in _.
54:2]!
9. H

":mm ...-1:!

. 3.1.4
Wfi. ..L
.

:
913w}

' .
3.35M”. awn.
@fi...‘ 3., .

1 . I .
‘C :I “. Dig
1.: A .2
I ..-...uaw
{VIvQT .Itscvu ,

. ...
at“
“z. ..- .. .51 .
M314??? ...mfi...‘1
. .. «11%.?

.11.... .....Hfihfiuflvl
. vs" audit”)?
. {nah no.»

i’
l. III- .‘
3 a:

c.
. .fmrl. .
....pvfiva’lxh
.ll - I; ...

V“ ’4‘ I

. jflfa;f,

‘15:! ....

 

 

rm

Z LIBRARY .
251799 Michigan State
University I

This is to certify that the
dissertation entitled

SCANNING PROBE RECOGNITION MICROSCOPY:
RECOGNITION STRATEGIES

presented by

QIAN CHEN

has been accepted towards fulfillment
of the requirements for the

Ph. D. degree in Department of Electrical &
Computer EngineerinL

 

 

7 Major Professor’sésalgnature

/2_ [9675/2698 2w?—

Date

 

MSU is an affinnative-action, equal-opportunity employer

 

_._ —.-.-.—.-----.--c—-.-o-o—.-.—

— — - -.—.-.-._.~

-—-rq.—--.-.-u--c—.-c—-.—.-.- —.—

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

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K:IProj/Acc&Pres/ClRC/DateDue.indd

SCANNING PROBE RECOGNITION MICROSCOPY:
RECOGNITION STRATEGIES

By

Qian Chen

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Department of Electrical & Computer Engineering

2007

ABSTRACT

SCANNING PROBE RECOGNITION MICROSCOPY:
RECOGNITION STRATEGIES

By
Qian Chen

Scanning Probe Recognition Microscopy (SPRM) is a specially modified Scanning
Probe Microscopy (SPM) system designed and developed by our research group in
partnership with Veeco Instruments. In Scanning Probe Recognition Microscopy, the
SPM system itself is given the ability to automatically track regions of interest during
scanning through incorporation of recognition—based tip control. The recognition
capability is realized by using algorithms and techniques in image processing, pattern
recognition and computer vision fields. Adaptive learning and prediction are also
implemented to make the object detection and tracking procedures quicker and more
reliable. The major contribution of SPRM includes: (1) decreasing overall operation time;
(2) providing the ability to quantitatively measuring properties while maintaining the
uniformity of experimental environment; (3) sequentially measuring topographic,
mechanical and chemistry properties of the same sample surface through repetitive high
resolution scanning in the appropriate mode; (4) properties measuring in situations which
are inaccessible by standard SPM. SPRM measurements of the surface roughness,
elasticity and surface chemistry of 2D nanoscale electrospun carbon nanofibers are given
in detail as an example of SPREM analysis in a situation that is not fully accessible by
SPM. The 3D nanoscale electrospun nanofibers are the components of a tissue scaffold
for regenerative medicine. These scaflolds are used for neural cells re-growth in a

damaged spinal cord. The candidate’s thesis research is therefore to design and

implement the recognition strategies that allow SPRM properties extraction along tissue
scaffold nanofibers, to perform the first quantitative measure of multiple properties that
have been shown to be important in neural cell re-growth and, by doing so, to contribute

significant understanding to the cure of presently incurable paralysis.

ACKNOWLEDGMENT

I would like to sincerely thank my major advisor, Dr. Virginia Ayres, for her support
all the way through the program. She not only gives me technical guidance, but also
teaches me about how to be face difficulties and tackle difficulties. I really appreciate her
encouragement and consideration. I would like to thank Dr. Lalita Udpa for her guidance
and support. She inspires me in many aspects and keeps me moving forward. I would
also like to express my gratitude to Dr. Marcos Dantus and Dr. Hayder Radha for their
precious advices and kindness as my committee members.

I would like to thank Yuan Fan for his consideration and fiiendship during the
cooperation. Thanks are also extended to Benjamin Jacobs for his help in nanocircuits
experiments.

I thank Dr. Sally Meiners, Ijaz Ahmed, Roberto Delgado-Rivera and Suzan Harris for
their generous help when I was visiting their lab to study for spinal cord repair.

I would like to express my appreciation to my husband, Xiangdong Peng, and my
parent, Jiwen Chen and Fengying Ma, for their love and support through all these years.
Without them, I can not get this far. Finally, to my lovely son, Kevin, he makes me

motivated and feels truly blessed.

iv

TABLE OF CONTENTS

LIST OF TABLES ........................................................................................................... viii

LIST OF FIGURES ........................................................................................................... ix

INTRODUCTION .............................................................................................................. 1

Part I SPRM INSTRUMENTATION DEVELOPMENT ................................................... 5

CHAPTER 1

IMAGING TECHNIQUES USED IN NAN OBIOLOGY ................................................. 7
1.1 Optical Microscopies ............................................................................................ 7
1.2 Electron Microscopies ........................................................................................ 10

1.3 Brief Introduction of Scanning Probe Microscopy in Biology and Medicine 12

CHAPTER 2
SCANNING PROBE MICROSCOPY IMAGING TECHNIQUES ................................ 14
2.1 Fundamentals of SPM ........................................................................................ 14
2.1.1 Block diagram ................................................................................................. 16
2.1.2 Operation modes ............................................................................................. 18
2.2 Advantages and Limitations of SPM ................................................................. 20
2.2. 1 Advantages ..................................................................................................... 20
2.2.2 Limitations ...................................................................................................... 20
2.3 Improvement of SPM Start of Art ...................................................................... 21

CHAPTER 3
SCANNING PROBE RECOGNITION MICROSCOPY ................................................. 23
3.1 SPRM System F unctionalities ............................................................................ 24
3.1.1 Recognition ability ......................................................................................... 24
3.1.2 Auto-tracking capability and region of interest .............................................. 24
3.1.3 Real-time result display capability ................................................................. 24
3.1.4 Property analysis capability ............................................................................ 25
3.2 SPRM Prototype System Structure .................................................................... 25
3.3 SPRM System Design & Software Development .............................................. 26
3.4 SPRM Recognition Algorithm ........................................................................... 28
3.4.1 Preprocessing: Wavelet-based image segmentation ....................................... 28
3.4.2 Feature detection ............................................................................................. 31
3.5 SPRM Operation Modes .................................................................................... 35
3.6 Novelty & Importance of SPRM versus State of the Art AF M .......................... 36
3.6.1 Novelty and importance of SPRM versus standard SPM ............................... 36
3.6.2 Novelty and importance of SPRM versus optical techniques ........................ 38
3.6.3 Novelty and importance of SPRM versus HRTEM ........................................ 39

V

Part II SPRM APPLICATIONS IN NANOBIOLOGY .................................................... 41
CHAPTER 4

MOTIVATION & BACKGROUND OF SPRM IN TISSUE ENGINEERING ............... 42
4.1 Introduction of Tissue Engineering: Spinal Cord Repair ................................... 42
4.2 Environmental Nanoscale Cues for Cell Growth ............................................... 44
4.3 Ultimate Goal ..................................................................................................... 50

CHAPTER 5

SPRM AUTO-TRACKING IMPLEMENTATION ON TISSUE SCAF F OLD

NANOFIBERS ................................................................................................................. 52
5.1 Importance of Auto-Tracking of Individual Nanofibers .................................... 52
5.2 SPRM Auto-Tracking Capability ....................................................................... 55
5.3 Examples of SPRM ............................................................................................ 60

CHAPTER 6

NAN OF [BER PROPERTIES ANALYSIS BY USING SPRM ........................................ 62
6.1 Tissue Scaffold Samples and Experimental Parameters .................................... 62
6.2 Nanofiber Property Analysis by Using SPRM ................................................... 64

6.2.1 Surface roughness ........................................................................................... 64
6.2.2 Elasticity ......................................................................................................... 71
6.3 Comparison of Nanofiber Diameter by Transmission Electron Microscopy ..... 77
6.4 Comparison of Nanofiber Properties by Transmission Electron Microscopy... 80
6.5 Conclusion & Discussion ................................................................................... 81

CHAPTER 7

CELL RESPONSE ANALYSIS ....................................................................................... 83
7.1 Tissue Scaffold Samples & Experimental Techniques ....................................... 86

7.1.1 Tissue scaffold samples .................................................................................. 86
7.1.2 SPRM Properties investigation of nanofibers ................................................ 87
7.2 First Model System: Fibroblasts on 2D versus Nanoscale 3D Nanofibers ........ 95
7.2.1 Cell culture ..................................................................................................... 96
7.2.2 Experimental investigation of fibroblasts response ........................................ 97
7.2.3 Comparison of fibroblasts response and conclusion .................................... 100
7.3 Second Model System: Astrocytes on 2D versus Nanoscale 3D Nanofibers .. 101
7.3.1 Cell culture ................................................................................................... 102
7.3.2 Experimental investigation of astrocytes response ....................................... 102
7.3.3 Conclusion .................................................................................................... 1 04

CHAPTER 8

8 ELECTROSPINNING PARAMETERS .................................................................. 105
8.1 Introduction ...................................................................................................... 106
8.2 Experimental Setting ........................................................................................ 108
8.3 Experimental Results ........................................................................................ 109
8.4 Discussion and Conclusion ............................................................................... 113

vi

Part HI CONCLUSION & FUTURE WORK ................................................................. 115
CHAPTER 9

9 CONCLUSION & DISCUSSION ............................................................................ 115
9.1 Conclusion ......................................................................................................... 1 15
9.2 Discussion ......................................................................................................... 116

CHAPTER 10

10 FUTURE WORK .................................................................................................. 117
10.1 Improvement of SPRM Recognition Capability ............................................... 117
10.2 Extended Work in Spinal Cord Repair .............................................................. 117
10.3 Extension of SPRM to Additional Research Problems: Nanocircuits .............. 118

REFERENCE .................................................................................................................. 123

vii

LIST OF TABLES

Table 2-1 Multimode SPM Scanner Specifications [32] .................................................. 18
Table 2-2 Operation Modes of AFM ................................................................................ 19
Table 6-1 Surface Roughness Statistical Analysis ............................................................ 70
Table 6-2 Elasticity/Area Map Statistical Analysis .......................................................... 77
Table 6-3 Nanofiber Diameter Measurements Using TEM and AF M Images ................. 79

viii

LIST OF FIGURES

Figure 2—1 Block diagram of Atomic Force Microscopy .................................................. 16
Figure 3-1 Block diagram of Scanning Probe Recognition Microscopy .......................... 26
Figure 3-2 Software design diagram ................................................................................. 27

Figure 3-3 Schematic of overall approach. Reproduced from Figure 2, Reference: Q.
Chen, Y. Fan, L. Udpa, and V.M. Ayres, Vol. 2, Issue 2, pp. 181-189, Int. J.
Nanomedicine (2007) ................................................................................................ 28

Figure 3—4 Edge detection based on wavelets (a) original image, (b) edge image when
scale = 2, (c) edge image when scale = 10 ................................................................ 30

F i gure3-5 Segmentation (a) the edge image after dilation, (b) the edge image after
threshold, (c) the segmented object .......................................................................... 31

Figure 3-6 The definition of the Edge Feature is based on (a) AF M images; (b)
Continuous wavelet transformed (CWT) images. (c) Pixel intensities on horizontal
and vertical lines clearly show the Edge Feature for each cell type. Reproduced from
Figure 4, Reference: Q. Chen, Y. Fan, L. Udpa, and V.M. Ayres, Vol. 2, Issue 2, pp.
181-189, Int. J. Nanomedicine (2007) ...................................................................... 35

Figure 4-1 Cellular constituents of the blood-brain barrier. Images from reference [55] 44

Figure 4-2 SEM micrograph showing the bone marrow cell morphology on substrates
with different surface roughness, Images from reference [59] ................................. 45

Figure 4-3 Movements of National Institutes of Health 3T3 cells on substrates with a
rigidity gradient (0) A cell moved fiom the soft side of the substrate toward the
gradient. The cell turned by ~90° and moved into the stiff side of the substrate. Note
the increase in spreading area as the cell passed the boundary. (b) A cell moved from
the stiff side of the substrate toward the gradient. The cell changed its direction as it
entered the gradient and moved along the boundary. Bar, 40 um. Images from
reference [60] ............................................................................................................ 46

Figure 4-4 Peptide modification enhances the ability of nanofibers to promote axonal
regrowth. Rostral is to the left. G, host-graft interface. Curves were adjusted in
Adobe Photoshop for each panel to enhance the visibility of the axons against the
autofluorescence of the nanofibers. (A) Neurofilament M-labeled axons were
observed within the implant of unmodified nanofibers. (B) Axonal growth was
more robust within the implant of nanofibers modified with the DS’ peptide. Bar,
150 um. (C) Axons were observed on the surface of the nanofibers following folds

ix

in the nanofibrillar fabric. The top of the image is ~ 300 um ventral to the top of
the spinal cord. Bar, 300 um. Single arrows in (B) and (C) denote axons growing
more or less parallel to the axis of the spinal cord. Double arrows denote axons
that deviated from a parallel path. Reference: S. Meiners, I. Ahmed, A. S. Ponery, N.
Amor, S. L. Harris, V. Ayres, Y. Fan, Q. Chen, R. Delgado-Rivera, and A. N. Babu,
Polymer International, vol. 56, pp. 1340-1348, 200 ................................................. 48

Figure 4-5 Astrocytes cultured on even bare nanofibers adopted a stellate morphology (a)
that is typical of their in vivo counterparts, whereas astrocytes cultured on
poly-L-lysine (PLL) coated plastic coverslips had a flat, cobblestone appearance (b)
which is never seen in the body. To date, PLL has been required for astrocyte
grth in culture. Reference: S. Meiners, I. Ahmed, A. S. Ponery, N. Amor, S. L.
Harris, V. Ayres, Y. Fan, Q. Chen, R. Delgado-Rivera, and A. N. Babu, Polymer
International, vol. 56, pp. 1340-1348, 2007 .............................................................. 50

Figure 5-1 Off-line segmentation results at scan size equals 15 microns a) AFM height
image; b) Recovered image by reading exported ASCH file; 0) Off-line
segmentation results using thresholding method. ..................................................... 53

Figure 5-2 Off-line segmentation results at scan size equals 3 microns (a) AF M height
image; (b) Recovered image by reading exported ASCII file; (c) Off-line
segmentation results by using thresholding method. ................................................ 53

Figure 5-3 Different types of regions on tissue scaffold nanofibers: Data point (1) is out
of region of interest; data point (2) is right on top of a nanofiber; data point (3) is on
the side of a nanofiber. Scan size: 10 microns [62]. ................................................. 54

Figure 5-4 SPRM auto-tracking capability to scan along individual nanofibers (a) AF M
height image; (b)-(g) A series of images shows the time sequence of scanned results
by using SPRM. [62] ................................................................................................. 55

Figure 5-5 Flow chart of auto-tracking scan ..................................................................... 56

Figure 5-6 SPRM scan along individual nanofibers in a tissue scaffold. Left image is the
standard AF M image, region in black box are target scan region by SPRM. Right
image shows the scan result of using SPRM, two individual nanofibers are scanned
one after the other, other regions not scanned are padded 0 for display. .................. 57

Figure 5-7 Scan plan to stay on individual nanofiber (a) without boundary prediction; (b)
with boundary prediction .......................................................................................... 58

Figure 5-8 Demonstration of Scan angle in imaging Tissue scaffold nanofibers using
SPRM (a) horizontal scan direction; (b)vertical scan direction ................................ 59

Figure 5-9 (a) Standard AFM image of tissue scaffolds with scan size equals 6 microns;
(b) simulation scan result of SPRM by using horizontal scan direction; (c) real-time
scan result of SPRM by using horizontal scan direction. ......................................... 60

Figure 5-10 (a) Standard AFM image of tissue scaffolds with scan size equals 6 microns;
(b) simulation scan result of SPRM by using vetical scan direction; (c) real-time
scan result of SPRM by using vertical scan direction ............................................... 61

Figure 6-1 AFM and SEM images of electrospun carbon nanofiber tissue scaffolds. (a)-(c)
AFM images of Sample A, B and C. The scan area of each image is 5 square
microns and the z-height projection is 1500 nanometers. (d)—(f) SEM images of
Sample A, B and C. The scan area of each image is 20 square microns. Reproduced
from Figure 1, Reference: Y. Fan, Q. Chen, et al. in press, Vol. 2, Issue 4, Int. J.
Nanomedicine (2007) ................................................................................................ 64

Figure 6-2 surface roughness calculation of Scanning Probe Microscopy ....................... 65

Figure 6-3 Circle fit based on Kasa method. Reproduced fi'om Figure 5, Reference: Y. Fan,
Q. Chen, et al. in press, Vol. 2, Issue 4, Int. J. Nanomedicine (2007) ...................... 67

Figure 6-4 (a) Characteristic tip—dilation artifacts (b) The cross section of a real AF M
nano fiber image and the best fit circle. Reproduced from Figure 6, Reference: Y.
Fan, Q. Chen, et al. in press, Vol. 2, Issue 4, Int. J. Nanomedicine (2007) .............. 68

Figure 6—5 SPRM analysis of surface roughness (a) AF M image; (b) surface roughness
map of an individual nanofiber ................................................................................. 69

Figure 6-6 Histograms of surface roughness (a) Sample A; (b) Sample B; (c) Sample C.
Reproduced fi'om Figure 3, Reference: Y. Fan, Q. Chen, et al. in press, Vol. 2, Issue
4, Int. J. Nanomedicine (2007) .................................................................................. 70

Figure 6-7 Force volume imaging of Atomic Force Microscopy ..................................... 72

Figure 6-8 Tip probing position (a) tip probes top of nanofiber, force is normal to sample
surface; (b) tip probes side of nanofiber, force is not normal to sample surface 73

Figure 6-9 Force distance records the indentation of tip into soft sample ........................ 73
Figure 6-10 (a) force curve, both approach part and retract part are displayed; (b) force
distance curve, only approach part curve is displayed, which is used to calculate
elasticity property ...................................................................................................... 74
Figure 6-11 Elasticity histograms for (a) Sample A; (b) Sample B; (c) Sample C.

Reproduced from Figure 4, Reference: Y. Fan, Q. Chen, et al. in press, Vol. 2, Issue
4, Int. J. Nanomedicine (2007) .................................................................................. 76

xi

Figure 6-12 TEM images of (a) Sample A; (b) Sample B; (c) Sample C, with
corresponding selective area electron diffraction (SAED) images shown beneath.
Reproduced from Figure 7, Reference: Y. Fan, Q. Chen, et al. in press, Vol. 2, Issue
4, Int. J. Nanomedicine (2007) .................................................................................. 81

Figure 6-13 Close up images (a) Sample B; (b) Sample C. Reproduced from Figure 8,
Reference: Y. Fan, Q. Chen, et al. in press, Vol. 2, Issue 4, Int. J. Nanomedicine
(2007) ........................................................................................................................ 81

Figure 7—1 (a) Characterization ofSZR+ cells: A, Morphology of cell. Many cells display
broad circumferential lamellipodia; B. Protrusive activity (time-lapse sequence).
Lamellipodium protrudes with regular and smooth outline. Dotted line shows the
initial position of the cell edge. Time in minzsec. C. Actin cytoskeleton (phalloidin
staining). Circmnferential lamellipodium is rich in actin. D. Structural
organization of lamellipodia; (b) Characterization of BG2 cells. A. Morphology of
cell population. Cells display polarized larnella. B. Protrusive activity
(time-lapse sequence). Leading edge moves forward by combination of filopodial
(arrowheads) and lamellipodial protrusion. Dotted line shows the initial position of
the cell edge. Time in minzsec. C. Actin cytoskeleton (phalloidin staining). D.
Structural organization of lamellipodia. Images from reference [80] ....................... 85

Figure 7-2 Four target substrates (a) 2D: planar plastic coverslip; (b) NAN S:
Non-activated nanofibers; (c) SANS: surface-activated nanofibers, functionalized
with amine group; ((1) SANS + XL + D5/FGF-2: crosslinked and modified with D5
peptide or FGF-2 grth factor. ............................................................................... 87

Figure 7-3 Histogram of surface rouglmess map for (a) NANS: unmodified nanofibers; (b)
SANS + XL + F GF -2: crosslinked and F GF -2 modified nanofibers. AF M images of
these two substrates are shown as the inside images, scan size: 5 pm. .................... 88

Figure 7-4 Box plots of elasticity for two different nanofiber substrates SAN+XL+FGF-2
and NAN. The top images are AFM images of these nanofibers, scan size: 5 pm.
Points shown as crosses on AF M images are the positions where force curves are
collected to calculate elasticity. ................................................................................ 89

Figure 7-5 In vivo astrocyte-capillary system: astrocytes are pointed by arrows in (a);
FGF-ZS are dots pointed by arrows in (b). S. Meiners, et al. UMDNJ ..................... 90

Figure 7-6 Repulsive regime versus attractive regime in Phase Imaging. ........................ 91

Figure 7-7 Force distance curve showing the division between the two regimes, this curve
was taken on a SAN S+XL+F GF-2 nanofiber sample. ............................................. 92

Figure 7-8 Characteristic morphologies of a—HSA antibody (a) phase attractive mode; (b)
phase repulsive mode [81] ........................................................................................ 93

xii

Figure 7-9 surface chemistry characteristics (a) height image of SAN S+XL+ FGF-2, Z
range for height image is 500 nm; (b) height image of NANS, Z range is 1000 nm;
(c) phase image of SAN S+XL+FGF-2 at repulsive mode, Z range is 75 °; ((1) phase
image of NAN S at repulsive mode, Z range is 20 °; (e) phase image of
SAN S+XL+FGF~2 at attractive mode, Z range is 75 °; (f) phase image of NAN S at
attractive mode, Z range is 20 °. Image scan size is 1 pm. ....................................... 95

Figure 7-10 Fibroblasts cultured on 2D surface. (a) Phase contrast microscopy of live
cells, and (b) Optical microscopy of fixed cells indicate similar morphologies.
Reproduced from Figure 1, Reference: Q. Chen, Y. Fan, et al, Mater. Res. Soc.
Symp FF. Proc. (2007) .............................................................................................. 97

