)
4.9 Supplemental Information
Table 4-4: Swelling of (PAA3/PEO3) in pH3 water as measured by wet ellipsometry
(PAA3.0/PEO3 .0)20.5
Film
(PAA3.0/PE03.0) .
16 5
(PAA3.0/PE03.0)
(
2 0 .5
exposed to CHI 60 min
Dry Thickness (nm)
Wet Thickness (nm)
Percentage Swelling
123.48 3.5
302.56 9.63
245.03 ± 4.26%
224.07 ± 0.95
580.00 ± 4.43
258.85 ± 0.87%
222.19 1.52
536.01 10.41
241.24 4.97
Swelling(%) = wet. thickness x100
dry _ thickness
For swelling experiments in pH3 water, a custom-built quartz cell was used to measure film
thickness using variable-angle spectroscopic ellipsometry as described in a previous publication
(4). All ellipsometry measurements were made using a Woolham Co. ellipsometer operating at a
70* angle of incidence. Measurements were made from 300 to I000nm, and all data analysis was
done using the WVASE32 software. The refractive index of the pH 3 water was assumed to be
the same as water (1.333).
45
15-
condition 1
condition 2
condition 3
condition 4
condition 5
o
0
0
C 10 -
.
C
8
5
0
0
0
50
100
X-ray exposure time (min)
150
Figure 4-6: Increasing X-ray Exposure Time Decreases Carboxyl Atomic Concentration:
To study the effect of X-ray exposure and sputtering conditions identical (PAA3/PEO3) 30.5
(PAH3/SPS3) 9.5 samples were exposed to a variety of X-ray exposure levels while depth
profiling. Data was analyzed with Multipak software (Physical Electronics, USA) and then
the carboxyl peak at 289eV was fitted using CasaXPS software. Different XPS data
acquisition parameters were tested along with different C60 + cluster ion sputtering
conditions to determine conditions with minimal decrease in carboxyl signal, high signal to
noise ratio and many data points while depth profiling. The carboxyl signal was chosen to
determine X-ray damage since it is highly sensitive to degradation. It was determined that
the condition 5 shown above provided a good balance of these demands and the final
acquisition parameters are found in Table 4-3. The degradation of the film due to extended
X-ray exposure or sputtering is a limitation that should be carefully considered by others
before using this technique. C6 0 deposition was not observed since the C-C bond at 285eV
did not systematically increase.
46
A
B
z
3000-
4
2000-
3
2
10001
11
E
Zo0
4 40 2 404O40 40
Thickness above glass (nm)
Binding Energy (WV)
I
400
300
200
100
0
Thickness above glass (nm)
Figure 4-7: XPS Depth profile in absence of chitosan. (A) N1s signal from a (PAA3/PEO3)
film before exposure to chitosan. The only N1s signal is from the (PDAC4/SPS4) adhesion
layer. (B) PDAC from (PDAC4/SPS4) adhesion layer does not diffuse into the hydrogen
bonded region (PAA3/PEO3) on top of it even after 15 hours in pH 3 water as shown by the
absence of nitrogen signal above the adhesion layer.
I
-
(PAA3'PFO3)
.
(PAA3
....
PE03) CHI 10 minutes
=1
6I-
--I----------
...
1
1900
1800
1600
11700
1500
4
140 0
Wavenumber(cm-1)
Figure 4-8: GATR-FTIR results showing ionization of PAA upon chitosan diffusion. Two
distinct peaks from the carboxylic acid functional group of PAA are shown: one at ~1550
cm-1 is associated with the asymmetric stretching of the ionized carboxylate (COO-) and
the other at -1710 cm~1 which is associated with the C=O stretching of the carboxylic acid
(COOH) groups. In comparison to the pure hydrogen bonded sample (PAA3/PEO3), the
sample exposed to chitosan for 10 minutes has a higher amount of ionized carboxylate
groups.
47
A
B
10min CHI
3000-
60min CHI
3000-
]1 2000.
2000
8 1000
1000
Ti 2kns 4o
Thickness above glass (nm)
20288
4 282
2
400
2
Thickness above glass (nm)
Binding Energy (eV)
g2290288
6284
Binding Energy (eV)
Figure 4-9: High resolution Cis XPS depth profiling of a hydrogen bonded film exposed to
chitosan solution. (A) 10 min and (B) 60 min exposure to chitosan. The color scheme is the
same as that of Figure 4-1A. Red spectra represent chitosan infused areas, yellow
spectra represent the hydrogen bonded (PAA3/PEO3) areas and black spectra represent
the (PDAC4/SPS4) adhesion layer.
A
Poly(ethylene oxide)
B
Chitosan
F'
I \
!\
10
19.
/
2
2.4
21
2
fiei
A
JIM
291
***
Figure 4-10: Pure component Cis XPS spectra. (A) Poly(ethylene oxide) (B) Chitosan (C)
Poly(acrylic acid)
48
A
B
No blocking layer
(PAH3/SPS3),
5
layer
3000-
2000
C
lS 1000,
00
Thickness above glass (nm)
C
404 42Binding Energy (eV)
Binding Energy (eV)
D
(PAH3/SPS3) 35 layer
Single (PAH3) layer
3000.
20001
2000]
0
1000]
C.)
0
1000
01]
98
398
300 4
Thickness above glass (nm)
398
Thickness above glass (nm)
3000.
8C
0
100
398
404 402
Binding Energy (eV)
402
404 Binding Energy (eV)
Thickness above glass (nm)
Figure 4-11: Nis spectra from depth profiling XPS of hydrogen bonded films with different
blocking layers. All samples were topped with (HA3/CHI3) 3.5 . (A) No blocking Layer (B)
(PAH3/SPS3)9 .5 layer (C) (PAH3/SPS3) 3 .s layer (D) Single (PAH3) layer
49
4.5
4
z 3.5
0
3
2.5
Z
E
0
2
1.5
1
0.5F
900
800
700
600
500 400
300
Thickness above glass (nm)
200
100
0
Figure 4-12: (CHI3/HA3) 3 .5 on top of a hydrogen bonded region. Independent of whether
CHI or HA is deposited first onto the hydrogen bonded region, the chitosan diffuses fully
through the film as seen by the nitrogen signal throughout the film.
4.5
CH110 after PBS
CH110 before PBS
4
z 3.51
0
3
2.51-
8
E
2
1.5 I-
1
0.5
Al
600
---- ----- -
500
400
300
200
Thickness above glass (nm)
100
0
Figure 4-13: Comparison of CHI10 nitrogen signal before and after PBS exposure. CHINO
sample did not dissolve after a 30 minute PBS soak. It was stabilized due to the diffusion
and electrostatic cross-linking of chitosan as seen by high levels of nitrogen throughout the
film.
50
5. DESIGNING ZWITTER-WETTABLE SURFACES
Given the capabilities of depth profiling XPS we could potentially explore a range of
nanostructured polymeric films. One application of interest to the Rubner/Cohen group was
developing a full understanding of the important characteristics in the design of 'zwitterwettable'
films. Developed by fellow labmate Hyomin Lee, these films use a unique
nanostructured system with a thin hydrophobic capping layer and a thick hydrophilic reservoir.
This system presents a hydrophobic surface to macroscopic water droplets, yet allows molecular
water to diffuse through to the reservoir and thus creates an anti-fogging surface. This chapter
represents a joint effort with equal contributions by Hyomin Lee and me to probe the molecular
structure and source of the 'zwitter-wettable' phenomenon. My experimental addition to the
project focused on the depth profiling analysis and creation of patterned films for controlling
condensation. The following paper is currently in preparation for publication.
Authors: Hyomin Lee, a, Jonathan B. Gilbert, a, Francesco E. Angile,b Rong Yang,a Daeyeon
Lee, Michael F. Rubner ' and Robert E. Cohen"*
aDepartment
of Chemical Engineering, Massachusetts Institute of Technology, Cambridge
02139
bDepartment
of Chemical and Biomolecular Engineering, University of Pennsylvania,
Philadelphia,
Department of Materials Science and Engineering, Massachusetts Institute of Technology,
Cambridge
51
5.1 Introduction
Control of water condensation from vapor on a surface is critical for a broad range of
applications including heat-transfer,(77, 78) fog-harvesting,(79) and high-performance optical
devices.(2, 80, 81) As a result, significant attention has been devoted to study the effect of
surface chemistry, and roughness on water condensation over the past few decades. Many
interesting new phenomenon such as jumping-droplets during coalescence,(82) and lowvolatility liquid impregnated surfaces with enhanced heat transfer(78) has been discovered and
expanded potential applications by engineering surfaces with optimal surface roughness and
surface wetting characteristics.
For example, in the area of controlling water condensation for optically clear antifogging
coatings, smooth hydrophilic(2) or nano-textured superhydrophilic coatings(8, 83, 84) were
found to significantly reduce light scattering by facilitating film-like condensation.(2, 85) While
many of these coatings perform satisfactorily in specifically defined condensing environments,
these hydrophilic surfaces (i) may exhibit frost formation or excess and non-uniform water
condensation under aggressive fogging conditions, which results in poor optical performance of
the coating,(80) and (ii) may have reduced durability due to fouling by contaminants.(86)
To overcome the limitations of the currently accepted hydrophilic antifogging method
new insights into the design of antifogging materials are needed. Recent developments by us(80)
and others(87, 88) have presented possible solutions by reporting surfaces that can be both
antifogging and hydrophobic. These counter-intuitive surfaces consist of polymeric networks of
hydrophilic as well as hydrophobic moieties that can imbibe a considerable amount of water
molecules from the vapor phase directly into the hydrogel-like film while the surface maintains a
hydrophobic appearance. We recently defined these surfaces as zwitter-wettable surfaces,(80)
52
due to the capability to simultaneously absorb molecular water from the environment while
exhibiting hydrophobic character to water droplets. It was also shown that incorporating
hydrophilic poly(ethylene glycol)(PEG) segments resulted in antifogging coatings with frostresisting capabilities. The presence of PEG increased the nonfreezing water capacity of the thin
film due to its strong hydrogen bonding capability.(80) While the major focus of these studies
was to achieve antifogging coatings that perform well in very harsh conditions (below the
freezing point of water), the detailed explanation correlating the abnormally high water contact
angles (60-70' (87), or even as high as 1100 (80) ) and the antifogging performance was rather
limited. Furthermore, the experimental complexity of these individual systems restricted their
thorough characterization which limited our understanding of the underlying mechanism.
In our previous work, we have shown that smooth hydrophilic antifogging surfaces result
in a sheet of water on the surface and thus additional droplets form an effective water contact
angle of 00. Yet with zwitter-wettable surfaces, a water droplet on top of a film containing
condensed water still exhibited a water contact angle above 90, indicating that the water
molecules in the vapor phase were directly imbibing into the film. Therefore, we hypothesized
that in addition to roughness and surface chemistry, a third factor, the capability of the film to
transport molecular water to the underlying reservoir must be considered in the design of
antifogging coatings. Thus, it was anticipated that a hydrophilic reservoir capped with a thin
hydrophobic layer with high molecular water permeability could limit the nucleation and growth
of water droplets on the surface during condensation and thus produce a hydrophobic and
antifogging surface.
53
5.2 Results and Discussion
a
b
RO
OR
O
0
NHa
i
*0
OR
J
Chitosan (CHI) (+)
n
120
100
R = II or C1 2 C0l2I
0)
Carboxymethyl cellulose
w
(CMC) (-)
C
0
60
C
U
40
so
U
-i
L-
20
(CH-2 )2
U,
Nafion (-)
0
9
CFF
So
w
o!
)
d
poly(perfluorodecyl acrylate)
(PPFDA)
C
Hydrophilic surface
d
Zwitter-wettable
PPFDA coated
surface
hydrophilic surface
Hydrophilic surface ((CHI/CMC)3)
6
a2
t Water molecule
a
Air
M
A,
V
a Condensed water layer
Swollen reservoir
4 ((CHI/CMC)0)
0
0 -20'
e
PPFDA only
(CHI/Naflon) 0
Zwitter-wettable surface ((CHI/CMC),(CHl/Nafion),)
4
a.
6
C
02
2
Hydrophobic capping layer
((CHI/Naflon)3 )
Swollen reservoir
((CHI/CMC), 0 )
(0
0
esv-105.
Figure 5-1: (a) Structures of the polymers described in this paper. Polycations and
polyanions are labeled with parenthesis (+), (-), respectively. (b) Water advancing contact
angle of the samples tested. (c) Photographs taken immediately after transfer to ambient
lab conditions (22 ± 1 *C, 40 ± 10% RH) from a -I*C refrigerator (1 hr). Only the
hydrophilic surface and zwitter-wettable surface resisted fog formation. (d) Schematic
54
representation of hydrophilic surface with sessile drop and during condensation. (e)
Schematic representation of zwitter-wettable surface with sessile drop and during
condensation.
To test this hypothesis, the layer-by-layer (LbL) self-assembly technique was chosen to
assemble various coatings with heterostructured architecture.(89) For the base platform,
hydrophilic polysaccharides (chitosan (CHI) as the polycation (+) and carboxymethyl cellulose
(CMC) as the polyanion (-) as shown in Figure 5-la) were LbL assembled electrostatically to
create a hydrophilic antifogging coating as reported earlier.(2) As shown in Figure 5-lb, 30bilayers of CHI/CMC on a glass substrate has a water advancing contact angle of -20'
. Then
the surface wetting characteristics of this 30- bilayer hydrophilic antifogging surface was altered
to become hydrophobic with either 3- bilayers of (CHI/Nafion) by LbL technique or chemical
vapor deposited (CVD) with poly(perfluorodecyl acrylate) (PPFDA). 3- bilayers of CHI/Nafion
increased the water advancing contact angle up to ~1100 and the PPFDA coating enhanced the
water advancing contact angle to -125. The number of 3- bilayers was chosen because fewer
bilayers would exhibit transient contact angle behavior due to surface rearrangement(90) which
complicates the analysis. All samples chosen for our analysis do not exhibit significant contact
angle changes with time (Supporting Information Figure 5-5).
As shown in Figure 5-1c, photographs were taken immediately after transfer to ambient
lab conditions (22 ± 1 'C, 40 ± 10% RH) from a -1 'C refrigerator (1 hr) and only the hydrophilic
surface and the CHI/Nafion coated hydrophilic surface resisted fog formation. It has been
reported earlier(2) that the hydrophilic CHI/CMC coating remains optically clear during
condensation due to their superior water-absorbing characteristics that form a uniform, non-lightscattering film of water on the surface as shown in Figure 5-1d. However, when this same
55
surface was subsequently coated with hydrophobic CHI/Nafion, this coating exhibited a zwitterwettable behavior where water molecules in the vapor phase directly imbibed into the film, while
the surface remains hydrophobic to water droplets as shown in Figure 5-le. Yet, when the same
hydrophilic surface was coated with PPFDA, the film exhibited fog formation. As revealed in
Table 5-1, film thickness change after subsequent assembly are similar in its extent (69 nm and
62 nm, respectively) and the water advancing contact angles are both well above 900. Two
control samples where (CHI/Nafion) 30 and PPFDA are solely coated on a glass substrate also
exhibited fog formation even though the surface wetting characteristics resemble the zwitterwettable (CHI/Nafion) 3 coated hydrophilic surface. These results clearly indicate that the surface
wetting characteristic does not solely determine whether a surface will fog and that the presence
of the underlying hydrophilic "reservoir" is indeed critical.
Table 5-1: Thickness of the multilayer films used in this work.
(CHI/CMC) 30
(CHI/CMC) 30
(CHI/Nafion) 3
(CHI/CMC) 30
PPFDA
(CHI/Nafion) 30
PPFDA
615± 16 nm
684 ± 5 nm
677 ± nm
18± 3 nm
72± 2 nm
(+ 69 m)a
(+ 62 nm)a
afilm thickness change after assembling a capping layer.
56
a
b
I Photomask, UV cure
su
t
Iut
1
'I
Glass substrate
,'
t e
Enlarged
SEM
Image
I
Hydrophilic coating deposition (CHI/CMC)
lasssu
Average
C Signal
e
Dissolve Photoresist
I
Gl1ass ubstrate
I Hydrophobic coating deposition (CHi/Naflon)
Average
IF Signal
Glasbstra
C
0
0
.
0
d
u.J
C',
Increase in water vapor pressure
Figure 5-2: (a) Schematic diagram of the polymer patch fabrication process. (b) Scanning
electron microscopy (SEM) image of the polymer patch. Scale bar is 50 pm. Outlet image
shows the enlarged SEM image and the average carbon and fluorine signal obtained from
energy dispersive X-ray spectrometry (EDS). (c) Optical microscopy image of the polymer
patch after exposure to humid air. (d) ESEM images of the condensation of water vapor on
a polymer patch. Water droplets start to form on the off-patch areas and at the edges of the
patches.
Previously, Varanasi et al., demonstrated that spatial control in the condensation of water
molecules from the vapor phase can be achieved by patterning a surface with significant contrast
in local intrinsic wettability to preferentially guide water molecules to nucleate and grow on the
57
hydrophilic regions.(91) To demonstrate the unique capability of our zwitter-wettable films to
spatially control condensation of water molecules from the vapor phase, patterned films with
zwitter-wettable characteristics were fabricated (Figure 5-2a). First, patches of the hydrophilic
(CHI/CMC) reservoir (10- bilayers) were fabricated using lift-off photolithography.(23, 29) 10bilayers were chosen because higher numbers of bilayers would result in incomplete patterning
during the fabrication process. Then, the hydrophobic thin film (CHI/Nafion) (3- bilayers) was
deposited, resulting in a surface with uniform (CHI/Nafion) over both the small patches of
(CHI/CMC) as well as the off-patch areas. The uniform fluorine EDS signal (Figure 5-2b)
confirms the consistent coverage of the Nafion, while the non-uniform carbon signal shows the
presence of the underlying hydrophilic reservoir ((CHI/CMC)) in the patch area. When the
patterned film is exposed to progressively higher water vapor pressure (Figure 5-2c and Figure
5-2d) the water condensation is spatially controlled. Initially the water condenses in the
hydrophilic reservoir of the patch, changing the refractive index and thus the color of the patch,
however minimal droplets are seen on the surface (Figure 5-2c). As the water vapor pressure
increases, progressively larger droplets start to form on the interstitial fogging areas, while no
droplets are seen on the patch areas with zwitter-wettable characteristics. The capability to
spatially control the condensation of water through patterning of the hydrophilic reservoir, while
still presenting a uniform hydrophobic surface is indeed unique to zwitter-wettable surfaces.
In addition, contact angle measurements were performed on the (CHI/Nafion) 3 coated
hydrophilic reservoir using both high and low surface tension nonpolar liquids, diiodomethane
(YLV = 50. 8 mN/m) and hexadecane(yLV
=
27.5 mN/m) to further investigate the uniqueness of
this zwitter-wettable surface. These results are shown in Supporting Information Figure 5-6.
Both liquids exhibited reasonably high advancing contact angles with low contact angle
58
hysteresis indicating that surface exhibits a Teflon-like behavior to nonpolar liquids. Unlike
typical hydrophilic surfaces where low surface tension liquids would completely spread, organic
contaminants could be removed relatively easily due to the weak retention force offered by the
hydrophobic capping layer.
a
x'_s
60
50
40
099
.5
30
0
w
+4-N
15
20
-is
10
t
w
(CHI/Nafion)3 coated
hydrophilic reservoir
*S120
600
700
500
400
300
200
0
100
Film Thickness (nm)
Sample preparation
b
A
0.8
Atomic Conc. Profile
d
C
CHI
1.0
0
Depth Profiling XPS
1.0
CMC
Nafion
-+-
C
CHI
10
+CMC
-0 - PPFDA
Nation
0.8
0.8
CHI
CMC
N
0.8
0.6
0.
0.4
4
0
0
0.2
0.2
* *
0.2
0.0
n
600
500
400
300
200
100
0
700
800
500
400
300
200
100
0
0.0
700
.....
600
500
400
300
200
100
Film Thickness (nm)
Film Thickness (nm)
Film Thickness (nm)
Hydrophilic reservoir
(CHI/Nafion),c ated
hydrophilic res ervoir
PPFDA coated
hydrophilic reservoir
0
Figure 5-3: (a) Schematic of depth-profiling x-ray photoelectron spectroscopy (XPS) used
to acquire atomic concentration profile and molar ratio profile of various samples tested.
Here, atomic concentration of (CHI/Nafion) 3 coated hydrophilic surface is shown as an
example. (b) Molar ratio of hydrophilic surface ((CHI/CMC) 30 ) with depth in the film. (c)
Molar ratio of (CHI/Nafion) 3 coated hydrophilic surface with depth in the film. (d) Molar
ratio of PPFDA coated hydrophilic surface with depth in the film.
To investigate the origin of the (CHI/Nafion) 3 coated hydrophilic reserviors favoring
direct imbibing of water molecules into the film, the axial chemical composition analysis was
performed. C60 cluster-ion depth profiling X-ray Photoelectron Spectropscopy (XPS) was used to
59
study the chemical 1 CHI:CMC molar ratio in the hydrophilic reservoir and that this ratio is
effectively maintained composition with depth on various coatings with heterostructured
architecture. Previously our group showed that this technique can provide atomic concentration
information as well as the chemical state of polymer thin films with a resolution around
15nm(92). This capability was only recently enabled by the development of cluster-ion C60
etching which is much less damaging to polymers than the single-ion etching methods
commonly used.(93) As shown in Figure 5-3a the depth profiling process uses iterative C6o0
etching and XPS acquisition to acquire an atomic concentration profile with depth. This atomic
concentration with depth can be further analyzed using a mass balance to determine the molar
repeat ratio of the various polymers with depth, allowing for the analysis of polymer interlayer
diffusion. From Figure 5-3b it is clear that there is a 2:when the reservoir is coated with either
the (CHI/Nafion) 3 (Figure 5-3c) or the PPFDA film (Figure 5-3d). Due to the lack of interlayer
diffusion, it is clear that the different antifogging performance can be attributed to the
characteristics of the thin capping layer alone.
60
a
b
0.25
250
0.20
200
~150
0.15
0.10-
100
-F1
c
C)
Low water dffusivity
cpiglayer
a A
High water diffusivity
Hydrophilic
rasrvoir
(CH Nafion) 3 coakid
hydropflic euervoIr
Hydrophilic
re rvoir
(CHi/Nafion) 3 coated
hydrophilc r-ervoir
ko'~"'
Opul
Figure 5-4: (a) Diffusivity (D) of water vapor in hydrophilic reservoir and (CHI/Nafion) 3
coated hydrophilic reservoir (b) Solubility of water vapor in hydrophilic reservoir and
(CHI/Nafion) 3 coated hydrophilic reservoir. (c) Schematic representation of the effect of
the ability of thin hydrophobic capping layer to transport water molecules: For low
diffusivity capping layer, water molecules nucleate and grow on the surface while for high
diffusivity capping layer, water molecules in the vapor phase prefer to directly imbibe into
the underlying hydrophilic reservoir.
With the insight gained from the depth-profiling XPS study, the ability of thin
hydrophobic capping layer to transport water molecules were investigated by permeability
measurements using quartz crystal microbalance with dissipation monitoring (QCM-D). It has
been previously demonstrated that QCM-D enables real-time measurements of water vapor
uptake by a thin film with high precision and provides quantitative information on water
diffusivity (D), and solubility (S).(94) Briefly, quartz crystals coated with various films are
introduced into a chamber where the changes in the frequency and dissipation are monitored
upon exposure to a steam of humidified nitrogen and fitted using the Voigt viscoelastic model to
predict the increase in thickness under humidfied nitrogen. Then, the diffusivity (D) of water
vapor in the film is determined by using Fickian diffusion model fitting and the solubility (S) by
calculating the change in mass between dry nitrogen and wet nitrogen. Detailed method of
measurement and assumptions used for one-dimensional diffusion of water molecules can be
found in the Supporting Information. As shown in Figure 5-4a and Figure 5-4b, the diffusivity
(D), solubility (S) were measured for the following samples: (1)
61
(CHI/CMC) 30 hydrophilic
reservoir and (2) (CHI/Nafion) 3 coated hydrophilic reservoir. Comparing the hydrophilic
reservoir alone to the hydrophilic reservoir coated with (CHI/Nafion) 3, both the diffusivity (D)
and the solubility (S) were similar in their values. However, a hydrophobic capping layer that
reduces the capability of the film to transport water molecules would thus favor nucleation on the
surface which is indeed the case for PPDFA coated hydrophilic reservoir. This phenomenon is
schematically shown in the top of Figure 5-4c. Since the diffusivity and solubility of the
hydrophilic reservoir did not change when coated with the (CHI/Nafion) capping layer this
indicates that a (CHI/Nafion) capping layer does not hinder water transport to the underlying
hydrophilic reservoir. However, the (CHI/Nafion) layer without a reservoir does not have
sufficient capacity to absorb water and eventually fogs after being saturated with water during
condensation, agreeing with the antifogging test results in Figure 5-1c.
