dawn: . x.
u . Marci ,
”WWW” . £1

a 9U.
.NCELQ

"ff,

 

r... .L .3
.1 . y
l . rd
MW..." . xv...t$73.

V éfizm
Jim

7 .
hl~(-~Ai

- u).
I‘M":

. s.

I i s... 2. V
3...}: .
affine ,:

Jawc...“ .
“ ...H..p_.£§ is. 55.. .333: u . ., .
a .. - .‘ u“ m ,w uésé 3.33%

 

TH“

JCM
51;). C *‘t 3-15

This is to certify that the
thesis entitled

THE VIABILITY OF CARBON NANOTUBES FOR SPACE-
RELATED TECHNOLOGY AND APPLICATIONS

presented by

SUSAN P. SONG

has been accepted towards fulfillment
of the requirements for the

MS. degree in Electrical Engineering

 

 

“7 I Major Professor's Signature
20 flfl/P/L 10 0 Lf

 

Date

MSU is an Affinnative Action/Equal Opportunity institution

 

LIBRARY
Michigan State
University

 

 

 

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

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c;/CIRC/DaIeDue.p65-p.15

THE VIABILITY OF CARBON NAN OTUBES FOR SPACE-RELATED
TECHNOLOGY AND APPLICATIONS

By

Susan P. Song

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
MASTER OF SCIENCE

Department of Electrical and Computer Engineering

2004

ABSTRACT
VIABILITY OF CARBON NANOTUBES FOR SPACE RELATED TECHNOLOGY
AND APPLICATIONS

By

Susan P. Song

Currently there is a critical need to develop non—Si based “radiation-bar ” devices and
nanotechnology for space applications. A “radiation-har ” device is a device able to
withstand a dosage up to 1000 Gray (Gy) before failing. Space technology must be
shielded from radiation encountered in space: heavy ions, protons, electrons, and
neutrons. Methods for radiation-hardening have included using silicon-on-sapphire [1—5]
or silicon-on-insulator substrates, which are however, expensive and fail after a period of
time. Silicon-based devices fail due to destruction of p-n junctions. Carbon nanotube
(CNT) based technology is not based on conventional p-n junctions with charge-
separated regions. CNT-based space technology could also bring the advantages of cost
and size reduction. At the time of this writing, only a few papers, which are theoretical
[6-7], have been published on the potential effect of radiation on CNTs. However, to the
best of this author’s knowledge, the experiments outlined in this thesis are the first such
experiments in which the effect of radiation on the properties of CNTs has been

investigated in such a well calibrated experiment.

This work is dedicated to my family and friends for all their love and support, to my
major professor Dr. Virginia Ayres, for her wonderful patience, guidance and support,
and to God, the maker of all creation and to those around me who helped in all the “little
ways”

iii

ACKNOWLEDGEMENTS
I would like to thank Dr. Barbara O’Kelly, Dr. Percy Pierre and Mrs. Linda Leon, for
their guidance and support during both my undergraduate and Master’s program; Dr.
Donnie Reinhard, Dr. Timothy Grotjohn and Dr. Timothy Hogan for their guidance and
support; Mr. Ben Jacobs, Dr. Reginald Ronningen and Dr. Albert Zeller for their
generous time and effort with the radiation portion of the project; The staff of the ECE
department: Mrs. Joyce Foley, Mrs. Sheryl Hulet, Mrs. Vanessa Mitehner, Ms. Marilyn
Shriver, Ms. Pauline Vandyke and Mrs. Margaret Conner, for their support; Dr. Corey
Collard, Dr. Mary Brake and Dr. John Holloway for their tremendous time and effort into
growth portion for this project; Dr. Ronald Gil genbach for the use of the GEC cell at the
University of Michigan; Dr. Martin Crimp for training me on the use of his TEM and
SEM, and for analyzing the TEM and SEM images; Dr. Ben Simkin, Mr. Steve Ng, Dr.
Krishna Boyapati and Dr. Liang Zeng, for their additional guidance on the SEM and
TEM; Mr. John Heckman for the HR-TEM images of University of Michigan samples
and guiding me with TEM sample preparation techniques; Dr. Melvin Schindler for
providing the electrospun carbon nanofibers. The staff of the ChEMS Department,
especially Mrs. Nancy Albright and Mrs. JoAnn Peterson, for their wonderful support;
Dr. Xudong Fan for the HR-TEM images of the NASA Goddard Space Flight Center
CNTs and University of Michigan samples; Ms. Shavesha Anderson, for showing me
how to weld CNTs; Dr. Jeannette Benavides, Mr. Harry Shaw and Dr. Jeannette Plante
for their guidance and support; NASA and United Negro College Fund Special Programs
Corporation for their financial support, without which, I would not have been able to

pursue my degree; Dr. Kendrick Curry, for his encouragement and moral support.