Figure 7-11 Fibroblasts cultured on amine coated nanofibers. (a) Phase contrast
microscopy of live cells, and (b) Optical microscopy of fixed cells indicate similar
morphologies. Reproduced from Figure 2, Reference: Q. Chen, Y. Fan, et a1. Mater.
Res. Soc. Symp FF. Proc. (2007) .............................................................................. 97

Figure 7-12 AFM images of 3T3 NIH fibroblast cultured on 2D surface. (a) Deflection
image of a typical fibroblast; (b) Close-up deflection image of top right region; (c)
Close-up deflection image of top left region; ((1) Close-up deflection image of
bottom right region; (e) Height image of vertex region for another cell showing

structures (close-up deflection image in inset). Reproduced from Figure 3, Reference:
Q. Chen, Y. Fan, et a1. Mater. Res. Soc. Symp FF. Proc. (2007) ............................. 98

Figure 7-13 3T3 NIH fibroblast cultured on amine coated nanofibers (SANS). (a)
Multiple cell-cell interactions on the SANS are observed in the optical microscopy
image. The cells in the black box are shown in close-up AF M images in (b) through
(g). (b) through ((1): increasing close-up AFM height images; (6) through (g):
corresponding increasing close-up AF M deflection images. Reproduced from Figure
4, Reference: Q. Chen, Y. Fan, et al. Mater. Res. Soc. Symp FF. Proc. (2007) ....... 99

Figure 7-14 Actin stain + Fluorescent Microscopy images of astrocytes on different
substrate (a) astrocytes on 2D substrate imaged at 24 hours; (b) astrocytes on NAN S
at imaged 24 hours; (c) astrocytes on modified NAN S imaged at 24 hours; ((1)
astrocytes on 2D substrate imaged at 48 hours; (e) astrocytes on NAN S imaged at 48
hours; (1) astrocytes on modified NAN S imaged at 48 hours ................................. 102

Figure 7-15 AF M images of astrocytes grown on different substrates for 24 hours (a) 2D
substrate; (b) NANS; (c) modified NAN S. Image scan size : 50 um ..................... 103

Figure 8—1Polymer monomers (a) poly (methyl methacrylate), and (b) poly
(e-caprolactone). (c) Experimental configuration of electrospinning experiment.. 108

Figure 8-2 (a) Field emission SEM of PMMA-based carbon nanofibers and (b) atomic

force microscopy of PMMA-based carbon nanofibers with 6% weight of single
walled carbon nanotubes added. A 200 micron bore radius resulted in micron-scale

xiii

diameter nanofibers. Reproduced from Figure 2, Reference: Rutledge, S.L., Shaw, H.
C., Benavides, J. B., Yowell, L. L., Chen, Q., Jacobs, B. W., Song, S. P., Ayres, V.
M., Diamond and Related Materials, 2006. 15(4-8): p. 1070-1074 [84] ................ 110

Figure 8-3 Field emission SEM of poly (e-caprolactone) carbon nanofibers spun at bore

radii (a) 152.4, (b) 254.0, and (c) 406.4 microns. As the bore size was increased, a
slight increase in fiber diameter with a large decrease in bubble size was observed.
Reproduced from Figure 3, Reference: Rutledge, S.L., Shaw, H. C., Benavides, J. B.,
Yowell, L. L., Chen, Q., Jacobs, B. W., Song, S. P., Ayres, V. M., Diamond and
Related Materials, 2006. 15(4-8): p. 1070-1074 [84] ............................................. 111

Figure 8-4 Mean force distance curves for three nanofibers samples electrospun with

different bore radii and the aluminum foil. The mean force distance curve areas
increased with increasing bore radius (insets), indicating a decrease in nanofiber
elasticity with increasing bore radius. Reproduced from Figure 5, Reference:
Rutledge, S.L., Shaw, H. C., Benavides, J. B., Yowell, L. L., Chen, Q., Jacobs, B.
W., Song, S. P., Ayres, V. M., Diamond and Related Materials, 2006. 15(4-8): p.

1070-1074 [84] ........................................................................................................ 112
Figure 10-1 Diagram of providing an automatic guide for tissue scaffold design ......... 118
Figure 10-2 SEM images of nano circuits ................................................................... 119

Figure 10-3 Scanning scheme implementation for GaN anocircuit: (a) Image captured by

standard SPM, the image side is 5 x 5 microns; (b) Scanning scheme simulation
showing lines of the scan plan; (c) Real-time image captured by SPRM system using
self-defined scanning scheme along the nanowire. Reproduced from Figure 4,
Reference: B.W. Jacobs, V.M. Ayres, M.A. Tupta, R.E. Stallcup, A. Hartman, J .B.
Halpem, M-Q. He, M.A. Crimp, A.D. Baczewski, N.V. Tram, Q. Chen, Y. Fan, S.
Kumar, L. Udpa , 2006 6th IEEE Conference on Nanotechnology Proceedings 120

Figure 10-4 SPRM scan strategy to locate nanowire in nanocircuits (a) AF M images of

nanocircuits, the arrow points to the nanowire; (b) simulation scan result shows
locating the nanowire by following edges of nanocircuits. Image Size: 50pm. ..... 121

xiv

INTRODUCTION

Scanning Probe Microscopy (SPM) is a newly emerging and powerful technique for
nanoscale science in biology and medicine. Its great advantages are its direct
investigative capability and its inherent resolution which easily and reliably reaches
nanometer level, the precise scale for significant investigations of macromolecular
biological units. SPM techniques can also function in a liquid ambient environment,
which means that biological entities can be investigated under nearly life-like conditions
[1-3].

But SPM has limitations in several aspects. One of them is the relatively slow rate of
scanning during SPM imaging. It usually takes several minutes to generate a typical
image. Another is the heavy dependence on an expert human both as an operator and as

an image interpreter. Still another set of issues comes from imaging artifacts.

Our research group designed and developed Scanning Probe Recognition
Microscopy (SPRM) system in partnership with Veeco Instruments, which integrates
recognition ability into the SPM system to allow the SPM system to automatically track
specific regions of interest. Also, additional powerful functionalities are also integrated
into the SPRM system, including feature recognition, object classification, auto-tracking
and property analysis abilities. Different fimctionalities can be realized by operating in
online or offline modes. Therefore, in addition to saving operation time, SPRM has the
advantages of providing more meaningful information about samples automatically,
efficiently and accurately. The power of SPRM allows the investigation of many

situations in which standard SPM is insufficient.

The candidates’ original research contributions to the development of SPRM are as

l

follows. One major contribution has been development of a set of recognition strategies
that provide mathematical criteria that guides the auto-tracking. The recognition
strategies are grounded on the imaging processing concept of feature definition. A feature
is any relevant aspect of a system. It must be probable of precise mathematical definition
in terms of a test criterion. The candidate’s contribution is to define the first set of
features within Scanning Probe Microscope data. These features were then used as part of
the feedback loop which guided the auto-tracking implementation. Another major
contribution by the candidate was the first adaptation of SPRM for investigation of a
significant nanobiomedical problem, analyzing nanoscale 3D cues for spinal cord repair
research. The candidate also made significant contributions to the development of the
auto-tracking implementation, and to material properties analyses.

The dissertation is divided into three parts. In the first part, the main focus is the
instrumentation design and development of the new Scanning Probe Recognition
Microscopy capability. The second part focuses on the successfirl application of SPRM
system in analyzing nanoscale 3D cues for spinal cord repair research. The third part
summaries the conclusions from the present research and identifies the future work.

In the instrumentation part: the first chapter introduces the commonly used imaging
techniques to set both SPM and the new SPRM within the context of techniques used for
nanoscale investigations, and particularly for nanobiological investigation. Then the
fundamentals of SPM, which is the base instrument of SPRM, are addressed in chapter 2.
Chapter 2 discusses the fundamental elements of SPM, its advantages, its limitations and
the current start of art. Chapter 3 discusses the new fimctionalities built in SPRM and the

basic design and development of SPRM and how these address issues in SPM.

In the second part, the main focus is using SPRM to investigate a very important
research topic: spinal cord repair. In Chapter 4, motivation and background of SPRM in
spinal cord repair are addressed. Chapter 5 discusses the auto-tracking ability of SPRM
which enables the system to scan only on individual nanofibers within the tissue scaffolds
used for regenerative medicine, thereby, collecting the first accurate nanofiber properties
analyses. In Chapter 6, the topographic, mechanical and chemical properties of
nanofibers which influence neural cell attachment are addressed and analysis of these
properties are done by using SPRM. Cell responses to different substrates are explored
and compared in Chapter 7. Chapter 8 discussed the correlation of electronspinning
parameters with nanofiber properties analysis. The last part of this dissertation
summarizes the conclusions reached from the candidate’s research, along with discussion

of the future research directions.

Part I SPRM

Instrumentation Development

Part I SPRM INSTRUMENTATION DEVELOPMENT

A new SPM-based technique, Scanning Probe Recognition Microscopy (SPRM), is
presented in this dissertation. The new SPRM system improves several aspects of SPM,
including speed, accuracy and flexibility. Also, the SPRM system enables the
investigation of situations in which SPM is insufficient. SPRM will be an outstanding
new tool in the multiple research fields that need nanoscale investigation with SPM
techniques.

Nanobiology especially needs techniques that can provide direct observation and
analysis of phenomenon at the nanoscale. SPRM has the potential to be widely used in
experimental nanobiology. The development of SPRM for applications in nanobiology is
multidisciplinary in two respects. One is the development of the new SPRM capability
within the overall field of scanning probe microscopy. Another is to set both SPM and the
new SPRM within the context of experimental techniques used in nanobiological
investigations. As the current state of art techniques used in nanobiology are also those
used in nanotechnology in general, discussion of these provides a setting for the
development of SPM as well as SPRM. Therefore, the experimental techniques used in
nanobiology will be discussed first, followed by discussion of SPM and finally by the
discussion of the development of SPRM.

Therefore, the first part of dissertation focuses on the development of the SPRM
system. It starts from the most commonly used experimental imaging tools in current
nanobiology researches. Then SPM techniques are introduced, the advantages and

limitations of SPM are emphasized, and the latest research works to improve SPM

performance are addressed. Finally, the design and development of the new SPRM

system are discussed in detail to enhance the performance of SPM.

CHAPTER 1

1 IMAGING TECHNIQUES USED IN NANOBIOLOGY

The present SPRM system is successfirlly used to investigate many research topics in
nanobiological field. Therefore, a survey of imaging techniques commonly used to
investigate nanobiological phenomenon is addressed and comparisons between them are

also discussed.

1. 1 Optical Microscopies

(1) Phase contrast microscopy

Phase contrast microscopy is an optical microscopy illumination technique in which
small phase shifts in the light passing through a transparent specimen are converted into
amplitude or contrast changes in the image [4]. It is preferable used to bright field
microscopy to provide contrast of transparent specimens such as living cells or small
organisms.

(2) Fluorescence microscopy

Fluorescence microscopy is an optical microscopy technique used to study the
properties of specimens using the phenomena of fluorescence and phosphorescence in
addition to reflection and absorption. It is one of the most ubiquitous tools in biomedical
laboratories. In fluorescence microscopy, the sample under investigation becomes itself
the light source. The fluorescence microscopy is based on the phenomenon that certain
material emits energy detectable as visible light when irradiated with the light of a
specific wavelength. The sample can either be fluorescing in its natural form like

chlorophyll and some minerals, or treated with fluorescing chemicals. Fluorescence

7

microscopy has 3 strengths. First, it has high biological specificity. Second, it is highly
sensitive in the imaging of cells and tissues. The exquisite sensitivity and image contrast
allows biological structures to be imaged on the submicron length scale. Third, it is a
minimally invasive imaging technique, which gives it the ability to study biological
structure and function in vivo[5].

But in a thick sample, the signals from multiple sample places are integrated to form
the final image. Since there is little correlation between the structures at different depths,
the final image becomes fuzzy.

Fluorescence microscopy can be implemented in other more advanced configurations
to enable novel imaging modes. Two particularly important configurations, confocal
microscopy and two-photon microscopy, are discussed.

(3) Confocal microscopy

Confocal microscopy is an optical imaging technique that can obtain 3D sections in
thick specimens by using a spatial pinhole to eliminate out-of—focus light or “flare” in
specimens that are thicker than the focal plane. It achieves blur-free images of thick
specimens at various depths, at the cost of obtaining fluorescence signal from only a
single point in the specimen.

There has been a tremendous explosion in the popularity of confocal microscopy in
recent years, due in part to the relative ease with which extremely high-quality images
can be obtained fi'om specimens prepared for conventional fluorescence microscopy. It
has been used extensively to investigate microstructures in cells and in the imaging of
tissues[6].

In laser scanning confocal microscopy, a fluorescent specimen is illuminated by a

point laser source, and each volume element is associated with a discrete fluorescence
intensity. Laser scanning confocal microscopy offers several advantages over
conventional fluorescence microscopy including controllable depth of field, the
elimination of image degrading out-of-focus information. The primary advantage is the
ability to serially produce thin optical sections through fluorescent specimens that have a
thickness ranging up to 50 pm or more[7]. Advances in confocal microscopy have made
possible multidimensional views of living cells and tissues that include image
information in the x, y and 2 dimensions as a function of time and presented in multiple
colors (using two or more fluorophores). The disadvantages of confocal microscopy are
primarily the limited number of excitation wavelengths available with common lasers,
which occurs over very narrow bands and are expensive to produce in the UV region.
Another downside is the harmful nature of high intensity laser irradiation to living cells
and tissues. Finally the best spatial resolution obtained to date is on the order of 400 nm,
which is still too large to explore macromolecular resolution processes which have l-10
nm resolution requirement.

(4) Two-photon microscopy

Two-photon microscopy is an alternative to confocal microscopy for the 3D imaging
of thick specimens. Two-photon microscopy uses non-linear absorption of two photons to
induce fluorescence that is confined to a very small region. A laser beam is scanned
laterally across the sample to generate 2D fluorescence images from an extremely thin
optical section within the sample. This optical section can be varied in depth, building up
a stack of images to produce a 3D rendering the sample. An excellent demonstration of

the ability of two-photon imaging for deep tissue imaging is in the neurobiology area[8].

The major advantage of two-photon microscopy is its ability to permit
high-resolution and high-contrast imaging from deep within intact living tissue.
Two-photon microscopy is particularly useful for live imaging of thick samples because it
has less photodamage compared with confocal microscopy. Since two-photon microscopy
obtains 3D resolution by the limitation of the region of excitation instead of the region of
detection as in a confocal system, photodamage of biological specimens is restricted to
the focal point. Photodamage at the focal plane will still occur, as with the confocal, but
there isn't damage above and below the plane of focus. The disadvantages of two-photon
microscopy lie in the following aspects: Pulsed lasers are generally much more expensive;
the microscope requires special optics to withstand the intense pulses; the two-photon
absorption spectrum of a molecule may vary significantly from its one-photon
counterpart; and wavelengths greater than 1400 nm may be significantly absorbed by the
water in living tissue. Finally the best obtained two-photon microscopy resolution to date

is on the order of 50 nm which is still above the desired 1-10nm resolution.

1.25Iectron Microscopies

The transmission electron microscope (TEM) operates on the same basic principles
as the light microscope but uses the quantum wavelike nature of electrons instead of light
pass through the object. It is an imaging technique whereby a beam of electrons is
transmitted through a specimen, then an image is formed, magnified and directed to
appear. In TEM, electrons are focused by two or more electron lenses to form a greatly
magnified image onto photographic film or a charge coupled device camera. The image
produced by TEM is 2D and the brightness of a particular region of the image is

proportional to the number of electrons that are transmitted through the specimen at that

10

position on the image. The resolution of light microscope is limited by the wavelength of

light. Since “light source” of TEM is electrons. The resolution of a TEM is limited by the

electron wavelength: A = fi— , where h = Plank’s constant 6.6 ><10'34 (J - s) and p
P

is the electron’s momentum. Because the electrons wavelengths are much shorter than

optical wavelengths, it is possible for TEM to get a much higher resolution. The

0—10 m) level.

resolution of a TEM easily reaches the Angstrom (1

In nanobiology, the TEM excels as a diagnostic tool with respect to the detection and
identification of both abnormal tissue anatomy and the pathogens responsible for the
disease [9]. The key problem in the use of TEM for nanobiology investigations is the
sample preparation. The specimens must be very thin and able to withstand the high
vacuum present inside the instrument. Biological specimens must be frozen, sectioned
while frozen into electron transparent slice around 10 nm thick, coated with a thin around
10 nm conductive film of gold and imaged in Cryo-TEM that maintains the biological
tissues in its frozen state. Therefore, there is always the question of how well the TM
section represents the living state. Also the field of view is relatively small, raising the
possibility that the region analyzed may not be characteristic of the whole sample. There
is also the potential that the frozen sample may be damaged by the electron beam.

The latest advancement in electron microscopy is 3D reconstruction of cellular
components at a resolution that is on the order of magnitude of atomic structures. The
method for reconstruction of 3D images of single, transparent objects recorded by TEM
is called electron tomography (ET). With improvements in instrumentation, data

collection methods and techniques for computation, ET may become a preferred method

for imaging isolated organelles and small cells [10]. Its resolution is excellent. Its main

ll

disadvantages are that the sectioned and coated biological specimens do not ideally
represent the dynamic living state, and that, as discussed, the sample preparation and

instrumentation requirement is intensive.

1.3 Brief Introduction of Scanning Probe Microscopy in Biology
and Medicine
The family of Scanning Probe Microscopy techniques has revolutionized studies of

nanostructures. The key capability of SPM is that, through a controlled combination of
feedback loops, detectors and piezoelectric actuation, it enables direct investigations of
atomic to nanometer scale phenomena. A key difference between scanning probe and
optical microscopies is that in scanning probe microscopies, a tip is in actual direct
interaction with a nanoscale object rather than viewing its image. The SPM record for
contributions in nanobiology and medicine includes an impressive number of firsts: the
first direct observations of DNA and RNA by atomic force microscopy [11], the first
direct investigations of receptor sites by atomic force microscopy [2, 12], the first direct
investigations of cellular water pore sites by atomic force microscopy [13, 14] , the first
direct investigations of ion channels by atomic force microscopy [15], the first direct
investigations of ligand docking by functionalized tip atomic force microscopy [16],
atomic force microscope investigations of surface and submembranous structures on
living cells [17, 18] including cancer cells [19-21], the first direct investigations of the
binding of fluorescently conjugated lectins to cell surface glycoproteins by near-field
optical microscopy [22], and many more important first direct studies.

Atomic Force Microscopy (AFM) is one of the most widely used techniques in SPM
family. It was developed specifically for nanoscale investigations of biological samples,

which are largely nonconductive [23]. It has been successfully used to image many

12

biological samples ranging from genetic material to cells to bone [24].

Studies can be done under physiological conditions in AFM allow us to image
biological processes [25]. Studies can also be performed at the macromolecular level of
resolution. For example, AF M images of transcription complexes have been obtained of
RNA polymerase in complex with DNA [26]. Such studies detailing the
structure-function relationship of transcription process are key in furthering
understanding of gene expression.

AFM can also be operated in force scan mode, which allows for the measurement of
adhesion forces between receptors and their corresponding binding partners, or ligands
[27]. It can also serve as a microindenter that probes cells revealing information about
their mechanical properties. The mechanical properties of cells play an important role in
such essential physiological processes such as cell migration and cell division [28].

Scanning tunneling microscopy (STM) has the highest resolution of all SPM family.
It is highly desirable for use in investigation of molecular biology and medicine [29].
DNA and RNA structures have been successfully investigated by using STM [30].

The above examples document why SPM is emerging as a powerful tool for
investigations in nanobiology and medicine. The drawbacks that limit the widespread use
of SPM in nanobiology, and therefore the new biomedical knowledge that it could
generate, are those previously given: slow scan speed, imaging artifacts and dependence
on an expert human operator. A detailed discussion of SPM operation and the root causes

of its drawbacks are given in Chapter 2.

13

CHAPTER 2

2 SCANNING PROBE MICROSCOPY IMAGING
TECHNIQUES

2. 1 Fundamentals of SPM

Scanning Probe Microscopy (SPM) is emerging as a powerful technique for
nanoscale science in biology and medicine because it provides direct real-time
information at the nanoscale with in-vitro level sample preparation. It can be used as an
imaging device, which allows for the acquisition of atomic-level images of biological
structures as well as to measure forces of interactions between two opposing surfaces
down to the single-molecule level. Its great advantages are its direct investigative
capability and its inherent resolution which easily and reliably reaches the nanometer
level, the precise scale for significant investigations of biological units. Therefore, SPM
resolution exceeds that of the optical techniques and reaches the correct level, and SPM
sample preparation is much less drastic than that required by the electron microscopy
techniques. SPM techniques can be operate in air, liquid or vacuum environment, which
means that biological entities can be investigated under nearly life-like conditions. In
addition to its use as an imaging technique, SPM can also be used as an interactive device
as shown in investigations of receptor-ligand binding. The interactive capabilities of SPM
are just beginning to be recognized and exploited.

Scanning Probe Microscopy (SPM) is a branch of microscopy that forms images of
surfaces using a sharp probe in near-field interaction with the surface to scan the
specimen. An image of the surface is obtained by mechanically moving the probe in a

raster scan of the specimen, line by line, and recording the probe-surface interaction as a

14

function of position. The SPM family includes Scanning Tunneling Microscopy (STM),
Atomic Force Microscopy (AFM), Near-Field Scanning Optical Microscopy (NSOM)
and others. STM was developed in the early 1980's. The significance of its invention was
recognized by the award of the Nobel Prize in Physics in 1986 [31], one of only four such
awards given for the development of a revolutionary new instrument. This invention was
quickly followed by the development of other techniques in SPM family. AFM is one of
the most important later techniques. Despite the number of types of SPM and modes in
which they can be operated, the underlying operation is the same for them all. A sharp
probe undergoes a near-field interaction with a sample surface during a piezoelectric
raster-motion scan of the surface, thereby producing a map of the interaction. Each
different type of SPM is characterized by the nature of the local probe and its interaction
with the sample surface. The inherent resolution is atomic level for STM and nanometer
level for AF M.

However, SPM has several limitations. The speed of SPM imaging is one key
limitation. In most SPM techniques, a sharp tip is scanned across a sample to image its
topography and material properties. It typically takes 1 to 100 minutes to obtain a
high-quality image. The use of SPM would increase dramatically if the speed of SPM
could match that of other scanning microscopes such as confocal and scanning electron
microscopes. Moreover, there are many experiments, such as watching biological
processes that simply cannot be done without faster imaging.