Previously, it has been reported(91) that a hydrophilic surface (6 -25'
) nucleates tens of
orders of magnitude faster than a hydrophobic surface (6 ~1100 ) at a typical saturation ratio of
p/pin=1.7. This allows water nucleation and subsequent growth to preferentially occur on more
hydrophilic regions. In our zwitter-wettable surface, on the other hand, the surface is
hydrophobic while highly diffusive to water molecules. Thus, the antifogging capability of the
surface is further enabled since surface hydrophobicity retards nucleation and growth, while the
CHI/Nafion capping layer allows facile diffusion of water molecules to the hydrophilic reservoir.
This unique combination allows a zwitter-wettable surface to be hydrophobic and also
antifogging.
Simple but careful design of this model system allowed the previous complex network of
hydrophilic and hydrophobic moieties to be decoupled into a heterostructured film consisting of
62
hydrophilic reservoir and a hydrophobic capping layer. While surface chemistry and roughness
were the main factors for controlling water condensation until now, a third factor, the capability
of the film to transport water molecules needs to be considered. Furthermore, no surface
rearrangement in our model zwitter-wettable surface allowed the film to exhibit truly
hydrophobic behavior to water droplets unlike the previously studied zwitter-wettable system(80,
95) where the film exhibited a hydrophobic to hydrophilic transition when contacted with water.
Enabled by the experimental techniques to determine diffusivity (D) and solubility (S) via
QCM-D, the design considerations discussed in this work can be further expanded to fabricate
optimal hydrophilic reservoirs and hydrophobic capping layers depending on the application
requirements. Furthermore earlier observations such as a critical film thickness requirement for
hydrophilic antifogging coatings(2) can be clarified in detail with respect to diffusivity (D) and
solubility (S). In addition, optimal material selection for a specific condensation environment,
governed by initial substrate temperature, and exposing environment (relative humidity, and
temperature), can be further fine-tuned allowing additional physicochemical properties such as
enhanced mechanical durability or frost-resisting capabilities to be taken into account while
designing future zwitter-wettable surfaces.
In summary, we showed that when a thin hydrophobic layer with sufficient permeability
is assembled on top of a hydrophilic reservoir, antifogging coatings with zwitter-wettable
characteristics can be prepared. Enabled by depth profiling XPS and QCM-D, the required film
properties were detailed for the design of future zwitter-wettable surfaces. Future work
incorporating new surface chemistry and roughness may further extend zwitter-wettable surfaces
to enable the fabrication superhydrophobic antifogging surfaces. Additionally we believe further
63
applications of directed condensation of molecules from the vapor phase include water
separation and purification.
5.3 Methods
Layer-by-Layer assembly of the polymer thin films
Sequential adsorption of polymer layers were performed using a StratoSequence VI spin dipper
(nanoStrata Inc.), controlled by StratoSmart v6.2 software, at 80 rpm. The concentrations of
chitosan (CHI, low molecular weight), carboxymethylcellulose (CMC, Mw = 250 000 g/mole)
and Nafion dispersion (Alfa Aesar, 5% (w/w) in water and 1-propanol) in the dipping solutions
were 1 mg/mL, 1 mg/mL and 0.25 % (w/w) respectively. For chitosan, 0.3 % (v/v) acetic acid
was added prior to dissolving the polymer and was filtered with 200 pm pore filter (VWR) after
stirring overnight. Distilled water (>18 M Q em, Millipore Milli-QTM) water (DI water) was used
in formulating the solution and in all rinsing procedures. The dipping time in the CHI, CMC, and
Nafion solutions were each 10 min followed by three sequential rinse steps (of 2, 1, and 1 min).
All solutions and their respective rinse solutions were adjusted to pH 4.0 with either NaOH or
HCl respectively. Glass substrates were first degreased by sonication in a 4% (v/v) solution of
Micro-90 cleaner (International Products Co.) for 15 min, subsequently sonicated twice in DI
water for 15 min. The substrates were blow-dried with dry air and treated for 2 min with oxygen
plasma (PDC-32G, Harrick Scientific Products, Inc.) at 150 mTorr before the LbL assembly.
Contact angle measurements
Transient contact angle measurements as well as advancing and receding contact angles of probe
liquids on various samples were performed using a Rame-Hart model 590 goniometer, by
dispensing liquid droplets of volume V
-
10 ptL.
64
Preparationofpoly(perfluorodecyl diacrylate) (PPFDA) coating using initiatedchemical vapor
deposition (iCVD)
iCVD polymerization of perfluorodecyl diacrylate (PFDA) (97%, Aldrich) was conducted
similar as previously described.(77) Briefly, the peroxide initiator, TBPO (98%, Aldrich), was
delivered into the reactor through a mass flow controller (MKS Instruments) at a constant flow
rate of 3.2 sccm. PFDA was vaporized in a glass jar that was heated to 80 'C. The flow rates
were controlled using needle valves and kept constant at 0.2 and 0.6 sccm. The filaments were
resistively heated to 230 'C using a DC power supply (Sorensen), and the temperature was
measured by a K-type thermocouple (Omega Engineering). The sample stage was backcooled at
30 'C using a recirculating chiller/heater (Neslab RTE-7). The working pressure was maintained
at 200 mTorr using a throttle valve (MKS Instruments). The reactor was covered with a quartz
top (2.5 cm) that allows in-situ thickness monitoring by interferometry with a 633 nm HeNe laser
source (JDS Uniphase). Final thickness of the polymer deposited on the hydrophilic surface
corresponds to ~70 nm.
Permeabilitymeasurements using QCM-D
Gold-coated QCM-sensors (Q-Sense Inc.) are cleaned with UV/ozone treatment for 10 min, then
immersed in 2% (v/v) Sodium Deodecyl Sulfate (SDS) for 30 min in room temperature, rinsed
with DI water, and finally blown dry with compressed air. Also, another UV/ozone treatment
was done for 10 min before the baseline measurement. The baseline frequency and dissipation
are first recorded with a blank crystal under dry nitrogen and calibrated to zero for the sorption
measurements. All the multilayer films are then formed on the crystals through the similar
procedure used for glass substrate. Here, gold crystals were immobilized on a plastic microscope
65
slide (P11011 P) in between two grooves made with a razor blade. The film on the back side of
crystal is carefully removed using 1.0 M NaOH with a cotton swab and then rinsed thoroughly
with DI water. After the multilayer coated sensors are loaded in the QCM chamber (E4 QCM-D
unit (Q-Sense Inc.)), the frequency and dissipation shifts are monitored at varying humidity
conditions.
The measured frequency and dissipation are fitted using the Voigt viscoelastic model
incorporated in Q-Sense analysis software (QTools) to predict the increase in film thickness
under wet nitrogen (~80% RH) due to the absorption and diffusion of water in the film similar to
previously described procedure.(94) Average density of chitosan, carboxymethyl cellulose and
Nafion was used to determine the film thickness from QCM-D measurements.
The diffusivity (D) of water vapor in the film is determined from the plot of the thickness
change versus time under wet nitrogen (-80% RH) using a numerical model for one-dimensional
diffusion of water into a thin film as shown previously.(94)
mt
mo
00
-
-n
1
[-D(2n + 1) 2 7r2 t
8
n=OI
(2n +1)
2
r
2
exp
,where m., m, and mo are the mass of the thin film and water at time
1
=
o, t, and 0, respectively. t
and 1 are the time and thickness of the film, respectively. Since the film mass is linearly
proportional to the thickness, above equation can be simplified to,
1 n(m
we2
- mt)ir2
8(m. - m)
D
a2 v
where the slope can be used to determine the diffusivity (D). The solubility (S) of water vapor in
66
the film is simply determined by calculating the change in the mass of the film between under
dry nitrogen and wet nitrogen, m. -m.
The water permeability (P) can be subsequently be
determined by the following relationship: P = D x S, the effect of water adsorption on blank
QCM crystals have been corrected as reported earlier.(94)
For this measurement, a humidified stream is introduced at 10 cm 3 (STP) min', with the
corresponding mass transfer analog Biot number to be in the order of 104, indicating that the
water-vapor transport process is diffusion-limited. It is known that the lateral diffusion is much
faster than the normal diffusion through the film(96) and thus our measurement will be
dominated by one-dimensional diffusion of water normal to the film and the film swelling occurs
dominantly in one dimension.
Scanning electron microscopy (SEM) measurements
Scanning electron microscopy micrographs were obtained using a JEOL 6010LA in back
scattered electron shadow image mode (BES) operated at an accelerating voltage of 10 kV.
Energy dispersive X-ray spectrometry (EDS) was performed as well to perform elemental
analysis of the Zwitter-wettable patch sample. Environmental scanning electron microscope
(ESEM, Zeiss Evo 55) was used to investigate the nucleation and condensation phenomena in
situ.
Depth ProfilingX-ray PhotoelectronSpectroscopy
Chemical composition of the surface was characterized using a PHI Versaprobe II X-ray
photoelectron spectrometer with a scanning monochromated Al source (1486.6 eV, 50W, spot
size 200 um). Depth profiling was accomplished using the instrument's C60 + ion source. The
67
takeoff angle between the sample surface and analyzer was 450 and the X-ray beam collected
CIs, Ols, Fls, NIs and Si2p elemental information while rastering over a 200 prm x 350 pm
area. Detailed XPS acquisition parameters are in the table below. Sputtering occurred in 1 minute
intervals while the sample was moved using concentric Zalar rotation at 1 rpm. The C60+ source
was operated at 10 kV and 10 nA and rastered over a 3x3 mm area at an angle 70' to the surface
normal. Atomic composition was determined based on photoelectron peak areas and the relative
sensitivity factors provided in PHI's Multipak processing software. All data were background
subtracted, smoothed using a five point quadratic Savitzky-Golay (S-G) algorithm and chargecorrected so that the carbon-carbon bond has a binding energy of 285.0 eV. The surface of the
glass substrate was defined as the point at which the atomic concentration of silicon reached 5%
in the depth profiling data. The molar ratio profile was calculated via an oxygen mass balance.
To minimize experimental bias, pure films of chitosan, CMC, Nafion and PPFDA were analyzed
via Depth Profiling XPS as the basis for the effective molecular repeats.
Table S1. Detailed XPS Depth Profiling Conditions
Element
Binding Energy (eV)
PassEnergy (eV)
Data Spacing (eV)
Sweeps
CIs
Ols
Si2p
FIs
NIs
280-298
526-538
95-107
682-695
393-405
117.4
117.4
117.4
117.4
117.4
0.5
0.5
1.0
0.5
0.5
2
2
1
2
2
Zwitter-wettable Patch Fabrication
Patch fabrication was based on a photolithographic lift off procedure. Briefly, the slides were
cleaned with 2% (v/v) solution of Micro-90 cleaner (International Products Co.), 1 M NaOH and
68
then rinsed with water. Next, the photoresist was spin coated onto the surface and developed as
previously described.(23) After deposition of the uniform hydrophilic layer via layer by layer
processing, the film was patterned by dissolving the remaining photoresist in sonicating acetone
for 10 seconds, leaving behind patches of hydrophilic (CHI/CMC). Finally the hydrophobic
(CHI/Nafion) film was deposited uniformly over the surface via layer by layer processing.
5.4 Supporting Information
a
b
120
.
.
.
.
.
.
.
.
10
100
0~
120 -
0100-
80 -
3
-(CHCMC)
0-
-(CHI/CMC)
0
30
(CHI/Naflon) 1
-
---+
+
60 -
(CHINaflon) 2
(CHI/Nafion) 3
0
0
-(CHI/Nafon),
-PPFDA
only
20
20 -.
100
30
(CHIlNafIon) 3
-- + PPFDA
0
200
300
400
500
600
0
Time (sec)
100
200
300
400
500
600
Time (sec)
Figure 5-5: (a) Water contact angle evolution over time for (CHI/CMC) 30 subsequently
coated with 1-, 2-, and 3- bilayers of (CHI/Nafion) by LbL technique. (b) Water contact
angle evolution over time for various samples tested in this work.
The linear decrease in water contact angle can be attributed to evaporation of water from
the static drop over the elapsed time of the experiment.(95)
69
120
,
100 -'
C"D3
80 --
60-
400
4200
Water
Dilodomethane Hexadecane
Probing Liquid
Figure 5-6: Advancing and receding contact angle of water, diiodomethane,
hexadecane on (CHI/Nafion) 3 coated hydrophilic reservoir.
70
and
6.
ION DISTRIBUTION IN BLOCK COPOLYMER ELECTROLYTE THIN FILMS
An additional application of depth profiling XPS is the analysis of the nanostructure in
block copolymer films. In collaboration with Ming Luo of the Epps Group at the University of
Delaware, I have studied the distribution of lithium-ions in aligned lamellar films. A better
understanding of the ion distribution in block copolymer films is important for the design of
solid polymer battery systems. My work is focused on the depth profiling characterization and
analysis while Ming focused on film fabrication and other characterization methods. This chapter
represents the joint effort and equal contributions of Ming Luo and me. The following paper is
currently in preparation for publication.
Authors Jonathan B. Gilbert, t Ming Luo, 1 Wei-Fan Kuan, 1 Cameron Shelton, 1 Thomas H.
Epps, III,* Michael F. Rubner,if Robert E. Cohen*t
T Department
of
Chemical
Engineering,
Massachusetts
Institute
of
Technology,
Cambridge, Massachusetts 02139, United States
1 Department of Chemical & Biomolecular
Engineering, University of Delaware, Newark,
Delaware 19716, United States
11Department of Materials Science and Engineering, Massachusetts Institute of Technology,
Cambridge, Massachusetts 02139, United States
71
6.1 Introduction
Over the past several decades, battery technology has been primarily driven by the need
for longer operating time, longer cycle life, smaller size, lighter weight, and improved safety and
stability. Lithium-ion batteries, a commonly used rechargeable system, have several desirable
features including a high energy density, a long lifespan, and minimum memory effects.(97, 98)
However, the electrolytes in most commercialized lithium-ion batteries are organic solvents
which are vulnerable to fires under mechanical deformation or overvoltage conditions.
Significant efforts have focused on designing solid-state lithium-ion batteries with a solvent-free
electrolyte.(99, 100) Amorphous poly(ethylene oxide) (PEO)-based materials doped with alkalimetal salts are favored candidates and have sufficient ionic conductivities for practical use (10-
4
S/cm).(101) These organic solvent free electrolyte systems enable the construction of lightweight batteries and are adaptable to continuous roll-to-roll processing which potentially lowers
the cost of manufacture. Yet, poor resistance to mechanical stress hampers their large-scale
production and commercialization.
Block copolymers (BCP) with microphase separated domain structures are an attractive
avenue to achieve the mechanical strength required for commercial use while maintaining the
high ionic conductivities needed. The BCP systems accomplish this by using one domain to
conduct ions while the other provides rigidity. Since conductivity and mechanical strength are
decoupled it becomes possible to design battery systems with appealing mechanical properties
and ion conductivities. One of the most extensively studied systems are Poly(styrene-blockethylene oxide) (PS-PEO) copolymers and several groups have achieved the desired mechanical
properties (shear modulus on the order of 108 Pa at 90 'C)(102) as well as ionic conductivity
levels similar to the corresponding homopolymer electrolytes. (103,
72
104) However, low
conductivity at room temperature due to PEO crystallization has limited the application of these
PS-PEO electrolytes. Comb-branched
polymers with short PEO side chains such as
poly(oligo(oxyethylene) methacrylate) (POEM) have been demonstrated to effectively eliminate
PEO crystallization and exhibit high ionic conductivity at room temperature.(105,
106)
Furthermore given the new design parameter of the PEO side chain length, the conductivity and
crystallization characteristics of the block copolymer could be further improved.
To optimize the solid BCP electrolyte, the effect of molecular weight on ionic
conductivity and mechanical properties needs to be clearly understood. It is known that the
stiffness of polymers increases with increasing molecular weight. However, the effect of
molecular weight on the ionic conductivity is more complex. Yuan et al. demonstrated that the
ionic conductivity of symmetric PS-PEO blended with lithium bis(trifluoromethanesulfone)imide
(LiTFSI) slightly decreased with increasing molecular weight of the PEO block at low molecular
weights (MPEO<10 kg/mol).(107) Additionally, Panday et al. showed an order of magnitude
increase in conductivity as the molecular weight of the PEO block (MPEO) increased from 7 to 98
kg/mol.(104) They ascribed the initial decrease in conductivity to hindered PEO segmental
motion near the interfacial zone between PS and PEO blocks. This caused the ionic conductivity
of the PEO to be affected by the glass transition temperature of PS block. As the molecular
weight of the PEO block increased above 10 kg/mol, the width of the conducting PEO channel
increases and the fraction of ions affected by the interfacial zone became negligible. This effect
was further corroborated by Gomez et al. who presented the first direct imaging of ions in a BCP
bulk using energy-filter transmission electron microscopy (EFTEM).(19) They showed that with
increasing molecular weight the lithium salt is progressively more localized near the middle of
73
the PEO lamellae. They also developed a theory that the increasing ionic conductivity can be
attributed to increased chain end and ion mobility in the middle of the PEO channels.
Experimentally quantifying the ion distribution is important to elucidate the relationship
between ionic conductivity and BCP domain structures. EFTEM is powerful; however
challenges remain due to the relatively small scattering cross section of lithium ions, tedious
sample preparation and cautious image analysis, especially for thin film systems. A new
potential method to analyze the ion distribution in BCP films is via cluster-ion depth profiling Xray Photoelectron Spectroscopy (XPS). This method recently showed its effectiveness in
resolving both chemical and atomic information with depth in nanostructured polymeric
films(92). The development of cluster-ion etching sources that induce much less damage than
single-ion sources has been essential to the accurate depth profiling of polymeric materials.
Single-ion sources like argon-ion cause severe damage in the film and change the chemical
composition. In comparison, cluster-ion sources like C60 + result in much less damage since the
energy transfer from the ion to the material occurs primarily at the surface, minimizing the
damage deep into the film.
In this work we studied a lamellar forming PS-POEM thin film system with lithium
trifluoromethanesulfonate (CF 3SO 3Li). Though the proven room temperature conductivity of
these PEO grafted BCP is very exciting, the lithium-ion distribution in the POEM domain has
not been explored and it is unclear whether a similar lithium concentration effect would be seen.
To accomplish this we analyzed the structure of the neat film and the lithium salt doped films
with 6:1 and 12:1 PEO monomer to lithium ions using atomic force microscopy (AFM) and Xray reflectivity (XRR). Furthermore we found that a new technique to analyze nanostructured
polymer films, XPS depth profiling with C6o0 sputtering, can resolve the chemical and atomic
74
structure of nanometer level domains found in block copolymers. Through this analysis we
confirmed the presence of the lithium salt in the POEM region and found that the lithium
distribution is directly correlated to the POEM concentration. Furthermore we find that depth
profiling XPS with C6o0 sputtering is readily applicable to analyze the nanostructure of block
copolymers. However when applying this technique one must account for the large area being
analyzed and roughness induced etching that may alter the results.
6.2 Results and Discussion
The PS-POEM BCP used in this study had an overall molecular weight of 36 kg/mol
(PS20k-POEM1 6k, determined by size exclusion chromatography using polystyrene as standards),
a polydispersity index of 1.09, and a bulk domain spacing (Lo) of 26.1 nm (determined from
primary peak in small angle X-ray scattering). As seen in Figure 6-1, blends were made in THF
solution in an Ar glovebox and methanol was added to the mixture to help dissolve the polymersalt complex. The number of lithium ions to PEO monomer was varied from 0, 1:12 to 1:6. Next
gradient thickness films (80-130 nm) were cast on toluene-rinsed, ultraviolet-ozone (UVO)treated silicon wafers using the flow coating technique.(108) The samples were then vacuum
oven annealed at 135 'C for 6 h to promote parallel oriented lamellar structure which is essential
for the characterization of the thin films.
75
PS-b-POEM
+
Lithium salt
THF/Methanol
Anneal 135 OC 6h
Lithium swollen
POEM
3.5
Figure 6-1: Fabrication of lamellar PS-POEM lithium containing films.
To characterize the domain structure of the sample we analyzed it using optical images,
atomic force microscopy (AFM), and X-ray reflectivity (XRR). From the optical images, we saw
the gradient thickness films exhibited cyclic changes of islands/holes structures and uniform free
surfaces (Figure 6-2A). As the free surface morphology is primarily determined by the
commensurability conditions (ratio of film thickness to BCP domain spacing, Lo)(109), we first
acquired the domain spacing from the height of the island/hole structures through AFM. We
further found the results are consistent with small angle X-ray scattering (SAXS) measurements
of the bulk system (Figure 6-2B). The growth in the domain spacing of the block copolymer
electrolyte with salt loading (from 26.8 nm to 36 nm to 41.3 nm) is expected as the salt is known
to preferentially swell the PEO domain. From commensurability calculations, the PS-POEM
films showed island/hole structures at film thickness of nL0 , and uniform surfaces at (n+0.5) Lo,
which indicated asymmetric wetting conditions. We also note that the bulk sample with 6:1 salt
developed a mixed lamellar and cylindrical structure (Figure 6-6); however, the thin film sample
clearly exhibited lamellar structure due to strong surface interactions.
76
A
97 nm
99.S nm
103 tim
PS-POEM
neat
124.5 nm
117 tim
114.5 tim
109.5 nm
'28 nm
_
97 nm
95 nm
91nm
102.5
nm
117.5 nm
122 nm
127 nm
114.5 nro
119 nm
12$
108 nm
PS-POEM
_
12:1 salt
86.5 nm
94 nm
93 nm
103
100 nm
nm
nrr
PS-POEM
6:1salt
B
PS-POEM neat
PS-POEM 12:1 salt
PS-POEM 6:1 salt
45
40
*
*
Measured
from SAXS
Measured fron
AFM
35
iI
L= 41.3 nm
Lo =36.0 nm
L= 26.8 nm
E30
25
I
\_j
--0
--4
~I0J
~I
---i
_.L
i
-
..
k
12:1
EO:Uk
ratio
6:1
0
Figure 6-2: (A) Optical images of gradient thickness PS-POEM films (neat, 12:1 salt and
6:1 salt) annealed at 135 *C for 6 h; (B) AFM height images and corresponding sections
show that the difference between the high and low regions is Lo, which is consistent with
SAXS measurement of the bulk system.
We further conducted XRR experiments on the thin film samples (12:1, 6:1 salt) to
confirm the layered structures. Figure 6-3 shows the X-ray reflectivity profiles for PS-POEM
thin films with 12:1 and 6:1 salt that are both 5LO/2 thick. The "o" symbols represent the
reflectivity profile calculated using the scattering length density profile shown in the inset. The
profile is characterized by a capping layer of PS at the air surface followed with POEM, PS,
POEM, PS and a wetting layer of POEM at the substrate. The lamellar unit cell thickness
77
(consisting of a POEM layer and a PS layer) is 35.8 nm (or 34.9 nm) for 12:1 salt sample and
40.0 nm (or 39.8 nm) for 6:1 salt sample, respectively. We compared these values with the
domain spacing measured from AFM and SAXS, and found they are in reasonable agreement.
We further note that the roughness of the interfaces for the domains is on the order of ~Anm
which is expected since the PS-POEM with lithium salt system is strongly segregating. It should
be emphasized, though, that the models used to calculate the reflectivity profiles may not be
unique. However, we believe the models presented herein should be very close to an accurate
description of the morphology in the films.
78
PS-PO EM 12:1 salt
0
20-
10
Material
Thickness
(nm)
]Roughness
(nm)
PS (capping)
8
1.45
POEM
22.9
1.7
PS
12.9
1.4
POEM
21.7
0.97
PS
13.2
1.13
POEM (wetting)
9.6
1.16
S10,
2.5
0.32
Material
Thickness
(nm)
Roughness
(nm)
PS (capping)
9.3
1.33
POEM
25
2.57
PS
15
0
POEM
25.2
1.23
PS
14.6
0.9
POEM (wetting)
10.09
1.03
s5o.
2.4
0.3
utim W
0
10
10
0
2
1
20C)
PS-POEM 6:1 salt
10 0
101
20
16*
.kIw16uIAI
0
C
10
10'
0
1
2
3
20(*)
Figure 6-3: X-ray reflectivity profile for PS-POEM films with salt ratio 12:1 and 6:1. The
solid line denotes measured profile and the "o" symbol denotes fitted profile.
For detailed chemical analysis of the BCP electrolyte we turned to depth profiling XPS.