TABLE OF CONTENTS

Abstract. ............................................................................................... ii
Acknowledgements .................................................................................. iv
List of Figures ....................................................................................... vii
List of Tables .......................................................................................... xi
INTRODUCTION .................................................................................... 1
CHAPTER 1

PHYSICAL STRUCTURE, MECHANICAL AND ELECTRONIC PROPERTIES, AND
THEORETICAL GROWTH MECHANISMS OF CARBON NAN OTUBES

Physical Structure of a Single-wall Carbon Nanotube ....................................... 4
Physical Structure of a Multi-wall carbon Nanotube ........................................ 7
Electronic Properties of Single-wall Carbon N anotubes .................................... 7
Mechanical Properties of Carbon Nanotubes ................................................ 17
Theoretical Growth Mechanisms ............................................................. 22
CHAPTER 2
ANALYSIS METHODS
Raman Spectroscopy/Surface Enhanced Raman Spectroscopy ........................... 25
Scanning Electron Microscopy ............................................................... 30
Transmission Electron Microscopy .......................................................... 31
Atomic Force Microscopy ..................................................................... 34
CHAPTER 3
SAMPLE SYNTHESES
Production of Single-wall Carbon Nanotubes .............................................. 37
Production of Multi-wall Carbon Nanotubes at the NASA Goddard Space Flight
center ............................................................................................. 38

Production of Multi-wall Carbon Nanotubes -Si nanowire Heterostructures.... . . . . ....39

CHAPIER 4

NEW HETEROSTRUCTURE FORMATION MECHANISM
Introduction ...................................................................................... 42
Scanning Electron Microscopy Results ...................................................... 43
TEM and Selected Area Diffraction Results and Micro-Raman Results ............... 43

High-Resolution Transmission Electron Microscopy and EDS Results... ... ... .. ......44

CHAPTER 5
HEAVY—ION INTERACTIONS IN NEW NANOSCALE MATERIALS. . . . . . . . ........63

Introduction ...................................................................................... 63

Experimental Procedure ........................................................................ 63
Sample Preparation ................................................................................. 64
Pre-irradiation Calculations .................................................................... 66
Post- irradiation Calculations .................................................................. 67
RadiationExperimentResults... ...67
Radiation Damage Effects on VGCF Samples ............................................. 68
Radiation Damage Effects on ESCNF Samples... ...69
Radiation Results on SWCNT Samples ...................................................... 70
Radiation Results on Multi-wall Carbon N anotubes ....................................... 71
Summary of Heavy Ion Irradiation Results .................................................. 71
CHAPTER 6
DISCUSSION ....................................................................................... 86
APPENDD( A ....................................................................................... 88
APPENDIX B ....................................................................................... 91
BIBLIOGRAPHY ................................................................................... 93

vi

LIST OF FIGURES

Figure 1 .............................................................................................. 10
Regions Where Heavy Ion and Other Forms of Radiation Are Encountered in Low-Earth

Orbit

Figure 2 ............................................................................................... 11

AFM, TEM, AND HR-TEM Images of Carbon Nanotubes

Figure 3 ............................................................................................... 12
HR TEM Image of a Typical Multi-wall CNT Showing

Figure 4.. ..................................................................................................................... 13
Definition of Chiral (Ch), Translational (T) and Symmetry (R) Vectors

Figure 5 ............................................................................................... 14
The Different Types of Chiral Structures of Carbon Nanotubes

Figure 6..... ................................................................................................................. 15
Relationship between Wave Vectors in the Carbon N anotube

Figure 7 ............................................................................................... 16
Indices vs. Conductivity in Carbon Nanotubes