SPM is human operator intensive and a long learning curve is required to become
truly expert operator. Insufficient instruments and teachers create a cycle of limited use of

a powerful capability.

15

2.1.1 Block diagram

The research work presented in this dissertation is mainly based on AFM, but it can be
extended to all the SPM family easily. Therefore, AFM is discussed in detail as a typical

rather than a specific example. The block diagram of AFM is shown in Figure 2-1.

 

 

 

 

 

 

 

 

 

  
 
    
 
 

 

 

 

 

 

 

 

 

 

 

Feedback Loop maintains constant

CONTROL tip-sample interaction
Controller
Electronics

Detector Laser

Electronics (J
ACTUATOR
SENSOR 4-cell 1

 

 

 

 

Photodiode

 

Piezoelectric
Scanner

Figure 2-1 Block diagram of Atomic Force Microscopy
(1) Cantilever & tip
The AFM consists of a microscale cantilever with a sharp tip (probe) at its end that is
used to scan the specimen surface. The cantilever is typically silicon or silicon nitride with
a tip radius of curvature on the order of nanometers. When the tip is brought into proximity
of a sample surface, forces between the tip and the sample lead to a deflection of the
cantilever according to Hooke's law. Depending on the situation, forces that are measured

in AFM include ion-ion repulsive forces, Van der Waals forces, capillary forces, chemical

l6

bonding, electrostatic forces, Casimir forces, solvation forces etc. The cantilever should
have a small spring constant, and the tip should have a radius of curvature less than
20~50nm and a cone angle between 10~20 degree. Material used to make AFM tips are
usually Silicon and Silicon Nitride for hardness greater than the majority of sample
surfaces, and for ease of fabrication using standard semiconductor industry micromachine

(2) Sensor & photodiode

Typically, the deflection of cantilever is measured using a laser spot reflected from
the top of the cantilever into a four-cell (quad) photodiode. The four elements of the quad
photodiode (position sensitive detector) are combined to provide different information
depending on the operating mode. The amplified differential signal between the top two
elements and the two bottom elements provides a measure of the deflection of the
cantilever. In this way, the photodiodes can detect the movement of the beam, and
therefore the movement of the cantilever and tip. The vertical movement of the tip is
thereby measured as a voltage change.

(3) Scanner

Traditionally, the sample is mounted on a piezoelectric scanner, which can move the
sample in the z direction for maintaining a constant force by receiving signal from
feedback control unit, and the x and y directions for scanning the sample. The scanners are
made from piezoelectric material, which expands and contracts proportionally to an
applied voltage. The scanner is constructed by combining independently operated piezo
electrodes for X, Y and Z into a single tube, forming a scanner which can manipulate

samples with extreme precision in 3 dimensions.

17

Table 2-1 Multimode SPM Scanner Specifications [32]

 

 

 

 

 

 

 

 

 

 

 

 

Model Scan Size Vertical Range

AS-0.5 (“A”) 0.4umx0.4um 0.4um

AS-12 (“E”) 10umx10um 2.5um

AS-12V(“E” vertical) 10umx10um 2.5 pm

AS-130(“J”) 125 umx125pm 5.0um

AS- 1 30V(“J” vertical) 1 25pmx125 pm 5 .Oum

AS-200 200umx200um 8.0 pm
(4)Feedback control

If the tip were scanned at a constant height, there would be a risk that the tip would
collide with the surface, causing damage. Hence, in most cases a feedback mechanism is
employed to adjust the tip-to-sample distance to maintain a constant relationship between
the tip and the sample. The type of relationship kept constant depends on the mode of
operation. The feedback control unit takes its input from the sensor and yield an output to

the scanner that maintains a fixed tip-sample interaction.
2.1.2 Operation modes

The AF M can be operated in a number of modes [33], depending on the requirements
of applications. Different operation modes have different force of interaction between the

tip and sample surface as shown in Table 2-2. The most commonly used operation modes

are contact mode and tapping mode.

18

Table 2-2 Operation Modes of AFM

 

Mode of Operation Force of Interaction

 

contact mode strong (repulsive) - constant force or constant distance

 

non-contact mode weak (attractive) - vibrating probe

 

tapping mode strong (repulsive) - vibrating probe

 

lateral force mode fi'ictional forces exert a torque on the scanning cantilever

 

 

 

 

Contact AFM measures topography by sliding the probe’s tip across the sample
surface. It can be operated in both air and fluids. It can have relatively high scan rate
(throughput). Rough samples with extreme changes in vertical topography can sometimes
be scanned more easily in Contact mode. However, Lateral (shear) forces may distort
features in the image. The forces normal to the tip-sample interaction can be high in air
due to capillary forces from the adsorbed fluid layer on the sample surface. The
combination of lateral forces and high normal forces can result in reduced spatial
resolution and may damage soft samples (i.e., biological samples, polymers, silicon) due
to scraping between the tip and the sample.

Tapping ModeTM AFM measures topography by tapping the surface with an
oscillating tip. This eliminates shear forces, which can damage soft samples and reduce
image resolution. Therefore, soft samples can be imaged and image resolution can be
improved by using Tapping Mode. Tapping Mode is available in air and fluids too. It has
relatively higher lateral resolution on most samples (1 nm to 5 nm), lower forces and less
damage to soft samples imaged in air. Lateral forces are virtually eliminated, so there is

no scraping. However, it has slightly slower scan rate than contact mode AF M.

19

2.2Advantages and Limitations of SPM

2.2.1 Advantages

The SPM has several advantages over the scanning electron microscope (SEM).
Unlike the electron microscope which provides a two-dimensional projection or a
two-dimensional image of a sample, the SPM provides a true three-dimensional surface
profile[34].

Additionally, samples viewed by SPM do not require any special treatments (such as
metal/carbon coatings) that would irreversibly change or damage the sample. While an
electron microscope needs an expensive vacuum environment for proper operation, most
SPM modes can work perfectly well in ambient air or even a liquid environment. This
makes it possible to study biological macromolecules and even living organisms.

In principle, SPM can provide higher resolution than SEM. It has been shown to give
true atomic resolution in ultra-high vacuum (UHV). UHV AFM is comparable in

resolution to Scamring Tunneling Microscopy and Transmission Electron Microscopy.
2.2.2 Limitations

A disadvantage of SPM compared with the scanning electron microscope (SEM) is
the image size. The SEM can image an area on the order of millimeters by millimeters
with a depth of field on the order of millimeters. The SPM can only image a maximum
height on the order of 10 micrometers. While a 12 inch water scan area capability has
been developed to serve the semiconductor fabrication community. The wide scan area is
paid for by a loss of resolution. SPM as it was developed was intended for high resolution

imaging of a nanoscale representative area.

20

Nonlinearity issue of piezo is also a problem. Nonlinearity means that when a linear
voltage ramp is applied to piezo, the peizo moves in a nonlinear motion. It will introduce
artifacts SPM images.

Another inconvenience is that at high resolution, the quality of an image is limited by
the radius of curvature of the probe tip, and an incorrect choice of tip for the required
resolution can lead to image artifacts, as can any tip damage that occurs while scanning.

SPM can not scan images as fast as an SEM, requiring several minutes for a typical
scan, while an SEM is capable of scanning at near real-time after the chamber is
evacuated. The relatively slow rate of scanning during SPM imaging often leads to
thermal drift in the image, making the SPM microscope less suited for measuring
accurate distances between artifacts on the image.

SPM images can be affected by hysteresis of the piezoelectric material and cross-talk
between the (x,y,z) axes that may require software enhancement and filtering. Such

filtering can "flatten" out real topographical features.

2.3lmprovement of SPM Start of Art

SPM speed is limited by the slowest components in its entire control loop.
High-speed AFM techniques developed to improve the speed [35] are an active research
area. Improvement of scanning speed has been made by using smaller cantilevers[36],
faster scanners[37] and better controllers[38, 39] [40].

In an independent but related development by the Veeco co-principal investigator of
the present project. Veeco's new Easy-AFM offers ability to automatically adjust scan
parameters by choosing scan parameters which make new parameter-based feedback

loops more stable. The scan rate can be increase when the feedback loop is stabilized.

21

Therefore, Easy-AFM reduces the time for initial setup and adjustment of parameters,
enabling a faster path to sophisticated, high-quality images[4l]. Easy-AFM was

developed with integration with SPRM in mind.

22

CHAPTER 3

3 SCANNING PROBE RECOGNITION MICROSCOPY

Our research group designed and developed the Scanning Probe Recognition
Microscopy in partnership with Veeco Instruments. Scanning Probe Recognition
Microscopy integrates recognition ability in Scanning Probe Microscopy to allow us to
adaptively follow and investigate specific regions of interest. Therefore, The SPM system
itself is given the ability to auto-focus on regions of interest through incorporation of
recognition-based tip control. The recognition capability is realized using techniques in
pattern recognition and image processing fields. Adaptive learning and prediction are also
implemented to make detection and recognition procedures quicker and more reliable.
The new SPRM system has the advantages of saving operation time, providing powerful
feature/property analysis ability etc. The powerful firnctionalities of the new SPRM
system will be explained in detail in the following section.

The candidate’s contribution is to define the first set of features within Scanning
Probe Microscope data. These features were then used as part of the feedback loop which
guided the auto-tracking implementation. Another major contribution by the candidate
was the first adaptation of SPRM for investigation of a significant nanobiomedical
problem, analyzing nanoscale 3D cues for spinal cord repair research. The candidate also
made significant contributions to the development of the auto-tracking implementation,

and to material properties analyses.

23

3. 1 SPRM System Functionalities

3.1.1 Recognition ability

The new SPRM system can recognize operator defined features of samples under
observation, such as edge, shape and so on. This is paradigm shift from the operator
defining scanning parameters to the operator defining significant aspects of their
problems in the term of features. The recognition ability gives the new SPRM system the
power to automatically recognize different types of objects and classify them into groups
based on given prior information. SPRM will increase ease of use and work efficiency at

the same time.
3.1.2 Auto-tracking capability and region of interest

Based on the recognition capability of the SPRM, auto-tracking capability is built
into the SPRM system by combining adaptive learning and prediction techniques.
Therefore, the SPRM system is able to automatically scan while focused on only the
region of interest instead of raster scanning the whole scan region which includes the
background as well as the region of interest. It is very usefirl especially when the
interested regions are sparse on the whole scan region, such as cells on substrate. With
SPRM, dramatic operation time can be saved and only meaningful information about the

interested region is extracted at the same time of scanning.
3.1.3 Real-time result display capability

The SPRM system provides a real-time scan result display, the same as a standard
SPM. The ability enables users to know the scan result in real-time. Then users are able

to adjust parameters in real-time according to the displayed scan result. Auto-adjustment

24

of scan parameters has also been implemented by the Veeco Co-Principal Investigator,

Veeco “Easy-AF M” TM .
3.1.4 Property analysis capability

Combined with auto-tracking capability, the SPRM system has the ability to
automatically analyze important properties of objects on sample surface, such as surface
roughness, elasticity, etc. In a standard SPM, this would involve manually applied
repetitive work which is very time consuming. Also a human operator applied probe may
actually access a region of unreliable data points. The new SPRM system has the obvious

advantages in convenience, efficiency and accuracy.
3. 2 SPRM Prototype System Structure

The Scanning Probe Recognition Microscopy system is built based on standard
Scanning Probe Microscopy, its block diagram is shown in the Figure 3-1. Compared
with standard SPM setting, the new SPRM system adds Signal Access Module [42], a
second computer and embeds new algorithms into the standard SPM computer.

The new SPRM system can get the real—time analog signals from standard SPM
through Signal Access Module (SAM). Signals from different ports of SAM include
multiple information: height of sample surface, cantilever deflection, and so on. The
SPRM software embedded in the controller also sends control signals to Signal Access
Module to assist real-time scan result display. All these signals are sampled by an A/D
card installed on the second computer. Scan results are displayed in real-time in the
second computer by analyzing these sampled signals.

The key recognition and implementation elements are sofiware embedded inside the

Veeco SPM controller. The module labeled “SPRM scan plan generation” automatically

25

generate scan plan in real-time in order to auto-track the region of interest. The scan plan
is generated and updated in real-time, and the system has the adaptive learning ability

based on previous information.

 

SPRM

Scanplan generation

Dt,Y coordinates X,Y coordinates
7. Holst!

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

   

 

 

 

 

 

 

Veeco SPM system
Signal
Access d—h Computer
Module
(2.05%? Laser ' '
0‘ \‘ Cantilever Congo! Display
3 Piezo
. dnve
Piezo signal )Z(,HYei ht Computer $P_RM
12 g ,, A/D Real-time display
Q—D
X,Y

 

 

 

Figure 3-1 Block diagram of Scanning Probe Recognition Microscopy
3.3SPRM System Design & Software Development
The Scanning Probe Recognition Microscopy software is divided into two parts. One
part is embedded in the main SPM controller to generate real-time SPRM scan plan. The
other part is responsible for analog signal sampling, data processing and real-time scan
result display on the second computer. Object oriented design is used in software design of

the first part. There are mainly 6 classes as shown in Figure 3-2.

26

 

Figure 3-2 Software design diagram

The algorithm class includes all algorithms used in SPRM system. It is the key part
of the system. It needs to communicate with all other parts.

The scan class has the interface with the SPM system to control the parameters
associated with scan and to control the real movement of tip.

The pointData class defines the information at a single point.

The scanRegion class stores information of the scan region.

The simulation class enables the software to be run under simulation mode instead of
run under real SPM. It is a helpful tool for debugging and testing of the system.

The trigger class provides the interface to Signal Access Module of Scanning Probe
Microscopy. It makes the analog signal output possible.

The second part of the SPRM software to display real-time scan results is developed
by Matlab. The DI sampling card is used in sampling real-time analog signals from Signal

Access Module.

27

3.4 SPRM Recognition Algorithm

Many different algorithms are used in implementation of Scanning Probe
Recognition Microscopy. Based on different applications, there are specific algorithms
that showed the best recognition results. But there are some common techniques which
can be applied to recognize most of observation targets and produce promising results.

The overall recognition/classification procedure consists of three steps as depicted in

Figure 3-3.

I

Step A. Preprocessing
- Segmentation
- Noise removal

I

Step B. Feature extraction
- Shape feature

- Edge feature

. Symmetry feature

I

Step C. Classification
- Classifier structure
- Decision rules

 

 

 

 

 

 

 

 

 

 

Figure 3-3 Schematic of overall approach. Reproduced from Figure 2, Reference: Q. Chen, Y. Fan, L.
Udpa, and V.M. Ayres, Vol. 2, Issue 2, pp. 18]-189, Int. J. Nanomedicine (2007)

3.4.1 Preprocessing: Wavelet-based image segmentation

The idea in using wavelet-based multi-resolution analysis is to detect the edges of the
target objects. After the edges of target objects are successfully detected, it will be much
easier to segment the region of the target objects from the background by calculating the
boundary box of the target objects.

Since wavelet analysis is multi-resolution analysis, it is very useful in detecting

28

edges under different levels of scale. A remarkable property of the wavelet transform is
its ability to characterize the local regularity of fimctions. For an image f (x, y), its edges
correspond to singularities, and thus are related to the local maxima of the wavelet
transform modulus. Therefore, the wavelet transform is an effective method for edge

detection [43].

Two wavelets are respectively the partial derivative along x and y of a 2D smoothing

function 0(x, y).

1 5900’)
x, = —— 3‘]
W ( y) 6x
60 x,
w2(x.y) = —( y) 3-2
0y
Let
1 x
W;(X,y)=—2I//l(—,X) 3'3
S S S
1 x
W3(x,y)=—2-w2(-.Z) 34
S S S

For any image f (x, y) , the wavelet transform has two components:

Wlf(s,x,y) = f * w; (x, y) and W2f(s,x, y) = f * 91/52 (x, y) , the two components of the
wavelet transform are the coordinates of the gradient vector of f (x, y) smoothed

by63 (x, y) at scales.

 

1 2 2 2
Mf(s,x,y) = \[IW f(s,x,y)| + ‘W f(s,x,y)| 3-5

Mf (s,x, y) is called the modulus of the wavelet transform at the scale s.

29

l
Af(s, x, y) = argtan(-W2f—(m—y—)—) 3-6
W f (sac. y)
Af(s,x, y) is the angle between the gradient vector and the horizontal. At each scale,

the edge points are the points (x, y) where the modulus image Mf (s,x, y) is local

maximum along the gradient direction given by Af(s, x, y) .

This edge detection method is used on an AFM image of cell grown on nanofiber
tissue scaffolding. The soothing fimction to construct wavelets is chosen to be Gaussian

function because it has good property of decreasing noise.

     

(a)

Figure 3—4 Edge detection based on wavelets (a) original image, (b) edge image when scale = 2, (c)
edge image when scale = 10

From the edge images, it is clear that when the scale parameter is chosen
appropriately, we can detect the edges of the target object, which is the cell in this figure.

The points on the edge image are still broken. We can use dilation operation to
connect those edge points. But there are still some noise points on the edge image. Since
these noise points have relatively small area compared with the edge points, a filter is
designed to count the area of the connected points and then threshold in order to remove

those noise points. Finally, the bounding box of the edge points is the boundary of the cell.

30

Therefore, the cell is successfirlly segmented as shown in Figure3-5.

   

(b)

F igure3-5 Segmentation (a) the edge image after dilation, (b) the edge image after threshold, (c) the
segmented object

3.4.2 Feature detection

The choice of discriminating properties as features is a key issue to any classifier being
successful. Most biological objects have their own shape. We define the shape feature using
moments-based methods and give an application of using shape feature to distinguish cells
from fibers. Another feature we found important is the edge feature. Many nanobiological
objects have inside structure, like nucleus of cells. Therefore, those objects will have both
inside edges and outside edges. Therefore, edge enhancement method is used and then edge

information is extracted as a feature to represent different inside structures.
3.4.2.1 Shape feature

Moment-based methods may be used to determine a feature representing[44]. A Shape

Feature was defined as the ratio of second moments along the major axis (M min) and

minor axis (M max ):

51% 3-7

min

ratio =

31

The Shape Feature defined in Equation above has the properties of translation, scale
and rotation invariance.

The nuclei of the two kinds of cells have different shapes. The differences in the shape
can be quantified by calculating the second moment along the major and minor axes of the
image which are orthogonal to each other. The major axis is found using Equation below.

2M

(9 = arctan( xy 2 2 ) 3-8
Mn—MW+fiMH-MW)+QMW)

 

 

Where

M... = ZZoc — a’f(x.y) 3—9
x y

Myy =ZZ(y-J7)2f(x,y) 3-10
x y

M.y = ZZe - no - f)f(x,y) 3-11

x y
The second moment along the major axis can be calculated using the following

equation.

Mm, =ZZ[(x—i)sin6—(y—y)cosa]2 3-12
x y

The second moment along the minor axis can be calculated using the orthogonality

property by the following equation.

Mmax = Zka — r)sin(0 + 90°) — (y — y)cos(l9 + 90°)]2 3-13
x y

3.4.2.2 Boundary enhancement & Edge feature

32

A continuous wavelet transform is a scale based two-dimensional transformation that

allows multiscale/multiresolution analysis of images. The transformed data set

W(0', 1'1, 2'2) is given by:

 

l .. x—r —r
W<o-,r1.r2)=— ]]f(x,y)v1 ( ‘3’ 2)dxdy 3—14
0' 0' 0'

Where 0 is the scale parameter, 11,12 are the translation parameters, * denotes the
complex conjugation, and w is the generating wavelet firnction. The scale parameter 0
plays a crucial role. Wavelet transforms process data at different scales or resolutions. At
each scale it computes the similarity of the image data with the wavelet kernel at the

selected scale 0. At large scales, the transform coefficients W(0', r] , 1'2) are dominant for

data from objects in the image which correspond with the larger scale wavelet kernel.
Similarly, for small scales, small features are highlighted in the W(o', 11,2'2) transform
domain. By choosing the appropriate scale, we can get the relevant details from the
images that are necessary to perform a particular recognition task across multiple scales.
The multiscale analysis procedure begins by defining a family of functions of varying

scales, expressed as

 

i/7(0',x, y) = I ) 3-15

 

 

 

where (p is a function used for systematic generation of images of slowly varying scale 0.
A Gaussian is a commonly used function whose scale is determined by the variance 0. A
multiscale edge detector can be derived from the spatial derivatives of (p.

Edges associated with abrupt transitions in the image were detected using a

33

multiscale Gaussian differential operator defined as below.

3-16

L de, ¢ de, -,
MPH—$7242

dy
We use a differential Gaussian firnction as the mother wavelet because this choice
enhances edge information.
The scaled decomposition of the image can be performed by transforming the image
data Z(x, y) with the differential Gaussian wavelet parameterized by scale 6. For a fixed

value of scale, we can calculate the 2D CWT in frequency domain:

W(0', w1,w2 ) =O'Z(w1,w2 )‘P*(ow1,ow2 ). 3-17

The resolution of the multiscale recognition technique corresponds to the resolution
of the features within the image. This is nanometer-scale for AFM images and
angstrom-scale for STM images.

Examples of AFM images and their corresponding CWTs are shown in Figure 3-6(a)
and Figure 3-6 (b). For each image, the mass center is calculated, which is the point where
the pixel intensities are balanced for any line through that point. By plotting the pixel
intensities along horizontal and vertical lines through the mass centers of the CWT
transformed images as shown in Figure 3-6 (c), it is clear that leukocytes and erythrocytes
have a different number of peaks. The peaks correspond to edge pixels and hence the
presence of a nucleus inside the cell is represented by the number of zero-crossings, which
is the number of times that the pixel intensities cross the zero level. We define the zero
level as the average value of the maximum and minimum pixel intensities along each line.

Two orthogonal lines are chosen for this analysis for robustness. We choose the

number of zero-crossings as the feature which represents discriminating information for

34

erythrocytes and leukocytes, and will refer to the feature thus defined as the Edge Feature.
Both leukocytes and erythrocytes have globular shapes. A distinguishing characteristic
between them is that erythrocytes are anucleate. Continuous Wavelet Transform (CWT)

method is used to successfully distinguish erythrocytes from leukocytes [45].