Enabled by the iterative etching and analysis of depth profiling XPS we are able to visualize the
discrete nanostructured lamellar regions (Figure 6-4A). To maximize the distinction between the
regions we chose film thicknesses that are free of island/hole structures (neat film thickness =
117 nm, 6:1 salt film thickness = 100 nm). Figure 6-4B shows the alternating Cl s and Ols signal
of the neat BCP lamellar thin film. Near the silicon substrate we see significant signal from the
Si2p peak as well as an increase in the Ols signal due to SiO 2 . Figure 6-4C shows the alternating
intensity of the Ols spectra due to the large amounts of oxygen in the red spectra representing
79
the POEM region. The Cl s signal in Figure 6-4D displays a similar alternating structure and two
Cls peaks (285 eV and 286.5 eV) can clearly be seen. In the blue spectra representing the PS
rich regions, the Cl s peak at 285 eV corresponds to carbon-carbon bonds and is dominant. In the
red spectra representing the POEM rich regions, the Cls peak at 286.5 eV corresponds to the
ether bonds in the PEO side chain. There is also a carbon-carbon bonding peak in the red spectra
at 285 eV due to the carbon-carbon backbone and some signal from the adjacent PS region. The
atomic profile also supports that the film is asymmetrically wetting, with PS at the air interface
and POEM in contact with the substrate.
80
Depth Profiling XPS
A
Acquire
Etch
Acquire
Etch
Acquire
Etch
Acquire
PS
POEM
C
B
01s
x10
100
80
._
0
C
60-
1
0
Carbon (C1s)
Oxygen (01s)
Silicon (Si2p)
100
Thc s0
Thickness above Silicon (nm)
536
2
Binding Energy (eV)
0
40
D
0
CIs
20
0
400
100
50
Thickness above Silicon (nm)
200
0
3
0
0
280
0
100
Thickness above Silicon (nm)
285
Binding Energy (eV)
Figure 6-4: Depth Profiling XPS of 36 kDa PS-POEM without lithium salt. (A) Schematic
of depth profiling XPS analysis. (B) Atomic concentration versus thickness above the
substrate. (C) and (D) Ols and Cis photoelectron spectra respectively, showing the
alternating intensity of a lamellar block copolymer film. The red and blue spectra are
spectra primarily in the POEM and PS region respectively.
To analyze the distribution of ions within BCP films, 6:1 salt was added to maximize the
lithium signal from the lamellar system. As seen in Figure 6-2, the addition of 6:1 salt
significantly increases the domain spacing of the film from ~26 nm to -41 nm. Figure 6-5A
81
shows the repeating structure of the lamellar film and it is clear that in the Ols rich POEM
regions there is significant Fis and Lils signal supporting the segregation of the lithium and
counter-ion into the POEM region. To further analyze the lithium-ion distribution within the
POEM region, Figure 6-5B shows only low concentrations. To directly compare the distribution
of the salt anion with the lithium cation, we can multiply the F I s signal by the molar ratio of Li Is
to Fls (1:3). The resultant atomic profile (Li(F)) overlays very closely with the Lils signal,
showing that the lithium cation and fluorine containing anion concentrations are similar in
distribution(l 9).
The lithium-ion distribution can be determined by further analyzing the atomic and
chemical composition of the film as seen in Figure 6-5C. In the system tested there are only two
sources of oxygen, first from the POEM region and second from the anion of the lithium salt; by
separating the two sources we can better analyze the material distribution. By subtracting the
lithium salt derived oxygen signal from the total oxygen signal we can determine the location of
the POEM block within the film (Figure 6-5C). This was accomplished by simply subtracting the
F Is signal from the Ols signal since it is known that for every fluorine in the salt anion there is
one oxygen. Another independent method to analyze the distribution of ions in the POEM block
is through component analysis of the Cls region. As seen in Figure 6-5D the Ci s region contains
a variety of peaks, each corresponding to different chemical bonding states present within the
BCP. For example the alternating photoelectron peaks found at 285eV and 286.5eV relate to
carbon-carbon bonds in the PS region and ether bonds in the POEM region respectively. Using
the known binding energy of the polymer components the multiple CIs peaks can be
deconvoluted and attributed to the lithium salt, PS region, or POEM region. An example of
fitting the spectra with their corresponding peaks in both a PS rich region and a POEM rich
82
region is shown in Figure 6-7. Comparing the results of the Ols subtraction method and the CIs
fitting method we see that the independent methods found very similar POEM distributions. We
suspect that the lower POEM signal in the domain closer to the substrate is due to cumulative Xray damage induced from the high energy settings needed to detect the lithium.
We can next compare the distribution of the POEM signal compared to the Li Is signal by
multiplying by the molar ratio of FIs to POEM (4.25:10.5, in 6:1 salt films Table 6-1). FIs was
used in place of the Lils signal since the Fis signal has much less noise and the previous
analysis showed that the Fis and Lils concentration are tightly correlated (Figure 6-5B). Since
the Fis signal follows the POEM amount closely, we can determine that the lithium salt
concentration corresponds with the POEM concentration. If the lithium concentration was nonuniformly distributed, we would not expect to see the coupling of the signals as seen in Figure
6-5C. It is also important to note that the unique capability to analyze the chemical composition
as well as the atomic concentration is not possible in other analysis techniques such as secondary
ion mass spectrometry (SIMS) or Energy-filtered TEM, displaying the unique potential of depth
profiling XPS to analyze BCP thin films.
83
100.
A
D
Cls
80.
0
*1
60-
~-Cs
40
--
01s
Fs
Liis
E
50 100 295
20
Thickness above silicon (nm)
0I
80
60
40
20
Thickness above Silicon (nm)
1 00
B
0
E
Lils
600-
20
-01s
Fls
Lils
-- -Li(F)
15
C
4000
8 200
10
C
0
F
E 5
60"
80
80
60
40'l
20
40
20
,"
0
Fls
1400082000
501
C
0
50
0
55
0
50
536
Thickness above silicon (nm)
BindingEnergy ()
6000-
Thickness above Silicon (nm)
C
-
8000-
I10
100
GFls
40
0
30
o
\
50
100
Thickness above silicon (nm)
Ols minus salt
Cls PEOM fit
Fl.s ratio to POEM
0
290
Binding Energy (eV)
692 690 M8338
Binding Energy (eV)
G
8000-
ZU
o000
20
000
0
0o
10
0
800
00
0
00
80
8
00
00
U-100
0
01
0
0
0
0
089
8
80
60
40
20
Thickness above Silicon (nm)
~00
0
03
100
Thickness above silicon (nm)
536 534 3
Binding Energy
(eV)
Figure 6-5: XPS depth profiling of 36kDa PS-b-POEM with 6:1 Li salt added. (A) Atomic
concentration profile with depth. (B) Zoom-in of low concentration and overlay of fluorine
atomic ratio comparison. (C) Distribution of PEO containing side block within the film
analyzed via Ols signal minus Ols signal from the salt (dashed green), fitting the Cis peak
(dashed blue) or Fis ratio to POEM (dashed red) (D), (E), (F) and (G) 3D spectra of the
Cis, Lils, Fis and Ols regions respectively. The Lils figure is rotated for clarity.
84
By analyzing other individual spectra we can garner further details on the film structure
and the limitations of depth profiling XPS to analyze BCP films. Figure 6-5E shows the Lils
spectra with depth and due to the low scattering factor of lithium, the signal intensity is relatively
low. Increasing the lithium concentration was not possible since higher concentrations of lithium
(4:1 salt) changed the morphology to a mixed cylinders and lamellae system (Figure 6-6) which
is not easily analyzed via depth profiling XPS. Figure 6-5F also shows the strong alternating
signal from Fls and also shows a small peak around 685eV. This minor peak at 685eV is
potentially due to X-ray damage from the high X-ray power needed to detect lithium. Figure
6-5F also shows the strong alternating pattern of the Ols signal.
As opposed to the sharp interface expected, the sinusoidal nature of the profile in Figure
6-5A can potentially be attributed to the material characteristics sources such as material defects,
interfacial roughness and interface wiggling. The profile could also be caused by experimental
effects like etching induced roughness and the inelastic mean free path (IMFP) of the
photoelectron being measured. To determine the effect of the IMFP of the photoelectron on the
atomic profile, the phenomenon was modeled (Figure 6-8 and Table 6-2) and shows this effect
contributes, but does not alone cause the sinusoidal profile see in Figure 6-5A.
The sinusoidal atomic concentration profile as opposed to the sharp interface expected
shows the limits of depth profiling XPS to analyze nanostructured films. However, one
advantage unique to depth profiling XPS is that the chemical state of the film can be analyzed
enabling new analysis never done before. Even though the technique requires a large area for
analysis (near 1mm 2 ) to minimize X-ray damage and has the potential for etching induced
roughness causing interface roughening, we believe the unique advantages of C60+ depth
profiling XPS will enable it to be a powerful tool for the analysis of BCP thin films.
85
6.3 Supporting Information
Figure 6-6: Bulk TEM image of salt-doped PS-POEM with EO:Li = 6:1. Dark domains
correspond to RuO 4-stained POEM; B. AFM height image of salt-doped PS-POEM thin
film with EO:Li = 4:1, indicating a mixed cylinder and lamella structure.
A
B
X10
20
C.C
50
16
14
3
12
E
6 10
20
10
0
2
2R4
26
288
Naft!Mqy(.Y
i85
294
2b2
Figure 6-7: Peak fitting to the different regions of the Cis
displaying peaks from the POEM backbone (C-C) PEO side
(Ester Carbon) and lithium Salt (CF3). B) PS region displaying
small peak from PEO side chain (PEO) and C-C bond shake up
86
.
Sishiaup
291
288
2528
spectra. A) POEM region
chain (PEO), ester linkage
Carbon bonds (C-C) a very
peak due to pi stacking.
1
0.9
I
I
0.8
0.7
0.6
0
0
C
0
F
Li
ideal
00
0
0
0
0
0
0
0
0
000
0
0
0 0
Oo
C
0
0.5
0.4 F
0.3
0.2
000
0
0
0.1
.
100
-
)
00
NA I I a A
-&a
a
Nmffiwqmp
60
40
20
80
thickness above silicon(nm)
0
Figure 6-8: Effect of Inelastic Mean Free Path (IMFP) on Depth Profiling Interface
Analysis. Comparing experimental data 'o' to the modeled system we see that though the
IMFP would affect the shape of the profile, but not to the extent seen in our data.
Table 6-1: Atomic Composition Ratios
PS
Component
C
F
Li
0
PS
8
0
0
0
%
100%
0%
0%
0%
Lithium salt (6:1,
PEO:salt)
8.5/6
8.5/6*3
8.5/6
8.5/6*3
POEM
5+2*8.5
0
0
2+8.5
% (total = 43.833)
23.4167/43.833
=53.4%
4.25/43.833
= 9.7%
1.42/43.833
14.75/43.833
=33.6%
domain
POEM
domain
87
=3.2%
Table 6-2: IMFP information used for modeling from Tanuma et al.(110). Also shown is the
relationship between the Intensity (I) of each electron from different depths (d) within the
film depending on its IMFP (k).
Photoelectron
Binding
Energy
Lils
CIs
Ols
F Is
55eV
285eV
533eV
688eV
IMFP
[nm]
(k)
% Signal From Top 3.5nm
99.9
98.0
92.0
88.4
0.69
1.26
1.96
2.37
I(d) = Ioed/IA(E)
6.4 Materials and Methods
The synthesis of PS-POEM block copolymers was achieved via atom transfer radical
polymerization (ATRP).
The PS block was synthesized in a mixture of copper (I) bromide
(CuBr, amount), N, N, N', N", N"' -pentamethyldiethylenetriamine (PMDETA, amount), and
anisole (amount) using propargyl 2-bromoisobutyrate (PgBiB, amount) as initiator at 100 0 C.
The reaction was allowed to proceed for 12 h and terminated by immediate cooling to room
temperature and exposing to air. The polymer was purified by passage through a neutral alumina
column and precipitated from methanol. The PS-Br was reinitiated as a macroinitiator for the
semibatch growth of POEM block. The final PS-POEM block copolymer was purified by
passage through a neutral alumina column and precipitated from cold isopropanol. (By Wei-fan
Kuan)
The PS-POEM solution with lithium salt was prepared as described in the main
manuscript. The PS-POEM thin film was cast using flow coating technique (cite Stafford's
paper). Film thickness was measured using a reflectance spectrometer (Filmetrics F20-UV).
Optical microscopy images were collected on a Nikon microscope equipped with a 5 MP CCD
camera (Nikon Eclipse LV100). The topologies of polymer films were assessed by atomic force
88
microscopy (Veeco Dimension 3100). Silicon probes (Tap 150G, BudgetSensors) were used in
tapping mode. A typical set point ratio was 0.9.
X-ray reflectivity (XRR) X-ray reflectivity (XRR) was performed for the polymer thin
films on a Ultima IV unit (Rigaku). A thin, parallel beam of Cu-Ka radiation, k = 0.154 nm, was
incident on the samples. The beam was sized so as to best capture the critical edge of the samples
for best results and fit accuracy. XRR profiles were collected by scanning a small incident angle
(0) of X-rays from the source and a detection angle (20) of reflected X-rays (0' < 20 < 30). The
electron density profiles across the film thickness were obtained by using Motofit software.
X-ray Photoelectron Spectroscopy (XPS): Chemical composition of the surface was
characterized using a PHI Versaprobe II X-ray photoelectron spectrometer with a scanning
monochromated Al source (1486.6 eV, 100W, spot size 200 um). Depth profiling was
accomplished using the instrument's C6o0 ion source. The takeoff angle between the sample
surface and analyzer was 45' and the X-ray beam collected Cls, Ols, Fls, Lils and Si2p
elemental information while rastering over a 200 pm x 1400 pm area. Detailed XPS acquisition
parameters are in Table 6-3 below. Sputtering occurred in 1 minute intervals while the sample
was moved using concentric Zalar rotation at 1 rpm. The C60+ source was operated at 10 kV and
10 nA and rastered over a 4x4 mm area at an angle 700 to the surface normal. Atomic
composition was determined based on photoelectron peak areas and the relative sensitivity
factors provided in PHI's Multipak processing software. All data were background subtracted,
smoothed using a five point quadratic Savitzky-Golay (S-G) algorithm and charge-corrected so
that the carbon-carbon bond has a binding energy of 285.0 eV. The surface of the silicon
substrate was defined as the point at which the atomic concentration of silicon reached 5% in the
89
depth profiling data. Spectra peaks were fit in CasaXPS and data was plotted and analyzed using
Matlab.
Table 6-3: Detailed XPS Acquisition Parameters for BCP Depth Profiling
Element
Binding Energy (eV)
Pass Energy (eV)
Data Spacing (eV)
Sweeps
CIs
278-296
11.75
0.2
3
Ols
528-537
11.75
0.2
1
Si2p
95-107
117.4
1.0
1
Fis
680-693
58.7
0.25
2
Lils
48-66
187.4
0.20
8
90
7. CELLULAR BACKPACKS
7.1 Introduction and Background
The potential of synthetic material designed to interact with biological materials for
therapeutic purposes has gained attention recently( 111). One type of material explored in this
thesis is a 'cell backpack'. Cell backpacks are flat, micron-sized polymeric particles assembled
through LbL and are nanostructed with each backpack containing distinct regions for release,
payload and attachment. The backpack was originally developed by Albert Swiston,who found
that it could be fabricated using a combination of PEMs and photolithography(23). In brief, a
glass slide is patterned with 7pm holes where the photoresist is removed. The different regions of
the backpack are then conformally deposited using layer by layer dipping. Next, the remaining
photoresist is dissolved in sonicating acetone. During this step, the PEM above the photoresist is
ripped and removed. This results in posts of PEMs on the surface as displayed in Figure 7-1.
These posts are less than one micron thick and 5-7 microns in diameter.
91
Glass substrate
Photomask, UV cure
lass substrate
Heterostructure Deposition
Wmelas, substrate
I.
Dissolve Photoresalst
Glass substrat
Cell-adhesive region
Payload region
- Release region
Figure 7-1: Fabrication of Backpacks on a Surface. Reprinted with permission from
Swiston, A.J., Cheng, C., Um, S.H., Irvine, D.J., Cohen, R.E., and Rubner, M.F. Nano Lett.
8, 4446-4453 (2008). Copyright 2008 American Chemical Society.
The backpacks have three regions with distinct purposes. The top region is the cell
adhesive layer. The composition of this region can be optimized for the particular cell of interest.
In particular, we investigated three options. One option developed by Albert Swiston and used in
the first generation of the backpack is the presentation of hyaluronic acid at the surface to
interact with B-cells through the CD44 receptor (112) (Figure 7-2A). This layer has been
thoroughly studied and optimized for its effectiveness at capturing CH-27 B-cells (10), and was
also used in the orientation controlled cell-tube study(1 13). To generalize the attachment to a
variety of cell types, two new attachment methods were developed. One method is a covalent
chemistry attachment method using either maleimide or activiated NHS esters to covalently react
with the surface proteins (Figure 7-2B). However, due to the short half-life of the reactive groups
and inability to target particular cell types to the necessary reactive groups, we considered other
92
attachment methods. An alternative method developed used the strong and specific biotin and
streptavidin interaction so that any biotinylated antibody could be used. To accomplish this,
poly(allylamine hydrochloride) was functionalized with biotin before incorporation into the film.
After backpack fabrication, the film was incubated in streptavidin, rinsed and then exposed to the
biotinylated antibody of choice (Figure 7-2C). For details of the attachment process see the
methods in section 7.3.2. In the majority of the backpack related work, we used IgG-b. However,
by simply changing the antibody, the backpack could apply to a wide range of cell types.
A
B
C
Figure 7-2: Potential Methods of Cell Attachment A) Attachment of red labeled B-cells
onto FITC labeled (hyaluronic acid/chitosan) patches (reproduced from Al Swiston). scale
bar 10pm B) Attachment of unlabeled Monocytes onto (PAA4/PAH4) patches activated
with EDC-NHS. scale bar 10pm C) Attachment of unlabeled Monocytes onto green labeled
IgG presenting patches. scale bar 50pm.
The next region is the payload region. To design effective in vivo backpacks, the system must
be loaded with a therapeutic agent. An advantage of the layer by layer system is that the payload
region can be separately designed since it is independent from the release region and cell
adhesive region. This flexibility allows the payload region of the backpack to be tuned to the
desired application while the other regions maintain the same structure. For example, the payload
region can contain magnetic nanoparticles which could be used for diagnostic tracking or to
induce hyperthermia at disease sights. Significant work was focused to expand the potential
loading of this region. However, the fabrication step where the backpack is sonicated in solvent
93
made this more difficult than expected. Small molecules were especially difficult to work with,
but Roberta Polak has recently discovered a way to incorporate liposomes containing the small
molecule doxorubicin into the backpack. Table 7-1 shows the materials successfully loaded into
the backpack during my work.
Table 7-1: Explored Payloads in a Cellular Backpack
Payload Type
Magnetic
nanoparticles
Quantum Dots
DNA/siRNA( 114)
Proteins
Purpose
Diagnostic
Tracking
Hyperthermia
Fluorescent
Specific Type
Ferrotech 10nm particles
Tracking
functionalized
Gene delivery
Therapeutic
Poly(I:C)
Enzymes, Anti-receptors
Figure Reference
Figure 7-4
Ocean Nanotech 665nm COOH Figure 7-19
Figure 7-18
Figure 7-6
Finally, the release region is a dissolvable region that releases the patch upon exposure to
stimuli (16). This region can be built of a variety of hydrogen bonding systems including
poly(acrylic
acid)
and
poly(ethylene
glycol),
or
poly(methacrylic
acid)
and
poly(vinylpyrrolidone). The poly(acrylic acid)/poly(ethylene glycol) hydrogen bonding film is
useful for analyzing interlayer diffusion (see Chapter 0) since the components lack nitrogen, thus
allowing the nitrogen signal to track the diffusion of components like poly(allylamine
hydrochloride) or chitosan. The poly(methacrylic acid)/poly(vinylpyrrolidone) film is stable up
to pH 6.5(16) and thus was used for the backpack studies to have the maximum flexibility in
payload and attachment region functionality. Another more advanced hydrogen bonding system
developed by Albert Swiston involves poly(N-isopropylacrylamide) and poly(methacrylic acid).
Poly(N-isopropylacrylamide) is a temperature sensitive polymer and using this effect and
hydrogen bonding, we have found that this system is only soluble above neutral pH and at low
temperature (4'C)(23).
94
The backpacks can then be attached to the cell in a variety of ways. In Figure 7-3, the
attachment method detailed allows for one to one cell to backpack ratios; however we have
found that larger scale attachment is possible when the backpacks are freed from the surface by
dissolving the release layer and mixed as a free floating solution with cells. Using this method
and using IgG-b as the attachment antibody, we can attach -50% of the incubated monocytes
(Figure 7-12).
Dissolve H-bonded region
(a)
Patches on Surface
ell-adhesive region
Payload region
4--
Release region
incubate w/cells
Functionalized
Surface
(b)
Releasable region
1 Payload region
1
Cell-adhesive region
Figure 7-3: Attachment Scheme of Cell Backpacks a) PEM posts with three regions
remains on the surface after the removal of the photoresist. b) The posts are incubated with
cells to attach them to the posts c) The hydrogen bonding region is then dissolved, resulting
in cells with backpacks attached. Reprinted with permission from Swiston, A.J., Cheng, C.,
Um, S.H., Irvine, D.J., Cohen, R.E., and Rubner, M.F. Nano Lett. 8, 4446-4453 (2008).
Copyright 2008 American Chemical Society.
A further question that arises is the composition of the backpack facing the extracellular
environment after attachment to a cell. Due to some interlayer diffusion it is expected that this
composition might be a combination of the release layer and also the payload region. We have
presented some preliminary work showing this effect in Appendix 11.
95
7.2 Phagocytosis Resistance of Cell Backpacks
Reproduced in part from: Doshi, N., Swiston, A.J., Gilbert, J.B., Alcaraz, M.L., Cohen, R.E.,
Rubner, M.F., and Mitragotri, S. Adv. Mater. 23, H105-H109 (2011). Copyright 0 2011
WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim. The work in this section was primarily
completed by Albert Swiston and Nishit Doshi with my contribution focusing on the drug release
study. I have included much of the original text and figures since the unique capabilities
discussed in this paper have driven our continued collaboration with UCSB.
7.2.1
Introduction
Cellular backpacks anisotropic and their round disc-like shape imparts unique functionalities
that are not see in most microparticles. Importantly, Champion et al.(24) showed that the local
curvature of the particle near a macrophage greatly affects particles' internalization and therefore
providing a theory of the unique capability of the BP. In previous studies it was found that the
cellular backpack can be immobilized on a cell without affecting cell functions. In particular the
backpacks have shown not to affect the acute vitality or mobility if they are attached to B-cells or
macrophages(23, 29). Further the backpack did not affect the distribution of CD44 receptors(23).
To study the phagocytosis resistance of backpacks we worked with macrophages.
Macrophages are ubiquitous phagocytic cells in the human immune system and they play a key
role in homeostatic, immunological and inflammatory processes(l 15, 116). Macrophages are
widely distributed in various tissues and play a central role in clearing invading pathogens, dead
cells and foreign entities through phagocytosis(l 17). Their wide presence in various organs and
tissues makes them particularly suited to provide an immediate defense against invading threats.
Moreover, macrophages are rapidly recruited to the diseased site by signaling molecules such as
cytokines. Hence, macrophages are involved in a wide repertoire of pathological conditions
96
including cancer, atherosclerosis, various inflammatory diseases such as vasculitis and asthma
and many others.
Since macrophages play an indispensable role in most pathological conditions, they
represent an ideal target for therapeutic applications. Several approaches seeking to use
macrophages for targeted therapies involve feeding therapeutic nanoparticles to macrophages ex
vivo, followed by re-injection of the macrophages to target the diseased site. This approach has
shown promising results for treating HIV infections( 118), brain disorders( 119) and solid
tumors(120). While this strategy is effective for certain conditions, its applications are limited by
the fact that the drug carriers are sequestered within the phagosome of macrophages, which
reduces the release rates, and in certain cases, degrades the drug. This limitation can be
potentially addressed by designing particles that: (i) attach to the macrophage surface, (ii) avoid
internalization, (iii) do not interfere with macrophage function and (iv) release the encapsulated
drugs in a controlled manner. However, development of materials that simultaneously fulfill
these requirements is a significant challenge. Herein, we report the ability of cellular backpacks
to successfully encapsulate and controllably release drugs, and avoid phagocytic internalization
while remaining on the macrophage's surface. These characteristics point to new possibilities in
creating cell-based bio-hybrid devices which leverage both the functions of the encapsulated
cargo (drugs, nanoparticles, etc.) and the native functions of the cell.