Figure 8 ............................................................................................... 21
Peierl’s Distortion

Figure 9 ............................................................................................... 24
Formation Mechanisms of Endcaps in Carbon Nanotubes

Figure 10 .............................................................................................. 28
Raw Raman Spectra Taken at 488 nm Deconvoluted Tangential Breathing Modes for
SWCNTs

Figure 11 ............................................................................................. 29
Schematic of Raman System

Figure 12 ............................................................................................. 48
SEMs Taken at 0° Tilt of Five Areas on Wafer on Which CNT-Si Nanowire
Heterostructures Formed in ICP Plasma

Figure 13 ............................................................................................. 49
Close-up SEMs of CNT-Si nanowire Heterostructures after 5 Hours

vii

Figure 14 ............................................................................................. 50
TEM Images of Carbon Structures Formed after 1, 3 and 5 Hours in ICP plasma

Figure 15 ............................................................................................. 51
TEM Image and Corresponding Selected Area Diffraction Pattern of a Carbon N anotube
Produced in ICP Plasma

Figure 16 ............................................................................................. 52
Micro-Raman Spectra Wafer Placed in ICP Plasma for 5 Hours, Indicating Well-
Graphitized Structures

Figure 17 ............................................................................................. 53
TEM Images of Tapered Nanostructures from 5-Hour Sample

Figure 18 ............................................................................................. 54
TEM Images comparing Lengths and Base Widths of Tapered N anostructures Produced
after Appx. (a) 1 Hour, (b) 3 Hours and (c) 5 Hours in ICP Plasma

Figure19 .............................................................................................. 55
Selected Area Diffraction Verified that the Nanostructures were Crystalline Silicon
(From 5 Hour Sample)

Figure 20 .............................................................................................. 56
Dark Contrast Dots at the Tips and Along the Sides of the Tapered Silicon Nanowires
Were Shown to be Iron.

Figure 21 ............................................................................................. 57
HR-TEM Images of CNTs and Vapor Grown Carbon fibers Produced after 5 Hours in
Inductively Coupled Plasma

Figure 22 (a)-—(b) .................................................................................... 58
HR-TEM Images of Silicon Components of CNT-Si nanowire Heterostructures and Iron
Nanodots

Figure 22 (c) ......................................................................................... 59
EDS Spectrum of Iron Nanodots
Figure 22 (d) ......................................................................................... 60

EDS Spectrum of Silicon Nanowires

viii

Figure 23 (a)-(b) ..................................................................................... 61
HR-TEM Images of an Intact CNT-Silicon Nanowire Heterojunction

Figure 23 (c) .......................................................................................... 62
EDS Spectrum of an Intact CNT-Silicon Nanowire Heterojunction

Figure 24 ............................................................................................. 73
SEM Images of Vapor Grown Carbon Fibers

Figure 25 ............................................................................................. 74
Damage Mechanisms in Graphite Due to Radiation

Figure 26 ............................................................................................. 75
Raman Spectra of Well-Graphitized Fibers Prior to Irradiation

Figure 27 ............................................................................................. 76
AFM Images of Electrospun Carbon Nanofibers

Figure 28 (a) .......................................................................................... 77
AFM Phase Image of Tubes@Rice Prior to Irradiation

Figure 28 (b) ......................................................................................... 78
Raman Spectra of Tubes@Rice Prior to Irradiation

Figure 29 .............................................................................................. 79
Raman Spectrum of Tubes@Rice After a Dose of 340 Gy

Figure 30 ............................................................................................. 8O
HR-TEM Images of Multi-wall CNTs after 90 seconds

Figure 31 ............................................................................................. 81
FE SEMs of Tubes@Rice Prior to Irradiation

Figure 32 .............................................................................................. 82
FE SEM Images of MWCNTs grown at the NASA Goddard Space Flight Center taken
after irradiation.