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 3-6 The definition of the Edge Feature is based on (a) AFM images; (b) Continuous wavelet
transformed (CWT) Images. (c) Pixel intensities on horizontal and vertical lines clearly show the
Edge Feature for each cell type. Reproduced from Figure 4, Reference: Q. Chen, Y. Fan, L. Udpa, and
V.M. Ayres, Vol. 2, Issue 2, pp. 181-189, Int. J. Nanomedicine (2007)

3.5SPRM Operation Modes

The SPRM system has multi—function, off-line and on-line operation modes. Some
applications, such as classification, need the whole information of objects, which can
only be done after the scan of the whole region of interest is finished. Other applications

need real—time capability. The SPRM system can be operated in either off-line mode or

35

on-line mode. In the off-line mode, powerful recognition and analysis ability of SPRM
enable the user to get more information of samples under observation conveniently and
efficiently. In the on-line mode, the SPRM system makes auto-focus/tracking on the
region of interest in real-time possible. This ability is the most important feature of the
SPRM system. It saves the operation time dramatically in many applications and
segments the meaningfirl information from the background automatically. Users can also
combine these two modes to be even more powerful. For example, after an object is
scanned in on-line mode, its property can be analyzed under off-line mode. Since on-line
auto-focus scan has already segments the meaningfirl data fi'om the background, it makes
any subsequent off-line analysis of the object more convenient, efficient and accurate.
The next several chapters will present examples that illustrate these points.

How to choose the operation mode and how to combine them together depends on
the applications. In the following four chapters, four important topics in biological and
biomedical fields are discussed separately. However, the SPRM system is used in all of
them to solve the application problems. Detail of techniques used in the SPRM system

and realizations of the SPRM system are also covered in these four different applications.

3.6Novelty 8. Importance of SPRM versus State of the Art AFM

3.6.1 Novelty and importance of SPRM versus standard SPM

Compared with standard SPM techniques, SPRM system improves the speed of SPM,
which is one of the most commonly observed problems of SPM techniques. In SPRM
system, only the region of interest is scanned, therefore, the overall time taken for
imaging can be decreased. Therefore, SPRM system has the advantage of saving

operation time.

36

Speed improvement for SPM is a state of the art research issue. Various groups have
recently reported successes in the overall improvement of conventional SPM speed
through improvements in the individual sub system components [46]. SPRM can take
advantage of the new improvement in overall system speed while contributing new
capabilities.

The SPRM system provides more powerful feature/property analysis abilities
compared with SPM. Auto-tracking specific regions of interest can provide more
meaningful information of samples automatically, efficiently and accurately. The power
of SPRM allows us to investigate situations in which standard SPM is insufficient to deal
with. Two examples are given in this thesis. One is the properties investigations of
individual nanofibers within a tissue scaffold used for regenerative medicine. This
example is presented in detail in this thesis. Another example is discussed in Chapter 10:
Future Work. This is an ongoing SPRM investigation of nanowire-based
nanoelectronics in which SPRM is used to access information in a previously unavailable
part of the circuit.

Another big advantage of SPRM system over SPM is that multiple data on same
sample area can be obtained in SPRM system by its recognition and auto-tracking
capability. It gives SPRM the potential for data fusion, which combines different types of
information into a composite picture more truly representative object under investigation.

The design and development of SPRM system includes many works from hardware
to software, from algorithms to data management, from tip control to result display. Basic
functionalities of SPRM are implemented at first, which is the kernel part of the system.

When SPRM is applied to investigate real nanobiological phenomenon, additional

37

functionalities specific to applications are added and integrated into SPRM as optional

functionalities.

3.6.2 Novelty and importance of SPRM versus optical

techniques

SPRM is an improved technique based on Scanning Probe Microscopy. Therefore, it
inherits all the advantages of SPM but improves some of the limitations of SPM.

Both SPM and advanced optical techniques, such as confocal microscopy and
two-photon microscopy can image living specimens and provide 3D information of
samples. But when SPM technique is compared with these advanced optical techniques, it
has the advantages of direct interacting with specimens rather than looking at them. The
direct interaction with sample surface enables SPM to serve as a microindenter that
probes soft samples to reveal information about the mechanical properties of samples. In
SPRM system, this advantage can be extended with flexibility to probe any region of
interest on sample surface and get mechanical properties of desired region. The
mechanical properties of living specimens are important for many biological
investigations. Thus, SPRM technique enhances the power of SPM for its application in
biological fields.

The disadvantage of SPM compared with optical techniques is that SPM is not
looking inside. This problem can be solved by combining the two techniques together.
Combination the two techniques can also bridge the gap of scales between optical
techniques and scanning probe microscopy techniques. Information gotten from optical
techniques can also be used to guide SPM investigation. Our group has the intention to be

able to integrate optical technique into SPRM system in the future for a powerful

38

. u.i'~—1

III-LR: r._ noun.
-.

combined instrument.

3.6.3 Novelty and importance of SPRM versus HRTEM

SPM and HRTEM techniques provide comparable image resolution. But SPM has
several advantages over HRTEM. Unlike the electron microscope which provides a
two-dimensional projection or a two-dimensional image of a sample, the SPM provides a
true three-dimensional surface profile.

Additionally, samples viewed by SPM do not require any special treatments that
would irreversibly change or damage the sample. While an electron microscope needs an
expensive vacuum environment for proper operation, which usually killed living samples.
Most SPM modes can work perfectly well in ambient air or even a liquid environment.
This makes SPM be able to study living biological samples. SPM studies can also be
done under physiological conditions allowing for the imaging of dynamic biological
processes taking place in real time.

In SPRM system, the same living biological object can be imaged repeatedly based
on recognition and auto-focus ability of SPRM. It makes possible to image the progress
of the same biological object under living condition as time elapse. This is very useful for

many biological experiments.

39

Part II SPRM

Applications in Nanobiology

4O

Part II SPRM APPLICATIONS IN NANOBIOLOGY

SPRM was developed with powerful recognition capability as explained in the first
part of this dissertation. Its successful application in analyzing nanoscale 3D cues for
spinal cord repair research is addressed in this part. The motivation and background of
SPRM in spinal cord repair are introduced first in Chapter 4. Chapter 5 discusses the
auto-tracking ability of SPRM which enables the system to scan only on individual
nanofibers within the tissue scaffolds used for regenerative medicine, thereby, collecting
the first accurate nanofiber properties analyses. In Chapter 6, the topographic, mechanical
and chemical properties of nanofibers which influence neural cell attachment are
addressed and analysis of these properties are done by using SPRM. Cell responses to
different substrates are explored and compared in Chapter 7. Chapter 8 discussed the
correlation of electronspinning parameters with nanofiber properties analysis.

The candidate developed the first adaptation of SPRM for investigation of a
significant nanobiomedical problem: analyzing nanoscale 3D cues for spinal cord repair
research. The candidate also made significant contributions to the development of the

auto-tracking implementation, and to material properties analyses.

41

CHAPTER 4

4 MOTIVATION & BACKGROUND OF SPRM IN TISSUE
ENGINEERING: SPINAL CORD REPAIR

4. 1 Introduction of Tissue Engineering: Spinal Cord Repair

The use of scaffolds composed of biocompatible materials as vehicles to deliver or
promote growth/differentiation of cells within damaged tissues of the body is an exciting
new tool for the field of regenerative medicine. Tissue scaffolds composed of artificial
nanofibers that mimic natural collagen nanofibers can enable the entrained re-growth of
cells into damaged areas and this is of great medical importance. Tissue scaffolds
fabricated from artificial nanofibers have recently been successfirlly used in kidney
repair[47], knee joint repair[48], and reproductive tract repair with fully restored natural
function [47-49].

Electrospun nanofibers in tissue scaffolds generated tremendous interest due to their
potential as scaffolds for regenerating tissue. The tissue scaffolds explored in this thesis
are: slowly degradable tissue scaffolds fabricated from electrospun [50, 51] polyamide
nanofibers by the Donaldson Company. The nanofibers were electrospun using an
adapted electrospinning probe procedure described in Reference [52].

The design of prosthetic implantable scaffolds to promote tissue repair in vivo will
require the accurate incorporation of the topographic, mechanical and chemical signaling
properties that best mimic endogenous support tissue. Sophisticated tissue scaffolds with
properties designed for the re-growth of specific cells types, while maintaining maximum
biocompatibility, would greatly increase the quality and number of applications. To

optimize the performance of these scaffolds for each type of tissue or application, new

42

analytical methods must be developed that can examine the interaction of these cells with
the surfaces of these scaffolds in vitro under a variety of biochemical and physical
conditions at a resolution in the nanometer range.

The support surface for cell organization in natural tissues is termed the extracellular
matrix/basement membrane (ECM/BM). This is especially necessary in the spinal cord
repair. Repair of damage to the spinal cord and optimal recovery of motor and sensory
functions can only occur when the specific physical and chemical cues are engineered into
an implantable system to provide proper guidance for oriented axonal regeneration that
bridges the spinal cord defective area.

During development of, and also following injury to, the nervous system, the growth
and regeneration of axons is strongly influenced by the extracellular matrix (ECM) [53]
and Schwann cell-derived basement membrane (BM) [54]. (The basement membrane is a
structurally compact form of the extracellular matrix). Therefore, artificial scaffolds that
are extracellular matrix/base membrane - mimetic might provide ideal substrates for
neuronal growth, particularly for in vivo applications that require regenerative scaffolds to
be compatible with the original tissues.

It is necessary to examine the natural in vivo environment in order to reproduce it.
Almost all cells are attachment-dependent, which means that they much first successfully
attach to substrate (cell adhesion) in order to begin cell growth (spreading) followed by
cell division (mitosis leading to proliferation) processes. (For comparison, a red blood
cell is one of the few non-attachment dependent cells, as its function requires that it float
freely within its environment). The neural cells: astrocytes and neurons investigated by

SPRM in this thesis are attachment-dependent cells. Astrocytes and neurons are a paired

43

system in which neurons attach to astrocytes which provide them with nutrients,
waste-removal and mechanical support, while astrocytes attach to a specialized
extracellular matrix (ECM)/basement membrane (BM), which provides them with
filtration and mechanical support. Astrocytes obtain nutrients and pass waste molecules
through exchange with capillaries which are encased by the ECM/BM filtration system
called the basal lamina as shown in Figure 4-1. Regulatory proteins embedded within the
ECM/BM assist these processes. Therefore, an artificial tissue scaffold must reproduce
the topographical, mechanical and surface chemistry cues of the natural basal lamina

environment.

Neuron
\

 

Figure 4-1 Cellular constituents of the blood-brain barrier. Images from reference [55]
4.2Environmental Nanoscale Cues for Cell Growth
To tap the potential of artificial slowly degradable nanofibers as scaffolds for many
different cell types, it is necessary to quantify the properties that provide all necessary

environmental cues for a truly biomimetic situation. A literature search was performed to

44

indentify specific properties that have been shown to influence cell attachment and
growth. Recent research has identified individual external triggers in the cell environment
that stimulate dynamic cell responses include surface roughness [56], elasticity [57] and
surface chemistry[58]. An additional trigger, 3D nanoscale topography has recently been
shown to influence cell attachment and growth. A brief review of examples of each is
provided.
(1) Surface roughness

Investigations of human bone marrow cells examined the effect of specifically
graded surface roughness of hydroxyapatite (HA) substrate on cellular adhesion and
proliferation as shown in Figure 4-2 [59]. For bone marrow cells, the results indicated

that high surface roughness best promoted cell adhesion and proliferation.

 

 

 

 

 

 

5.9

o l I: ,.

E

.C t. .. _ .. --....“ -.-...

.‘sfl I:

3 H

E .-

e 'I

Eel 1.-
4h 18h 4h 18h 4h 18h 4h 18

 

 

Control Fine Medium Coarse

Figure 4-2 SEM micrograph showing the bone marrow cell morphology on substrates with
different surface roughness, Images from reference [59]

45

(2) Elasticity
Research work was performed to investigate the response of NIH 3T3 Fibroblasts
cultured on polyarnine substrates with diflerent rigidities, provided by different amounts
of cross-linking. Experimental results indicated that fibroblasts cells move in favor of
rigid substrate areas as shown in Figure 4-3 [60]. This means elasticity properties of

substrate will influence cell in growth.

 

Figure 4-3 Movements of National Institutes of Health 3T3 cells on substrates with a rigidity gradient
(a) A cell moved from the soft side of the substrate toward the gradient. The cell turned by ~90° and
moved into the stiff side of the substrate. Note the increase in spreading area as the cell passed the
boundary. (b) A cell moved from the stiff side of the substrate toward the gradient. The cell changed
its direction as it entered the gradient and moved along the boundary. Bar, 40 pm. Images from
reference [60]

46

(3) Surface chemistry

It is well known that regulatory proteins are complexed with the extracellular matrix
and that these provide chemical cues for healthy cell function. Therefore, the surface
chemistry property of implantable tissue scaffold nanofibers is important issue in tissue
engineering. Research in our collaborator’s laboratory (S. Meiners) has focused on the
growth-promoting actions of bioactive neuro-regulatory molecule embedded in the
extracellular matrix in vivo. One of these is the fibronectin type HI region of tenascin-C.
The active site of this peptide for neurite outgrowth for cerebellar granule, cerebral
cortical, spinal cord motor, and dorsal root ganglion neurons has been localized to an
amino acid sequence VFDNFVLKIRDTKK [58], called the D5 peptide. A D5 peptide
was covalently complexed to the electrospun nanofibers investigated in this dissertation.
The results provide an example of the importance of chemical cues for neural cells.

Astrocytes cultured on even bare nanofibers adopted a stellate morphology that is
typical of their in vivo counterparts, whereas astrocytes cultured on poly-L-lysine (PLL)
coated plastic coverslips had a flat, cobblestone appearance which is never seen in the
body. The nanofibers were then covalently modified with the D5 peptide and this
additional chemical cue resulted in aligned axonal regeneration by neurons in vivo as

shown in Figure 4-4 [58].

47

A B

- D5 peptide + 05 peptide

 

Figure 4-4 Peptide modification enhances the ability of nanofibers to promote axonal regrowth.
Rostral is to the left. G, host-graft interface. Curves were adjusted in Adobe Photoshop for each
panel to enhance the visibility of the axons against the autofiuorescence of the nanofibers. (A)
Neurofilament M-labeled axons were observed within the implant of unmodified nanofibers. (B)
Axonal growth was more robust within the implant of nanofibers modified with the DS’ peptide.
Bar, 150 pm. (C) Axons were observed on the surface of the nanofibers following folds in the
nanofibrillar fabric. The top of the image is ~ 300 pm ventral to the top of the spinal cord. Bar,
300 pm. Single arrows in (B) and (C) denote axons growing more or less parallel to the axis of the
spinal cord. Double arrows denote axons that deviated from a parallel path. Reference: S. Meiners,
I. Ahmed, A. S. Ponery, N. Amor, S. L. Harris, V. Ayres, Y. Fan, Q. Chen, R. Delgado-Rivera, and A.
N. Babu, Polymer International, vol. 56, pp. 1340-1348, 2007

48

Therefore, we present a quantitative investigation of surface roughness, elasticity,
surface chemistry, as these may be expected to influence the cell regeneration in its
search for a biomimetically attractive environment.

SPRM-enhanced atomic force microscopy (SPRM-AFM) was used by the
candidate to investigate the curvature, surface roughness and elasticity of individual
nanofibers. The detail of neural cell response to a set of well quantified properties was
then investigated. This provided the first data obtained along individual nanofibers using
a technique that maintains uniformity of experimental conditions.

(4) Nanoscale topography.

Recent work strongly suggests that substrate topography influences cell
attachment and that this influence is nanoscale in nature. The recent investigations are
based on newly available nano patterning techniques such as electron beam, photo and
ink jet soft lithography, and oriented grth of nanowires/ nano pillars [61]. Experiments
performed using these as guides indicate a strong cell response to nanoscale topography
for many cell types. Quantification and understanding of the observed cell responses is
now an active research area.

As previously discussed and as shown in Figure 4-5, astrocytes cultured on even
bare nanofibers adopted a stellate morphology (a) that is typical of their in vivo
counterparts, whereas astrocytes cultured on poly-L-lysine (PLL) coated plastic
coverslips had a flat, cobblestone appearance (b) which is never seen in the body. Until
this work, PLL had been required for astrocyte growth in culture. This indicates the
strong positive influence of 3D nanoscale curvature in itself as a topographic trigger for

astrocyte growth.

49

a" .

l
. > .
n . '{I .
'4‘?

50.0 pm

 

Figure 4-5 Astrocytes cultured on even bare nanofibers adopted a stellate morphology (a) that is
typical of their in vivo counterparts, whereas astrocytes cultured on poly-L-Iysine (PLL) coated
plastic coverslips had a flat, cobblestone appearance (b) which is never seen in the body. To date, PLL
has been required for astrocyte growth in culture. Reference: S. Meiners, I. Ahmed, A. S. Ponery, N.
Amor, S. L. Harris, V. Ayres, Y. Fan, Q. Chen, R. Delgado-Rivera, and A. N. Babu, Polymer
International, vol. 56, pp. 1340-1348, 2007

4.3 Ultimate Goal

In spinal cord repair, individual external triggers in the cell environment that
stimulate dynamic cell responses include surface roughness, elasticity and surface
chemistry. Other additional triggers, such as porosity and mesh density, are also shown to
influence cell attachment and growth. The ultimate goal is to find the best tissue scaffold
with most appropriate topographical, mechanical and chemical properties for specific
cells or cell classes. These properties of nanofibers can be analyzed by using the new
Scanning Probe Recognition Microscopy technique. And cell response can also be
investigated by SPRM and other optical techniques. Information can be combined to
provide a composite picture representative of a cell’s perception of its environment. Then,
the electrospinning materials or parameters can be adjusted adaptively until the properties
of the produced nanofibers are found to be best for cell regeneration. The whole
procedure will provide an automatic guide to design tissue scaffold for regenerative
medicine.

50

 

 

K

1

I Tissue Sc‘affold I

 

j-

I Curvature I

 

 

 

 

I Surface rbughness I
IElasticity I IProperties I

I SurfaceChemistry I

 

 

 

 

 

 

 

 

I Mesh Density I J

 

 

 

IBio-data fusion I ] Ilnterpretation I

I Cell growth I
I J I I Outcomes I

 

 

 

 

 

 

I Electrospinning parameters

5

Figure 4-6 Application of SPRM in spinal cord repair

 

51

CHAPTER 5

5 SPRM AUTO-TRACKING IMPLEMENTATION ON
TISSUE SCAFFOLD NANOFIBERS

As explained in previous chapter, electrospun tissue scaffold nanofibers structurally
mimic the extracellular matrix on which cells attach and grow in vivo. This property
gives electrospun nanofibers great potential for use in tissue engineering applications.
However, much fundamental understanding is still needed to design tissue scaffolds with
the most appropriate mechanical, topographical and chemical properties for particular
cells or cell classes. SPM methods are essential to measure the nanoscale triggers that
cells respond to. It is also essential for individual nanofibers within a tissue scaffold to
be scanned for accurate mechanical or topographical properties investigations. SPRM

provides the first ability to do this.
5. 1lmportance of Auto-Tracking of Individual Nanofibers

In a standard SPM of a candidate tissue scaffold, an extra image segmentation
technique would be necessary after getting the image of the whole scan region in order to
get the information along each individual nanofiber. This would be extremely difficult to
realize using off-line segmentation techniques as shown in Figure 5-1 and Figure 5-2. By
contrast, when nanofibers only are automatically tracked during an SPRM scan, the
segmentation is already done in the scan step. The useful information on nanofibers is

extracted from the environment automatically.

52

 

Figure 5-1 Off-line segmentation results at scan size equals 15 microns a) AFM height image; b)
Recovered image by reading exported ASCII file; c) Off-line segmentation results using thresholding
method.

 

(a) ’ (b) ” (c)

Figure 5-2 Off-line segmentation results at scan size equals 3 microns (a) AFM height image; (b)
Recovered image by reading exported ASCH file; (c) Off-line segmentation results by using
thresholding method.

A second important feature of auto-tracking is data collection over a reliable region of
interest. An attempt to measure the elasticity of a tissue scaffold nanofiber using standard
SPM illustrates this point. In tissue scaffold nanofibers, there are three types of regions as
shown in Figure 5-3. Regions as data point (1) are not region of interest; regions as data
point (2) are right on top of a nanofiber; regions as data point (3) are on the side of a
nanofiber. Only regions as data point (2) contain reliable data in these three types. SPRM

has the capability to collect information only along these reliable regions.

53

 

Figure 5-3 Different types of regions on tissue scaffold nanofibers: Data point (1) is out of region of
interest; data point (2) is right on top of a nanofiber; data point (3) is on the side of a nanofiber. Scan
size: 10 microns [62].

A third important feature of auto-tracking each individual nanofiber is that it
maintains the uniformity of experiments. When each nanofiber is automatically scanned
by tracking fi'om beginning to end, the information on nanofiber is collected
independently to other parts of environment which are not scanned. This will provide a
uniformity of scanning probe experimental condition as long as other operation
parameters are same. The experimental uniformity is extremely important when the
research interest is to compare different samples.

The new Scanning Probe Recognition Microscopy is designed to provide
auto-tracking ability. As shown in Figure 5-4 (and also as a movie at
http://wwwegrmsucdu/ebnl, Research, Scanning Probe Recognition Microscopy),
SPRM can scan along individual nanofibers in tissue scafl‘olds by tracking them one by
one, which will automatically provides reliable information of individual nanofiber in

real-time.

54

lllpm

 

Figure 5-4 SPRM auto-tracking capability to scan along individual nanofibers (a) AFM height image;
(b)-(g) A series of images shows the time sequence of scanned results by using SPRM. [62]

5.28PRM Auto-Tracking Capability

The flow chart to realize auto-tracking capability of SPRM for auto-tracking along
nanofibers is shown in Figure 5-5.

A coarse scan and fine scan is used alternatively to scan all the nanofibers in scan
range with high efficiency. The basic idea of tracking nanofiber is to use real-time
information to detect the boundary of a single nanofiber while using coarse scan. The
coarse scan is set to have relatively large step size and a relatively high scan velocity. On
the contrary, fine scan has relatively small step size and relatively low scan velocity.
Once a boundary of nanofiber is detected, it triggers the initiation of a fine scan set to
have a small step size and a relatively low scan velocity. The new auto-tracking scan is
combined with boundary prediction when enough information is accumulated in previous
lines of scanning, usually 10-20 lines. With boundary prediction, SPRM is able to stay
along the nanofiber even though there are other nanofibers crossing it. Adaptive learning

ability is also integrated in our algorithms to improve performance.