7.2.2
Results
Cellular backpacks are fabricated using a standard photolithography lift-off technique of a layerby-layer and spray deposited film. Briefly, a positive photoresist is patterned with regularly
spaced 7 tm diameter holes that extend down to the substrate. Next, a layer-by-layer deposited
film consisting of alternating hydrogen bond donor-acceptor pairs is deposited, and this layer
97
comprises the release region that tethers the rest of the backpack to the substrate. Two hydrogenbonded regions were used, and details can be found elsewhere(23). Next, a PEM of either (FITCPAH/MNP) or (PDAC/SPS) is deposited to provide sufficient mechanical rigidity for the
backpack to survive the final acetone sonication step. In the protein-eluting backpack
construction, a PLGA/FITC-BSA film is then sprayed onto the existing two layers, and forms the
hydrolytic payload region of the backpack. An optional second (PDAC/SPS) or (PAH/SPS) film
is deposited, and finally the cell-adhesive (HA/CHI) region is added to the top of the
heterostructure. A schematic of the three final backpack films is found in Figure 7-4(i-iii). The
last fabrication step is sonicating the film in acetone, which dissolves the photoresist and
simultaneously lifts off the deposited heterostructured film in all areas except those adsorbed
directly to the substrate. A SEM micrograph of an as-fabricated backpack may be found in
Figure 7-4b.
98
(a)
PEMa*.sp LbL
(i)
MWAAMA~t
PEU sean. Lb.
PAWM pan: Lb
ARMsftown LW.
(ii)
PEMd. regbn L
!!!
-
(iii)
.
PEWtown. Lt
PEM roplon: Lb.
lishsa
span LbL
(d)
(c)
80
C
N
Lo
60
40
C)
20
0
HA coated HA coated
backpacks spheres
Amine modified
spheres
(e)
Figure 7-4: Backpacks for cell-based drug delivery devices: (a) schematic of the different
backpack layers, all of which deposited on top of a photoresist-patterned glass substrate
and (b) scanning electron micrograph of a backpack with a PMAA/PVPON release region
fabricated by the above strategy. The average diameter is ~6 gm and the backpacks have a
flat disk shape. (c) Time lapse microscopy image sequence of macrophages interacting with
HA/CHI coated backpacks. The arrow indicates an internalized backpack. All other
backpacks (n=7) are attached to macrophage surfaces for the entire time lapse sequence of
50 minutes, exhibit strong resistance to macrophage phagocytosis. (d) Quantitative analysis
of 6 pm diameter particles internalization by macrophages. Data is presented in terms of
internalization percentage (percentage of particles observed that were internalized by
phagocytosis). At least 100 particles were observed for each condition. Amine modified
spheres act as a positive control and show a high internalization percentage. HA coated
backpacks show significantly lower internalization percentage compared to both HA
99
coated spheres and amine modified spheres. (e) Scanning electron micrographs showing the
interaction of backpacks and spheres with macrophages after 3 hours of incubation in
standard cell culture conditions: (i) HA coated backpack attached to the surface of a
macrophage, and (ii) three 6 gm diameter HA coated spheres internalized within a
macrophage.
Backpack Interactionswith Macrophages
Attachment and phagocytosis of backpacks was studied using J774 mouse macrophages as a
model cell, and 6 pm diameter HA-coated spheres were used as control particles. Around 95% of
backpacks added to the cells attached to the macrophage surfaces (n=100 backpacks). Once
attached, macrophages did not release backpacks from their surface, suggesting a strong
attachment between the two. Time lapse video microscopy provided more insights into the cellbackpack interactions Figure 7-4c. Backpacks remained attached to macrophage surfaces without
internalization even 50 minutes following introduction. In contrast, HA-coated spheres were
phagocytosed within 30 minutes of attachment to the macrophage surface (see Supporting
Information for time lapse microscopy video).
These qualitative observations were quantified in terms of percent internalization.
Particles were observed for 3 hours following attachment to the macrophage surface. At least
100 particles were observed for each condition for statistical significance. Amine modified
(positively charged at physiological pH) and HA-coated (negatively charged) spheres (diameter
= 6 pm) were used as positive controls(121). As expected, amine-modified particles exhibited
high internalization with almost 80% of attached particles being phagocytosed. HA-coated
spheres were internalized to a lesser extent (35% of attached particles), however a large
proportion was still internalized. The difference between HA-coated and amine-modified spheres
likely originates from a difference in their surface charge. HA was used for coating backpacks
since interactions between HA and macrophages are specific, mediated through the cell surface
100
receptor CD44(122). Backpacks exhibited almost 30-fold lower extent of internalization
compared to positive controls and about 14-fold lower extent of internalization compared to HAcoated spheres (Figure 7-4d). It was found that backpacks typically attach to macrophages by
their flat face (Figure 7-4e).
Effect of Backpacks on Macrophages
We assessed whether backpack attachment alters the cellular functions of macrophages. Two
behaviors of macrophages with attached backpacks were assessed; (i) the ability to internalize
spherical particles (i.e., easily phagocytosed targets) and (ii) motility. We measured the
internalization capability of macrophages with attached backpacks. PS spheres (diameter = 3
ptm) were chosen as the target since particles of this diameter have been shown to be
preferentially phagocytosed by macrophages(123). Macrophages were first incubated with
backpacks to allow sufficient attachment and then 3 ptm PS spheres were added to the
suspension. Time-lapse video microcopy clearly shows that backpack-laden cells were able to
internalize 3 ptm spheres while simultaneously remaining attached to a backpack (Figure 7-5a).
Macrophage motility was measured by tracking their center of mass for 3 hours. No statistical
difference was found between the mobility of macrophages with and without backpacks (Figure
7-5b).
101
(a)
(c)
(b)
40 -
2
-
30 -
6
E ;0
t120 -
1010
0
~0 A
20
--
Cells with
Backpacks
0.2
0
Cells Wo
backpacks
Backpacks
6 urn
Spheres
Control
Figure 7-5: Backpack attachment to macrophages does not affect cellular health and
functions. (a) Time lapse video microscopy showing that a backpack-laden macrophage can
still efficiently internalize 3 gm PS spheres (MP: Macrophage, BP: Backpack, PS-S: 3 gm
polystyrene sphere). The backpack is not internalized during the entire duration of the
time lapse sequence. (b) Differences in migration distance of macrophages with and
without backpacks 3 hours were found to be insignificant. (c) MTT assay results in terms
of proliferation potential, where 1 indicates healthy proliferation capacity and 0 indicates
complete cell toxicity.
Backpack Cytotoxicity
An MTT assay was performed to determine whether the attachment of backpacks adversely
affects the health and proliferation of macrophages. The proliferation index of backpackassociated macrophages was not statistically different from the native population (Figure 7-5c),
indicating that backpacks are not acutely toxic to the cells and they do not interfere with a cell's
ability to reproduce. As a control, macrophages exposed to 6 pm spheres also did not induce any
noticeable toxicity.
102
Drug Release from Backpacks
The backpacks reported in this study have significant biomedical potential as multi-modal
therapeutic and diagnostic platforms that may carry a variety of drugs, imaging/diagnostic
agents, or nanoparticles. Specifically, backpacks can be loaded with drugs which can be released
in a controlled manner. As a proof of concept, we loaded the backpacks with fluorescein
conjugated Bovine Serum Albumin (FITC-BSA) within a poly(lactic-co-glycolic acid) (PLGA)
polymer matrix and measured subsequent protein release.
The fabrication method for drug releasing backpacks was slightly different from
previously described due to the inclusion of PLGA (Figure 7-6a), but the backpack still
presented the same cell-adhesive and pH labile release regions as backpacks used in the toxicity
and mobility studies. FITC-BSA was chosen since it is commonly used as a model protein and it
could be replaced with a variety of therapeutic or diagnostic materials. We see that the backpacks
released FITC-BSA in a burst manner with ~40% released by 2 hours and >90% released by 24
hours (Figure 7-6b). Also of note is that very similar release profiles were seen independent of
the weight percentages of FITC-BSA within the polymer solution before spraying.
103
aa
e aw
(a)
p ho kTwr"s pa n
-bonded PWsse
a md
wnd PEM LbL depodbed
Spvuy PLOAMTC1SA COW
I
LyopNkmze ovmniglE
LbL depost PEM "nd cIl-adhesvo rgon
Fffiw dipo n omp
sonicml In 4001"n
ltas and incubato inPBS
/
Meesure FITC signal of supernatan
1.2
(b)
1 --
0.8
0.6
v
0.4
0.2
20 40
60
80 100 120 140
Hors
104
Figure 7-6: Controlled release of FITC-BSA from therapeutic backpacks. (a) Fabrication
schematic (b) Controlled release profile of FITC-BSA in-vitro. The normalized release of
FITC-BSA occurred over multiple hours and was not dependent on the loading percentage
of the spray emulsion used. Red and blue lines correspond to 7.5 wt% and 25 wt%,
respectively.
7.2.3
Discussion
Macrophages are pervasive cells in the mammalian immune system, capable of identifying and
neutralizing non-native agents found in the body. In addition, macrophages are routinely
recruited to diseased sites, potentially making them highly efficient targeting devices. If nonnative entities could be attached to these macrophages without being phagocytosed, these cells
could serve as ideal chaperones for delivering drugs or other therapeutic tools to various
pathological tissues. This is particularly attractive for the treatment of cancer and inflammatory
disorders, where macrophages are very strongly recruited(124-126). This approach would offer
much greater spatial resolution and selective uptake of therapeutic or diagnostic components,
essentially using an individual cell as a cargo-carrying device.
Cellular
backpacks
are
nanoscale-thickness
microparticles
fabricated
using
photolithography where polyelectrolyte multi-layers are assembled using layer by layer
technique, one of the most versatile methods to incorporate a variety of materials within film
structures for biochemical applications(127, 128). The backpack are engineered to (1)
contain
therapeutic or diagnostic materials, such as small molecule drugs, proteins, nanoparticles, or
functional polymers, and (2) attach to the surface of a cell using a non-toxic mechanism relying
upon a natural ligand-receptor interaction such as the HA-CD44-mediated adhesion used in this
study(129, 130). Because of their unique geometry and capacity to contain therapeutic and
diagnostic materials,
cellular backpacks
present unique opportunities
phagocytosis-resistant microparticle.
105
as a functional
Long-term immobilization of any particle on the surface of phagocytic cells like
macrophages is extremely difficult. Recently, Matthias et al. showed the therapeutic potential of
covalently attaching nano-particles to the surface of cells, but phagocytic immature dendritic
cells posed an internalization problem(131, 132). The primary challenge of attachment to
phagocytic cells, like macrophages, is their propensity to engulf any material attached to their
surface. Various properties of particles, particularly their size, surface chemistry, shape and
mechanical flexibility have been shown to play a critical role in phagocytosis(24, 121, 133-136).
Particles in the range of 2-3 pm are known to exhibit high attachment to macrophages(24, 133,
137), and particles in this size range are very efficiently phagocytosed. Although the rate of
internalization depends on the size of the particle, most particles with 2-3 ptm diameters are
internalized within 30 minutes of attachment to the macrophage surface.
Recent literature has shown that the local shape, orientation, and mechanical properties of
a particle significantly influence its susceptibility to be internalized by phagocytosis(134).
Elongated, high aspect ratio particles and very flat disk-shaped particles are shown to exhibit
reduced phagocytosis compared to spherical particles(138). The low internalization rate of diskshaped backpacks reported here is consistent with these observations. Though not considered in
this study, a backpack's flexibility and low modulus may also contribute to low internalization
frequency, a phenomenon seen in previous work(139). Indeed, the right blend of three factors
(shape, orientation, and mechanical properties) mentioned may make backpacks an ideal system
for avoiding the phagocytosis behavior of macrophages.
Avoiding internalization by macrophages is a key feature of the backpack, but equally
important is the observation that backpack attachment does not interfere with the macrophage's
native cellular functions. The results of the MTT proliferation assay suggest that backpacks are
106
not toxic and do not affect cell proliferation capacity. Furthermore, backpack attachment did not
adversely affect a macrophage's ability to internalize particles that are otherwise phagocytosable.
Further investigation is needed to understand whether or not backpack attachment affects other
macrophage functions, including the release of chemokines and cytokines.
Cellular backpacks provide a unique opportunity for drug delivery and therapy since the
polyelectrolyte multilayers which comprise much of the backpack are well studied for their
biomedical applications(140) and drug delivery properties(28). In particular, polyelectrolyte
multilayers have shown the ability to deliver DNA( 114), vaccines( 141), proteins(6) and small
molecules(7). The results presented here show that cellular backpacks are capable of delivering a
model protein in a controlled and sustained manner in vitro. Another strategy for finely
controlled drug release from therapeutic backpacks could be to trigger release by thermal
ablation. Chemotherapy agents or particles capable of ablation via RF heating (i.e., gold or iron
oxide) may be loaded into a backpack, which can then ride on a monocyte being recruited into a
solid tumor. Using a recruited monocyte as an active therapeutic-carrying device may allow
access to the relatively inaccessible hypoxic region of a tumor(142), offering clear advantages
over passive particle uptake through "leaky" vasculatures. Such a strategy has been adopted for
gold nanoparticles with promising results(120).
7.2.4
Experimental Section
Backpack Fabrication
The backpacks were formed through a combination of photolithography and polyelectrolyte
multilayers (PEMs). Photolithographically patterned (PDAC4.0/SPS4.0) 15 .5(where PDAC is poly
(diallyl dimethyl ammonium chloride) and SPS is poly(styrene sulfonate) coated glass substrates
107
were sequentially dipped in dilute polymer or nanoparticle solution using an automated Zeiss
programmable slide stainer or nanoStrata dipping unit, the details for which have been
previously described(23, 130, 143). The fully-charged polyelectrolyte and hydrogen-bonded
release PEMs were built in the Zeiss dipper; the cell adhesive region was built in the nanoStrata
unit.
The
formula
for
used
backpacks
in
internalization
studies
is
(PMAA3.0/PNIPAAM3.0) 80.5(FITC-PAH3.0/MNP4.0) 1 0(PAH3.0 /SPS4.0) 30 , where the number
following each species abbreviation indicates the solution pH and subscripts are the number of
bilayers (where a half bilayer is indicated as 0.5). PMAA stands for poly(methacrylic acid),
PNIPAAM
for poly(N-isopropylacrylamide),
FITC-PAH for fluorescein poly(allylamine
hydrochloride) and MNP for magnetic nanoparticles. An optional (CHI3.0/HA3.0) 3 (where CHI
is chitosan and HA is hyaluronic acid) region was used as the cell adhesive region for both
backpacks and PS spheres. The purpose of each region is described in the backpack fabrication
results section. See Supporting Information for backpack compositions in toxicity and migration
studies.
MTT assay
J774 mouse macrophages were seeded at 104 cells/well onto a 96-well plate. Backpacks and 6
pm spheres at a concentration of 10 5/well (200 pl media) were exposed to the cells for 4 hours.
Cells not exposed to any particles were used as negative control. Ten microliters of reagent from
an MTT kit (Invitrogen, Carlsbad, CA) was applied to each well for 5 hours. The liquid from
each well was then replaced by DMSO and allowed to incubate in the dark at room temperature
for about 10 minutes to lyse the cells. Absorbance was read at 570 nm (kmax of MTT dye) and
proliferation index was calculated as the ratio of absorbance of the sample to the absorbance of
108
negative control. The proliferation index values range from 0 to 1, with 0 indicating maximum
mitochondrial toxicity and 1 representing high proliferation capacity and healthy cells.
ControlledRelease of BSA from Backpacks
Backpacks for drug release studies were slightly different in structure to enable the inclusion of a
layer comprised of PLGA and FITC-BSA. The layer was applied by pressure spraying a dilute
sonicated mixture of PLGA solubilized in chloroform and FITC-BSA solubilized in water. (See
Figure 3a and Supporting Information for details). Backpacks were released from the surface
using PBS pH 7.4. BSA release was performed in pH 7.4 PBS in static conditions at 37*C using
fluorescence analysis (see Supporting Information for details). Prior to analysis, the sample was
centrifuged at 10,000 relative centrifugal force and 0.8 mL of the supernatant was removed for
analysis. Protein concentration was determined using a Fluorolog-3 spectrofluorometer (Horiba
Jobin Yvon, Japan) exciting at 494nm and the emission was measured at 520nm. Immediately
following, 0.8mL of fresh PBS was added to replace the removed solution. Data was measured in
triplicate and each sample had one microscope slide of backpacks present. Approximately 13
2
5
cm2 per slide was covered with backpacks with an approximate backpack density of 2 x 10 /cm .
After 5 days of elution the backpacks were isolated into 200 pL of PBS, frozen and lyophilized
overnight. The samples were then dissolved in 250 pL of dichloromethane and 200 gL of DI
water and 800 pL of PBS were added. The samples were then sonicated for 90 minutes to
increase extraction surface area. Finally, samples were centrifuged as before and the aqueous
phase was removed for fluorescence analysis.
Materials
Poly(methacrylic
acid) (PMAA, PolySciences,
PA, USA
M=100kDa),
poly(allylamine
hydrochloride) (PAH, Aldrich, MO, USA M=70kDa), poly(N-isopropylacrylamide) (PNIPAAm,
109
Polymer Source, Quebec, Canada M=258kDa), poly(vinylpyrrolidone) (PVPON, Aldrich,
M=1.3MDa),
fluorescein-labeled
poly(allyl
amine hydrochloride)
(FITC-PAH,
Aldrich,
M=70kDa), poly (diallyl dimethyl ammonium chloride) (PDAC, Aldrich, M=200-35OkDa in
20% aqueous solution), poly(styrene sulfonate) (SPS, Aldrich, M=70kDa), hyaluronic acid (HA,
from Streptococcus equi, Sigma, MO, USA, M-1.58 x 106 Da), and low MW chitosan (CHI,
Aldrich, DS=.75-.85, M-5 x 104 Da) were used without purification. Iron oxide magnetic
nanoparticles stabilized with a proprietary anionic surfactant were used (MNP, Fe 3 0 4 , 1 Onm
diameter, Ferrotec EMG 705, NH, USA). 3 pm and 6 pm polystyrene spheres were purchased
from Polysciences. FITC-Bovine serum albumin (Sigma), Chloroform (Sigma), poly(DL-lactideco-glycolide) ester terminated (50:50) M, 7,000 to 17,000, RESOMER* RG 502 (Sigma) was
used.
Cells
The J774 mouse monocyte macrophage cell line (ECACC products, Sigma Aldrich, MO, USA)
was used in this study. Cells were cultured with Dulbecco Eagle media (ATCC, VA, USA)
which was supplemented with 1% penicillin/streptavidin (Sigma Aldrich) and 10% fetal bovine
serum (Invitrogen, CA, USA). Cells were grown and passaged in standard culture conditions (37
C and 5%CO2 ).
Backpack releasefrom fabricationsubstrate
Backpacks were fabricated with a thermally-controlled release region to give control over the
number of backpacks attached to each macrophage, Briefly, this system is based upon the lower
critical solution temperature (-29C) of PNIPAAm, above which the PMAA/PNIPAAm film is
insoluble and cells may be attached, and below which the film is unstable and the backpacks
110
delaminate from the fabrication substrate. These thermally controlled backpacks were used for
all internalization studies, where ensuring a single backpack per macrophage was critical.
For all other studies, including toxicity and mobility, non-thermally released backpacks with a
PMAA/PVPON release region were used. Here, a change in solution pH above 6.0 is sufficient
to trigger release region dissolution and concomitant backpack release from the surface. The
composition of these backpacks is similar to the thermally-released backpacks except for the
release
region,
and
may
be
represented
as:
(PMAA2.0/PVPON2.0)
20 .s(FITC-
PAH3.0/MNP4.0) 1 0(PAH3.0/SPS4.0) 30, with an optional (HA3.0/CHI3.0)3 cell-adhesive region.
These backpacks were harvested in PBS and added to a macrophage suspension, where cells
randomly associate with backpacks. While this method does not afford control over the
association between backpacks and cells (ie, more than 1 backpack per cell is possible), postexposure processing may be used to adjust this ratio.
Backpacksfor drug release
In chronological order of deposition (starting at the lithographically patterned substrate), films
used were: (PMAA2.0/PVPON2.0) 40 .5 (PDAC4.0/SPS4.0) 10.5 (PLGA + FITC-BSA)(PDAC4.0/
SPS4.0)10 .5 (HA3.0/CHI3.0) 2.5 . A BSA solution in water (40 mg/mL) was added to a PLGA
solution in chloroform (2 mg/mL) for the final BSA concentration, and the mixture was
sonicated for 10 minutes. The solution was sprayed using an air brush (Badger, IL, USA) at
15cm with a N2 pressure of 50 psi and a flow rate of ~7 mL/min. After spraying, the film was
lyophilized overnight before any further dipping. Next, the final layers of the backpack were
deposited and the sample was sonicated.
Optical and FluorescenceMicroscopy
111
Cells were cultured overnight in glass bottom delta T dishes (Bioptechs, PA, USA) at a density
of around 105 cells per dish (3.8 cm 2 ). The cells were washed with PBS the following day and
the medium was replaced with HEPES (Sigma Aldrich) medium (DMEM + 25mM HEPES) to
maintain pH regardless of atmospheric carbon dioxide. Experiments were performed with
various particle types, concentrations, and length of incubation. Cells were imaged on an
Axiovert 25 microscope (Zeiss, NY, USA) fitted with a Delta T temperature controller
(Bioptechs) that maintains 37 0 C and observed using a Neo-Fluor oil immersion, 100x, 1.3 NA
objective (Zeiss). Brightfield images were taken using a cooled CCD camera (CoolSnapHQ,
Roper Scientific, GA, USA).
112
7.3 Monocyte Hitchhiking: Delivery of Phagocyte Resistant Backpacks to Lungs
Building on the success of backpack's resistance to phagocytosis presented in section 7.2
we desired to test the in vivo application of the backpack. In collaboration with Aaron Anselmo
at UCSB, we determined that a new attachment method was needed since the HA/CHI topped
backpacks did not interact with monocytes. As a result, I developed antibody labeled backpacks
and validated their monocyte attachment. The in vitro study of BPs effects on monocyte
differentiation and transmigration as well as the in vivo biodistribution study were performed by
Aaron at UCSB. The following text was written in collaboration with Aaron Anselmo as equal
contributors and is currently in the process of being submitted for publication.
Authors: Aaron C. Anselmol*, Jonathan B. Gilbert 2 *, Sunny Kumari, Vivek Gupta', Robert E.
Cohen 2 , Michael F. Rubner 3, Samir Mitragotril
1
Department of Chemical Engineering, University of California Santa Barbara, CA 93106
2
Department of Chemical Engineering 3 Department of Materials Science and Engineering,
Massachusetts Institute of Technology, Cambridge MA 02139
The demand for the controlled delivery of therapeutics to desired locations in the body
has resulted in tremendous advances in particle technology. These advances include the use of
particles with various surface chemistries, targeting ligands, shapes and sizes to direct polymeric
drug carriers to the desired location in the body.(21, 144, 145) Despite these advances, polymeric
drug carriers have seen limited success in the clinic due to poor circulation times, stemming from
the rapid clearance of foreign particles by the immune system, and to the overall inefficient
targeting to diseased tissues.(146, 147) While polymeric carriers have been unable to overcome
113
these challenges, circulating cells can remain in the blood for extended periods and can target
and accumulate at diseased tissues in cases of cancer,(148) atherosclerosis,(149) and
arthritis,(150) to name a few. Recent studies have combined polymeric particles with mammalian
cells to transfer the natural delivery capabilities of circulatory cells to these particles. Indeed, it
has been shown that circulatory cells functionalized with polymeric particles, either by
internalization or by surface attachment, have been able to improve both the circulation and
delivery of polymeric carriers to disease sites.( 118, 151) In this work, we have developed
phagocytosis resistant polymeric cell backpacks that attach to the surface of monocytes through
antibody interactions. Following attachment to monocytes, essential cell functions such as
differentiability and extravasation through an endothelial cell layer are preserved. Backpack
functionalized monocytes were then tested in vivo and shown to target backpacks to sites of
inflammation. Cellular backpacks represent a new type of polymeric carrier that is designed to
work synergistically with cells to improve the targeted delivery of therapeutic carriers.
Monocytes are a type of white blood cell that can target, and cross the endothelium of
inflamed tissues before terminally differentiating into macrophages. One example of this is in the
case of tumors, where monocyte-differentiated macrophages potentially make up 50% of the cell
tumor mass,(148) illustrating the potential of monocytes to carry drug containing particles into
inflamed tissues. Macrophages have already been used for a variety of cellular hitchhiking
therapies, including reducing HIV replication in the brain,( 119) preventing tumor growth,(152)
targeting brain metastases(153) and reducing inflammation by providing neuroprotective
capabilities in a Parkinson's disease model.(154) While similar, macrophages and monocytes are
distinct cells. Unlike macrophages which are terminally differentiated and reside in the tissue,
monocytes are circulatory cells and thus can navigate the vasculature and target typically elusive
114
diseased tissues. However, monocytes are also capable of internalizing foreign materials.