Figure 33 ............................................................................................. 83

HR-TEM Images of NASA Goddard Space Flight Center Multi-wall CNTs Showing
Distinct Wall Layers

Figure 34 ............................................................................................. 84
Electrospinning Setup

ix

Figure 35 ............................................................................................. 85
HR-TEM Images of Electrospun Carbon Nanofibers Showing Progressive Damage from
0- 900 seconds

LIST OF TABLES

Table 1 ................................................................................................ 67
Calculations Performed to Determine Irradiation Times

Table 2 ................................................................................................ 67
Calculations Performed to Determine Actual Doses

xi

INTRODUCTION
This work described in this thesis investigates the effect of space radiation, specifically at
low-earth orbit (LEO), on the properties of carbon nanotubes (CNTs) in order to
determine their viability for space applications. LEO is the region ranging from 10 km to
1000 km above the earth. Potential space applications of carbon nanotubes include:
self-repairing materials for the skins of shuttles or space suits, new space-resilient

electronics, flat-panel displays and even space elevators.

Heavy ions are a significant source of radiation and are encountered in LEO at high
declination. The heavy-ion, and trapped particle regions are illustrated in Figure 1.
Current transistor-based electronics technology requires radiation shielding against heavy
ions, which are massive, charged particles. The amount of shielding required for
effective defense against heavy ions would be mass-prohibitive. Radiation effects are
classified into non—destructive errors, which can be overcome by device reset and
destructive errors, which cause permanent device malfunction. Examples of radiation
effects include: Single Event Upset (SEU), a change of state or transient induced by an
ionizing particle such as a cosmic ray or proton in a device; Single Event Functional
interrupt (SEFD/Single Event Transient (SET), a condition where the device stops
operating in its normal mode, and usually requires a power reset or other special
sequence to resume normal operations; Single Hard Error (SHE), an SEU which causes a
permanent change to the operation of a device; Single Event Latchup (SEL), a
potentially destructive condition involving parasitic circuit elements forming a silicon

controlled rectifier (SCR); Single Event Gate Rupture (SEGR), the burnout of a gate

insulator in a power MOSFET and a destructive condition; Lastly, Single Event Burnout
(SEB) which is a highly localized burnout of the drain-source in power MOSFETs, which
is also a destructive condition. All the previously mentioned damage mechanisms are

serious problems for current transistor-based electronics in space applications.

Since their discovery over a decade ago by lijima [8], CNTs have spawned a wave of
research into their potential applications. NASA is particularly interested in the
development of CNT-based electronics and textiles which are space-qualifiable.
Investigations into the potential use of CNTs as cold cathode sources have also been

conducted [9-11].

CNTs are essentially self—closed sheets of graphite. CNTs are either single-wall or
multi-wall. Single-wall CNTs can be either separate or bundled in a rope containing
twenty to fifty individual nanotubes in a triangular lattice (hexagonal close-packed
stacking); the latter is shown in Figure 2 (a)-(b). Individual single-wall CNTs are
generally on the order of 0.7-2 nm in diameter and several microns in length. The mean
diameters of sin gle—wall CNTs (both individual and bundled), and the inter-tube distances
for bundled CNTs, are dependent on the catalyst used and the growth conditions.
Individual, separate single—wall CNTs, to best of this author’s knowledge, have been
synthesized only in the presence of any one of the transition metals such as iron, nickel,

copper or cobalt or lanthanides, such as lanthanum, yttrium, gadolinium and neodymium.

l0

Multi-wall CNTs consist of several layers, with the diameter of the outermost wall
ranging from 20 to 30 nm, are also generally up to several microns in length. An example
of a multi-wall CNT of about 15 layers is shown in Figure 3. Various unique forms,

curly or Y-junction, of CNTs have been reported as well [12].

In the course of the research presented in this thesis, first time studies of heavy ion
interaction in carbon nanotubes and other new nanoscale materials have been conducted.
Our results show high radiation resilience for single and multi—wall CNTs, which indicate

that carbon nanotubes may have good potential for space applications.

The organization of the thesis is as follows: In Chapter 1, we review the basic properties
of CNTs. In Chapter 2, the main experimental techniques used in the research are
described. In Chapter 3, synthesis of the carbon nanotubes investigated in this thesis is
discussed. The synthesis research included system components at Michigan State
University, at the University of Michigan, and at the NASA Goddard Space Flight
Center. Parts of the research were done in collaboration with these two institutions. The
heavy ion radiation experiments and their results are described in Chapter 5. Our

conclusions are discussed in Chapter 6.