55

I

Set coarse scan

wnned object?

Y

Set fine scan and remember
coarse stop point

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I Finescan object I

I

I Go next line I

\ Y
. Find left DOUDW

N
Set coarse scan go coarse stop point

Figure 5-5 Flow chart of auto-tracking scan

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(1) Coarse scan versus fine scan

A coarse scan is used to detect individual nanofiber while fine scan is used to
actually scan along the detected nanofiber. This scheme enables SPRM to scan nanofiber
one by one as shown in Figure 5-6, which will increase overall imaging efficiency

without losing resolution of imaging.

56

951.3nm

 

Figure 5-6 SPRM scan along individual nanofibers in a tissue scaffold. Left image is the standard
AFM image, region in black box are target scan region by SPRM. Right image shows the scan result
of using SPRM, two individual nanofibers are scanned one after the other, other regions not scanned
are padded 0 for display.

(2) Statistics analysis

After boundary of the next line is predicted, the scan range in the next line will be
controlled only from the left predicted boundary to the right predicted boundary, which
will make the scan stays within the targeted region of interest. In order to increase the
accuracy of the system, boundary detection technique is applied after each line of scan.
With both predicted boundary and detected boundary at hand, a criterion is needed to
decide whether the detected boundary matches the predicted boundary. When the
detected boundary matches predicted boundary within a reasonable error range, the
detected boundary is used to predict boundary of next line since it records the newest
information. When the detected boundary is too far away predicted boundary, which

means that there is special situation which makes the detected boundary not the boundary

57

of the target nanofibers, detected boundary will not be used to predict boundary
information in the future because it doesn’t record the correct boundary information of
the target nanofiber. In the tissue scafl'old application, this problem normally happens

when there are several nanofibers crossing over one another as shown in Figure 5-7.

     

(a) (b)

Figure 5-7 Scan plan to stay on individual nanofiber (a) without boundary prediction; (b) with
boundary prediction

 

Hypothesis test in statistics[63] are applied to give a criterion of how far away the
distance between detected boundary and predicted boundary is acceptable. Therefore,
accuracy of boundary detection can be analyzed statistically. When the detected boundary
is not the real boundary, the predicted boundary will be used instead.

(3) Adaptive learning

It is important for SPRM system to have adaptive learning ability in real-time
because there are many types of noise when SPRM is imaging in tissue scaffolds, for
example, sample tilting issue will cause the height information of the sample surface to
have tilting too. With adaptive learning algorithm, important parameters in algorithms

can be adjusted automatically in real-time in order to capture changes of sample surfaces.

58

The newest information is updated into the information database and used to replace the
oldest information in algorithms. Therefore, the SPRM system can still stay within
individual nanofibers in spite of complex situation of sample surface.

The key parameter needed to be adjusted adaptively is the thresholding value to
detect whether scan is on nanofibers or not. The adaptive adjustment considered multiple
factors that define the vertical placement of an individual nanofiber within its 3D mesh.

(4) Scan direction

In tissue scaffolds, nanofibers exist in random directions to produce the mesh shown
in Figure 5-8. When the scan direction is almost same as the orientation of the nanofiber
as shown in Figure 5-8 (a), there will be fewer lines of data. Prediction will have
relatively larger error because of limitation of data. This error will be decreased when we
change scan angle by 90 degree as shown in Figure 5-8 (b). Therefore, SPRM gives the

flexibility of changing scan angle to get best performance.

 

Figure 5-8 Demonstration of Scan angle in imaging Tissue scaffold nanofibers using SPRM (a)
horizontal scan direction; (b)vertical scan direction

59

5.3Examples of SPRM

Samples of tissue scaffolds fabricated from electrospun carbon nanofibers were
obtained from the Donaldson Company. The SPRM is used to selectively scan only along
individual nanofibers.

The standard AFM contact mode was used first to get an image of the tissue scaffolds
sample as shown in Figure 5-9 (a). The SPRM system can be and was run under
simulation mode to confirm that the SPRM can automatically scan only on the nanofibers
one by one as shown in Figure 5-9 (b). Then a real-time scan along individual nanofibers
was performed. The real-time scan result of the SPRM is shown in Figure 5-9 (0). It is
clear that the SPRM system can auto-track on the nanofibers as desired.

The scan direction can also be changed in the SPRM depending on requirements. The
scan result of the same tissue scaffold sample at the same region but with a difi‘erent scan

direction is shown in Figure 5-10.

-----

..ia‘

  

(a) (b) (c)

 

Figure 5-9 (a) Standard AFM image of tissue scaffolds with scan size equals 6 microns; (b) simulation
scan result of SPRM by using horizontal scan direction; (c) real-time scan result of SPRM by using
horizontal scan direction.

60

 

(a) ' (bi ' (c)

Figure 5—10 (a) Standard AFM image of tissue scaffolds with scan size equals 6 microns; (b)
simulation scan result of SPRM by using vetical scan direction; (c) real-time scan result of SPRM by
using vertical scan direction.

From the scan results shown in the Figure 5-9 and Figure 5-10, it is clear that SPRM
scanned most of the nanofibers in the tissue scaffolds one by one. The experimental
results show that operation time can be saved dramatically by applying SPRM to
selective scan only on nanofibers. Only meaningful data is captured in SPRM; the
background regions were not scanned at all. It makes the analysis of properties of
nanofibers much easier after this step of useful information extraction. Another advantage
of using SPRM is that users have flexibility in imaging the samples. SPRM can repeat the
scan of same nanofiber because of its on-line recognition ability.

In conclusion, the auto-tracking ability of SPRM is realized based on algorithms and
techniques in pattern recognition and image processing fields. Adaptive learning and
prediction are implemented to make detection and recognition quicker and more reliable.
Error analysis and hypothesis test techniques in statistics are also used to improve
performance of SPRM. Experimental results show that SPRM can auto-track on
individual nanofibers one by one efliciently and accurately, which makes the property
analysis of nanofibers very convenient. The detailed analysis of topographic and
mechanical properties of nanofibers will be explained in the next chapter.

61

CHAPTER 6

6 NANOFIBER PROPERTIES ANALYSIS BY USING

SPRM

As discussed in Chapter 4, spinal cord repair can only occur when the specific physical
and chemical cues are engineered into an implantable device to provide proper guidance
for the axons across the spinal cord defect. Therefore, it is necessary to be able to analyze
physical and chemical cues, including curvature, surface roughness, elasticity and surface
chemistry, which are all known to affect cell attachment and growth. As explained in
chapter 5, Scanning Probe Recognition Microscopy allows us to adaptively follow along
individual nanofibers within a tissue scaffold. The ability of SPRM to auto-track on
individual nanofibers was also demonstrated. In this chapter, the result of a physical and
mechanical property analysis of electrospun polyamide nanofibers by using SPRM is
presented in detail. Statistically significant data for multiple properties can be collected
and calculated by scanning only over individual nanofibers, and information can be

combined by repetitively scan over the same nanofibers if necessary.

6. 1 Tissue Scaffold Samples and Experimental Parameters

Three samples of tissue scaffolds fabricated from electrospun [50, 51] carbon
nanofibers were obtained from the Donaldson Company. These will be referred to as
samples A, B, and C. The nanofibers were electrospun using an adapted electrospinning
probe procedure described in Reference [52]. No further information about the tissue

scaffolds samples was provided.

62

The new Scanning Probe Recognition Microscopy was performed in atomic force
microscopy contact mode in ambient air. Other instrumental parameters include the use of
a J scanner with a maximum 125x125 square micron x-y scan range and silicon nitride tips
with a nominal 20 nm tip radius of curvature. Transmission electron microscopy (TEM) of
the nanofibers was performed to provide an independent check of nanofiber
diameter/curvature. TEM with selected area electron diffraction (SAED) was performed
using a JEOL lOOCXII TEM. Scanning electron microscopy (SEM) was performed to
ensure that the nanofiber mesh scanned by SPRM was representative of the tissue scaffold
as a whole. SEM was performed using a Hitachi S-4700H field emission SEM.

Atomic force microscope (AFM) images of tissue scaffold samples A, B, and C are
shown in Figure 6-1 (a)-(c). Scanning electron microscope (SEM) images, shown in Figure
6-1 (d)-(f), were also used to ensure that the AFM results were representative over larger
scaffold areas. The general appearance of the nanofibers was similar for samples A, B, and
C. Tissue scaffold samples A, B, and C were then analyzed using Scanning Probe
Recognition Microscopy for the more specific environmental triggers surface roughness

and elasticity.

63

 

Figure 6-1 AFM and SEM images of electrospun carbon nanofiber tissue scaffolds. (a)-(c) AFM
images of Sample A, B and C. The scan area of each image is 5 square microns and the z-height
projection is 1500 nanometers. (d)-(f) SEM images of Sample A, B and C. The scan area of each
image Is 20 square microns. Reproduced from Figure 1, Reference: Y. Fan, Q. Chen, et al. in press,
Vol. 2, Issue 4, Int. J. Nanomedicine (2007)

6.2Nanofiber Property Analysis by Using SPRM
6.2.1 Surface roughness

The surface roughness of the substrate has been shown to influence cell attachment. It
is therefore important to obtain this information accurately and conveniently along
individual tissue scaffold nanofibers.

In the majority of studies, the surface roughness is the Root Mean Square (RMS) of

height values in a region of interest as defined in equation below.

6-1

 

64

For AFM-based measurements, N is the number of total pixilated data points in this

N
region; Z ,- is the height value at each pixel and Z ave = %ZZ,— is the average value of
i=1

all height values in this region.
The AFM nanoscope software provides the ability to calculate surface roughness as
shown in Figure 6-2. The user can draw any rectangular box in the image, and the surface

roughness in the box is calculated based on the above equation.

10.0
7.5

5.0

Image Statistics

 

Img. Rms (Rq) 189.70 nm

 

 

 

 

2.5 Img. Ra 151.94 nm
Box Statistics
Rms (Rq) 8.108 nm
Mean roughness (Ra) 6.202 nm
0
0 2.5 5.0 7.5 10.0 “n”

Figure 6—2 surface roughness calculation of Scanning Probe Microscopy

First, this method will introduce some error when it is used to measure the surface
roughness of nanofibers. Height data, Z, is used to calculate the surface roughness in this
way. This is acceptable for samples whose surfaces are relatively flat because then the
variation in height data will reflect the variation in their surface roughness. This
assumption is no longer appropriate for tissue scaffolds because the nanofibers have a

cylinder shape. Therefore, the variation of height data, Z, includes not only its surface

65

roughness variation, but also the variation caused by the non-planar shape.

Second, the surface roughness information is acquired through manual application
of a rectangular region of interest box[64]. The surface roughness within the box is then
calculated. There are several problems with the conventional approach to surface
roughness investigation when the sample is a tissue scaffold nanofiber. The shape of the
region of interest (ROI) may not be rectangular, necessitating the application of several
small ROI boxes which follow the curvature of the nanofiber. Only a single value is
provided for each time of operation. Therefore, in order to get surface roughness along a
nanofiber, this operation would need to be repeated for many times, which is
time-consuming and inefficient.

In Scanning Probe Recognition Microscopy, the surface roughness property
measurement ability was designed to be able to do accurate and efficient measurement
along individual nanofibers, i.e. along non-planar substrates. These two discussed problem
will be solved.

In SPRM, distance data, D, instead of height data, Z, was used to get the real
roughness of the sample surface. The Kasa circle fit method [65] was implemented to get

the center (X center, Zcemer) of the most fitted circle as shown in Figure 6-3. Then the

distance D,- between each point on the surface and the center was evaluated as equation

below and used to calculate the surface.

 

2 2
Di = l/(Zi —Zcenter) +(Xr' TXcenter) 6‘2

66

 

 

 

. {I k
'- 49"? “‘9 .4
”HM \h‘ -~ a
..\c ’ \-‘
I: .“a Z . ‘W‘ -4
f" I \.
(I
i- ll, " g. .1
s
R
..(f i \\ "
r "r.
" 'r ‘4

 

 

 

(X0811 161', ZCCl‘l ter)

Figure 6-3 Circle fit based on Kisa method. Reproduced from Figure 5, Reference: Y. Fan, Q. Chen,
et al. in press, Vol. 2, Issue 4, Int. J. Nanomedicine (2007)

Only the center part of the height data was used to find the best fit circle. The
boundary regions are unreliable due to tip-shape dilation effects. Dilation, or broadening of
the image, is a result of the side of the tip coming into contact with the curved side of the
nanofiber [66, 67]. When an SPRM scan is tracking an individual nanofiber, both the edge
and center data is acquired. However, only the center part of the nanofiber provides
reliable data because of the tip convolution effect [68, 69]. The ideal AFM would have an
infinitely sharp tip to reach as much of the surface as possible and an infinitely sharp
impulse response in its feedback system to instantly adjust the height of the tip as it is
scanned over the surface. In reality, the tip has a pyramidal or conical shape with some
finite end radius so it is durable enough to withstand the surface interaction forces. The
effects of the tip shape cannot be avoided and these result in characteristic tip-dilation
artifacts, as shown in Figure 6-4. We restricted the properties evaluations to regions of

reliable data through the use of an erosion operation. This is an important consideration

67

when analyzing tissue scaffold nanofiber geometries using AFM-based methods (instead of

planar substrates).

 

(a)

 

 

200

Height (nm)
0
\\

 

. '
O " x
r— ...ogood‘ ‘~\
.,

-1 5” L l 1 J 1 1
o 40 80
Distance along profile (nm)

 

 

 

Figure 6-4 (a) Characteristic tip-dilation artifacts (b) The cross section of a real AFM nano fiber
image and the best fit circle. Reproduced from Figure 6, Reference: Y. Fan, Q. Chen, et al. in press,
Vol. 2, Issue 4, Int. J. Nanomedicine (2007)

In the SPRM system, a recognition-based scan plan can be generated to
automatically control tip motion along an individual nanofiber [70]. Adaptive scanning
enables SPRM to follow along an individual nanofiber even when it crosses another.
Therefore, the whole nanofiber, or whole nanofiber mesh becomes the region of interest.

For the second problem, the surface roughness was calculated on each pixel based
on a local neighborhood region by using data obtained by SPRM along individual
nanofibers. The shape and size of the local neighborhood region can be adjusted by the
user, which makes the system adaptable to different samples. We chose a rectangular box
around each pixel as the local neighborhood region, with a box size close to the nanofiber

diameter. Multiple sets of overlapping surface roughness information were generated,

68

with the provision that any box that extended outside the nanofiber boundaries was
automatically truncated. A roughness map along individual nanofibers was then generated
efficiently.

Figure 6-5 (b) shows a typical surface roughness map along an individual nanofiber
in tissue scaffold. This is the first time that statistically meaningful information has been
extracted along individual nanofibers using an automatic procedure that maintains

uniformity of experimental conditions.

 

Figure 6-5 SPRM analysis of surface roughness (2) AFM image; (b) surface roughness map of an
individual nanofiber

The histograms of the surface roughness for samples A, B and C are shown in
Figure 6-6. From the histograms, the mode, the mean value, the range, and the variance of
the surface roughness were investigated for several nanofibers for each tissue scaffold
sample. The distribution was also approximately analyzed. Table 6-1 shows the results

from the statistical analysis.

69

    

moor 2500{—~~ , 7-
§ (C)
c 800* 2000»
I!
2 600 1500»
.2
o
9 400 1000»
o
.o
E 200- 500'
3 .
Z 075." 5.9 0- a a .. 3;: , ,7, .
0 5 10 5 20 0 5 10 20 O 5 10 15 20
Surface roughness (nm) Surface roughness (nm) Surface roughness (nm)

Figure 6-6 Histograms of surface roughness (a) Sample A; (b) Sample B; (c) Sample C. Reproduced
from Figure 3, Reference: Y. Fan, Q. Chen, et al. in press, Vol. 2, Issue 4, Int. J. Nanomedicine (2007)

Table 6—1 Surface Roughness Statistical Analysis

 

The mode values were close for all three samples. The mean values might indicate
a progression between the samples: Sample C (least) < Sample A < Sample B (greatest).
However, the surface roughness distributions of samples A and B were wide spread with
prominent tails and irregular values while the surface roughness distribution for samples C
was narrow, peaked and smoothly connected. Analysis of the surface roughness
distribution therefore indicated that sample C was different from samples A and B. The
corresponding range and variance values further indicated that samples A and B were
similar to each other but difierent from Sample C.

The histogram analysis shows the possibilities for misinterpretation of data using
conventional AFM methods. The surface roughness mode values for the three samples

were all very close. An individually applied ROI box would be most likely to return the

70

mode value. However, the variances and distributions differed substantially between the
samples. This was the true difference between the surface roughness of the samples, and it

specifically indicated that sample C was different from samples A and B.

6.2.2 Elasticity

Getting one (or a series of) force curves at the desired position (or region) is very
important for measuring mechanical properties of samples. Atomic force microscopy can
be used to measure elastic properties by collecting force curves over points on the surface
of the sample. A single force curve records the force felt by the tip as it approaches and is
then drawn away fiom the sample [71, 72]. It is more useful to collect arrays of force
curves across the sample surface at regular intervals and this is known as force volume
imaging [73]. A force volume data set can be used to generate a 3-D map of interaction
forces between the sample and tip as shown in Figure 6-7.

The problem of using force volume imaging mode to get force curves to analyze
elasticity property lies in the following two aspects:

First, force volume imaging takes about hours to get force curves at each pixel as
shown in Figure 6-7. But most of the force curves captured on the regions outside
nanofibers are not necessary when the only interest is the elasticity of nanofibers. Time is
wasted in getting unnecessary information. Therefore, applying SPRM to accurately

recognize region of interest and then get force curves only along nanofibers is vital.

71

FV Image

 

 

 

 

 

 

0 10000.0 nm
Data type Height
2 range 102.27 nm
2 scan 10.00 pm
Z position 15.24nm/div
Tip
' Deflection
' 10.23
0 10000.0 nm . nm/dlv
Data Type Height j .
2 range 3.80 um

 

 

 

.....
......

”Molll'llllylla'soum ForcePlot
FVO 102.27nm

Figure 6-7 Force volume imaging of Atomic Force Microscopy

 

The second problem is that force volume imaging technique probes sample surface at
given interval. The probe position might be anywhere, on top of nanofibers or at the side of
nanofibers. But in order to using Hertz model [74] to calculate elasticity property, the force
curves are required to be on top of the nanofibers as shown in Figure 6-8 (a). When tip is
probing side of nanofibers as shown in Figure 6-8 (b), the force will not be normal to
sample surface. This won’t satisfy one of the assumptions of using Hertz model, which
makes Hertz model not appropriate to be applied to calculate elasticity under this situation.
Therefore, it is important to use SPRM to automatically track the top position of individual
nanofrbers and guarantee that force is normal to the sample surface.

72

   

(a) (b)

Figure 6—8 Tip probing position (a) tip probes top of nanofiber, force is normal to sample surface;
(b) tip probes side of nanofiber, force is not normal to sample surface

The new SPRM system solved these problems by integrating recognition ability into
the AFM system. Same as force volume imaging, tip will probe sample surface to generate
force curve, but only meaningful and useful probes are done by recognition the top of
individual nanofibers. Therefore, operation time will be decreased dramatically because
only force curves on the top of nanofibers are collected. And forces are normal to the

sample surface in these force curves is guaranteed at the same time.

 
  

AFM tip t Distance

"'22- 0-2-11
4

    
   

SUBSTRATE

Figure 6-9 Force distance records the indentation of tip into soft sample

The force curves record the cantilever deflection (d) versus the height of the sample

73

surface (Z) as shown in Figure 6-10 (a). The elasticity analysis is based on the approaching
part of the force curves. Rather than using the sample surface position (Z), it is more usefiil
to obtain an absolute distance (D) that is relative to the separation between the tip and the
sample surface which records the indentation of the tip into soft sample as shown in Figure
6-9. A force-distance curve can be obtained from the force curve which is defined by the
cantilever deflection versus the absolution distance (D) between the tip and sample surface

(D = Z— d) as shown in Figure 6-10 (b).

nn

 

 

 

 

 

 

 

r r v 00 1' 7 v x r
1 ............ Approach Approach
l __ Retract *\
A ' - \
E A =
5 i E 250- \\
5400 l . g \\
‘6 .1 a3 \
CD ~ \
r: l. 0 \
0 ' g \
9 l ‘ o \
s l a \
.2 l “NW” E 100 \
c -600' \ ' *3 \
m to \
o 1 ,1 o \.‘
a r - «a.
[J . x”.fi ,
.R_n r r l r -59 .. r .. ,.....L. W 1 . A. r r L A .
‘ 0 100 200 300 400 500 603 -300 -100 100 300 500
Z (nm) Separation Distance (nm)
(a) (b)

Figure 6-10 (a) force curve, both approach part and retract part are displayed; (b) force distance
curve, only approach part curve is displayed, which is used to calculate elasticity property

The Force Integration to Equal Limits (FEEL) [75] mapping method is used to produce
a robust measurement of relative elasticity. This method has the advantage of being
independent of the tip-sample contact point, and of not requiring calibration of the AFM
cantilever’s spring force constant. Using the Hertz model[74], which has been widely
applied to AFM data, if the tip of an AFM is approximated by a sphere, then the force on

the cantilever (F) is decided by indentation (8), elastic modulus (E), Poisson ration (v) and

74

radius of the probe sphere(R).

F: 4BR 53/2 6-3
3(1—v)
Let k=1—"—", 17:53:53” 6-4
7rE 37r-k

To compare the elastic properties at two different positions, a pair of force-distance
curves is collected at positions P1 and P2 using the relative trigger mode in which the
sample is probed until a preset same cantilever deflection (relative to the zero deflection)
position is reached [32]. Therefore, force Fl equals force F2 at these two points. It results

in the following equation.

4x/E 3/2 4R 3/2
—-—5 =——§2

37r-k1 l 37r-k2

which reduces to

3/2
E. = fl. 6'6
k2 52

The work done by the cantilever at each position is the area under a force-distance

curve, and is given by

W1 = El F1d6=££615/2 6-7
15 fl'kl
and
w2 = [:52 F2d6=-8—£—625/2 6-8
15 fl‘kz

Therefore, the relationship between the elasticity and the force-distance curve at two

different data positions is

75

5/2
fl=k_2 fl 6-9
W2 k1 52

From equation 6-6, the following equation is derived.