Avoiding this internalization for cell-mediated therapies is of paramount importance since
internalization can potentially lead to endosomal degradation and subsequent loss of therapeutic
use. However by engineering the backpack shape using recently discovered design parameters as
a guideline, we developed polymeric particles capable of avoiding phagocytosis.(24, 133, 138)
By taking advantage of a monocyte's natural targeting abilities, monocyte hitchhiked backpacks
can potentially target tissues that backpacks, on their own, would find challenging.
Cellular backpacks are phagocytosis resistant particles designed to attach to the surface of
immune cells.(23, 29) These flat, anisotropic polymeric particles are fabricated by layer-by-layer
(LbL) processing and are hundreds of nanometers thick and microns wide. The LbL technique
has the ability to constrain the location of the desired materials within the backpack simply by
controlling the order of fabrication;(92) thus allowing the backpack to have a distinct region
designed for cell attachment and other regions for functional materials. Cellular backpacks are
fabricated using a combination of photolithography and polyelectrolyte multilayers as described
previously.(23, 29) Photolithographically patterned substrates control the shape and size, while
polyelectrolyte multilayers allow for control over the composition, nanostructure(92) and stimuli
responsiveness of the cellular backpack.(155) Backpacks are ~7pm in diameter and are made of
three polyelectrolyte multilayer regions with a total thickness less than 500nm (Figure 7-7a). The
region shown in green is hydrogen bonded and consists of the donor poly(methacrylic acid)
(PMAA) and the acceptor poly(vinylpyrrolidone) (PVPON). This region attaches the backpack
to the fabrication substrate and when this layer is exposed to a solution with a pH > 6.5 the
region dissolves by ionizing the acid.(16) Thus by controlling pH, the backpacks can be released
from the surface when desired. The next region, shown in yellow, consisting of electrostatically
115
interacting poly(allylamine hydrochloride) (PAH) and anionic iron oxide magnetic nanoparticles
(MNP, 1 Onm diameter) provides both mechanical rigidity for fabrication and functional activity
in the magnetic responsiveness of the backpack.(23) The final region, shown in blue, is the
attachment region consisting of electrostatically interacting poly(acrylic acid) (PAA) and PAH
functionalized with biotin (PAH-b). The attachment region is exposed to streptavidin before a
biotinylated antibody of choice is used to attach to the desired cell. For this work, we used
biotinylated mouse IgG to ensure interaction between the backpack and cell (Figure 7-10). The
activity of the mouse-IgG present on the surface of the backpack, was validated using secondary
antibody labeling (Figure 7-7 b, c and method show in Figure 7-11). The resultant backpack is
chemically anisotropic with an attachment region found only on one side and thus can control its
orientation on the surface of living cells.(1 13)
(a)
(b)
FITC
(c)
Brightfield
FITC
Brightfield
Overlay
TRITC
Overlay
SMouse IgG-b
STRITC-Streptavidin
100nm -
(PAA/PAH-b)e
-
100nm -
- (PAH/MNP)o.5
300nm
-
(PMAA/PVPON)30.5 TRITC
5-7 Urn
Figure 7-7: Design of cellular backpacks and confirmation of antibody presentation. (a)
Schematic of a cellular backpack. (b) TRITC-Streptavidin and mouse-IgG-biotin are
localized to the backpack and activity is shown by secondary antibody FITC-anti-mouseIgG labeling. (c) If mouse-IgG-biotin is not added to the backpack, no specific FITC-antimouse-IgG labeling is seen.
Backpacks (BP) were then removed from glass slides and attached to WEHI-265.1
murine monocytes by incubation at a 1:1 (BP:monocyte) ratio. BP attachment was mediated by
IgG-Fc receptor interactions.(156) The attachment was confirmed by confocal microscopy in
116
solution (Figure 7-8ai) and on the surface of backpack arrays without the pH sensitive
(PMAA/PVPON) region (Figure 7-8aii). Attachment was quantified using fluorescent activated
cell sorting (FACS) (Figure 7-12). At a BP:monocyte ratio of 1:1, -50% of monocytes carried at
least 1 BP. MTT assays determined that BP attachment to monocytes does not impair health or
proliferation of macrophages. The proliferation index of BP-conjugated monocytes at 24 hours
was not statistically different than that of monocytes interacting with plain 3pm polystyrene
particles or 3pm polystyrene particles coated with IgG-b, which offer roughly the same surface
area of the IgG-b functionalized BP surface (Figure 7-13).
(a)
(b)
Figure 7-8: Attachment of cellular backpacks to WEHI-265.1 monocytes and effect on
monocyte ability to differentiate into macrophages. (a) Confocal and brightfield overlay
images of: (i) a BP (green) wrapping around a stained monocyte (red) and (ii) monocytes
(unstained) attaching to an array of patterned BPs (green). (b) Brightfield images of: (i)
WEHI-265.1 monocytes in normal culture media at 48 hours, (ii) WEHI-265.1 monocytes
differentiated into macrophages via incubation with phorbol 12-myristate 13-acetate
117
(PMA) treated culture media at 48 hours, (iii) FACS sorted monocytes with BPs (green)
attached prior to incubation for 48 hours with 50nM PMA, and (iv) J774 macrophages.
Scale bar = 10 prm.
To determine whether monocyte's ability to cross an endothelial monolayer was affected
by BP attachment, we evaluated the transmigration of BP-conjugated monocytes through a
confluent human umbilical vein endothelial cells (HUVEC) monolayer. HUVECs were grown to
a stable monolayer (30 ± 5 Qcm 2 ) and ability of monocytes alone, BPs alone, and BP-conjugated
monocytes was assessed for transmigration through the monolayer. BP-conjugated monocytes
and unconjugated monocytes passed through the endothelial layer in similar amounts at both 24
and 48 hours (Figure 7-9a). Independently using the BP fluorescence the percentage of BPs
transmigrated for BP-conjugated monocytes and free BPs was then assessed. At 24 and 48 hours,
-35% and -60%, respectively, of BPs attached to monocytes were able to cross the monolayer
while free BPs were unable to cross the monolayer (Figure 7-9b). Altogether, the in vitro
experiments show that monocytes maintain many of their important functions following BP
attachment.
Monocytes were next investigated for their ability to home to, and carry BPs to, inflamed
tissues in vivo in an acute lipopolysaccharide (LPS) lung inflammation model. This widely
established model(157) results in increased lung expression of markers VCAM(158) and
ICAM.(159) Together, VCAM which mediates the initial endothelial adhesion(160) and ICAM
which mediates leukocyte spreading across the endothelium,(161) should enhance adhesion of
monocytes in the inflamed lungs (Figure 7-14). Inflammation was confirmed by measuring
increased expression of ICAM-1 in lungs (Figure 7-15). For in vivo studies, WEHI-3 monocytelike cells were used as they have been proven to circulate and migrate to tissues in BALB/c mice
when injected intravenously, or even intraperitoneally. 3H-labeled BPs were attached to
monocytes and were intravenously injected in both LPS lung inflamed mice and healthy mice. At
118
6 hours, backpacks accumulated ~2-fold higher in the lung tissue of mice that were challenged
with LPS (Figure 7-9c). All other organs and blood showed no statistical differences.
(a)
(b)
350
Media (No Cells)
nomines
-e-
300
---
80
70
I
* 250
40
(0
1
4D
-o 200
0!
150
30
- M
100
4W
20
so
10
0
ft.
72
Ill
46 111%
24 FWS
Tim. (hours)
20
(C)
U Normal Mice
17.5
|
LPS Induced Lung Inflammation
I-
15
E 12.5
Is-
a
10
a 7.5
5
2.5
0
r
Liver
Spleen Kidney
Heart
Lunas
6M 89 tl,Brain
Skin
Blood
Figure 7-9: Transmigration of WEHI-265.1 monocytes and in vivo targeting. (a) Analysis of
monocyte migration (MCP-1 initiated) through HUVEC layer for monocytes alone (black
squares), monocytes with backpacks attached (black circles) and negative control of media
alone without monocytes (black triangles). Values represent mean : SD (n = 3). (b)
Analysis of BP migration through HUVEC layer when attached to monocytes (black) and
in free solution without monocytes present (hatched) at 24 and 48 hours. Values represent
mean ± SD (n = 3). (c) Percent of injectable dose per gram (%ID/g) 3H-backpacks attached
to monocytes in normal mice (black bars) and LPS induced lung inflamed mice (hatched
119
bars) at 6 hours. Values represent mean ± SEM (n = 4). Statistical difference (P < 0.01) is
seen only in lungs.
Attachment of backpacks to the surface of monocytes for cellular hitchhiking delivery
presents a promising method to achieve targeted drug delivery. The results presented here
demonstrate the design and synthesis of a novel particle-based system with the ability to target
inflamed endothelium via monocyte-mediated hitchhiking. To accomplish this intricate task, the
cell backpack was designed as a multifunctional material that binds to monocytes strongly,
contains a payload and resists phagocytosis. Given the high material requirements the LbL
technique was chosen since it allows for the precise control over the shape and structure of the
backpacks as well as flexibility in the backpack composition. Alternate methods including
PRINT(162, 163), particle stretching(24) and template based methods(164, 165) have fabricated
non-spherical polymeric particles for drug delivery purposes, but it is difficult to match the
inherent material flexibility found with LbL. By combining lift-off photolithography and
conformal coatings a variety of shapes could be explored but given the proven phagocytosis
resistance of circular backpacks, we chose to maintain the design from our previous work.(29)
Monocyte hitchhiked backpacks represent one of many cellular hitchhiking formulations that can
utilize the backpack platform. Furthermore, due to the freedom afforded by LbL fabrication,
these backpacks can be modified to present any biotin functionalized antibody for the adhesion
to the appropriate circulatory cell. Cellular hitchhiking of drug carrying particles can potentially
address the two main limitations preventing the widespread use of nanoparticles in the clinic,
notably the poor circulation times and limited targeting to diseased tissues. This is further
evidenced by the fact that there are currently no active targeting nanoparticle drug delivery
systems in the clinic, as all rely on passive targeting to diseased tissue. Future research will be
focused on incorporating drug containing backpacks for therapeutic treatment and utilizing
120
primary monocytes. Collectively, these results display the huge potential of monocyte mediated
backpack delivery for the treatment of a numerous inflammation based diseases.
7.3.1
Supporting Figures
ii)
i)
iii)
0
b
+ TRITC
0L
z
+ mouse
a_
streptravidin
IgG-biotin
z
b
0.
z
Released
backpacks
ready for cell
attachment
b
U
(PAA4/PAH-b4)
Figure 7-10: Preparation of Cellular Backpacks i) incubation of Cell Backpacks with
streptavidin (can be TRITC labeled) ii) Exposure to mouse IgG-biotin to dock onto
streptavidin iii) Backpacks washed and ready for attachment.
i)
iii)
+ Anti-mouse
'U
IgG-FITC
0.
z
a.
z'
~0~
iv)
ii)
+ Anti-mouse
0.
IgG-FITC
z
a.
z
Figure 7-11: Secondary Labeling to Check Activity of the Antibody on Backpack Array (no
pH sensitive release region). i) and ii) show backpacks exposed to IgG-b or not respectively.
iii) After exposure to anti-mouse IgG-FITC, if the antibody is active and present the sample
will fluoresce (iii), if not there will be no signal (iv).
121
(I)
(i)
OkqfvcyelAMQne
4,
'U
ii~1
II
Fluorescest IntensIty
-
fluesn____-i
-_.
FIvovc~cent In1,ns*j~
(iv)
Conjugates. Monocytes and Backpacks
w4=
M94
0.1
Flurecn nest
T
I -;
10
ip
ii
I,
Fluorescent Intensity
Figure 7-12: FACS plots and approximate gating for backpack attachment to WEHI-265.1
monocytes. (i) Representative FACS output for WEHI-265.1 monocytes alone with
approximate gate. (ii) Representative FACS output for backpacks alone with approximate
gate. (iii) Representative FACS output for solution containing conjugates, unmodified
monocytes and unattached backpacks with approximate gates. (iv) Representative FACS
data of monocyte-bound BPs via FACS, showing unmodified monocytes (red), unattached
BPs (purple), and conjugates (green) showing both >50% BPs attached to monocytes and
>50% of monocytes with at least 1 BP attached at a 1:1 (cell:BP) incubation ratio. Inset
shows representative histogram as a function of fluorescent intensity from the same
experiment.
122
MTT Assay
24 Hours
1.2
T
1
T
0.8
0.6
L
0.4
0.2
0
Backpacks
Negative
Control
3 pm Plin tgG-8 3pm
PS Paliclos PS Pawlicles
Figure 7-13: Toxicity assay for backpacks and WEHI-265.1 monocytes. 24 hour MTT assay
showing limited toxicity to monocytes incubated with BPs functionalized with IgG-b.
Values represent mean ± SD (n = 3).
Inflamed Tissue
With Monocyte Targeting
le ele
[ele le *
Healthy Tissue
Without Monocyte Targeting
ele*
3: 3:
le
e
.Ie-e1elee1e1
it 1:
'
let
le
1e le l e
e
Figure 7-14: Schematic of Monocyte Targeting in vivo Experiment. Acute inflammation
caused by LPS will cause the monocytes to localize to the lung (left), while in the noninflamed case (right) minimal accumulation will occur.
123
a
3
CL
0.
0
Oug LPS
5Wg LPS
I10ug LPS
Figure 7-15: Expression of ICAMI-1 following LPS induction in BALB/c mice. ICAM-1
expression in lungs 24 hours after intranasal administration of 0 ptg, 5 ptg, and 10ptg LPS.
7.3.2
Experimental Section
Materials:
PMAA (Aldrich, M=100kDa) PVPON (Sigma, M=1300kDa), PAH (Polysciences, M=150kDa),
EMG 705 anionic MNP (Ferrotec), PAA (Polysciences, M=450kDa) N-hydroxysuccinimidebiotin (Sigma) was used as received. To fabricate PAH-biotin, 25mg of PAH was dissolved in
50mM MES buffer at pH 8. 15.3mg of NHS- biotin was dissolved in DMSO at 22mg/mI before
adding to the PAH solution resulting in ~15% amine substitution. The solution reacted for 4
hours before it was purified by dialysis overnight into water.
Backpack Fabrication:
The nomenclature for LbL follows (poly IX/poly2Y)z, where X and Y are the pH of the pol ymer
solutions and z is the number of bilayers deposited (I bilayer = polyl+poly2). A non-integer
value of z indicates the assembly was terminated with polyl. Backpacks are formed through a
combination of photolithography and polyelectrolyte multilayers as previously described.(23) All
124
backpacks tested had the structure of (PMAA2/PVPON2)
30 .5
(PAH3/MNP4)1 0 .5 (PAA4/PAH-
biotin4)8 . The attachment region (PAA4/PAH-biotin4) had 150mM NaCl added to the polymer
solutions.
Backpack Preparation
Fluorescent (TRITC, FITC) streptavidin (Pierce Thermo) was diluted with 0.1 M (pH 5) sodium
acetate to 1 00ug/ml. 600ul streptavidin solution was pipetted and spread uniformly onto glass
slides containing BPs for 30 minutes at room temperature. Glass slides were then washed in 0.1
M sodium acetate (pH 5) for 30s to 1 min and allowed to air dry. Mouse IgG functionalized with
biotin, IgG-B (Santa Cruz Biotechnologies), was diluted to 100ug/ml with PBS and pipetted onto
glass slide containing BPs as above. For 3H studies, 31H-labeled biotin (American Radiolabeled
Chemicals) was mixed with IgG-B to radiolabel BPs. After 30 minutes, BPs (in PBS) were
gently scraped off of the glass slide surface and centrifuged at 50OOg for 10 minutes. IgG-B
containing supernatant was removed and BPs were washed once in PBS.
Attachment of BPs to Monocytes
A known number of BPs, in suspension, were incubated with monocytes at 37 0 C for 30 minutes
in PBS. Unbound BPs were separated from conjugates via centrifugation at 200g. Conjugates
were imaged on an Olympus Fluoview 500 (Olympus America Inc.).
Differentiationof Monocytes
WEHI-265.1 monocytes without BPs attached were incubated in 6 well plates with 50nM
phorbol 12-myristate 13-acetate (Sigma), PMA, in standard cell culture media. In parallel,
125
WEHI-265.1 monocytes were incubated with FITC labeled BPs and sorted for using FACS.
FACS sorted BP-monocyte conjugates were incubated in 6 well plates at identical conditions.
TransmigrationAssay
HUVECs were plated in the upper chamber of Coming Transwell plates (3pm, 6 well) at 6x10 5
cells per well in M200 media and allowed to attach and grow to form a complete monolayer
(confirmed with TEER measurements), after which media was changed to DMEM in all
chambers.6x10 5 WEHI-265.1 monocytes (also in DMEM), either with none or a known number
of FITC-labeled BPs, were pipetted on top of the HUVEC monolayer in the top chambers, which
was then suspended above the below chamber containing DMEM with 1 OnM Monocyte
Chemoattractant Protein-1 (Fisher). Monocytes which passed through the HUVEC monolayer
were counted manually on a hemocytometer. BP migration was quantified via fluorescent
reading (normalized to applied dose) of BPs that had migrated into the lower chamber.
Inflammation Model and Biodistribution
Healthy female BALB/c mice (18-20g) were administered 1 Ogg of lipopolysaccharide (LPS) in
60ul saline intranasally. 3x10 6 WEHI-3 monocytes conjugated with 1.5x106 3H-labeled
backpacks (2:1 cell:BP ratio) in saline were then intravenously injected into either healthy
female BALB/c mice (1 8-20g; n=4) or LPS challenged female BALB/c mice (1 8-20g; n=4). At 6
hours, mice were sacrificed via CO 2 overdose and known organ weights were harvested and
dissolved at 60*C in 5ml Solvable (Perkin Elmer) overnight. 5ml Ultima Gold was then added to
each sample, allowed to cool, and measured for 3H content in a Packard TriCarb 2100TR
126
scintillation counter. All animal protocols were approved by the Institutional Animal Care and
Use Committee (IACUC) at the University of California, Santa Barbara.
Cell culture
WEHI-265.1 monocytes (ATCC) were incubated at 37 0 C in 5% CO 2 in high glucose DMEM,
10% FBS, 1%
penicillin/streptomycin, and 0.05 M 2-mercaptoethanol. HUVECs (Invitrogen)
were incubated at 37 0 C in 5% CO 2 in Medium 200 with low serum growth supplement (LSGS)
and 1% penicillin/streptomycin. J774 macrophages were incubated at 37 0 C in 5% CO 2 in high
glucose DMEM, 10% FBS, 1% penicillin/streptomycin. WEHI-3 monocytes were incubated at
37 0 C in 5% CO 2 in Iscove's modified DMEM, 10% FBS, 1% penicillin/streptomycin, and 0.05
M 2-mercaptoethanol.
Fluorescence-activatedcell sorting (FACS)
BPs, monocytes, and conjugates were suspended in PBS. A FACSAria Flow Cytomer with a
70um nozzle was used to quantify attachment of BPs to monocytes. Mammalian cells were
filtered prior to analyzing sample. BD FACSDiva software 6.0 was used to analyze data.
MTT Assay
WEHI-265.1 monocytes were seeded in a 96-well plate at 1.5x1 04 cells per well for 24 hours.
BPs, 3pm polystyrene (PS) particles (Polysciences), and 3pm IgG-B coated PS particles were
incubated with monocytes at a 1:1 (Cell:Particle) ratio. 1 Opl of MTT reagent was added to each
well and allowed to incubate in standard cell culture conditions for 2 hours. 100ul of Detergent
127
Reagent was then added and the plate was incubated in the dark at room temperature for 2 hours.
Absorbance was read at 570nm using a Tecan Safire plate reader.
Inflammation Model
To induce lung inflammation, healthy female BALB/c mice (18-20g) were administered 0-10ptg
of lipopolysaccharide (LPS) in 60ul saline intranasally. 24 hours later, the lungs were harvested
and mechanically homogenized and the expression of ICAM-1 was measured using an ELISA
kit (Ray Biotech).
128
7.4 Future Applications
Due to the unique phagocytosis resistant property of the backpack, the applications of this
material for cell-mediated drug delivery is exciting. We believe this could be applied to areas
ranging from delivering material across the blood-brain barrier to responding to chronic or acute
inflammation in the lungs. It is very difficult for nanoparticles or free floating drugs to cross the
blood-brain barrier but it has been found that macrophages traffic to inflammation regions in the
brain and they may help alleviate diseases like Parkinson's disease (166). Currently macrophages
are loaded with enzymes that are designed to escape endosomes and be excreted by the
macrophage near the desired location (166). However, a macrophage with an attached phacytosis
resistant backpack could allow for efficient delivery of proteins to the brain and other previously
inaccessible regions of the body.
129
7.5 Tissue Engineering Applications
This section will discuss how the size of immune cell aggregates depend on the backpack
size and ratio of cells to backpacks in solution. This section will also discuss methods to create
controllable and stable bilayers of living cells.
7.5.1
Controlled Aggregate Size using Cell Backpacks
Reproduced in part with permission from Swiston, A.J., Gilbert, J.B., Irvine, D.J., Cohen, R.E.,
and Rubner, M.F. Biomacromolecules 11, 1826-1832 (2010). Copyright 2010 American
Chemical Society.
We investigated whether by simply controlling the size of the backpack, and the ratio of
cells to backpacks, we could control the size of the resultant cell clusters. Controlling cell cluster
size is of value for creating aggregates of reproducible and consistent size for tissue engineering.
One could potentially imagine injecting cells and backpacks into a tissue void and using the
design rules discovered to create aggregates of controllable size. Alternatively for drug delivery
applications a single cell and single backpack would be desired and thus since this interaction
occurs in solution, the design rules presented here can help design the material. This work was
primarily completed by Albert Swiston, but relevant to the future of backpacks for tissue
engineering or drug delivery. In this work we have shown through a systematic study that the
backpack size and the ratio of cells to backpacks determines how many cells attach to create a
cluster(167).
130
20
(a)
15
R=10
C
R=5
+R=3
S
C
e
10
-
-R=1
micron Oft
05
10
100
Diameter (pm)
25
.5
10R0
Diameter (Mm)
Figure 7-16: Cluster Size Dependence on Patch Diameter A) Using a 7pim backpack B)
Using a 15pm backpack. R= (# cells)/(# Backpacks) (167)
In Figure 7-16 we show that the ratio of cells to backpacks and also the backpack diameter
influences the size of the small aggregates that occur(167). At lower R, meaning more backpacks
per cell, the CH27 B-cells will start to form multiple cells per backpack clumps or multibackpack, multi-cell aggregates. This effect is particularly seen with 1 5pm backpacks. We
believe that this close proximity of multiple immune cells might have uses in tissue engineering
areas such as creating lymphoid-like tissue(168).
7.5.2
Immune Cell Bilayers Assembled Using Layer-by-Layer Processing
*This work was done in collaboration with Girma Endale. Girma completed the experiments
while I helped image the results and design the experiments.
131
The specific hyaluronic acid-CD44 cell receptor interaction that attaches the backpack to a
cell can also create bilayers of non-adherent immune cells on a surface. These multilayers of Bcells are stable, fabricated from biopolymers and relatively simple to produce. They are exciting
since the surface immobilization of immune cells such as T-cells and B-cells in close proximity
is extremely important for development of biosensors(169). Furthermore this may allow for the
creation of free-floating films of immune cells.
A
B
Figure 7-17: Bilayers of B-cells stabilized by HA and CHI on a (HA3/CHI3) substrate A)
First layer of the cell structure and the structure is primarily green. B) The second layer of
the cellular bilayer and the structure is primarily red. Scale bar is 20 gm.
To build a bilayer of immune cells, the first step is to settle single layer of green labeled
B-cells onto a substrate of (HA3/CHI3) 3 .5. The fabrication conditions of this layer have been
previously optimized(lO). The second layer of red labeled cells were added after an intermediate
CHI, HA or (HA/CHI), polyelectrolyte layer. The resulting bilayer of cells was imaged with
confocal microscopy. Figure 7-17 shows the first layer is primarily composed of green labeled
B-cells and the second layer is primarily composed of red labeled B-cells. It was found that
intermediate CHI and HA/CHI were able to form a bilayer of cells, while intermediate HA was
132
unable to form a second layer. This result suggests that HA alone was antagonistic to the first
layer of cells and competitively bound the B-cell surface receptor, inhibiting the creation of a
second layer.
Furthermore, immune cell bilayers that were built with an intermediate (HA/CHI)
polyelectrolyte layer showed strong resistance to detachment during incubation at 100 and
200rpm in HBSS for 20 minutes. On the other hand, immune cell bilayers assembled on an
intermediate CHI layer were unable to resist agitation at 200rpm (data not shown). Bilayers
assembled only with an intermediate CHI layer fell off during the robustness test due to a weaker
electrostatic bond and the absence of CD44 based interactions. In contrast, the HA/CHI
intermediate based bilayers resisted detachment during the robustness test due to the combination
of electrostatic and CD44 based interactions. One role of CHI may be to stop soluble HA from
saturating the B-cell surface receptor of the already deposited cells and thus allow for strong
adhesion of the second layer of B-cells. The result suggests that robust cellular bilayers need
both HA-CD44 interactions as well as electrostatic attractions driven by CHI.