CHAPTER 1
PHYSICAL STRUCTURE AND IDEAL ELECTRONIC AND MECHANICAL
PROPERTIES OF CARBON NANOTUBES

Ideal properties of carbon nanotubes (CNTs), for better understanding of their
applications potential, and also for understanding of what a radiation-induced deviation
from ideal would entail, are described in this Chapter. Theoretical growth mechanisms
are briefly discussed, since part of the research presented in this thesis involved the

growth of multi-wall carbon nanotubes and nanotube-nanowire heterostructures.

An excellent reference for theoretical and some experimental properties of carbon
nanotubes is “Physical Properties of Carbon Nanotubes”, by R. Saito, G. Dresselhaus,
and MS. Dresselhaus, published by Imperial College Press (1998). This reference is

extensively quoted in this section.

Physical structure of a single-wall CNT

A single—wall CNT is a single-layer seamless cylindrical graphene tube. There are three
key vectors that define the structure of a single-wall CNT: the chiral (which also defines
the “handedness” of a CNT, i.e., whether the nanotube is “rolled” clockwise or counter-

clockwise), translational and symmetry vectors. These vectors are shown in Figure 4.

The chiral vector is denoted as 23—},— , and is represented by 5A in Figure 3. The vector

0A is a unit vector in the direction perpendicular to the nanotube axis, or equator, of the

nanotube and OB is in the direction of the nanotube axis. T , the translational vector, is a
unit vector in the direction of ‘0—B in the Figure, and-R , in the direction of a in Figure

4, is the symmetry vector. The chiral vector,-C; is defined by
C—h-=n;1—+mzaa(n,m) (1)
where n and m are integers, aand a—E are the unit vectors in real space of the CNT, and

OSImISn.

The unit cell of a single-wall CNT is given by rectangle formed by the translation vector,
T and the chiral vector, Ch- shown in Figure 4. Single-wall CNTs can be chiral (non-
symmorphic) or achiral (symmorphic). Achiral single-wall CNTs, have a mirror plane,
whereas chiral or non-symmorphic, nanotubes have no mirror plane. There are two types
of achiral single-wall nanotubes: armchair and zigzag (Figure 4). The names armchair
and zigzag are derived from the shape of the cross-section formed when a nanotube is

“cut” along its equator. Each of the end caps of the CNTs contains six pentagons and an

appropriate number and location of hexagons. Additionally, the armchair nanotube

corresponds to the condition when nzm, i.e., when C7: (n,n) and the zigzag nanotube

corresponds to the case when m=0, or C: =(n,0). All other (n,m) chiral vectors

correspond to chiral nanotubes.
The chiral angle, 0, determines the angle at which the graphitic sheet is “rolled up” to

make the CNT. It is defined by the following relationship:

cost9=£:h_—._—E-l— (2)
lChla]

 

The diameter of a single—wall CNT, (It, is given by

d,=%,whereL=|al=‘/C;0C;=a\/n2+m2+nm (3)

where n and m are the indices of the CNT, L is the length of the nanotube and

 

a] -a] =02 ~a2 :02. For the CNT, a=.144 nm.

The translational vector is defined to be the unit vector of a 1D CNT and is parallel to

the nanotube axis. The translational vector is defined by:

T : I121— +i§21§ 2 (11,12) , where t] and 12 are integers (4)
Finally, the symmetry vector R is described by the following,

E=p3i+p332mq1 (5)
where p and q are integers and do not have a common divisor except unity. The
symmetry vector R is defined as the site vector having the smallest component in the

direction of CI.

The brillouin zone of a single-wall CNT is the same as the brillouin zone of the single

graphene sheet. It is defined by two unit vectors b1 and 192, where

_?-_7r_*_1__ £151.-
bl—a (fifiy), b2 (1(ng y)

and a: J3 * the bond length between carbon atoms. This is a standard derivation of the

reciprocal space unit vectors b, and b2 from the real space unit vectors.

3 1 3 I
:a*—x+— , a =a—x-—’
0] (2 2y) 2 (2 2.1)

and
a—i OI)? 5 27:5,]

but it is implemented within the appropriate 2 — D space. The real and reciprocal space

unit vectors are shown in Figure 6 (a)-6 (b).