2/3
K: k_1 6-10
W2 k2

where k is inversely proportional to E, the elastic constant which represents the local
elasticity of the sample. The area under the force distance curve can be calculated and used
to represent the inverse relative elasticity of the tissue scaffolding. Similar to reference
[75], this area value was used to compare different samples.

In order to compare one tissue scaffold with another, it is required to maintain the
maximum relative deflection to be a constant, which means the force (F) will be a
constant. Constant maximum relative deflection should be provided.

Each force curve in all three samples was triggered to exhibit the same maximum
relative deflection of 45m. The histograms of elasticity for samples A, B and C are shown
in Figure 6-11. As before, the mode, the mean value, the range, and the variance were
investigated for each tissue scaffold sample, and the distribution was approximately

analyzed. The results are shown in Table 6-2.

 

    

           

 

250r *r —r 300_ .v. if 600r~—— ~~
i
. 250 500+
g 200i
7;,- r 200 400»
c rsor
‘8 150 300*
u, 100‘-
E 100» 200
g 50* : 50. 100" [I
r 1 1| :
Z 0..- , ,i, _ 3152. : ilal .. -_,,, 0,,_v DH i ti 5 a. of ill .2251: it EM,
1400 1600 1800 2000 1400 1600 1800 2000 1400 1600 1800 2000
Relative elasticity (nN-nm) Relative elasticity (nN-nm) Relative elasticity (nN-nm)

Figure 6-11 Elasticity histograms for (a) Sample A; (b) Sample B; (c) Sample C. Reproduced from
Figure 4, Reference: Y. Fan, Q. Chen, et al. in press, Vol. 2, Issue 4, Int. J. Nanomedicine (2007)

76

The distribution of sample C was almost a normal distribution, while samples A and
sample B had a symmetric but irregular distribution at both sides.

Two points were identified from the histogram analysis. The first was that samples A
and B had similar distributions, but sample C had a different distribution even though the
mode and mean of all three were close. The second was that sample B had the widest range,

the largest variance and the most irregular distribution.

Table 6-2 Elasticity/Area Map Statistical Analysis

 

 

 

 

Sample Data Points Mode Mean Range Variance2
(nN- nm) (“N‘ nm) (nN' nm) ((nN- nm) )

A 1739 1780.0 1777.8 1199533: 4463.3

B 2757 1770.0 1747.5 11596739: 5760.5

C 6325 1730.0 1770.7 116535; 4144.3

 

 

 

 

 

 

 

 

6.3 Comparison of Nanofiber Diameter by Transmission Electron

Microscopy

TEM images of several nanofibers were taken to provide independent verification
of nanofiber diameter, and from this, an estimate of the severity of tip-dilation artifacts
present in the AFM images. Representative TEM images are shown in Figure 6-12. The
nanofiber diameter measurements within the TEM images were taken using Scion Image
software (http://www.scioncorp.com/). Scion Image was used to obtain a conversion
factor relating pixels to nanometers using the scale bar in the lower right comer of the
TEM images. Scion Image uses this conversion factor to measure lines drawn in the

program’s image editor in units of physical length. For each TEM image, three

measurements of the nanofiber diameter were taken at distinct points along the length of

77

the wire. Care was taken to identify fluctuations in diameter that would be reflected in the
mean and variance of the measurements.

Nanofiber diameter measurements within the AFM images were taken using the
Nanoscope software. The software can be used to take a series of cross sections of the
image orthogonal to the longitudinal axis of the nanofiber being measured. These cross
sections are returned to the user as l-dimensional signals representing height versus
distance parallel to the cross section. The software can then be used to position markers at
various points along the signal, allowing the user to measure the distance between them.
As in the TEM images, three measurements were taken at a variety of points along the
length of the chief nanofiber in each image. Variations in diameter were not as apparent in
the AFM images, so care was taken to space the measurements evenly along the length of
the nanofiber.

The discrepancy between the diameters measured using the TEM images and the
diameters measured using the AFM images are given in table below. Based on these
measurements, a dilation factor of over 100% was estimated for AF M measurements of all
three samples. These results are demonstrative of the severity of tip dilation effects in AFM
techniques. As previously discussed, an erosion operation was applied to both the surface

roughness and elasticity data to restrict the properties analysis to the most reliable data.

78

Table 6-3 Nanofiber Diameter Measurements Using TEM and AFM Images

 

 

 

Section 1

Image (nm) Section 2 (nm) Section 3 (nm) Mean (nm) Standard Deviation (nm)
TEM al 120 100 130 116.6667 12.47219
TEM 02 110 120 150 126.6667 16.99673
TEM a3 109.86 99.78 105.74 105.1267 4.137933
TEM b] 117.67 117.02 106.88 113.8567 4.94038
T EM 64 111.3 77.52 97.52 95.44667 13.86834
TEM b9 65.91 111.96 98.03 91.96667 19.28252
TEM b12 110.01 94.34 91.89 98.74667 8.026939
T EM c3 69.14 82.53 84.76 78.81 6.898063
TEM c4 390 380 350 373.3333 16.99673
TEM c5 390 380 350 373.3333 16.99673
TEMc7 312.96 313.21 309.12 311.7633 1.871903
TEM c8 351.19 348.64 329.44 343.09 9.707986
TEM c9 345.83 325.06 327.96 332.95 9.184164
TEM c] I 345.71 346.33 345.36 345.8 0.401082
TEM c14 415.95 388.83 345.04 383.2733 29.21432
TEMc/6 377.23 369.2 337.71 361.38 17.05524
AFM A 647 589.91 494.77 577.2267 62.79142
AFMB 781.25 722.66 742.19 748.7 24.35819
AFM C 761.72 800.78 644.53 735.6767 66.39381
Average Dilation of A 509.5275%

Average Dilation of B 739.6577%

Average Dilation of C 220.8495%

 

79

 

6.4 Comparison of Nanofiber Properties by Transmission

Electron Microscopy

Typical TEM images of samples A, B, and C with corresponding selective area
electron diffraction (SAED) images are shown in Figure 6-12. Samples A and B both
showed a dark contrast outer layer surrounding a light contrast inner core, possibly hollow,
while sample C was solid throughout. Diffraction images for samples A and B showed
prominent rings typical of disordered structures while diffraction images for sample C
showed spots typical of an ordered (crystalline) structure. These results indicated that the
atomic arrangement of sample C nanofibers was substantially different from sample A and
B nanofibers.

Close-up TEM images of samples B and C are shown in Figure 6-13. These images
are consistent with the normal and narrow surface roughness distribution of sample C as
well as the wide variance and irregular surface roughness distribution of sample B.
Therefore the TEM results were consistent with the results obtained by SPRM which
consistently indicated 1) that sample C was different from both samples A and B, and 2)
that samples A and B, while similar, had differences with sample B having the more

extreme values.

80

(C)

 
 
 

  

. V‘figfir
.r.

.4
Fl‘.”

1N;

goo—rim—

 

Figure 6-12 TEM images of (a) Sample A; (b) Sample B; (c) Sample C, with corresponding selective
area electron diffraction (SAED) images shown beneath. Reproduced from Figure 7, Reference: Y.
Fan, Q. Chen, et al. in press, Vol. 2, Issue 4, Int. J. Nanomedicine (2007)

 

Figure 6-13 Close up images (a) Sample B; (b) Sample C. Reproduced from Figure 8, Reference: Y.
Fan, Q. Chen, et al. in press, Vol. 2, Issue 4, Int. J. Nanomedicine (2007)

6.5Conclusion & Discussion

Using SPRM, we have performed investigations of tissue scaffold properties directly
along individual nanofibers. We have investigated surface roughness and elasticity
properties that influence cell attachment. This is the first time that statistically meaningful

information has been extracted along individual nanofibers using an automatic procedure

81

that maintains uniformity of experimental conditions.

The SPRM approach provided a wealth of data. Statistical methods based on
histograms were developed to analyze the surface roughness and elasticity properties of the
tissue scaffold nanofibers. The mode, mean range variance and distribution of surface
roughness and elasticity were analyzed for tissue scaffold samples A, B and C. Sample C
consistently showed properties that differed from samples A and B. The most prominent
differences were observed in the surface roughness and elasticity distributions rather than
in the individual mode values.

TEM with electron diffraction analysis confirmed that both the structure and surface
properties of sample C differed from those of samples A and B. The electron microscopy
results were consistent with the results of the histogram analyses using SPRM generated
data, including the large property variances and wide irregular distributions observed for
sample B as well as for the small variances and narrow smooth distributions observed for
sample C.

Preliminary investigations indicated that normal rat kidney cells (NRKZ) showed good
adhesion to sample C and poor adhesion to sample B tissue scaffolds [76]. Systematic cell
studies with SPRM-characterized tissue scaffolds are reported in next two chapters.

SPRM has the obvious advantage of saving operation time by seaming only regions of
interest. This is especially useful for investigations of tissue scaffold properties, as a
scaffold is composed of many individual nanofibers. SPRM also enabled us to discriminate

and use only the most reliable data from each nanofiber for the properties evaluations.

82

CHAPTER 7

7 CELL RESPONSE ANALYSIS

As discussed in Chapter 5, SPRM has been used for first-time investigations of
surface roughness and elasticity directly along individual nanofiber within a tissue
scaffold. The true nanofiber curvature was also identified by using deconvolution
techniques to identify the “missed” artifact regions, and the true nanofiber diameter was
estimated by using the Kasa circle fit techniques. Surface roughness, elasticity and 3D
nanoscale curvature are all parameters which have been demonstrated to evoke strong
cell attachment responses.

As discussed in Chapter 4, specific combination of these three properties plus
regulatory protein surface chemistry should mimic the natural environment for successful
regrowth of neural cells. The candidate’s dissertation research explored the response of
astrocytes to a series of 2D planar substrates, 3D unmodified nanofiber substrates and 3D
peptide-modified nanofiber substrates designed to mimic the basal lamina. The
unmodified and peptide-modified nanofiber substrates were characterized by SPRM.

Fibroblasts were investigated as a first model system for astrocytes. Then the
astrocyte response was analyzed as a second model system. Fibroblasts and astrocytes are
both actin-based cells, which means that they actively probe their environment through
their extension of protrusions called larnellipodia (broad) and fi10podia (narrow). The

extensions form through the polymerization of actin filaments within the cells as a

83

reconfiguration endo-skeleton.

For actin-based cells, such as fibroblasts and astrocytes, cell motility towards and
adhesion to tissue scaffolds results from the extension of cell protrusions due to actin
polymerization. Motility and adhesion are triggered through a complex interaction of
receptors at the leading edge of the protrusion with the external environment ahead, as
well as with the cell internal enviromnent behind. Cell sensing of surface roughness
(haptotaxis), elasticity (durotaxis) and surface chemistry (chemotaxis)[61, 77, 78] are all
known to trigger cell motility towards conditions which promote adhesion. Leading edge
formation corresponds to a 10’s of nanometers sensing area at the cell’s extending tip.
The environmental triggers should therefore be assessed on a comparable scale. Atomic
force microscopy is an investigative tool with the correct nanometer-scale resolution and
which may be used to investigate minimally conductive biological surfaces[79].

In a situation that reproduces its natural biological environment, cells will extend
protrusions towards the scaffold. The initial attachment of the cell with scaffold surface is
triggered through a complex interaction of chemical and mechanical receptors at the
leading edge of the protrusion. Actin-based cells develop the protrusions through a
dynamic cycle of assembly and disassembly of intracellular actin filaments. Recent
research has provided new insight into the internal signaling cascades that promote the
reorganization of actin filaments into aligned and branched internal structures that result
in the extension of the two different types of protrusions, narrow filopodia and broad

lamellipodia as shown in Figure 7-1 [80].

84

The use of scafl‘olds composed of biocompatible materials as vehicles to deliver or
promote growth of cells within damaged tissues of the body is an exciting new tool for
the field of regenerative medicine. To optimize the performance of these scaffolds for
each type of tissue or application, new analytical methods must be developed that can
examine the interaction of these cells with the surfaces of these scaffolds in vitro under a

variety of biochemical and physical conditions at a resolution in the nanometer range.

 

 

Figure 7-1 (a) Characterization ofSZR+ cells: A, Morphology of cell. Many cells display broad
circumferential lamellipodia; B. Protrusive activity (time-lapse sequence). Lamellipodium protrudes
with regular and smooth outline. Dotted line shows the initial position of the cell edge. Time in
minzsec. C. Actin cytoskeleton (phalloidin staining). Circumferential lamellipodium is rich in actin.
D. Structural organization of lamellipodia; (b) Characterization of BG2 cells. A. Morphology of cell
population. Cells display polarized lamella. B. Protrusive activity (time-lapse sequence). Leading
edge moves forward by combination of filopodial (arrowheads) and lamellipodial protrusion. Dotted
line shows the initial position of the cell edge. Time in mlnzsec. C. Actin cytoskeleton (phalloidin
staining). D. Structural organization of lamellipodia. Images from reference [80]

The topographic, mechanical and chemical properties of a series of tissue scaffolds

fabricated from electrospun carbon nanofibers tissue scaffolds were investigated by

85

SPRM. Atomic force microscopy, phase contrast microscopy, selective staining &
fluorescence microscopy were used to observe cell responses to these tissue scaffolds
nanofibers. Phase Contrast Microscopy was performed on live cells using an Ohnpus
1X70 inverted microscope. The images were captured by using IP lab scientific imaging
software. The desiccated samples were optically examined during AF M imaging using a
Sony XC-999P CCD color video camera microscope with a VCL-12S12XM lens (f = 12
mm). AF M was operated based on contact mode in ambient air. Other instrument
parameters included the use of a J scanner with a maximum 125x125 square micron x-y
scan range and silicon nitride tips with a nominal 20 nm tip radius of curvature.

The differences in cell responses corresponding to the differences in substrate

properties are discussed.

7. 1 Tissue Scaffold Samples & Experimental Techniques

7.1.] Tissue scaffold samples

Samples of tissue scaffolds fabricated from electrospun [50, 51] carbon nanofibers
were obtained from Donaldson Co., Inc. (Minneapolis, MN). The nanofibers were
electrospun using an adapted electrospinning probe procedure described in Reference[52].
They were electrospun onto plastic coverslip for in-vitro study.

Four different surfaces: 2D, NANS, SANS and SANS + XL + D5/FGF-2, as

shown in Figure 7-2, are target substrates to be investigated. Where 2D is planar plastic

86

coverslip; NANS are non-activated unmodified nanofibers; SANS are surface-activated
nanofibers, functionalized with an amine group; and SANS + XL + D5/FGF-2 are SANS
crosslinked and modified with D5 peptide or Fibroblast Growth Factor 2 (FGF-2). These
nanofibers were randomly oriented polyamide nanofibers (a continuous fiber that collects
as a nonwoven fabric). They were electrospun from a blend of two polymers

[(C2804N4H47)n and (C2704.4N4H50)n] onto plastic coverslips.

 

L M... ...M. .............. 2 I“ _. 2.2,

Figure 7-2 Four target substrates (a) 2D: planar plastic coverslip; (b) NANS: Non-activated
nanofibers; (c) SANS: surface-activated nanofibers, functionalized with amine group; (d) SANS + XL
+ D5/FGF-2: crosslinked and modified with D5 peptide or F GF—2 growth factor.

7.1.2 SPRM Properties investigation of nanofibers
The new Scanning Probe Recognition Microscopy was used to analyze topographic,
mechanical and chemical properties of NANS and SANSS+XL+FGF-2 nanofibers as

discussed in chapter 6.

87

7. 1.2. 1 Surface roughness

The Surface Roughness property of unmodified and FGF-2 modified nanofiber
substrates was analyzed as discussed in Chapter 6, with the creation of surface roughness
maps and histograms, created from the collection of reliable data points from SPRM scan
along individual nanofibers. The comparison between distributions of surface roughness
property is shown in Figure 7-3. The result shows that NAN S nanofibers have a slightly
smaller surface roughness mode value. ~3 nm versus ~5 nm. The FGF-2 modified '

nanofibers have a larger surface roughness variance: more ~5-10 nm points.

   

0 — a...
0 5 10 15 20

(a)
Figure 7-3 Histogram of surface roughness map for (a) NANS: unmodified nanofibers; (b) SANS +
XL + FGF-Z: crosslinked and FGF-2 modified nanofibers. AFM images of these two substrates are

shown as the inside images, scan size: 5 pm.

 

7.1.2.2 Elasticity
The force curves are collected at the top of the nanofiber where the force vector is
exactly normal to the sample surface as shown in Figure 7-4 by using the new Scanning

Probe Recognition Microscopy as explained in chapter 6. Comparison of the results

88

shows that the FGF-Z modified nanofibers have a lower median Young’s modulus and

their variance is less too.

SAN+XL+FGF2 NAN

 

1000'

800
700

600»

Relative elasticity

l ,,_ ,

SAN+XL+FGF2 NAN

Figure 7-4 Box plots of elasticity for two different nanofiber substrates SAN+XL+FGF—2 and NAN.
The top images are AFM images of these nanofibers, scan size: 5 pm. Points shown as crosses on
AFM images are the positions where force curves are collected to calculate elasticity.

89

7.1.2.3 Surface chemistry

Chemistry as well as geometry of the extracellular matrix is critical for proper
neuronal function. It is necessary to optimize chemical cues of a biomimetic surface for
spinal cord repair. Fibroblast Growth Factor 2 (FGF-2) has demonstrated activity in the
regulation of pathways that control regeneration and repair. FGF-2 is also observed to
exist in astrocyte-capillary system in-vivo as shown in Figure 7-5. Therefore, it can be
attached to nanofibers to recreate chemical cues in tissue engineering and regenerative
medicine. FGF-2 provides an example of a peptide-modification that might enhance the

function of a nanofibrillar scaffold.

 

Figure 7-5 In vivo astrocyte-capillary system: astrocytes are pointed by arrows in (a); FGF-Zs are
dots pointed by arrows in (b). S. Meiners, et al. UMDNJ

(1) Phase Imaging technique to obtain chemical information
In tapping mode AFM, a stiff cantilever, which has a sharp tip at its free end, is

excited at or near its free resonance frequency. The oscillation amplitude is used as the

90

setpoint parameter. The feedback loop adjusts the tip-sample distance to maintain
constant oscillation amplitude in order to measure the topography of the sample surface
[33]. Additionally, material properties variations could be mapped by recording the phase
shift between the driving force and the tip oscillation. Tapping mode AFM allows us to
obtain both topographic and compositional information in samples by recording the phase
angle difference between the external excitation and the tip motion, which is called phase
imaging. It has been shown in work by Garcia et a1 [81] that two tip-sample interaction
regimes exist in tapping mode. These are designated the phase-repulsive and the phase
attractive regime. The difference between them is shown in Figure 7-6 .

Attractive regime

Repulsive regime V... d... Waals

  
 
  

Van der Waals

  

Ion-ion

 

Figure 7-6 Repulsive regime versus attractive regime in Phase Imaging.

In the repulsive regime, the tip-sample interaction includes the strong ion-ion
repulsion Coulomb force. In the phase attractive regime, scanning takes place with the tip
at a distance from the sample surface which corresponds to a tip—sample interaction by
Van der Waals Forces only. The height above the sample surface that corresponds to the

division between the two regimes can be identified through analysis of force distance

91

curves as shown in Figure 7-7. The removal of the strong ion-ion repulsion has proved to
be very important for imaging biological samples. An example taken from Garcia’s work

is shown in Figure 7-8.

Force Calibration Plot

§ E i—rrriirw

-— Fetr‘titirrn

£11.12] i (11.1%

_ . .11

Digital Ir'r:tr‘lirnerrt;
F‘ 1111]“ clrttrm-el
Fi-‘lllli _ _-

 

Figure 7-7 Force distance curve showing the division between the two regimes, this curve was taken
on a SANS+XL+FGF—2 nanof'iber sample.

The morphology and dimensions of the fragments that form antibodies are clearly
visible in the image obtained at phase attractive mode as shown in Figure 7-8 (a), while
the image obtained at phase repulsive mode has no clear evidence of the domain structure

as shown in Figure 7—8 (b) [81].

92

 

Figure 7-8 Characteristic morphologies of a-HSA antibody (a) phase attractive mode; (b) phase
repulsive mode [81]

(2) Experimental results of using Phase Imaging technique

In our tapping mode experiments, imaging in the phase attractive mode is used to
investigate surface chemical properties of unmodified nanofibers (NANS) and modified
nanofibers (SANS+XL+FGF-2). Atomic force microscopy is performed in tapping mode in
ambient air. Other instrumental parameters include the use of an E scanner with a
maximum 10x10 square micron x-y scan range and silicon tips with a nominal 10 nm tip
radius of curvature.

The phase images at attractive regime shows in Figure 7-9 (e)-(f) are recording the
phase angle difference between the external excitation and the tip motion. Comparison
between the phase images of unmodified nanofibers and modified nanofibers shows that

this technique provided macromolecular resolution. The unmodified nanofibers have a

93

smooth phase image, while structures at nanometer level (about 15 nm) appear on the
modified nanofibers as shown in Figure 7-9. 15 nm is consistent with the expected size ‘of
FGF-2 protein. Therefore, imaging in the phase attractive regime provided first-time
direct quantification of the FGF-2 peptide modification. Based on these images, the
peptide-modified nanofiber appears to be densely covered with FGF-2 macromolecules.
Coverage density is a very important property to quantify and should reproduce in vivo or
wound healing overages. For comparison, phase images taken in phase repulsive mode
are shown in Figure (c)-(d). The results are consistent with those obtained by Garcia: no
clear evidence of peptide domain structures could be obtained in phase repulsive mode.

Phase imaging of modified nanofibers in attractive mode provided detailed
information of the structures on the nanofiber as shown in Figure 7-9 (e), but it is hard to
maintain the weak tip-sample interaction when there are large variations at the sample
surface. Therefore, Phase Imaging in the attractive regime displayed loss of contact when
imaging boundary regions of nanofibers.

SPRM is currently being adapted for use in Tapping Mode/Phase Imaging along
nanofibers. SPRM auto-tracking will enable phase attractive imaging through decrease of

surface variations.