Experimental Details
Hyaluornic acid (HA,Fluka, from Streptococcus equi, MW 150kDa), chitosan (CHI, sigma,
deacetelated 0.85, MW~390KDa) 0.1%CHI (w/V) and 0.1% HA(w/V) solution were dissolved
in deionized water. Poly (diallyldimethylammonium chloride (PDAC, molecular weight 200-300
kDa), in 20 wt % in water solution), poly (styrene sulfonate) (SPS, MW 70 kDa) were purchased
from Sigma-Aldrich and used without purification. Immune Cells, CH27 B lymphocyte cell
lines, were passage and maintained in RPMI
133
cell culture media with L-glutamine,
streptomycin/penicillin (P/S), and Fetal calf serum were purchased from mediatech. Hank's
balanced salt solution Ix was purchased from Gibco. Clean glass slides (VWR) were used as
substrates. All polymer and biopolymer solutions were prepared with deionized water.
Glass Substrate Preparation
For single layer HA and CHI deposition, glass slides were sonicated in a 3% detergent aqueous
solution for 15 min followed by a 10 min wash in 1 M NaOH and two consecutive 5 min rinses
in deionized water and then air-dried before use.
Polyelectrolyte Solutions
Poly (diallyldimethylammonium chloride) (PDAC) and poly (styrene sulfonate) (SPS) solutions
(10-M) were prepared by dissolving the polymer in water with 100mM NaCl at concentrations
of 1 mg/ml and adjusted at pH 4.0. The biopolymers CHI and HA solution were also prepared
with 100 mM NaCl and adjusted at pH 3.0.
Polyelectrolyte Multilayer Film Assembly
Polyelectrolyte multilayers (PDAC4/SPS4)15.5 and (HA3/CHI3)3.5 were deposited on clean
glass substrates using an automatic dipping procedure. (PDAC4/SPS4)15.5 bilayers were
deposited with a Zeiss HMS Series Programmable Slide Stainer by immersing the substrate
alternatively in the polymer solution for 10 min followed by 2min and 1 min pH 4.0 water rinse.
The biopolymers (HA3/CHI3)3.5 bilayers were deposited on the (PDAC4/SPS4)15.5 layers with
spinning dipper by immersing the substrate alternately in the polymer solution for 10 min,
followed by 2,1,1 min pH 3.0 water rinse. For good immune cell adhesion, the prepared samples
were stored overnight at ambient conditions before performing immune cell adhesion.
134
Cell Culture
The immune cells (CH27 B lymphocytes) were maintained at 37 *C, 5% C02 and passaged 1:10
once every 3 days in RPMI-1640 culture media containing 10% fetal calf serum (FCS), and 1%
penicillin/ streptomycin (mediatech).
Immune Cells Multilayer Assembly
The immune cells were washed two times with HBSS and then, the cells resuspended at 106
cell/ml in warm cell culture media (RPMI-1640). Next, gently pipette 2ml of the suspension
directly onto the prepared HA/CHI multilayered surfaces in 35mm Petri dishes. Samples were
incubated for 15 min, followed by 15min agitation at 100 rpm for 15 min at 37'C and 5% C02
for a total time of 1 or 2hrs and then samples removed from the incubator and gently washed in
warm fresh HBSS to remove unbound cells. Samples were placed in warm RPMI media and
analyzed. Prior to the second and subsequent immune cell layer adhesion, the first layer immune
cells were incubated either in an intermediate HA/CHI mixture, or only in HA or CHI solution.
The above solutions were prepared in Hank's balanced salt solution Ix ( HBSS) at a
concentration of 0.01%CHI and 0.1% HA solution or 0.1%CHI and 0.01%HA. In the former
case, the first layer immune cells were placed in pH 7.4 HA solution and incubated for 10 min
and gently rinsed in HBSS and then transferred into pH 6.3 CHI solution and incubated for 1 min
followed by HBSS rinse. Then 2ml of 1 x 106 cell/ml immune cells were gently pipetted on the
sample, and incubated for 15 min followed by a gentle agitation at 100 rpm for 15 min. The
above steps were repeated for subsequent immune cell adhesion. After each polyelectrolyte layer
deposition, samples were rinsed in warm HBSS.
Robustness Test
135
The robustness of the immune cells polyelectrolyte multilayers were tested by intensive rinsing
and dipping of the sample in HBSS. In addition, samples were placed in HBSS solution in Petri
dish and agitated, for 1, 2, 5, 10 and 20 min, at 100 and 200rpm
136
7.6 Supplemental Information
(Poly(I:C)5/LPEI5) on top of
(PDAC4/SPS4) 1s.5
M2
S1.5
UD
8 0.5
3
0
5
10
15
20
25
Bilayers
Figure 7-18: Inclusion of synthetic RNA into a film: A) Film containing (Poly(I:C)5/LPEI5)
was built and is stable through acetone sonication. Thickness of (Poly(I:C)5/LPEI5) 20 was
350nm.
In the figure above a possible payload of synthetic double stranded RNA (200-600 basepairs)
was incorporated into a film which retained its thickness and UV absorbance even after acetone
sonication. Poly(I:C) is an immunostimulant molecule that interacts with the toll receptor 3 of
the immune system(170). This could serve to enhance the potency of a vaccine by acting as an
adjuvant.
137
A
D
Quantum Dot Backpacks
543nm excitation
160000
140000
B
-- PBS
120000
-MNP
100000
-QD
'QD,
80000
BPs
BPs
MNP BPs
.S60000
C
40000
20000
-
0
600
O
*
650
700
750
800
850
Wavelength (nm)
Figure 7-19: 7pm Quantum Dot full BPs as excited by 543nm HeNe Laser A) Backpack
Structure (PMAA2/PVPON2) 30. 5 (PAH3/MNP4) 10.5 (PAA4/PAH-b4)s B) Backpack
Structure (PMAA2/PVPON2) 30.5 (PAH5/QD5) 30.5 (PAA4/PAH-b4)s C) Backpack Structure
(PMAA2/PVPON2) 30.5 (PAH3/MNP4) 105. (QD5/PAH5) 30 (PAA4/PAH-b4) 8 D) Emission of
backpacks shown in A-C. Backpacks were released into PBS and excited at 543nm on a
spectrofluorometer.
138
8. ORIENTATION-SPECIFIC ATTACHMENT OF POLYMERIC MICROTUBES ON
CELL SURFACES
Reproduced with permission from: Gilbert, J. B., O'Brien, J. S., Suresh, H. S., Cohen, R. E. and
Rubner, M. F. (2013), Adv. Mater., 25: 5948-5952. C 2013 WILEY-VCH Verlag GmbH & Co.
KGaA, Weinheim
This work was profiled in Nature "Materials Science: To bind or not to bind" 502, 313-314 (17
Oct 2013)
8.1 Introduction
Understanding and controlling the interaction of living cells with synthetic micro or
nanoparticles is becoming increasingly important for a number of biomedical applications,
driven largely by the desire to limit cytotoxicity and produce more efficient drug delivery
systems. To enhance drug delivery efficiency, for example, material characteristics such as
particle shape and surface chemistry have been manipulated to tune the half-life, biodistribution
and cellular uptake of synthetic particles.(20, 135, 162, 164, 171) From such studies it has been
established that the local shape or orientation of a particle when it contacts a cell surface
drastically alters the internalization rate.(24, 29, 172, 173) Therefore of particular interest are
strategies to control the orientation of a particle on a cell surface.(174, 175) Chemically nonuniform, Janus or 'patchy' particles have been utilized to control local orientation in colloid
systems by presenting heterogeneously defined regions of attractive or repulsive chemistry.(176,
177) These simple building blocks can also self-assemble into a variety of higher order structures
depending on the placement and interaction strength of the heterogeneous regions.(25, 178-180)
In this work, we demonstrate a simple method to create functionally heterogeneous hollow
139
microtubes whose interactions with living cells exploit both the effects of particle shape and
chemistry. The net result is an ability to control the orientation of a tube shaped microparticle on
cell surfaces, thus opening up new design space for the fabrication of novel cell-biomaterial
hybrids.
In addition to considering the characteristics and potential of synthetic particles as
vehicles for drug delivery, the functionalization of a cell surface with synthetic materials to
create cell-biomaterial hybrids is an evolving new method to control cell phenotype, fate and
function.(22, 29, 111, 131, 181-183) A logical extension of this approach is to attach synthetic
particles directly to the surface of living cells to enhance their native functionalities and provide
additional capabilities not found in nature.(22, 29, 111, 184) Toward this goal, prior work in our
group has investigated the attachment of payload carrying cellular backpacks onto immune
system cells for cell-mediated drug delivery, imaging and magnetic manipulation.(23, 29) These
disc-shaped cellular backpacks, which are microns wide and a few hundred nanometers thick, are
able to resist internalization by macrophages due to their unique shape.(29) The ability to attach
functional, cargo carrying backpacks onto living immune cells without inducing phagocytosis
provides new opportunities for extracellular targeted drug delivery. In this work, we explore the
use of hollow microtubes with a controllable, non-uniform presentation of functional groups as a
new type of biomaterial with the means to control particle orientation on cell surfaces.
Controlling the orientation of a microparticle on a cell surface requires the use of an
anisotropic particle with distinct surface regions that promote and resist cellular attachment.
Surfaces that are highly hydrated are commonly used to resist cell attachment,(l 85) while
surfaces presenting particular ligands promote cell attachment.(l 62) Polyelectrolyte multilayers
(PEMs) can achieve such surface characteristics simply by choice of polyelectrolytes and the
140
assembly conditions.(10, 185) In this work, poly(acrylic acid)/poly(allylamine hydrochloride)
(PAA/PAH) multilayers assembled at pH 3 were used to create microtube surfaces that resist cell
interactions due to their highly hydrated nature.(185) To create regions of the microtubes that
strongly interact with the lymphocyte B-cells used in this study, multilayers containing the
biopolymers hyaluronic acid (HA) and chitosan (CHI) were used. Previous studies have shown
that the B-cell CD44 receptor binds strongly and specifically to HA present in a film.(10) HA
containing multilayers also could be used to interact with other cell types including macrophages
and T-cells.(23, 29) In all cases rhodamine (Rhod) labeled PAH and fluorescein (FITC) labeled
CHI were used to identify the two different multilayer systems (depicted as red and green colors
respectively in all figures).
8.2 Fabrication of Cellular Tubes
To fabricate anisotropic, chemically non-uniform microtubes, we used a simple template
approach involving the use of a sacrificial track-etched membrane.(4, 186-188) The PEM
assemblies were designated as follows: (polyl/poly2)z, where polyl and poly2 represent the
polymers used to assemble the multilayer and z is the total number of bilayers deposited (1
bilayer = polyl+poly2). The fabrication sequence for microtubes is shown in Figure 8-1. To
fabricate chemically uniform tubes, the entire membrane, including the pores, was coated
conformally with a PEM (Figure 8-1 step i).(189, 190) For completely cell-resistant tubes,
multilayers of (PAA/PAH-Rhod) were assembled.(l 85) The coated membrane was then plasma
treated to remove the polymer selectively from the top and bottom surfaces of the membrane;
previous work has demonstrated that the polymers within the pores are protected from the
plasma etching process (Figure 8-1 step ii).(4) Carbodiimide chemistry was then used to crosslink the PAH amine groups to the PAA carboxylic acid groups to form stable amide bonds. The
141
membrane was then selectively dissolved and single component cell-resistant tubes filtered out
(Figure 8-1 step iii).
Track-etch
poly(carbonate) membrane
(v) Plasma etch
and cross-link if desired
(i) Deposit polyelectrolyte
multilayer
vi) Dissolve
membrane and collect
heterostructured tubes
(iv) Deposit polyelectrolyte
multilayer
(Ii) Plasma etch()
and cross-link if desired
Dissolve membrane
nd collect single
component tubes
Figure 8-1: Fabrication Scheme of Orientation Controlled Microtubes
To fabricate orientation-specific tubes, a cell-adhesive multilayer containing HA and CHI
was deposited onto a membrane previously coated with (PAA/PAH-Rhod) (Figure 8-1 step iv).
Another plasma treatment step removed the polymer film from the exposed surfaces, leaving the
polymers only in the pores (Figure 8-1 step v). The resultant heterostructured tubes in the
membrane pores have cell-adhesive multilayers on the inside of the tube and therefore will be
largely shielded from interacting with cells. However, as will be shown, when a cell approaches
the end of the tube, the polymer from the interior of the tube is sufficiently exposed to participate
in cell adhesion. The tubes were then cross-linked before the polycarbonate membrane was
dissolved and the tubes filtered out (Figure 8-1 step vi). The fabrication method yields millions
of stable tubes per membrane and aqueous suspensions containing around 106 tubes/mL.
Scanning electron micrographs of typical microtubes are seen in Figure 8-2.
142
Figure 8-2: Scanning electron micrographs of cell tubes with (PAA/PAH-Rhod) 40 outside
and (HA/FITC-CHI)40 inside on the surface of a syringe filter.
8.3 Effect of Chemistry on the Cell Surface Orientation
To investigate how the structure of the microtubes affects the orientation on cell surfaces,
three microtube groups were fabricated and imaged with confocal microscopy (Figure 8-3). The
first group (Figure 8-3a) consists of red labeled cell-resistant (PAA/PAH-Rhod) microtubes that
were fabricated by ending at step (iii) in Figure 1. The second group (Figure 8-3b) consists of
heterostructured microtubes with red labeled cell-resistant (PAA/PAH-Rhod) multilayers on the
outside and green labeled cell-adhesive (HA/FITC-CHI) multilayers on the interior which were
fabricated by continuing to step (vi) in Figure 1. The third group (Figure 8-3c) is produced by
reversing the order of film deposition such that green labeled cell-adhesive (HA/FITC-CHI)
multilayers are on the outside and red labeled cell-resistant (PAA/PAH-Rhod) multilayers are on
the interior of the tube. Figure 8-3b shows the green FITC-CHI presence at both ends, and in
most cases localized at the ends of the tubes. The clumping at the ends of the tubes is believed to
143
be caused by the (HA/FITC-CHI) multilayer clogging or plugging the tube during fabrication. In
Figure 8-3c the overlay images do not distinctly show the green FITC-CHI around the tube
exterior surface since the layer containing FITC labeled chitosan is thin, but as seen in the inset,
upon splitting into independent FITC and rhodamine channels, the full coating of chitosan is
displayed. Additional images of the three tube types are shown in Figure 8-6.
a)
c)
0
0
C'
b)
0_
yU
gi
TT
Figure 8-3: Confocal microscopy images of heterostructured tubes dried on a slide along
with cartoons of the structure. Red represents the cell-resistant rhodamine conjugated
(PAAIPAH) multilayers and green represents the cell-adhesive FITC conjugated (HA/CHI)
multilayers. a) (PAA/PAH-Rhod) 40 b) (PAA/PAH-Rhod) 40 outside and (HA/FITC-CHI)
40
inside c) (HA/FITC-CHI)4 outside and (PAA/PAH-Rhod) 40 inside. Full image scale bars are
10pm and inset scale bars are 5pm.
To study the cell-microtube interaction, tubes suspended in water were added to a
solution of B-cells. B-cells were chosen since they are non-adherent lymphocytes that are
144
integral to the adaptive immune response and thus are therapeutic targets in immune system
diseases. (191,
192) After 1-2 hours confocal imaging was used to analyze attachment
orientation.
The quantitative results shown in Figure 8-4a were obtained by analyzing over 100 tubes
each in at least 3 different experiments and assigning each interaction to one of three possible
categories; i) unattached, ii) end-on attachment or iii) side-on attachment. The first group of
(PAA/PAH-Rhod) tubes minimally associated with cells due to the highly hydrated, nonadhesive nature of the (PAA/PAH-Rhod) multilayer and therefore >80% of the tubes were
unattached to any cell. The second group of tubes with (HA/FITC-CHI) multilayers on the
interior attached preferentially end-on, consistent with the cell interacting with HA on either end
of the microtubes. The third group of tubes with (HA/FITC-CHI) multilayers on the outside
preferentially attached side-on, consistent with the presentation of cell-adhesive regions on the
outside of the tube. Confocal microscopy images of cell-tube interactions are shown in Figure
8-4b with larger images in Figure 8-5. Approximately 50% of the tubes in groups 2 and 3 did not
attach after 1-2 hours of agitation and incubation. This time was chosen as the analysis point
since longer intervals of agitation and incubation led to increased attachment and the formation
of large aggregates, limiting the ability to determine the attachment orientation. Conditions were
not optimized to achieve the highest levels of tube attachment; future work focuses on
immobilizing cells on a surface to minimize the aggregation potential. Nevertheless, these data
clearly show a preference for microtube attachment based on the location of the cell binding
(HA/FITC-CHI) multilayers. Figure 8-4c compares the ratio of tubes end-on to side-on and
further supports the notion that the presentation of HA on the ends of the tubes favors end-on cell
surface orientation. Some of the side-on orientation observed with the (HA/FITC-CHI)
145
multilayers on the inside of microtubes may simply be due to the cell binding the HA from both
ends of the tube at the same time. We are confident that HA is not diffusing through the
(PAA/PAH-Rhod) outside multilayer, since previous work has documented the effective
blocking properties of PAH-containing layers.(92) Regarding the issue of cytotoxicity, Figure
8-4d shows that the microtubes did not decrease cellular viability.
a)
1
2
c)
0.9-
0le
0.8-
1natahed
0.7-
C
*
FZ7 End-on
1.8
MSide-on
2 1.4
C
9 1.2
0
0.8-
C
w
0.5I-
0.4-
'U
0.2-
0
b)
r+li
(PA.AIPAH) only
Ii
(HA/CHI) inside
0.8
0.6
11
2
0.3
0.1 .
1'
0.4
0.2
0
(HAJCHI) outside
(PAA/PAH) only
(HMACHI)
inside
(HA/CHI) outside
d) 0.9
0.8
0.7
> 0.6
0.5
8
0.4
U. 0.3
0.2
0.1
0
uontro
Ue- Iuoes
Figure 8-4: Controlled Orientation of Microtubes on the Surface of B-cells: a) Samples
were analyzed using confocal microscopy and microtubes with cell-adhesive (HA/CHI) on
the inside caused more end-on attachment. b) Confocal images of cell tube interactions i)
(PAA/PAH-Rhod)40 ii) (PAA/PAH-Rhod) 40 outside and (HA/FITC-CHI)4 0 inside iii)
(HA/FITC-CHI) 4 outside and (PAA/PAH-Rhod)40 inside. Scale bar 10 Pm. c) Ratio of 'Endon' to 'Side-on' showing effect of tube structure d) Cellular viability of the B-cells after 24
hr incubation with tubes. Viability was not altered by the addition of (HA/CHI) outside
tubes added at a 1:1 cell to tube ratio.
146
a)
C)
b)
Figure 8-5: Representative confocal images of cell-tube orientation study after 1-2 hours a)
(PAA/PAH-Rhod) 40 b) (PAA/PAH-Rhod) 40 outside and (HA/FITC-CHI)40 inside c)
(HA/FITC-CHI)4 outside and (PAA/PAH-Rhod)40 inside.
8.4 Conclusion and Future Outlook
In conclusion, we have developed a novel and reproducible fabrication method that
allows for the presentation of a selected polymer preferentially on the ends of microtubes. Our
method provides a general platform that could be expanded to many applications in biomaterials
and responsive materials. In particular, we have shown in the present work that by limiting the
cell-adhesive polymer to the end of the microtube, we can control the orientation of an
anisotropic microtube on cell surfaces. These results point to the possibility of using microtubes
as part of a bottom-up cell scaffold in the future. Controlling the interaction points on the
microparticle may allow for structures such as linear or branched 'polymerized' chains of cells.
In addition, regions of the microtube can be designed to respond to stimuli such as pH, salt or
temperature, opening the possibility of stimuli sensitive polymer tubes for drug delivery or
controlled assembly.(155, 186, 188, 193-197) In general, we believe that controlling the
147
orientation of microparticles on the cell surface could open new design space for applications in
tissue engineering and drug delivery.
8.5 Experimental Details
Materials: PAA (Aldrich, M=450kDa), PAH (Aldrich, M=56kDa), HA (from Streptococcus
equip, Fluka, M~1580 kDa), I-Ethyl-3-[3-dimethylaminopropyl] carbodiimide hydrochloride (EDC,
Sigma), N-hydroxylsuccinimide (NHS, Sigma), NHS-Rhodamine (Pierce), Nuclepore track-etch
polycarbonate membrane 3.0pm (Whatman), Fluorescein isothiocyanate (FITC, Sigma), MillexLCR 0.45ptm hydrophilic PTFE 25mm membranes (Millipore) and low molecular weight CHI
(deacetylation 0.9, M=50-190kDa, Sigma) were used as received. FITC-chitosan was made from
in a 1%
aqueous acetic acid-methanol (1:2 v/v) solution of pH 4.5 at room temperature as
described.(198) Rhodamine-PAH was synthesized by reaction between PAH and NHSrhodamine in 50mM MES buffer at pH 8 for 2hrs, after the solution was dialyzed into water.
Detailed tube fabrication steps are given in supporting information.
Fabrication of Microtubes: Polymers solutions were made from Milli-Q 18.2MO water.
Solutions of PAA and PAH-Rhod were 0.01M and solutions of HA and FITC-CHI were 0.1%
and 0.05% (w/v) respectively, FITC-CHI had 0.1M of acetic acid added to aid dissolution. HA
and CHI solutions had 100mM NaCl added to the solution. All solution pHs were adjusted to pH
3.0 with IM HCl. Track-etch polycarbonate membranes were immobilized on teflon holders and
sequentially dipped in the polymer solutions using a nanoStrata dipping unit. Polymer
depositions were done for 10 minutes, followed by three pH3 Milli-Q water rinses: one for 2
minutes and two for 1 minute, while the substrate was rotating at 100rpm. Both sides of the
membrane were exposed to plasma etching simultaneously. The chamber was flushed with 02
148
twice before igniting the plasma at 150 mTorr. (PAA/PAH-Rhod) 4o and (HA/FITC-CHI) 40 layers
required 20 min and 60 min of etching respectively. Following plasma etching, the membrane
was gently rubbed with a moist q-tip to fully remove the damaged film. The tubes were crosslinked by EDC/NHS at a concentration of 60mM EDC and 24mM NHS in 50mM MES buffer at
pH 5 for 15 minutes. The membrane was dissolved in dichloromethane and the tubes filtered
with a Millex-LCR filter before rinsing with 1OmL of dichloromethane. The filter surface was
wet with water and gently rubbed with a cell scraper for 5 minutes to transfer the microtubes.
The concentration of tubes were measured using a standard hemocytometer on a fluorescent
microscope. If aggregation occurs, light sonication can disperse the tubes.
Cell Study: CH27 B-cells were grown in RPMI media with 10% fetal calf serum and 1%
penicillin/streptomycin. Before incubation with microtubes the cells were washed with Hanks
Buffered Salt Solution and resuspended at 106 cells/mL in complete media. The tubes and cells
were iteratively incubated on a shaker plate for 15 minutes at 100rpm and then incubated for 15
minutes for 1-2 hours before analysis. Confocal images were taken on a Zeiss LSM 510 using a
488nm Argon laser for FITC excitation and a 543nm HeNe laser for rhodamine excitation. The
images were visually analyzed to determine the orientation of the microtube. A minimum of 100
interactions were analyzed per sample. Cell viability was determined by trypan blue staining.
149
8.6 Supplemental Figures
a)
c)
0
0
U'
Y
b)
0
I
U
-C C)
II
T T
Figure 8-6: Confocal microscopy images of heterostructured tubes dried on a slide along
with cartoons of the structure. Red represents the cell-resistant rhodamine conjugated
(PAA/PAH) multilayers and green represents the cell-adhesive FITC conjugated (HA/CHI)
multilayers. a) (PAA/PAH-Rhod)40 b) (PAA/PAH-Rhod)40 outside and (HA/FITC-CHI)
40
inside c) (HA/FITC-CHI)4 outside and (PAA/PAH-Rhod)40 inside. Image scale bars are
10pm.
150
9. FUTURE WORK
Polymeric nanostructured thin films have a wide variety of applications and through new
experimental techniques, we can begin to understand the underlying design principles involved.
Depth Profiling XPS using C60+ sputtering is only beginning to be developed for analyzing
nanostructured films. I hope future members of the Rubner and Cohen group and many others
will continue to exploit its strengths to determine the chemical and atomic structure of polymeric
thin films. This technique could help define the necessary characteristics to achieve the desired
nanostructure. In particular, the technique is suited to help determine the true mechanism behind
exponential film growth by analyzing the diffusion and exchange of material throughout the film
during growth. Another area in which I believe depth profiling XPS will be utilized is in lamellar
block copolymer films. We have chosen to look at lithium-ion distribution block copolymer
films, but given the exceptionally wide ranging applications of block copolymers, I believe that
once this technique is clearly displayed, many other areas of investigation will begin.
The biomedical applications of structured thin films are equally broad, and I believe that the
surface of the cell presents new ground for scientists to manipulate and combine synthetic
materials with natural cell functions. This combination can be applied to a variety of applications
including diagnostic and therapeutic uses. Currently, our synthetic backpack material is
uninvolved, with no major effect on the attached cell; the cell simply serves as a cargo carrying
device. In particular, we saw that one can use the natural homing ability of the cell to carry the
therapeutic cargo to a disease site, thus reducing general toxicity and allowing for more efficient
therapeutic delivery. The cargo could also be diagnostic since the cell could be labeled and
tracked as it migrates through the body. In general, if synthetic materials can be attached to a cell
without deleteriously altering it, the material may serve as a platform for a variety of molecules,
151
particles, ligands and/or responsive devices. Alternatively, the material could be designed to
promote artificial homing capabilities, signaling capabilities and other new functionalities that
are not biologically natural. These new capabilities could greatly expand the flexibility of the
cellular 'platform' and open up a new level of cellular engineering. With so many possibilities, I
am excited for the many new applications of nanostructured polymeric materials in the future.
152
10. CITATIONS
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
Lee D, Cohen RE, Rubner MF (2005) Antibacterial properties of ag nanoparticle loaded
multilayers and formation of magnetically directed antibacterial microparticles. Langmuir
21(21):9651-9659.
Nuraje N, Asmatulu R, Cohen RE, Rubner MF (2010) Durable antifog films from layerby-layer molecularly blended hydrophilic polysaccharides. Langmuir 27(2):782-791.
Nolte AJ, Rubner MF, Cohen RE (2004) Creating effective refractive index gradients
within polyelectrolyte multilayer films: Molecularly assembled rugate filters. Langmuir
20(8):3304-3310.
Chia KK, Rubner MF, Cohen RE (2009) Ph-responsive reversibly swellable nanotube
arrays. Langmuir 25(24):14044-14052.
Lee D, Nolte AJ, Kunz AL, Rubner MF, Cohen RE (2006) Ph-induced hysteretic gating
of track-etched polycarbonate membranes: Swelling/deswelling behavior of
polyelectrolyte multilayers in confined geometry. JAm Chem Soc 128(26):8521-8529.
Macdonald M, Rodriguez NM, Smith R, Hammond PT (2008) Release of a model protein
from biodegradable self assembled films for surface delivery applications. J Control
Release 131(3):228-234.
Kim BS, Smith RC, Poon Z, Hammond PT (2009) Mad (multiagent delivery) nanolayer:
Delivering multiple therapeutics from hierarchically assembled surface coatings.
Langmuir 25(24):14086-14092.
Lee D, Rubner MF, Cohen RE (2006) All-nanoparticle thin-film coatings. Nano Lett
6(10):2305-2312.
Choi J, Rubner MF (2005) Influence of the degree of ionization on weak polyelectrolyte
multilayer assembly. Macromolecules 38(1):116-124.
Vasconcellos FC, Swiston AJ, Beppu MM, Cohen RE, Rubner MF (2010) Bioactive
polyelectrolyte multilayers: Hyaluronic acid mediated b lymphocyte adhesion.
Biomacromolecules 11 (9):2407-2414.
Zhang JT, Chua LS, Lynn DM (2004) Multilayered thin films that sustain the release of
functional DNA under physiological conditions. Langmuir 20(19):8015-8021.
Lutkenhaus JL, Hammond PT (2007) Electrochemically enabled polyelectrolyte
multilayer devices: From fuel cells to sensors. Soft Matter 3(7):804-816.
Decher G (1997) Fuzzy nanoassemblies: Toward layered polymeric multicomposites.
Science 277(5330):1232-1237.
Boudou T, Crouzier T, Ren KF, Blin G, Picart C (2010) Multiple functionalities of
polyelectrolyte multilayer films: New biomedical applications. Adv Mater 22(4):441-467.
Ono SS, Decher G (2006) Preparation of ultrathin self-standing polyelectrolyte multilayer
membranes at physiological conditions using ph-responsive film segments as sacrificial
layers. Nano Lett 6(4):592-598.
Kharlampieva E, Sukhishvili SA (2006) Hydrogen-bonded layer-by-layer polymer films.
Polymer Reviews 46(4):377-395.
Hammond PT (2011) Engineering materials layer-by-layer: Challenges and opportunities
in multilayer assembly. AIChE Journal57(11):2928-2940.
Hallinan DT, Balsara NP (2013) Polymer electrolytes. Annual Review of Materials
Research, Vol 43 43:503-+.
153
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
Gomez ED, Panday A, Feng EH, Chen V, Stone GM, Minor AM, Kisielowski C,
Downing KH, Borodin 0, Smith GD, Balsara NP (2009) Effect of ion distribution on
conductivity of block copolymer electrolytes. Nano Lett 9(3):1212-1216.
Doshi N, Mitragotri S (2009) Designer biomaterials for nanomedicine. Advanced
FunctionalMaterials 19(24):3843-3854.
Champion JA, Katare YK, Mitragotri S (2007) Particle shape: A new design parameter
for micro- and nanoscale drug delivery carriers. J Control Release 121(1-2):3-9.
Stephan MT, Irvine DJ (2011) Enhancing cell therapies from the outside in: Cell surface
engineering using synthetic nanomaterials. Nano Today 6(3):309-325.
Swiston AJ, Cheng C, Um SH, Irvine DJ, Cohen RE, Rubner MF (2008) Surface
functionalization of living cells with multilayer patches. Nano Lett 8(12):4446-4453.
Champion JA, Mitragotri S (2006) Role of target geometry in phagocytosis. Proc Natl
AcadSci USA 103(13):4930-4934.
Chen Q, Bae SC, Granick S (2011) Directed self-assembly of a colloidal kagome lattice.
Nature 469(7330):381-384.
Cebeci FC, Wu ZZ, Zhai L, Cohen RE, Rubner MF (2006) Nanoporosity-driven
superhydrophilicity: A means to create multifunctional antifogging coatings. Langmuir
22(6):2856-2862.
Hiller J, Mendelsohn JD, Rubner MF (2002) Reversibly erasable nanoporous antireflection coatings from polyelectrolyte multilayers. Nat Mater 1(1):59-63.
Zelikin AN (2010) Drug releasing polymer thin films: New era of surface-mediated drug
delivery. ACS Nano 4(5):2494-2509.
Doshi N, Swiston A, Gilbert JB, Alcaraz M, Cohen R, Rubner M, Mitragotri S (2011)
Cell-based drug delivery devices using phagocytosis- resistant backpacks. Adv Mater
23(12):H105-H109.
Wood KC, Chuang HF, Batten RD, Lynn DM, Hammond PT (2006) Controlling
interlayer diffusion to achieve sustained, multiagent delivery from layer-by-layer thin
films. Proc Natl Acad Sci USA 103(27):10207-10212.
Lavalle P, Voegel JC, Vautier D, Senger B, Schaaf P, Ball V (2011) Dynamic aspects of
films prepared by a sequential deposition of species: Perspectives for smart and
responsive materials. Adv Mater 23(10):1191-1221.
Zhai L, Nolte AJ, Cohen RE, Rubner MF (2004) Ph-gated porosity transitions of
polyelectrolyte multilayers in confined geometries and their application as tunable bragg
reflectors. Macromolecules 37(16):6113-6123.
Wu Z, Lee D, Rubner MF, Cohen RE (2007) Structural color in porous, superhydrophilic,
and self-cleaning sio2/tio2 bragg stacks. Small 3(8):1445-1451.
Kim JK, Park HS, Rhee DK, Ham SJ, Lee KJ, Yoo PJ, Park JH (2012) Ultrathin nanoclay
films with tunable thickness as barrier layers in organic light emitting devices. Journalof
Materials Chemistry 22(16):7718-7723.
Kim KH, Moon JH, Kim EY, Kim HJ, Jang SH, Lee WI (2012) Interface control with
layer-by-layer assembled ionic polymers for efficient low-temperature dye-sensitized
solar cells. Journalof Materials Chemistry 22(22):11179-11184.
Cho J, Caruso F (2003) Polymeric multilayer films comprising deconstructible hydrogenbonded stacks confined between electrostatically assembled layers. Macromolecules
36(8):2845-2851.
154
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
Podsiadlo P, Michel M, Critchley K, Srivastava S, Qin M, Lee JW, Verploegen E, Hart
AJ, Qi Y, Kotov NA (2009) Diffusional self-organization in exponential layer-by-layer
films with micro- and nanoscale periodicity. Angewandte Chemie-InternationalEdition
48(38):7073-7077.
Yoo PJ, Zacharia NS, Doh J, Nam KT, Belcher AM, Hammond PT (2008) Controlling
surface mobility in interdiflusing polyelectrolyte multilayers. Acs Nano 2(3):561-571.
Francius G, Hemmerle J, Ball V, Lavalle P, Picart C, Voegel JC, Schaaf P, Senger B
(2007) Stiffening of soft polyelectrolyte architectures by multilayer capping evidenced by
viscoelastic analysis of afm indentation measurements. Journal of Physical Chemistry C
111(23):8299-8306.
Picart C, Mutterer J, Richert L, Luo Y, Prestwich GD, Schaaf P, Voegel JC, Lavalle P
(2002) Molecular basis for the explanation of the exponential growth of polyelectrolyte
multilayers. Proc Natl Acad Sci U S A 99(20):12531-12535.
Lavalle P, Vivet V, Jessel N, Decher G, Voegel JC, Mesini PJ, Schaaf P (2004) Direct
evidence for vertical diffusion and exchange processes of polyanions and polycations in
polyelectrolyte multilayer films. Macromolecules 37(3):1159-1162.
Haynie DT, Cho EH, Waduge P (2011) "In and out diffusion" Hypothesis of exponential
multilayer film buildup revisited. Langmuir 27(9):5700-5704.
Porcel C, Lavalle P, Decher G, Senger B, Voegel JC, Schaaf P (2007) Influence of the
polyelectrolyte molecular weight on exponentially growing multilayer films in the linear
regime. Langmuir 23(4):1898-1904.
Garza JM, Schaaf P, Muller S, Ball V, Stoltz JF, Voegel JC, Lavalle P (2004)
Multicompartment films made of alternate polyelectrolyte multilayers of exponential and
linear growth. Langmuir 20(17):7298-7302.
Richert L, Lavalle P, Payan E, Shu XZ, Prestwich GD, Stoltz JF, Schaaf P, Voegel JC,
Picart C (2004) Layer by layer buildup of polysaccharide films: Physical chemistry and
cellular adhesion aspects. Langmuir 20(2):448-458.
Lundin M, Blomberg E, Tilton RD (2010) Polymer dynamics in layer-by-layer
assemblies of chitosan and heparin. Langmuir 26(5):3242-325 1.
Lee L, Johnston APR, Caruso F (2012) Probing the dynamic nature of DNA multilayer
films using forster resonance energy transfer. Langmuir 28(34):12527-12535.
Zacharia NS, Modestino M, Hammond PT (2007) Factors influencing the interdiffusion
of weak polycations in multilayers. Macromolecules 40(26):9523-9528.
Soltwedel 0, Ivanova 0, Nestler P, Muller M, Kohler R, Helm CA (2010) Interdiffusion
in polyelectrolyte multi layers. Macromolecules 43(17):7288-7293.
Xu L, Ankner JF, Sukhishvili SA (2011) Steric effects in ionic pairing and
polyelectrolyte interdiffusion within multilayered films: A neutron reflectometry study.
Macromolecules44(16):6518-6524.
Jomaa HW, Schlenoff JB (2005) Salt-induced polyelectrolyte interdiffusion in
multilayered films: A neutron reflectivity study. Macromolecules 38(20):8473-8480.
Arys X, Laschewsky A, Jonas AM (2001) Ordered polyelectrolyte "Multilayers". 1.
Mechanisms of growth and structure formation: A comparison with classical fuzzy
"Multilayers". Macromolecules 34(10):3318-3330.
Nobuta T, Ogawa T (2009) Depth profile xps analysis of polymeric materials by c-60 (+)
ion sputtering. Journalof MaterialsScience 44(7):1800-1812.
155
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
Szakal C, Sun S, Wucher A, Winograd N (2004) C-60 molecular depth profiling of a
model polymer. Applied Surface Science 231:183-185.
Hollander A, Haupt M, Oehr C (2007) On depth profiling of polymers by argon ion
sputtering. PlasmaProcessesand Polymers 4(9):773-776.
Chen YY, Yu BY, Wang WB, Hsu MF, Lin WC, Lin YC, Jou JH, Shyue JJ (2008) X-ray
photoelectron spectrometry depth profiling of organic thin films using c-60 sputtering.
Anal Chem 80(2):501-505.
Yu BY, Chen YY, Wang WB, Hsu MF, Tsai SP, Lin WC, Lin YC, Jou JH, Chu CW,
Shyue JJ (2008) Depth profiling of organic films with x-ray photoelectron spectroscopy
using c(60)(+) and ar(+) co-sputtering. Anal Chem 80(9):3412-3415.
Postawa Z, Czerwinski B, Szewczyk M, Smiley EJ, Winograd N, Garrison BJ (2004)
Microscopic insights into the sputtering of ag{111} induced by c-60 and ga
bombardment. JournalofPhysical Chemistry B 108(23):7831-7838.
Mahoney CM (2010) Cluster secondary ion mass spectrometry of polymers and related
materials. Mass Spectrom Rev 29(2):247-293.
Zacharia NS, DeLongchamp DM, Modestino M, Hammond PT (2007) Controlling
diffusion and exchange in layer-by-layer assemblies. Macromolecules40(5):1598-1603.
Mertz D, Vogt C, Hemmerle J, Mutterer J, Ball V, Voegel JC, Schaaf P, Lavalle P (2009)
Mechanotransductive surfaces for reversible biocatalysis activation. Nat Mater 8(9):73 1735.
Hong J, Shah NJ, Drake AC, DeMuth PC, Lee JB, Chen JZ, Hammond PT (2012)
Graphene multilayers as gates for multi-week sequential release of proteins from
surfaces. Acs Nano 6(1):81-88.
Peralta S, Habib-Jiwan JL, Jonas AM (2009) Ordered polyelectrolyte multilayers:
Unidirectional fret cascade in nanocompartmentalized polyelectrolyte multilayers.
Chemphyschem 10(1):137-143.
Salomaki M, Kankare J (2009) Influence of synthetic polyelectrolytes on the growth and
properties of hyaluronan-chitosen multilayers. Biomacromolecules 10(2):294-301.
Mertz D, Hemmerle J, Mutterer J, Ollivier S, Voegel JC, Schaaf P, Lavalle P (2007)
Mechanically responding nanovalves based on polyelectrolyte multilayers. Nano Lett
7(3):657-662.
Okamura Y, Kabata K, Kinoshita M, Saitoh D, Takeoka S (2009) Free-standing
biodegradable poly(lactic acid) nanosheet for sealing operations in surgery. Adv Mater
21(43):4388-4392.
DeLongchamp DM, Hammond PT (2004) Highly ion conductive poly(ethylene oxide)based solid polymer electrolytes from hydrogen bonding layer-by-layer assembly.
Langmuir 20(13):5403-5411.
Shard AG, Green FM, Gilmore IS (2008) C(60) ion sputtering of layered organic
materials. Applied Surface Science 255(4):962-965.
Beamson G, Briggs D (1992) High resolution xps of organic polymers: The scienta
esca300 database(Wiley and Sons, Chichester).
Crank J (1975) The mathematics of diffusion (Clarendon Press, Oxford) second Ed.
Kharlampieva E, Kozlovskaya V, Ankner JF, Sukhishvili SA (2008) Hydrogen-bonded
polymer multilayers probed by neutron reflectivity. Langmuir 24(20):11346-11349.
156
72.
73.
74.
75.
76.
77.
78.
79.
80.
81.
82.
83.
84.
85.
86.
87.
88.
89.
90.
Hubsch E, Fleith G, Fatisson J, Labbe P, Voegel JC, Schaaf P, Ball V (2005) Multivalent
ion/polyelectrolyte exchange processes in exponentially growing multilayers. Langmuir
21(8):3664-3669.
Shiratori SS, Rubner MF (2000) Ph-dependent thickness behavior of sequentially
adsorbed layers of weak polyelectrolytes. Macromolecules 33(11):4213-4219.
Yuan WY, Lu ZS, Wang HL, Li CM (2012) Stimuli-free reversible and controllable
loading and release of proteins under physiological conditions by exponentially growing
nanoporous multilayered structure. Advanced Functional Materials22(9):1932-1939.
Lutkenhaus JL, Hrabak KD, McEnnis K, Hammond PT (2005) Elastomeric flexible freestanding hydrogen-bonded nanoscale assemblies. JAm Chem Soc 127(49):17228-17234.
Sukhishvili SA, Granick S (2002) Layered, erasable polymer multilayers formed by
hydrogen-bonded sequential self-assembly. Macromolecules 35(1):301-3 10.
Paxson AT, Yagtie JL, Gleason KK, Varanasi KK (2014) Stable dropwise condensation
for enhancing heat transfer via the initiated chemical vapor deposition (icvd) of grafted
polymer films. Advanced Materials26(3):418-423.
Xiao R, Miljkovic N, Enright R, Wang EN (2013) Immersion condensation on oilinfused heterogeneous surfaces for enhanced heat transfer. Sci. Rep. 3.
Zhai L, Berg MC, Cebeci FQ, Kim Y, Milwid JM, Rubner MF, Cohen RE (2006)
Patterned superhydrophobic surfaces: Toward a synthetic mimic of the namib desert
beetle. Nano Letters 6(6):1213-1217.
Lee H, Alcaraz ML, Rubner MF, Cohen RE (2013) Zwitter-wettability and antifogging
coatings with frost-resisting capabilities. ACS Nano 7(3):2172-2185.
Gao X, Yan X, Yao X, Xu L, Zhang K, Zhang J, Yang B, Jiang L (2007) The dry-style
antifogging properties of mosquito compound eyes and artificial analogues prepared by
soft lithography. Advanced Materials 19(17):2213-2217.
Miljkovic N, Preston DJ, Enright R, Wang EN (2013) Electrostatic charging of jumping
droplets. Nat Commun 4.
Cebeci FQ, Wu Z, Zhai L, Cohen RE, Rubner MF (2006) Nanoporosity-driven
superhydrophilicity: A means to create multifunctional antifogging coatings. Langmuir
22(6):2856-2862.
Park K-C, Choi HJ, Chang C-H, Cohen RE, McKinley GH, Barbastathis G (2012)
Nanotextured silica surfaces with robust superhydrophobicity and omnidirectional
broadband supertransmissivity. ACS Nano 6(5):3789-3799.
Briscoe BJ, Galvin KP (1991) The effect of surface fog on the transmittance of light.
Solar Energy 46(4):191-197.
Howarter JA, Youngblood JP (2008) Self-cleaning and next generation anti-fog surfaces
and coatings. MacromolecularRapid Communications 29(6):455-466.
Zhao J, Meyer A, Ma L, Ming W (2013) Acrylic coatings with surprising antifogging and
frost-resisting properties. Chemical Communications 49(100):11764-11766.
Hsiao E, Barnette AL, Bradley LC, Kim SH (2011) Hydrophobic but hygroscopic
polymer films - identifying interfacial species and understanding water ingress behavior.
ACS Appl Mater Interfaces 3(11):4236-4241.
Lee H, Sample C, Cohen RE, Rubner MF (2013) Ph-programmable sequential dissolution
of multilayer stacks of hydrogen-bonded polymers. ACS Macro Letters 2(10):924-927.
Kleingartner JA, Lee H, Rubner MF, McKinley GH, Cohen RE (2013) Exploring the
kinetics of switchable polymer surfaces with dynamic tensiometry. Soft Matter.
157
91.
92.
93.
94.
95.
96.
97.
98.
99.
100.
101.
102.
103.
104.
105.
106.
107.
Varanasi KK, Hsu M, Bhate N, Yang W, Deng T (2009) Spatial control in the
heterogeneous nucleation of water. Appl Phys Lett 95(9):094101.
Gilbert JB, Rubner MF, Cohen RE (2013) Depth-profiling x-ray photoelectron
spectroscopy (xps) analysis of interlayer diffusion in polyelectrolyte multilayers. Proc
NatlAcadSci USA 110(17):6651-6656.
Sanada N, Yamamoto A, Oiwa R, Ohashi Y (2004) Extremely low sputtering degradation
of polytetrafluoroethylene by c-60 ion beam applied in xps analysis. Surface and
Interface Analysis 36(3):280-282.
Lee MH, Lim B, Kim JW, An EJ, Lee D (2012) Effect of composition on water
permeability of model stratum corneum lipid membranes. Soft Matter 8(5):1539-1546.
Kleingartner JA, Lee H, Rubner MF, McKinley GH, Cohen RE (2013) Exploring the
kinetics of switchable polymer surfaces with dynamic tensiometry. Soft Matter
9(26):6080-6090.
Girard KP, Quinn JA, Vanderlick TK (1999) The directional dependence of water
penetration into langmuir-blodgett multilayers. J Colloid Interface Sci 217(1):146-153.
Tarascon JM, Armand M (2001) Issues and challenges facing rechargeable lithium
batteries. Nature 414(6861):359-367.
Goodenough JB, Kim Y (2010) Challenges for rechargeable li batteries. Chem Mater
22(3):587-603.
Agrawal RC, Pandey GP (2008) Solid polymer electrolytes: Materials designing and allsolid-state battery applications: An overview. Journal of Physics D-Applied Physics
41(22).
Quartarone E, Mustarelli P (2011) Electrolytes for solid-state lithium rechargeable
batteries: Recent advances and perspectives. Chem Soc Rev 40(5):2525-2540.
Lascaud S, Perrier M, Vallee A, Besner S, Prudhomme J, Armand M (1994) Phasediagrams and conductivity behavior of poly(ethylene oxide) molten-salt rubbery
electrolytes. Macromolecules 27(25):7469-7477.
Singh M, Odusanya 0, Wilmes GM, Eitouni HB, Gomez ED, Patel AJ, Chen VL, Park
MJ, Fragouli P, Iatrou H, Hadjichristidis N, Cookson D, Balsara NP (2007) Effect of
molecular weight on the mechanical and electrical properties of block copolymer
electrolytes. Macromolecules 40(13):4578-4585.
Young WS, Epps TH, III (2012) Ionic conductivities of block copolymer electrolytes
with various conducting pathways: Sample preparation and processing considerations.
Macromolecules 45(11):4689-4697.
Panday A, Mullin S, Gomez ED, Wanakule N, Chen VL, Hexemer A, Pople J, Balsara
NP (2009) Effect of molecular weight and salt concentration on conductivity of block
copolymer electrolytes. Macromolecules 42(13):4632-4637.
Ruzette AVG, Soo PP, Sadoway DR, Mayes AM (2001) Melt-formable block copolymer
electrolytes for lithium rechargeable batteries. JElectrochem Soc 148(6):A537-A543.
Niitani T, Shimada M, Kawamura K, Dokko K, Rho YH, Kanamura K (2005) Synthesis
of li+ ion conductive peo-pst block copolymer electrolyte with microphase separation
structure. Electrochemical and Solid State Letters 8(8):A385-A388.
Yuan R, Teran AA, Gurevitch I, Mullin SA, Wanakule NS, Balsara NP (2013) Ionic
conductivity of low molecular weight block copolymer electrolytes. Macromolecules
46(3):914-921.
158
108.
109.
110.
111.
112.
113.
114.
115.
116.
117.
118.
119.
120.
121.
122.
123.
124.
125.
126.
Stafford CM, Roskov KE, Epps TH, III, Fasolka MJ (2006) Generating thickness
gradients of thin polymer films via flow coating. Rev Sci Instrum 77(2):023908.
Albert JNL, Epps TH, III (2010) Self-assembly of block copolymer thin films. Materials
Today 13(6):24-33.
Tanuma S, Powell CJ, Penn DR (1994) Calculations of electron inelastic mean free paths
.5. Data for 14 organic-compounds over the 50-2000 ev range. Surface and Interface
Analysis 21(3):165-176.
Zhao WT, GS; Kumar, N (2010) Chemistry and material science at the cell surface.
Materials Today 13(4):14-21.
Culty M, Nguyen HA, Underhill CB (1992) The hyaluronan receptor (cd44) participates
in the uptake and degradation of hyaluronan. J Cell Biol 116(4):1055-1062.