94

 

Figure 7-9 surface chemistry characteristics (a) height image of SANS+XL+ FGF-2, Z range for
height image is 500 nm; (b) height image of NANS, Z range is 1000 nm; (e) phase image of
SANS+XL+FGF—2 at repulsive mode, Z range is 75 °; (d) phase image of NAN S at repulsive mode, Z
range is 20 °; (e) phase image of SANS+XL+FGF-2 at attractive mode, Z range is 75 °; (f) phase
image of NANS at attractive mode, Z range is 20 °. Image scan size is 1 pm.

7.2First Model System: Fibroblasts on 20 versus Nanoscale 3D
Nanofibers
The responses of 3T3 NIH fibroblasts to 2D planar surfaces versus their response to
nanofiber mat surfaces, which have 3D nanoscale curvature effects, were investigated by

our group as first model system for astrocytes. The interactions of filopodia and

95

lamellipodia emanating from the NIH 3T3 fibroblasts with 2D planar and 3D
nanofibrillar surfaces were investigated by using phase contrast microscopy, optical
microscopy and atomic force microscopy. The 3D nanofibrillar surfaces were previously

investigated using SPRM.

7.2.1 Cell culture

NIH 3T3 fibroblast cells were cultured for 24 hours on two different surfaces, planar
(2D) tissue culture plastic and amine coated nanofibers (SANS). The amine coated
nanofibers (SANS) were randomly oriented polyamide nanofibers electrospun by
Donaldson Co., Inc. (Minneapolis, MN). They were electrospun from a blend of two
polymers [(C2804N4H47)n and (C2704.4N4H50)n] onto plastic coverslips. The
polymeric nanofiber mat was cross-linked in the presence of an acid catalyst and formed
a dense network of filaments 50-80 nm in diameter with minimal porosity. The
nanofibers were covalently coated with a proprietary polyamine polymer by Surmodics,
Inc. (Eden Prairie, MN) to provide functional groups for further covalent modification of
the nanofibers with bioactive peptides.

Samples for optical and AFM microscopy were fixed with 2.5% glutaraldehyde in
0.1% phosphate buffer for 15 minutes, briefly rinsed with 0.1M phosphate buffer and
washed 3 times with triple distilled water. The samples were then permitted to desiccate

in air for 48 hours prior to optical and AFM imaging.

96

7.2.2 Experimental investigation of fibroblasts response
3T3 NIH fibroblasts are cultured on two different surfaces, tissue culture plastic (2D)
and amine coated nanofibers. In order to investigate cell response, fibroblasts cultured on

these two surfaces were observed and compared by using different techniques.

 

(a) (b)
Figure 7-10 Fibroblasts cultured on 2D surface. (a) Phase contrast microscopy of live cells, and (b)
Optical microscopy of fixed cells indicate similar morphologies. Reproduced from Figure 1,
Reference: Q. Chen, Y. Fan, et al, Mater. Res. Soc. Symp FF. Proc. (2007)

 

Figure 7-11 Fibroblasts cultured on amine coated nanofibers. (a) Phase contrast microscopy of live
cells, and (b) Optical microscopy of fixed cells indicate similar morphologies. Reproduced from
Figure 2, Reference: Q. Chen, Y. Fan, et al. Mater. Res. Soc. Symp FF. Proc. (2007)

Phase contrast images of living fibroblasts at the end of the first 24 hours are shown

97

in Figure 7-10 (a) and Figure 7-11 (a). Bright field images of the desiccated cells shown
in Figure 7-10(b) and Figure 7-11 (b) show similar morphologies. A difference in cell

morphology was observed using phase contrast and bright field microscopy.

r
1 > :
r .-
i ’ '

3011111

Data Type Deflection ' Data Type Height
Z range 19 '1711m Z range 700 0 nm

 

Figure 7-12 AFM images of 3T3 NIH fibroblast cultured on 2D surface. (a) Deflection image of a
typical fibroblast; (b) Close-up deflection image of top right region; (c) Close-up deflection image of
top left region; (d) Close-up deflection image of bottom right region; (e) Height image of vertex
region for another cell showing structures (close-up deflection image in inset). Reproduced from
Figure 3, Reference: Q. Chen, Y. Fan, et al. Mater. Res. Soc. Symp FF. Proc. (2007)

AFM images shown in Figure 7-12 and Figure 7-13 provided more detailed
information about size, and shape of fibroblasts growing on the two different surfaces.
The cell body size of fibroblasts was about 50pm in both cases. However, fibroblasts
grown on the tissue culture plastic 2D surface showed fine filopodia-like structures at the
ends of pointed projections, as shown in Figure 7-12. Fibroblasts grown on the amine
coated nanofibers showed blunted lamellipodia—like projections extending over several
nanofibers. These projections lacked the fine filopodia-like end structures. Fibroblasts

grown on the amine coated nanofibers were also observed to show a great number of

98

cell-cell interactions between the blunted vertiees as shown in Figure 7-13. Both factors
contributed to the changed appearance of the 2D and nanofiber cultured fibroblasts

observed in the images obtained using phase contrast and bright field microscopy.

 

(a)

 

Figure 7-13 3T3 NIH fibroblast cultured on amine coated nanofibers (SANS). (a) Multiple cell-cell
interactions on the SANS are observed in the optical microscopy image. The cells in the black box are
shown in close-up AFM images in (b) through (g). (b) through (d): increasing close-up AFM height
images; (e) through (g): corresponding increasing close-up AFM deflection images. Reproduced from
Figure 4, Reference: Q. Chen, Y. Fan, et al. Mater. Res. Soc. Symp FF. Proc. (2007)

99

7.2.3 Comparison of fibroblasts response and conclusion

3T3 NIH fibroblasts cultured on two different types of surfaces: tissue culture
plastic and amine coated nanofibers (SANS) are observed by using phase contrast
microscopy, optical microscopy and atomic force microscopy. The cell body size of
fibroblasts was about 50m in both cases. However, fibroblasts grown on the tissue
culture plastic 2D surface showed fine filopodia—like structures at the ends of pointed
vertices while fibroblasts grown on the amine coated nanofibers showed blunted
lamellipodia-like vertices extending over multiple nanofibers. Fibroblasts grown on the
amine coated nanofibers (SANS) were also observed to show a greater number of
cell-cell interactions between the blunted vertices. Both factors contributed to the
changed appearance of the 2D and nanofiber cultured fibroblasts observed. The
appearance of the vertices and the greater number of observed cell-cell interactions
suggests the possible formation of a greater number of adherens junctions. If this is so, it
further suggests that amine-coated nanofibers (SANS) may stimulate Rae activation since
Rae activation stimulates the formation of adherens junctions.

The remarkably different response of the actin based cells to a nanoscale 3D
versus the planar 2D situation may prove to be medically important. 3T3 and embryonic
fibroblasts are being used by our group as first model system for astrocytes which are

another actin based cell system.

100

7.3Second Model System: Astrocytes on ZD versus Nanoscale

3D Nanofibers

The astrocyte system was next investigated, which is another actin based cell system.
Astrocytes perform multiple functions for brain and spine neuronal systems, including
biochemical support of endothelial cells which form the blood-brain barrier, providing
food and waste removal for the nervous tissue, and a principal role in the repair and
scarring process in the brain [82]. Astrocytes can provide direct cell body-cell body
mechanical support for neuron cell bodies. Astrocytes can also provide indirect
mechanical support for neuronal systems through the generation of astrocyte-derived
extracellular matrix molecules [53].

Astrocytes have environmental requirements which enable healthy firnction. Healthy
versus pathological astrocyte development on nanoscale 3D nanofibers versus planar
substrates are observed. Promising result of in vivo implantation of slowly degradable
nanofiber scaffolds composed of electrospun polyamide nanofiber random meshes in the
rat confirms these observations [83]. We will focus on the response of astrocyte, such as
attachment and division, to 2D planar substrate and 3D nanofibrillar substrate, as this step
is important for nerve system regeneration. Surface chemistry of nanofibers would affect
astrocyte regeneration in its search for an attractive enviromnent. Therefore, both

unmodified nanofibers and FGF-2 modified nanofibers are included in the analysis.

101

7.3.1 Cell culture

Astrocyte cells were cultured on three different surfaces, 2D, NANS and
SANS+XL+FGF-2 for 24 hours in one group and on the same three substrates for 48
hours in another group. Samples for optical and AFM microscopy were fixed with 2.5%
glutaraldehyde in 0.1% phosphate buffer for 15 minutes, briefly rinsed with 0.1M
phosphate buffer and washed 3 times with triple distilled water. The samples were then
permitted to desiccate in air for 48 hours prior to optical, AF M imaging.

7.3.2 Experimental investigation of astrocytes response

 

Figure 7-14 Actin stain + Fluorescent Microscopy images of astrocytes on different substrate (a)
astrocytes on 2D substrate imaged at 24 hours; (b) astrocytes on NANS at imaged 24 hours; (c)
astrocytes on modified NANS imaged at 24 hours; (d) astrocytes on 2D substrate imaged at 48 hours;
(e) astrocytes on NANS imaged at 48 hours; (1') astrocytes on modified NANS imaged at 48 hours

Astrocytes are cultured on 2D, NANS and peptide-modified SANS+XL+FGF-2 for

24 hours in one group and for 48 hours in another group. Both groups were observed by

102

fluorescent microscopy after actin stained as shown in Figure 7-14. At 24 hours, all cells
are probing their environments. By 48 hours, astrocytes on nanofibers have a more in

vivo-like morphology indicating a good environment as shown in Figure 7-14.

E; 1pm

   

Figure 7-15 AFM images of astrocytes grown on different substrates for 24 hours (a) 2D substrate; (b)
NANS; (c) modified NANS. Image scan size : 50 pm

Atomic force microscopy is performed in contact mode in ambient air. Other
instrumental parameters include the use of a J scanner with a maximum 125x125 square
micron x-y scan range and silicon nitride tips with a nominal 20 nm tip radius of curvature.

Even at 24 hours, AFM images of astrocytes on different substrates show differences
as shown in Figure 7-15. Astrocytes on nanofibers develop very thin ( ~1 actin layer thick)

broad processes to probe their environments. This was similar to the cell response

103

observed for NIH 3T3 fibroblasts previously discussed. Also astrocytes on nanofibers
show more cell-cell interactions. The number of cell-cell interactions appeared to
increase for FGF-Z-modified nanofibers.
7.3.3 Conclusion

The new Scanning Probe Recognition Microscopy has been successfully used to
quantify 3D nanoscale cues: elasticity, surface roughness and peptide modification in
tissues scaffold nanofibers. Cell responses to these different nanofibers are also reported.
Therefore, suecessfirl assessment cell response to variations in nanofiber properties
accurately quantified using SPRM may provide vital insights that lead to successful

spinal cord repair techniques.

104

CHAPTER 8

8 ELECTROSPINNING PARAMETERS

As discussed in Chapter 4, In spinal cord repair, individual external triggers in the
cell environment that stimulate dynamic cell responses include surface roughness,
elasticity and surface chemistry. The ultimate goal is to find the best tissue scaffold with
most appropriate topographical, mechanical and chemical properties for specific cells or
cell classes. These properties of nanofibers can be analyzed by using the new Scanning
Probe Recognition Microscopy technique. Cell responses can also be investigated by
SPRM and other optical techniques. Information can be combined to provide a composite
picture representative of a cell’s perception of its environment. Then, the electrospinning
materials or parameters can be adjusted adaptively until the prOperties of the produced
nanofibers are found to be best for cell regeneration. The whole procedure will provide an
automatic guide to design tissue scaffold for its application in regenerative medicine.

The new Scanning Probe Recognition Microscopy was used to investigate of the
influence of droplet size on electrospun carbon nanofiber properties. The key property
under investigation is the effect of electrospinning conditions on the resulting carbon

nanofiber elasticity. [84]

105

 

[ i

[Tissue Scaffold I

 

 

.... I-

I Curvature I

 

 

 

I Surface roughness I

I Properties I

I SurfaceChemistry I

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I Mesh Density I J
IBio-data fusion I ] Ilnterpretation I
ICell growth I
I IOutcomes I

 

 

 

 

\ IElectrospinningparameters

5

Figure 4-6 (repeated) Application of SPRM in spinal cord repair

 

8. 1 Introduction

Electrospun carbon nanofibers are hollow core carbon structures produced by
self-assembly within an evaporating liquid jet. They are of great interest as biocompatible
carbon-based structures for tissue scaffold applications.

Electrospinning-based self-assembly is achieved by the application of an electrostatic
force between a charged droplet containing polymer monomers in liquid suspension and a
collecting metal electrode [85, 86]. A charged droplet of monomer in liquid suspension
is formed at a capillary tip, conveniently a metal syringe tip of known bore radius. A

20-30 kV potential difference is applied between the syringe tip and a metal collecting

106

plate, held 10-20 cm apart. At a critical field, the force due to the electric field overcomes
the surface tension forces holding the droplet, and the solution starts flowing towards the
collecting electrode in the form of a charged jet. Initially, the charged droplet deforms
into a cone, which then separates into splayed charged jets due to mutual repulsion. As
the liquid in each splayed jet evaporates, the jet diameter shrinks rapidly, creating the
conditions for self assembly of the polymers into nanofibers. The diameters of the
collected electrospun carbon nanofibers may range from a few microns to as low as 10
nanometers depending on the complicated interplay of Coulomb forces. Electrospun
carbon nanofibers used in tissue scaffold applications typically have diameters on the
order of ~100 nm.

Results of an investigation of monomer choice, to control the initial droplet
charge state, and syringe bore size, to control the initial droplet radius/surface tension, are
presented in this chapter. In Series 1, 15% weight of poly methyl methacrylate suspended
in tetrahydrofuran (THF)/Dimethylformamide (DMF) was electrospun using a 200
micron fixed bore radius, without and with the addition of 6% single walled carbon
nanotubes produced by NASA-GSFC Cooled Welding Method added to the suspension.
In Series 2, 15% weight of poly (e-caprolactone) suspended in Methylene Chloride (MC)/
Dimethylformamide (DMF) was electrospun using bore radii of 152.4, 254.0 and 406.4
microns. The tip to collector plate distance was held constant at 10 cm. The voltage was
held constant at 25kV. A key property under investigation in our group is the effect of

electrospinning conditions on resulting carbon nanofiber elasticity.

107

8.2 Experimental Setting

Poly (methyl methacrylate) (PMIVIA) with a molecular weight of 120,000 and poly
(e-caprolactone) with a number average molecular weight (lVIn) of 80,000 from Aldrich
Chemical were used in these experiments Figure 8-1 (a—b). Single walled carbon
nanotubes, 6 % by weight, produced by NASA-GSFC Cooled Welding Method were
added to one batch of PMMA experiments. The electrospinning set-up consisted of a
Sorensen 230-3/12P R&D high voltage DC power supply, with a syringe and an 8x12”
aluminum collecting foil held 10 cm apart. The experiments were horizontally configured,

with the syringe tilted about 300 to facilitate droplet formation Figure 8-1 (0).

(a) one. (b) CH
1 / \c/ 3

—CH2 c— \ I
I o
c

 

 

 

 

\C/

%\o
O \

 

CH3 0
(c) 20 — 30 kV

 

 

 

l ....

- /
. Pipette Wrth

Polymer Solution
(capiilafY)

 

 

 

 

 

10—25 cm

 

 

Collector Plate

 

Figure 8-1Polymer monomers (a) poly (methyl methacrylate), and (b) poly (e-caprolactone). (c)
Experimental configuration of electrospinning experiment.

108

Field emission scanning electron microscopy (FESEM) of gold coated samples was
used to assess the initial non-woven mat topography and nanofiber diameter, and to
confirm that subsequent atomic force microscopy results were truly representative. Field
emission scanning electron microscopy was performed using a Hitachi S-47OOII Field
Emission SEM operated at 1.0 kV.

Atomic force microscopy (AFM) was performed using a Veeco Nanoscope IIIA
Scanning Probe Station, in contact mode with a nominal tip radius of 25 nm, operated in
ambient air. A silicon nitride tip was used. The silicon nitride tip (hardness z 35 GPa),

was assumed to be much harder than any of the tissues scaffold samples.

8. 3 Experimental Results

The PMMA monomers with and without single walled carbon nanotubes were
electrospun under conditions of 200 microns bore radius,10 cm between the tip and the
collector foil, and 25kV electrostatic potential difference. The results are shown in Figure
8-2. Electrospinning of both resulted in micron-scale diameter fibers. This is too large for
tissue scaffold applications and PMMA was not investigated further. Anomalous beam
and tip deflections observed during the electron and atomic force microscopy
experiments indicated highly charged surface states.

The poly (s-caprolactone) monomers were electrospun under conditions of 10 em
between the tip and the collector foil and 25kV electrostatic potential difference. The

bore radius was varied through (a) 152.4, (b) 254.0, and (c) 406.4 microns. The results

109

are shown in Figure 8-3. The nanofiber average diameter was ~ 90-110 nm, increasing
slightly as the bore size was increased. Nuisance bubbles were observed in addition to
the nanofibers. The average diameter of the nuisance bubbles deceased by ~50% as the

bore size was increased.

      

0 25pm

Figure 8-2 (a) Field emission SEM of PMMA-based carbon nanofibers and (b) atomic force
microscopy of PMMA-based carbon nanoflbers with 6% weight of single walled carbon nanotubes
added. A 200 micron bore radius resulted in micron-scale diameter nanofibers. Reproduced from
Figure 2, Reference: Rutledge, S.L., Shaw, H. C., Benavides, J. B., Yowell, L. L., Chen, Q., Jacobs, B.
W., Song, S. P., Ayres, V. M., Diamond and Related Materials, 2006. 15(4-8): p. 1070-1074 [84]

110

A?

‘1‘

 

Figure 8-3 Field emission SEM of poly (e-caprolactone) carbon nanofibers spun at bore radii (a)
152.4, (b) 254.0, and (c) 406.4 microns. As the bore size was increased, a slight increase in fiber
diameter with a large decrease in bubble size was observed. Reproduced from Figure 3, Reference:
Rutledge, S.L., Shaw, H. C., Benavides, J. B., Yowell, L. L., Chen, Q., Jacobs, B. W., Song, S. P., Ayres,
V. M., Diamond and Related Materials, 2006. 15(4-8): p. 1070-1074 [84]

SPRM was used to measure the relative elasticity between the three poly

(e-caprolactone) nanofiber samples and the metal foil. As the AFM tip is vertically

111

indenting a cylindrical object, only data collected along the top and center of each
nanofiber matches the model. Then new SPRM technique can collect reliable force
curves along this region by its recognition capability. Elasticity property of these three
nanofibers electrospun with different bore sizes can be investigated by using SPRM
technique as explained in chapter 6. Results show that the mean force distance curve
areas increased with increasing bore radius, which indicate a decrease in sample elasticity

with increasing bore radius as shown in the following Figure.

 

 

 

 

       

60:" ~~— ~ 7' e e ~ ~7~ * — A - 7+ A w: u -;
o bore radius 406.4 microns
-——-—-i-—-— bore radius 254 microns
50 ~ ——-~—— bore radius 152.4 microns
' ‘ ' " foil
40 - ,1 1 ~
: l
E . i
5 30 ' ‘ i1: ,
C I
.9 1- .1
t5 1‘" . ‘.._________.__._._--_.._.__ 4:!- -‘—‘ -' -i . ..-... j,
g 20 ‘17, ,. . ;.
o , *1}; Area = 3427.2 Area = 2976.8 Area = 2839.1 Area = 2728.81
0 -- “ “-1:va 37;: :; aw :; ;: :1: 3351414“: 52-. :-—~‘-;-‘
-19- . L L __.’ , .L_LL -Ll L _.L._LL_L1_LLLL.m LL Ll. _ L L L‘ L L L-
- 00 0 100 200 300 400 500
Distance(nm)

Figure 8-4 Mean force distance curves for three nanofibers samples electrospun with different bore
radii and the aluminum foil. The mean force distance curve areas increased with increasing bore
radius (insets), indicating a decrease in nanofiber elasticity with increasing bore radius. Reproduced
from Figure 5, Reference: Rutledge, S.L., Shaw, H. C., Benavides, J. B., Yowell, L. L., Chen, Q.,
Jacobs, B. W., Song, S. P., Ayres, V. M., Diamond and Related Materials, 2006. 15(4-8): p. 1070-1074
1841

112

8.4 Discussion and Conclusion

A 15% weight of poly (e-caprolactone) suspended in Methylene Chloride
(MO/Dimethylformamide (DMF) was shown to produce electrospun carbon nanofibers
with average diameters ~ 90-110 mm. The average nanofiber diameter was observed to
increase slightly as the bore size was increased. There is a large body of research on
control of electrospun nanofiber diameter, which indicates that the monomer charge
density and polymer solution concentration have the greatest influence, by varying the
charge concentrations, and hence the Coulomb forces. For electrospun carbon
nanofibers used in tissue scaffold applications, diameters that mimic the structural portion
of the extracellular matrix (ECM) are desirable. The ECM is composed of fibrillar
collagen 50-300 nm in diameter. Therefore, the Series 2 nanofibers had acceptable
diameters for potential tissues scaffold applications.

The mean value force distance curves for data analyzed along several nanofibers
indicated a decrease in relative elasticity with increasing bore radius. This is a new
result. The correlations between electrospinning parameters (bore radius) and properties
of produced nanofibers (diameter, elasticity) may lead to dynamically adjusting
electrospinning parameters in order to get desired nanofibers properties. It will be very
important for many applications of nanofibers, for example, applying electronspun

nanofibers as implantable artificial tissues in regenerative medicine.

113

Part 111 Conclusion & Future Work

114

Part III CONCLUSION & FUTURE WORK
CHAPTER 9

9 CONCLUSION 8: DISCUSSION

9. 1 Conclusion

A new scanning probe microscopy modality: Scanning Probe Recognition
Microscopy system has been developed. The new SPRM system gives the scanning probe
microscopy itself the ability to auto-track regions of interest through incorporation of
recognition-based tip control. A low resolution scan to identify regions of interest is
followed by auto-track high resolution scans of just these areas. SPRM has the
advantages of providing more meaningful information about samples automatically,
efficiently and accurately. SPRM enables investigations in many situations in which
standard SPM is insufficient. Investigations of biomimetic properties along individual
nanofibers in actual tissue scaffolds used for regenerative medicine in spinal core repair
showcases the power of SPRM within a significant biomedical investigation.