Gilbert JB, O'Brien JS, Suresh HS, Cohen RE, Rubner MF (2013) Orientation-specific
attachment of polymeric microtubes on cell surfaces. Adv Mater 25(41):5948-5952.
Jewell CM, Lynn DM (2008) Multilayered polyelectrolyte assemblies as platforms for
the delivery of DNA and other nucleic acid-based therapeutics. Advanced Drug Delivery
Reviews 60(9):979-999.
Lasser A (1983) The mononuclear phagocytic system: A review. Human pathology
14(2):108.
Burke B, Lewis C (2002) The macrophage(Oxford Univeristy Press, USA) 2 Ed.
Aderem A, Underhill D (1999) Mechanisms of phagocytosis in macrophages. Annu Rev
Immunol 17(1):593-623.
Dou H, Destache CJ, Morehead JR, Mosley RL, Boska MD, Kingsley J, Gorantla S,
Poluektova L, Nelson JA, Chaubal M, Werling J, Kipp J, Rabinow BE, Gendelman HE
(2006) Development of a macrophage-based nanoparticle platform for antiretroviral drug
delivery. Blood 108(8):2827-2835.
Dou H, Grotepas CB, McMillan JM, Destache CJ, Chaubal M, Werling J, Kipp J,
Rabinow B, Gendelman HE (2009) Macrophage delivery of nanoformulated
antiretroviral drug to the brain in a murine model of neuroaids. J Immunol 183(1):661 669.
Choi MR, Stanton-Maxey KJ, Stanley JK, Levin CS, Bardhan R, Akin D, Badve S,
Sturgis J, Robinson JP, Bashir R, Halas NJ, Clare SE (2007) A cellular trojan horse for
delivery of therapeutic nanoparticles into tumors. Nano Lett 7(12):3759-3765.
Thiele L, Merkle H, Walter E (2003) Phagocytosis and phagosomal fate of surfacemodified microparticles in dendritic cells and macrophages. PharmaceuticalResearch
20(2):221-228.
Peach R, Hollenbaugh D, Stamenkovic I, Aruffo A (1993) Identification of hyaluronic
acid binding sites in the extracellular domain of cd44. Journal of Cell Biology
122(1):257.
Tabata Y, Ikada Y (1988) Effect of the size and surface charge of polymer microspheres
on their phagocytosis by macrophage. Biomaterials9(4):356-362.
Luster A (1998) Chemokines--chemotactic cytokines that mediate inflammation. The
New Englandjournalofmedicine 338(7):436.
Hansson G (2005) Inflammation, atherosclerosis, and coronary artery disease. The New
Englandjournalof medicine 352(16):1685.
Coussens L, Werb Z (2002) Inflammation and cancer. Nature 420(6917):860-867.
159
127.
128.
129.
130.
131.
132.
133.
134.
135.
136.
137.
138.
139.
140.
141.
142.
143.
144.
145.
146.
Ariga K, Hill J, Ji Q (2007) Layer-by-layer assembly as a versatile bottom-up
nanofabrication technique for exploratory research and realistic application. Phys Chem
Chem Phys 9(19):2319-2340.
Caruso F (2001) Nanoengineering of particle surfaces. Advanced Materials 13(1):11-22.
Goa K, Benfield P (1994) Hyaluronic acid: A review of its pharmacology and use as a
surgical aid in ophthalmology, and its therapeutic potential in joint disease and wound
healing. Drugs 47(3):536-566.
Swiston A, Gilbert J, Irvine D, Cohen R, Rubner M (2010) Freely suspended cellular
"backpacks" lead to cell aggregate self-assembly. Biomacromolecules 11(7):4920-4925.
Stephan MT, Moon JJ, Um SH, Bershteyn A, Irvine DJ (2010) Therapeutic cell
engineering with surface-conjugated synthetic nanoparticles. Nat Med 16(9):1035-1041.
Stephan M, Moon J, Um S, Bershteyn A, Irvine D (2010) Therapeutic cell engineering
with surface-conjugated synthetic nanoparticles. Nature Medicine 16(9):1035-1041.
Champion J, Walker A, Mitragotri S (2008) Role of particle size in phagocytosis of
polymeric microspheres. PharmaceuticalResearch 25(8):1815-1821.
Champion J, Mitragotri S (2006) Role of target geometry in phagocytosis. Proceedingsof
the NationalAcademy ofSciences 103(13):4930-4934.
Mitragotri S, Lahann J (2009) Physical approaches to biomaterial design. Nat Mater
8(1):15-23.
Martinez G, Csaba N, Fischer S, Sichelstiel A, Kundig T, Gander B, Johansen P (2008)
Surface coating of plga microparticles with protamine enhances their immunological
performance through facilitated phagocytosis. Journal of controlled release: official
journal of the ControlledRelease Society 130(2):161.
Doshi N, Mitragotri S (2010) Macrophages recognize size and shape of their targets.
PLoS One 5(4):doi:10.1371/joumal.pone.0010051.
Champion J, Mitragotri S (2009) Shape induced inhibition of phagocytosis of polymer
particles. PharmaceuticalResearch 26(1):244-249.
Beningo K, Wang Y (2002) Fc-receptor-mediated phagocytosis is regulated by
mechanical properties of the target. JCell Sci 115(4):849.
Tang ZY, Wang Y, Podsiadlo P, Kotov NA (2006) Biomedical applications of layer-bylayer assembly: From biomimetics to tissue engineering. Adv Mater 18(24):3203-3224.
Su XF, Kim BS, Kim SR, Hammond PT, Irvine DJ (2009) Layer-by-layer-assembled
multilayer films for transcutaneous drug and vaccine delivery. Acs Nano 3(11):37193729.
Primeau A, Rendon A, Hedley D, Lilge L, Tannock I (2005) The distribution of the
anticancer drug doxorubicin in relation to blood vessels in solid tumors. Clin Cancer Res
11(24):8782.
Swiston A, Cheng C, Um S, Irvine D, Cohen R, Rubner M (2008) Surface
functionalization of living cells with multilayer patches. Nano Lett 8(12):4446-4453.
Allen TM, Cullis PR (2004) Drug delivery systems: Entering the mainstream. Science
303(5665):1818-1822.
Ruoslahti E (2012) Peptides as targeting elements and tissue penetration devices for
nanoparticles. Adv Mater 24(28):3747-3756.
Sanhai WR, Sakamoto JH, Canady R, Ferrari M (2008) Seven challenges for
nanomedicine. Nat Nanotechnol 3(5):242-244.
160
147.
148.
149.
150.
151.
152.
153.
154.
155.
156.
157.
158.
159.
160.
161.
162.
163.
Riehemann K, Schneider SW, Luger TA, Godin B, Ferrari M, Fuchs H (2009)
Nanomedicine--challenge and perspectives. Angew Chem Int Ed Engl 48(5):872-897.
Murdoch C, Giannoudis A, Lewis CE (2004) Mechanisms regulating the recruitment of
macrophages into hypoxic areas of tumors and other ischemic tissues. Blood
104(8):2224-2234.
Moore KJ, Sheedy FJ, Fisher EA (2013) Macrophages in atherosclerosis: A dynamic
balance. Nat Rev Immunol 13(10):709-721.
Szekanecz Z, Koch AE (2007) Macrophages and their products in rheumatoid arthritis.
Currentopinion in rheumatology 19(3):289-295.
Anselmo AC, Gupta V, Zem BJ, Pan D, Zakrewsky M, Muzykantov V, Mitragotri S
(2013) Delivering nanoparticles to lungs while avoiding liver and spleen through
adsorption on red blood cells. A CS Nano 7(12):11129-11137.
Choi J, Kim HY, Ju EJ, Jung J, Park J, Chung HK, Lee JS, Lee JS, Park HJ, Song SY,
Jeong SY, Choi EK (2012) Use of macrophages to deliver therapeutic and imaging
contrast agents to tumors. Biomaterials 33(16):4195-4203.
Choi MR, Bardhan R, Stanton-Maxey KJ, Badve S, Nakshatri H, Stantz KM, Cao N,
Halas NJ, Clare SE (2012) Delivery of nanoparticles to brain metastases of breast cancer
using a cellular trojan horse. Cancer Nanotechnol 3(1-6):47-54.
Brynskikh AM, Zhao Y, Mosley RL, Li S, Boska MD, Klyachko NL, Kabanov AV,
Gendelman HE, Batrakova EV (2010) Macrophage delivery of therapeutic nanozymes in
a murine model of parkinson's disease. Nanomedicine (Lond) 5(3):379-396.
Delcea M, Mohwald H, Skirtach AG (2011) Stimuli-responsive ibi capsules and
nanoshells for drug delivery. Advanced Drug Delivery Reviews 63(9):730-747.
Ravetch JV, Bolland S (2001) Igg fc receptors. Annu Rev Immunol 19:275-290.
Szarka RJ, Wang N, Gordon L, Nation P, Smith RH (1997) A murine model of
pulmonary damage induced by lipopolysaccharide via intranasal instillation. J Immunol
Methods 202(1):49-57.
Skerrett SJ, Liggitt HD, Hajjar AM, Ernst RK, Miller SI, Wilson CB (2004) Respiratory
epithelial cells regulate lung inflammation in response to inhaled endotoxin. American
Journalof Physiology-Lung Cellularand Molecular Physiology 287(1):L143-Li52.
Beck-Schimmer B, Schimmer RC, Warner RL, Schmal H, Nordblom G, Flory CM,
Lesch ME, Friedl HP, Schrier DJ, Ward PA (1997) Expression of lung vascular and
airway icam- 1 after exposure to bacterial lipopolysaccharide. Am J Respir Cell Mol Biol
17(3):344-352.
Alon R, Kassner PD, Carr MW, Finger EB, Hemler ME, Springer TA (1995) The
integrin vla-4 supports tethering and rolling in flow on vcam-1. The Journal of cell
biology 128(6):1243-1253.
Kaplanski G, Marin V, Fabrigoule M, Boulay V, Benoliel A-M, Bongrand P, Kaplanski
S, Farnarier C (1998) Thrombin-activated human endothelial cells support monocyte
adhesion in vitro following expression of intercellular adhesion molecule-i (icam-1;
cd54) and vascular cell adhesion molecule-I (vcam-1; cdI06). Blood 92(4):1259-1267.
Petros RA, DeSimone JM (2010) Strategies in the design of nanoparticles for therapeutic
applications. Nature Reviews Drug Discovery 9(8):615-627.
Merkel TJ, Chen K, Jones SW, Pandya AA, Tian SM, Napier ME, Zamboni WE,
DeSimone JM (2012) The effect of particle size on the biodistribution of low-modulus
hydrogel print particles. J Control Release 162(l):37-44.
161
164.
165.
166.
167.
168.
169.
170.
171.
172.
173.
174.
175.
176.
177.
178.
179.
180.
181.
Shimoni 0, Yan Y, Wang YJ, Caruso F (2013) Shape-dependent cellular processing of
polyelectrolyte capsules. Acs Nano 7(1):522-530.
Kohler D, Schneider M, Kruger M, Lehr CM, Mohwald H, Wang DY (2011) Templateassisted polyelectrolyte encapsulation of nanoparticles into dispersible, hierarchically
nanostructured microfibers. Adv Mater 23(11):1376-1379.
Brynskikh AM, Zhao YL, Mosley RL, Li S, Boska MD, Klyachko NL, Kabanov AV,
Gendelman HE, Batrakova EV (2010) Macrophage delivery of therapeutic nanozymes in
a murine model of parkinson's disease. Nanomed 5(3):379-396.
Swiston AJ, Gilbert JB, Irvine DJ, Cohen RE, Rubner MF (2010) Freely suspended
cellular "Backpacks" Lead to cell aggregate self-assembly. Biomacromolecules
11(7):1826-1832.
Suematsu S, Watanabe T (2004) Generation of a synthetic lymphoid tissue-like organoid
in mice. Nat Biotechnol 22(12):1539-1545.
Kim H, Cohen RE, Hammond PT, Irvine DJ (2006) Live lymphocyte arrays for
biosensing. Advanced FunctionalMaterials 16(10):1313-1323.
De Haes W, De Koker S, Pollard C, Atkinson D, Vlieghe E, Hoste J, Rejman J, De Smedt
S, Grooten J, Vanham G, Van Gulck E (Polyelectrolyte capsules-containing hiv-1 p24
and poly i:C modulate dendritic cells to stimulate hiv-1-specific immune responses. Mol
Ther 18(7):1408-1416.
Geng Y, Dalhaimer P, Cai SS, Tsai R, Tewari M, Minko T, Discher DE (2007) Shape
effects of filaments versus spherical particles in flow and drug delivery. Nature
Nanotechnology 2(4):249-255.
Gratton SEA, Ropp PA, Pohlhaus PD, Luft JC, Madden VJ, Napier ME, DeSimone JM
(2008) The effect of particle design on cellular internalization pathways. Proc Natl Acad
Sci USA 105(33):11613-11618.
Florez L, Herrmann C, Cramer JM, Hauser CP, Koynov K, Landfester K, Crespy D,
Mailander V (2012) How shape influences uptake: Interactions of anisotropic polymer
nanoparticles and human mesenchymal stem cells. Small 8(14):2222-2230.
Yoshida M, Roh KH, Mandal S, Bhaskar S, Lim DW, Nandivada H, Deng XP, Lahann J
(2009) Structurally controlled bio-hybrid materials based on unidirectional association of
anisotropic microparticles with human endothelial cells. Adv Mater 21(48):4920-4925.
Wu LY, Ross BM, Hong S, Lee LP (2010) Bioinspired nanocorals with decoupled
cellular targeting and sensing functionality. Small 6(4):503-507.
Kaewsaneha C, Tangboriboonrat P, Polpanich D, Eissa M, Elaissari A (2013) Janus
colloidal particles: Preparation, properties, and biomedical applications. ACS Appl Mater
Interfaces 5(6):1857-1869.
Roh KH, Martin DC, Lahann J (2005) Biphasic janus particles with nanoscale anisotropy.
Nat Mater 4(10):759-763.
Lattuada M, Hatton TA (2011) Synthesis, properties and applications of janus
nanoparticles. Nano Today 6(3):286-308.
Bianchi E, Blaak R, Likos CN (2011) Patchy colloids: State of the art and perspectives.
Phys Chem Chem Phys 13(14):6397-6410.
Glotzer SC, Solomon MJ (2007) Anisotropy of building blocks and their assembly into
complex structures. Nat Mater 6(8):557-562.
Wilson JT, Cui WX, Kozovskaya V, Kharlampieva E, Pan D, Qu Z, Krishnamurthy VR,
Mets J, Kumar V, Wen J, Song YH, Tsukruk VV, Chaikof EL (2011) Cell surface
162
182.
183.
184.
185.
186.
187.
188.
189.
190.
191.
192.
193.
194.
195.
196.
197.
198.
199.
engineering with polyelectrolyte multilayer thin films. J Am Chem Soc 133(18):70547064.
Sarkar D, Spencer JA, Phillips JA, Zhao W, Schafer S, Spelke DP, Mortensen LJ, Ruiz
JP, Vemula PK, Sridharan R, Kumar S, Kamik R, Lin CP, Karp JM (2011) Engineered
cell homing. Blood 118(25):E1 84-E 191.
Gartner ZJ, Bertozzi CR (2009) Programmed assembly of 3-dimensional microtissues
with defined cellular connectivity. Proc Natl Acad Sci USA 106(12):4606-4610.
Moon JJ, Huang B, Irvine DJ (2012) Engineering nano- and microparticles to tune
immunity. Adv Mater 24(28):3724-3746.
Mendelsohn JD, Yang SY, Hiller J, Hochbaum Al, Rubner MF (2003) Rational design of
cytophilic and cytophobic polyelectrolyte multilayer thin films. Biomacromolecules
4(l):96-106.
Lee D, Cohen RE, Rubner MF (2007) Heterostructured magnetic nanotubes. Langmuir
23(1):123-129.
Yang Y, He Q, Duan L, Cui Y, Li JB (2007) Assembled alginate/chitosan nanotubes for
biological application. Biomaterials28(20):3083-3090.
Perry JL, Martin CR, Stewart JD (2011) Drug-delivery strategies by using templatesynthesized nanotubes. Chemistry-a European Journal17(23):6296-6302.
DeRocher JP, Mao P, Han JY, Rubner MF, Cohen RE (2010) Layer-by-layer assembly of
polyelectrolytes in nanofluidic devices. Macromolecules 43(5):2430-2437.
Lazzara TD, Lau KHA, Abou-Kandil Al, Caminade AM, Majoral JP, Knoll W (2010)
Polyelectrolyte layer-by-layer deposition in cylindrical nanopores. Acs Nano 4(7):39093920.
St Clair EW (2009) Novel targeted therapies for autoimmunity. Curr Opin Immunol
21(6):648-657.
Levesque MC (2009) Translational mini-review series on b cell-directed therapies:
Recent advances in b cell-directed biological therapies for autoimmune disorders. Clin
Exp Immunol 157(2):198-208.
Lee KJ, Yoon J, Rahmani S, Hwang S, Bhaskar S, Mitragotri S, Lahann J (2012)
Spontaneous shape reconfigurations in multicompartmental microcylinders. Proc Natl
AcadSci USA 109(40):16057-16062.
Peyratout CS, Dahne L (2004) Tailor-made polyelectrolyte microcapsules: From
multilayers to smart containers. Angewandte Chemie-InternationalEdition 43(29):37623783.
Wang L, Kim M, Fang Q, Min J, Jeon WI, Lee SY, Son SJ, Joo SW, Lee SB (2013)
Hydrophobic end-gated silica nanotubes for intracellular glutathione-stimulated drug
delivery in drug-resistant cancer cells. Chem Commun 49(31):3194-3196.
Ariga K, Lvov YM, Kawakami K, Ji QM, Hill JP (2011) Layer-by-layer self-assembled
shells for drug delivery. Advanced Drug Delivery Reviews 63(9):762-771.
Hu SH, Gao XH (2010) Nanocomposites with spatially separated functionalities for
combined imaging and magnetolytic therapy. JAm Chem Soc 132(21):7234-7237.
Tikhonov VE, Lopez-Llorca LV, Salinas JS, Monfort E (2004) Endochitinase activity
determination using n-fluorescein-labeled chitin. J Biochem Biophys Methods 60(1):2938.
Olivier GK, Shin D, Gilbert JB, Monzon LAA, Frechette J (2009) Supramolecular ionpair interactions to control monolayer assembly. Langmuir 25(4):2159-2165.
163
200.
Willey TM, Vance AL, van Buuren T, Bostedt C, Nelson AJ, Terminello U, Fadley CS
(2004) Chemically transformable configurations of mercaptohexadecanoic acid selfassembled monolayers adsorbed on au(1 11). Langmuir 20(7):2746-2752.
164
11. APPENDIX
11.1
Chemistry of 'Released' Side of Free Floating Films
To study the chemistry of the released side of a free floating film, a region of hydrogen
bonded polymers, topped with a region of electrostatically bonded polyelectrolytes was created.
The hydrogen bonded region was made of poly(acrylic acid) and poly(ethylene glycol) or
(PAA3.0/PEG3.0) which has been found to be stable up to pH 3.6(16) and therefore pH 7.4 PBS
solution was used to dissolve the region. The electrostatically bonded region was built with the
weak cation poly(allylamine hydrochloride) and magnetic nanoparticles coated with an anionic
surfactant (PAH3.0/MNP4.0).
.
(
Electrostatic Region :(PAH3.0/MNP4.0) 9.5
- Hydrogen Bonding Region: (PAA3.0/PEG3.0)x
- Substrate: Glass coated in (PDAC4.0/SPS4.0) 15.5
Figure 11-1: PEM System studied for analysis of released free floating film chemistry
XPS measurements
(PAA3.0/PEG3.0)20.5,
(PAH3.0/MNP4.)
10
were performed
2)
on films of the following compositions:
(PAA3.0/PAH3.0)1o
and
3)
flipped
1)
(PAA3.0/PEG3.0)20.5
. Flipping the third sample is detailed in part A of the figure below. In part B
of the figure below, the flipped carbon region is compared to the other two films, a
(PAA3.0/PEG3.0) 20.s hydrogen-bonded film, and a (PAA3.O/PAH3.0)
10 electrostatic
A)
X-rays
Flip for XPS
analysis
PBS
pH 7.4
165
film.
B)
16
14
(PAA3. 0/PAH3.0)10
12
10~
Flipp ed -in pH 7.4
Ionized (COO-):
Electrostatic
Not Ionized (C)OH): Bonded
(PAA: 3.O/PEG3.0)20. 5
Hydrogen Bon ded
4-
2-
294
291
288
Big Energy (eV)
285
282
Figure 11-2: A) Method for flipping of the film prior to XPS, pH 7.4 PBS dissolves the
hydrogen bonding region, allowing the film to be flipped. B) Cis XPS spectra of an asdeposited (PAA3.0/PEG3.0) 20 .5 , (PAA3.0/PAH3.0) 10 and flipped (PAA3.0/PEG3.0) 20.5
(PAH3.0/MNP4.0) 10 film. PAA's carboxyl Cis peak in the flipped film shifts to lower
binding energy as would be expected when paired with an amine. PEG's Cis peak also
significantly shrinks in the flipped sample, suggesting the ejection of PEG upon PAH
disrupting the PAA-PEG hydrogen bonds.
The PAA/PAH film's a broad peak in the binding energy of 288eV to 290eV is due to the
carboxyl groups. This peak can be deconvoluted into two regions, one peak centered around
289eV that represents a protonated carboxyl group and one peak around 288eV that represents an
166
ionized carboxyl group that has been electrostatically paired with a primary amine. The lower
binding energy displayed when a carboxyl is paired with an amine has been previously seen
(199, 200). It has been hypothesized that it is due to the delocalization of the negative charge of
the oxygen groups due to the presence of the nitrogen group (199, 200). Thus the carbon's
electrons are not pulled to the oxygen as strongly and have a lower binding energy. Previous
studies have shown that (PAA3.0/PAH3.0) films have around 50% of the carboxylic acid groups
ionized and the deconvolution of the figure above qualitatively supports this conclusion(9).
We see that the carboxyl groups of the hydrogen bonded film (PAA3.0/PEG3.0) are all in the
protonated form since there are no electrostatic interactions occurring. However, the flipped
sample has a large amount of ionized groups showing that the carboxyl groups that were once
hydrogen bonding are now electrostatically interacting with amine groups of PAH. This
percentage is consistent with a prior study since the film was last placed in a pH 7.4 PBS(9).
A second qualitative change is that the PEG O-C-C peak at 286.5 eV(69) significantly
shrinks in the flipped film. This is consistent with PAH disrupting PAA-PEG hydrogen bonds,
allowing the PEG to exit the film. To detect if PEG remained in the film, a thiolated PEG was
used and as seen by the presence of sulfur in Figure 11-3 we see some PEG is still kinetically
entangled in the multilayer. Therefore part of the sacrificial region is tethered to the outer face of
the released film. These XPS results show that the outer surface of the released film is composed
of PAA-PAH electrostatic pairs with a small amount of PEG trapped.
167
Study Details
To directly measure if the hydrogen bonding region is attached to the released film,
hydrogen-bonded films were prepared with a 20kDa MW thiol-end group PEG (PEG-SH/PAA)
layers and topped with (PAH3.0/MNP4.0)
10.
The film was scored and released in pH 7.4 PBS,
flipped over and placed on a glass slide so that the film's face originally closest to the glass
surface was exposed. The (PAA/PEG-SH) system was chosen over other amine-containing
hydrogen-bonded systems so that any N peak must be due to PAH. If the polycation PAH was
present in the hydrogen bonding region, a nitrogen peak would be found. Also if part of the
hydrogen bonding region remains attached to the film after release, XPS results will show a
small sulfur peak. Further, if no iron peak was seen, then the superparamagnetic Fe 2O 3
nanoparticles are not within -10nm (the penetration depth of XPS) of the outer face. The figure
below shows that both S and N peaks were found, but no magnetic nanoparticle signal, proving
that some PEG-SH from the sacrificial region and PAH are present on the released face of the
film and that the magnetic nanoparticles are >10nm beneath the surface. It is expected that the
PAH has electrostatically paired with the PAA and thus formed a stable region that has
entrapped some of the PEG-SH.
168
Atomic %
0
20-
15.
'j
A~
C
173.22
0
24.15
N
2.38
C
le
10.
N
.5-
$
I
I
1000
800
600
400
200
Bidu Energy (eV)
Figure 11-3: XPS spectra and atomic percentages for a flipped (PAA3.0/PEG-SH3.0) 20 .5
(PAH3.0/MNP4.) 10 film. The peaks for C, N, 0, and S are as indicated - all other peaks
are due to residual salt from PBS. Inset shows the S peak at 168eV.
169