Tissue scaffolds for spinal cord repair are a specific example of regenerative
medicine, with national importance and current emphasis. Electrospun polyamide
nanofibers are nanoscale 3D platforms for regenerative medicine. Nanoscale curvature
is indicated as a key biomimetic aspect. Tissue scaffolds made from nanoscale 3D
nanofibers must reproduce the physical properties and the functions of the basal
lamina/basement membrane. The SPRM system developed by our group enables SPM to

have recognition ability and be able to auto-track regions of interest, which are individual

115

nanofibers. Therefore, investigations of tissue scaffold properties directly along
individual nanofibers are possible. We have performed investigations of curvature,
surface roughness, elasticity and surface chemistry properties that have been shown to

influence cell attachment in statistically meaningful detail.

Investigations of cell attachment to the 2D and nanoscale 3D nanofiber surfaces,
which have different topographical, mechanical and chemical properties, are observed.
The remarkably different response of the actin based cells to planar 2D versus nanoscale
3D situation provide vital insights that head to use tissue scaffolds made of electrospun
nanofibers for successful spinal cord repair techniques. The correlation of cell responses

with nanofiber properties analysis are discussed.

9.20iscussion

The new Scanning Probe Recognition Microscopy has great potential to be used in
many research fields where macro/nano-scale resolution is required. SPRM also has the
flexibility to be able to extend into different applications. As one reviewer of one of our
successful grant applications expressed it, “This is enabling technology for nanoscience”.
Initial research work is done in applying SPRM into the following research topics. More

future work at these aspects will be discussed in detail in the next chapter.

116

 

CHAPTER 10

10 FUTURE WORK

The new Scanning Probe Recognition Microscopy has been successfully used to
solve problems in tissue engineering that require macro/nano-scale resolution. It also has
great potential to be applied to many other research fields because of its flexibility to
extend into different applications. Several directions of future work that are under
investigation are listed.

10. 1 Improvement of SPRM Recognition Capability

More image processing and pattern recognition techniques can be integrated in
SPRM to make it more powerful in feature recognition and information processing.
Additional algorithms are under investigation to make SPRM to run more efficiently,
accurately and reliably.

Different types of information are available from Scanning Probe Microscopy, such
as height information, deflect information and fiiction information in contact AFM. How
to combine and fuse all these information to get the best feature recognition ability and
optimal algorithms is also an ongoing work.

10. ZExtended Work in Spinal Cord Repair

In spinal cord repair, several other aspects can also be investigated by the new

Scanning Probe Recognition Microscopy technique. Future work will develop

SPRM-based porosity and mesh density investigations, as this is known to be an

117

important property for cell motility and adhesion[5 8].

The ultimate goal is to find the tissue scaffold with most appropriate topographical,
mechanical and chemical properties from cell regeneration. These properties of
nanofibers can be analyzed by using the new Scanning Probe Recognition Microscopy
technique. The next step is to adjust the electrospinning materials or parameters
adaptively until the properties of the produced nanofibers are best for cell regeneration as

shown in Figure 10-1. It will provide an automatic guide in tissue scaffold design.

 

 

 

{ I . I Cellculture on 7
TrssueScaffold = tissue scaffold ]

Properties 1

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

,
I CurvatureI IElasticityI Mesh I Cellresponse I
7‘ " D-ensity ~ SPRM
Surface Surface investigation
Roughness Chemistry
1
[“1 _ 810 data fusron J ]I InterpretationI

 

 

l.

.'5‘-‘r" - '- . 11823.5. ' "I 1«.;-.,;.-,-.;. Lair-aw '.'.~:- -:.:+-.— 91:. «It 9.2.11: - 1.": .~ .. .

‘ Ad‘tlct 'n' armtr

JUS e e rospm mg p a e e s ,.
K ”at. RAJ-U". “it.“ A i It“ ”13‘ ‘r3‘17.‘:':'P"~.‘I"11‘.“¢“HP?‘t‘i‘J,3":'.‘T‘5 .'.-t'.—.t‘\.i‘-".'.*r -' ‘eui ~.- ”1". ‘1 1' .-

Figure 10-1 Diagram of providing an automatic guide for tissue scaffold design

 

I Outcomes I

 

 

 

 

10.3Extension of SPRM to Additional Research Problems:

Nanocircuits

Nanowire made from a wide variety of materials has demonstrated extraordinary

electronic, mechanical and chemical characteristics. The new Scanning Probe

118

Recognition Microscopy can be used to detect the contact pad and nanowire radial

boundary.

5.1.1 Uum

 

Figure 10-2 SEM images of nano circuits
For the nanocircuits investigation, several aspects of investigation require the ability
to auto-track the scan path to proceed from the conducting contact pad onto the

semiconducting nanowire, while avoiding the insulating oxide layer.

Initial experiments are done to auto-track a nanowire by using SPRM working under
contact mode AFM setting. Experimental results indicate the nanowire can be tracked by
using recognition based tip control of SPRM as shown in Figure 10-3. The image
captured by standard AFM is shown in Figure 10-3 (a). Figure 10-3 (b) shows the

simulation result, which indicate that the scanning scheme can detect the edges very well

119

and predict the boundaries accurately. Therefore, the nanowire can be auto-tracked
reliably. Experimental result in real-time are displayed in Figure 10-3 (c) where the

regions that are not really scanned are padded with 0 (dark region) for display.

Figure 10-3 Scanning scheme implementation for GaN anocircuit: (a) Image captured by standard

 

SPM, the image side is 5 x 5 microns; (b) Scanning scheme simulation showing lines of the scan plan;
(c) Real-time image captured by SPRM system using self-defined scanning scheme along the
nanowire. Reproduced from Figure 4, Reference: B.W. Jacobs, V.M. Ayres, M.A. Tupta, R.E. Stallcup,
A. Hartman, J.B. Halpern, M-Q. He, M.A. Crimp, A.D. Baczewski, N.V. Tram, Q. Chen, Y. Fan, S.
Kumar, L. Udpa , 2006 6th IEEE Conference on Nanotechnology Proceedings

Another potential application of the new Scanning Probe Recognition Microscopy
is to use its recognition ability to locate the nanowire in the nanocircuits automatically
with a fast scan with low resolution. Once the nanowire is detected, SPRM can start a
relatively low scan scheme with high resolution to get more detail information of the
nanowire. The initial experimental results shown in Figure 10-4 demonstrate the ability of

SPRM to automatically locate the position of nanowire and scan along the nanowire.

120

 

:9

Figure 10-4 SPRM scan strategy to locate nanowire in nanocircuits (a) AFM images of nanocircuits,

the arrow points to the nanowire; (b) simulation scan result shows locating the nanowire by following
edges of nanocircuits. Image Size: 50pm.

Since both ends of the nanowire are attached to nanocircuits, the scan strategy of the
new SPRM is to follow the edge of nanocircuits once an edge is detected. The scan angle
can turn 90 degree automatically when the edge of the nanocircuits changes. After the
nanowire is found, SPRM scan scheme is to start high resolution scan on nanowire region
with relatively low scan rate, which will provide detail information about the nanowire.

Many nanocircuits investigation require the scan path to proceed from the

121

conducting contact pad onto the semiconducting nanowire, while avoiding the insulating
oxide layer. One example is imaging nanocircuits under scanning tunneling microscopy
mode. Scanning tunneling microscopy require sample surface to be conductive or
semi-conductive, otherwise, there will be no tunneling current, and the feedback loop will
fail to work, then there is no way to establish constant interaction between tip and sample
surface.

With the recognition ability presented in the new Scanning Probe Recognition
Microscopy, the scan can stay only within the region of interest which can be defined as
the contact pad and nanowire region. Therefore, the nanocircuits can be imaged under
scanning tunneling microscopy mode even though there are insulating region which is
oxide layer in the sample surface.

A new set of SPRM feature definitions involving highly planar surfaces with

nanoscale height variation is under development at the present time.

122

REFERENCE

1. Dufrene, Y.F., Atomic Force Microscopy, a Powerful Tool in Microbiology.
Journal of Bacteriology, 2002. 184(19): p. 5205-5213.

2. Fotiadis, D., et al., Atomic—force microscopy: Rhodopsin dimers in native disc
membranes. Nature, 2003. 42: p. 127-128.

3. Gadegaard, N., Atomic force microscopy in biology: technology and techniques.
Biotechnic & Histochemistry, 2006. 81(2 - 3): p. 87-97.

4. Phase contrast microscopy. Wikipedia 2007 [cited.

5. Wang, X.F. and B. Herman, Fluorescence Imaging Spectroscopy and Microscopy.
1996, New York: Wiley.

6. Amos, W.B. and J .G White, How the confocal laser scanning microscope entered
biological research. Biology of the Cell 2003. 95: p. 335—342.

7. Sandison, DR. and W.W. Webb, Background rejection and signal-to-noise
optimization in the confocal and alternative fluorescence microscopes. Appl. Opt. ,
1994. 33: p. 603-610.

8. Kleinfeld, D., et al., Fluctuations and Stimulus-Induced Changes in Blood Flow
Observed in Individual Capillaries in Layers 2 through 4 of Rat Neocortex. Proc
Natl Acad Sci U. S. A., 1998. 95: p. 15741-15746.

9. Bozzola, J .J . and L.D. Russell, Electron Microscopy: Principles and Techniques
for Biologists. 1998, Sudbury (MA): Jones and Bartlett.

10. Bonetta, L., Zooming in on electron tomography. Nature methods, 2005. 2(2): p.
139—144.

11. Hansma, H.G, Surface biology of DNA by atomic force microscopy. Annual Rev.
ofPhys. Chem., 2001. 52: p. 71-92.

12. Liang, Y., et al., Organization of the G-Protein-coupled receptor rhodospin and
opsin in native memebranes. J. Biol. Chem, 2003. 278: p. 21655-21662.

13. Moller, C., et al., Determining the molecular forces that stabilize human
aquaporin-I. J. Struct. Biol., 2003. 142: p. 369-378.

123

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

Engel, A., Microscopic assessment of membrane protein structure and function.
Histochem Cell. Biol., , 2003. 120: p. 93-102

Shao, D.C.a.Z., Submolecular Resolution of Single Macromolecules with Atomic
Force Microscopy. FEBS Letters, 1998. 430: p. 51-54.

Florin, E.L., V.T. Moy, and HE. Gaub, Adhesion Forces between individual
ligand-receptor pairs. Science, 1994. 265: p. 415-417.

Hoh, J .H. and PK. Hansma, Atomic force microscopy for high-resolution imaging
in cell biology. Trends in Cell Biology, 1992. 2(7): p. 208-213.

F.Braet, et al., Imagining surface and submembranous structures with the atomic
force microscope: a study on living cancer cells, fibroblasts and macrophages. J.
Microscopy, 1997. 190: p. 328-338.

Fillmore, H.L., et al., Atomic force microscopy observations of tumour cell
invadopodia: novel cellular nanomorphologies on collagen substrates.
Nanotechnology, 2003. 14: p. 557-561.

Chasiotis, I., H.L. Filhnore, and GT. Gillies, Atomic force microscopy
measurement of cytostructural elements involved in the nanodynamics of tumour
cell invasion. Nanotechnology, 2003. 14: p. 73-76.

Leighton, X., et al., Significant allelic loss of ANX 7 region (10q21) in hormone
receptor negative breast carcinomas. Cancer Letters, 2004. 210(2): p. 239-244.

Subramaniam, V., A.K. Kirsch, and TM. Jovin, Cell biological applications of
scanning near-field optical microscopy (SNOM). Cellular and Molecular Biology,
1998. 44: p. 689-700.

Binnig, G, C.F. Quate, and C. Gerber, Atomic Force Microscope. Physical
Review Letters, 1986. 56(9): p. 930.

Pesen, D. and J.H. Hoh, Modes of Remodeling in the Cortical Cytoskeleton of
Vascular Endothelial Cells. FEBS Lett., 2005. 579: p. 473-476.

Hansma, H.G and J.H. Hoh, Biomolecular Imaging with the Atomic Force
Microscope. Annual Review of Biophysics and Biomolecular Structure, 1994.
23(1): p. 115-140.

Rivetti C, C.S., Dieci G, Bustamante C., Visualizing RNA extrusion and DNA

124

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

wrapping in transcription elongation complexes of bacterial and eukaryotic RNA
polymerases. J Mol Biol 2003. 326(5): p. 1413-1426.

Lee, GU., D.A. Kidwell, and RI. Colton, Sensing discrete streptavidin-biotin
interactions with atomic force microscopy. Langmuir, 1994. 10(2): p. 354- 361.

Wojcikiewicz, E.P., X. Zhang, and V.T. Moy, Force and Compliance
Measurements on Living Cells Using Atomic Force Microscopy(AFM). Biol
Proced Online, 2004. 6: p. 1-9.

Ayres, V.M., Microscopy, Scanning Tunneling in Encyclopedia of Medical
Devices and Instrumentation (Second Edition), J. Webster, Editor. 2006, John
Wiley and Sons Inc.: Hoboken, NJ. p. pp.516-523.

Tanaka, H. and T. Kawai, Visualization of detailed structures within DNA. Surface
Sci Lett, 2003. 539: p. 531-536.

The Nobel Prize in Physics 1986. nobelprize 1986 [cited.
MultiMode SPM Instruction Manual, Veeco Instruments Inc.

Howland, R. and L. Benatar, A Practical Guide to Scanning Probe Microscopy.
1996, Sunnyvale CA: Park Scientific Instruments.

Giessibl, F.J., Advances in atomic force microscopy. Reviews of Modern Physics,
2003. 75(3): p. 949-35.

Hansma, P.K., et al., APPLIED PHYSICS: High-Speed Atomic Force Microscopy.
Rev. Sci. Instrum., 1996. 314(5799): p. 601-602.

Walters, D.A., et al., Short cantilevers for atomic force microscopy. Review of
Scientific Instruments, 1996. 67(10): p. 3583-3590

Humphris, A., M. Miles, and J. Hobbs, A mechanical microscope — high speed
atomic force microswpy. Applied Physics Letters, 2005. 86: p. 034106.

Schitter, G, F. Allgower, and F. Stemmer, A new control strategy for. high-speed
atomic force microscopy. Nanotechnology, 2004. 15: p. 108—114.

Kodera, N., H. Yarnashita, and T. Ando, Active Damping of the Scanner for
High-speed Atomic Force Microscopy. Rev. Sci. Instrum., 2005. 76: p. 053708.

Hobbs, J.K., C. Vasilev, and A.D.L. Humphris, Real time observation of

125

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

crystallization in polyethylene oxide with video rate atomic force microscopy.
Polymer, 2005. 46(23): p. 10226-10236.

Wire, B. 2006 [cited.

NanoScope Signal Accesss Module (SAM)-Description & Use. 1997, Santa
Barbara, CA: Digital Instruments.

Xu, Y., et al., Wavelet transform domain filters: a spatially selective noise
filtration technique. Image Processing, IEEE Transactions on, 1994. 3(6): p.
747-758.

Chen, Q., V.M. Ayres, and L. Udpa, Biological Investigation Using Scanning
Probe Recognition Microscopy. Proceedings of the 3rd IEEE Conference on
Nanotechnology (IEEE-NANO 2003), 2003. 2: p. 863-865.

Goolsby, B., et al., Scanning probe microscopy with landmark referenced control
for direct biological investigations. J Nanosci Nanotechnol, 2003. 3(4): p. 347-50.

Fantner, GE., et al., Components for high speed atomic force microscopy.
Ultramicroscopy, 2006. 106(8-9): p. 881-887.

Atala, A., Technology insight: Applications of tissue engineering and biological
substitutes in urology. Nat Clin Pract Urol, 2005. 2(3): p. 143-9.

Hu, J .C. and K.A. Athanasiou, A self-assembling process in articular cartilage
tissue engineering. Tissue Eng, 2006. 12(4): p. 969-79.

Xu, M., et al., Iissue—Engineered F ollicles Produce Live, Fertile Offspring. Tissue
Eng, 2006. 12(10): p. 2739-2746.

Dzenis, Y., Spinning Continuous Fibers for Nanotechnology. Science, 2004.
304(5679): p. 1917-1919.

Taylor, G, Electrically Driven Jets. Proceedings of the Royal Society of London.
Series A, Mathematical and Physical Sciences, 1969. 313(1515): p. 453-475.

Chung, H.Y., et al., Polymer; polymer microfiber; polymer nanofiber and
applications including filter structures. 2004, Donaldson Company, Inc United
States Patent No. 6743273.

Silver, J. and J .H. Miller, Regeneration beyond the glial scar. Nat Rev Neurosci,

126

54.

55.

56.

57.

58.

59.

60.

61.

62.

63.

64.

65.

2004. 5(2): p. 146-156.

Chemousov, M.A. and DJ. Carey, Schwann cell extracellular matrix molecules
and their receptors. Histol Histopathol, 2000. 15(2): p. 593.

Abbott, N.J., L. Ronnback, and E. Hansson, Astrocyte-endothelial interactions at
the blood-brain barrier. Nat Rev Neurosci, 2006. 7(1): p. 41-53.

Deligianni, D.D., et al., Efl'ect of surface roughness of hydroxyapatite on human
bone marrow cell adhesion, proliferation, dzflerentiation and detachment strength.
Biomaterials, 2001. 22(1): p. 87-96.

Discher, D.E., P. Janmey, and Y.-l. Wang, Iissue Cells Feel and Respond to the
Stiffness of Their Substrate. Science, 2005. 310(5751): p. 1139-1143.

Ahmed, 1., et al., Three-dimensional nanofibrillar surfaces covalently modified
with tenascin-C-derived peptides enhance neuronal growth in vitro. Journal of
Biomedical Materials Research Part A, 2006. 76A(4): p. 851-860.

Deligianni, D., et al., Eflect of Surface Roughness of Hydroxyapatite on Human
Bone Marrow Cell Adhesion, Proliferation, Diflerentiation and Detachment
Strength. Biomaterials, 2001. 22: p. 2001.

Lo, C.-M., et al., Cell Movement Is Guided by the Rigidity of the Substrate.
Biophys. J ., 2000. 79(1): p. 144-152.

Arnold, M., et al., Activation of Integrin Function by Nanopatterned Adhesive
Interfaces. ChemPhysChem, 2004. 5(3): p. 383-388.

Fan, Y., Image Processing Enhancements for Scanning Probe Recognition
Microscopy, in Electrical & Computer Engineering. 2007, Michigan State
University.

Chatterjee, S. and AS. Hadi, Influential Observations, High Leverage Points, and
Outliers in Linear Regression. Statistical Science, 1986. 1(3): p. 379-393.

Chapter [4.6, in Command Reference Manual. 2003, Digital Instruments p.
381-400.

Kasa, I., A curve fitting procedure and its error analysis. IEEE Trans. Inst. Meas.,
1976. 25(1): p. 8-14.

127

66.

67.

68.

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

Keller, D., Reconstruction of STM and AFM images distorted by finite-size tips.
Surface Science, 1991. 253(1-3): p. 353-364.

Villarrubia, J .S., Morphological estimation of tip geometry for scanned probe
microscopy. Surface Science, 1994. 321(3): p. 287-300.

Villarrubia, J .S., Algorithms for scanned probe microscope image simulation,
surface reconstruction, and tip estimation. Journal of Research of the National
Institute of Standards and Technology, 1997. 102(4): p. 425-454.

Udpa, L., et al., Deconvolution of atomic force microscopy data for cellular and
molecular imaging. Signal Processing Magazine, IEEE, 2006. 23(3): p. 73-83.

Fan, Y., et al., Scanning Probe Recognition Microscopy Investigation of Cell
Elastic Properties, in 2005 APS March Meeting. 2005: Los Angeles, CA.

Butt, H.-J., B. Cappella, and M. Kappl, Force measurements with the atomic force
microscope: Technique, interpretation and applications. Surface Science Reports,
2005. 59(1-6): p. 1-152.

Cappella, B. and G Dietler, Force-distance curves by atomic force microscopy.
Surface Science Reports, 1999. 34(1-3): p. 1-104.

Hoh, J .H., W.F. Heinz, and E. A-Hassan, Force Volume, in Support Note No. 240.
1997, Digital Instruments.

Hertz, H., Uber den Kontakt elastischer Ko'per. J. Reine Angew. Mathematik,
1881.92:p.156.

A-Hassan, B., et al., Relative Microelastic Mapping of Living Cells by Atomic
Force Microscopy. Biophys. J ., 1998. 74(3): p. 1564-1578.

Schindler, M.S., Private Communication:M.S. Schindler. 2006.

Stevens, M.M. and J.H. George, Exploring and Engineering the Cell Surface
Interface. Science, 2005. 310(5751): p. 1135-1138.

Murray, J.D., Chapter 4, in Mathematical biology 11: Spatial Models and
Biomedical Applications. 2003, Springer: New York. p. 192-248.

Braga, RC. and D. Ricci, Atomic Force Microscopy: Biomedical Methods and
Applications. Methods in Molecular Biology. Vol. 242. 2003, New Jersey:

128

80.

81.

82.

83.

84.

85.

86.

Humana Press. 408.

Biyasheva, A., et al., Cascade pathway of filopodia formation downstream of
SCAR. J Cell Sci, 2004. 117(6): p. 837-848.

Garcia, R. and R. Perez, Dynamic atomic force microscopy methods. Surface
Science Reports, 2002. 47(6-8): p. 197-301.

Astrocyte. Wikipedia 2007 [cited; Available from:
http://en.wikipedia.org/wikifT issue engineering.

Meiners, S., et al., Engineering electrospun nanofibrillar surfaces for spinal cord
repair: a discussion. Polymer International, 2007. 56(11): p. 1340-1348.

Rutledge, S.L., et al., Self assembly and correlated properties of electrospun
carbon nanofibers. Diamond and Related Materials, 2006. 15(4-8): p. 1070-1074.

Bibekananda, S., et al., Electrospinning of continuous aligned polymer fibers.
Applied Physics Letters, 2004. 84(7): p. 1222-1224.

Lee, K.H., et al., Characterization of nano-structured poly([ var

epsilon] -caprolactone) nonwoven mats via electrospinning. Polymer, 2003. 44(4):
p. 1287-1294.

129

-;.'~ :::::::

         
 

M A

Milli:

T U l

Hill/H H

973 0'27

5

         

in Elli”

2

 

g