Physical structure of a Multi-wall CNT

Multi-wall CNTs are usually formed by ten to twelve concentric graphene layers. In
general, each of the individual concentric layers will keep their respective electronic
properties and not be commensurate, i.e., the carbon atoms of each of the comprising
nanotubes will not be aligned with each other. The interlayer spacing is slightly less than

that of graphite, being 0.34 nm.

Electronic properties of Single-wall CNTs

The diameter and chirality of single-wall CNTs determine whether they are metallic or
semi-conducting. Two-thirds of all CNTs are semi-conducting, while the other third are
metallic [l3]. Semi-conducting CNTs have a band gap between of about 1 eV. More
specifically, the amt-chair nanotubes, denoted by (n,n) are always metallic and the zigzag
nanotubes, denoted by (n,0) are metallic when n is a multiple of 3, semi—conducting
otherwise. This results in two categories for metallic CNTs. Metallic zigzag CNTs are
all in the metal—1 category. The arm-chair single-wall CNTs are in the metal-2 category.
Figure 6 shows the indices of the CNTs in relation to their conductivity and illustrates

how the indices are determined.

The relationship between the diameter and chirality of single-wall CNTs and their

electronic properties is as follows. E]— is the reciprocal lattice vector in the direction of
the chiral vector,C; , of a single-wall CNT. The condition for metallic energy bands is

that the ratio of the length vector R to that of K—l is an integer. I? = (2n+m)/3 x-K_1 ,

where n and m are integers that represent the number of carbon atoms in the same
corresponding positions in the hexagonal rings of carbon which are crossed horizontally
and downwards, respectively, to form the chiral vector on the honeycomb-like lattice of

an unrolled CNT. The wave vectors are shown in Figure 6 (c).

Another important property of CNTs (both multi-wall and single-wall) is ballistic or
quasi—ballistic transport. Ballistic transport refers to single electron or quantum transport,

in which there is no scattering of electrons in the nanotube. Mathematically the quantum

conductance, G0, can be expressed as

Go = — (6)
where e is the charge of an electron and h is Planck’s constant, and the factor of two in

the expression comes from the fact that single wall nanotubes have two conducting

channels. Ballistic transport represents the special case when L<< LW<<Lm, where L is

the length of the wire, Lm is the momentum relaxation length and Lw is the phase

relaxation length. Momentum relaxation refers to the average distance that an electron

wave function travels in a conductor or semiconductor before its momentum is

randomized due to a collision, and phase relaxation refers to the average distance that an
electron wave travels in a conductor or semi-conductor before its phase is randomized
due to scattering. In other words, nanotubes exhibit ballistic transport because of their
extremely small size, and in ballistic transport there is little or no scattering which results

in very high current flow.

Semiconducting CNTs have been successfully doped p-type and n-type. Hence, diodes,
which are a junction between p and n-type semiconductors, have been developed using
crossed CNTs [14]. However, the role of the p-n doping versus the effect of the contacts

in affecting the diode action is currently under investigation.

.cfioouamm 9558 5:253

:mE .560 32 ..8 E0395 aflsofiam a 2m >05.
.35. 20: 28:92: Eon 2: mac—a E 0E8
mane 2:58 282% Eoc 22 Exam 3: oEmE

  

\.

2.2 .93: 2.3. bio:

Sud .35 saw 3.: Risa... 6m; .35 £5 25..
..9; a :20

:EO Eamis E uoaoucsooco 2a
cosflcmm Lo Eton .550 was =0T>>mom 22;? 2.23% ”2 ocsmi

35 .x E 59...? .2; .353 5.22 x32}? 5;: _.

v.5: >23: washes: ....EzEL

83:55 3:95.

 

10

 

0 500 nm

Data type Phase
Z range 180 deg
(a) (b)

Figure 2: Examples of ropes of single-wall carbon nanotubes (a) Tapping Mode AFM
image of ropes dispersed on a gold substrate. Each “strand” is actually 15-20 close-
packed single-wall carbon nanotubes. (b) Cross sectional HRTEM from the Smalley
group photo gallery showing how the individual tubes are arranged within a typical rope
(http://smalley.1ice.edu/index.php?topgroupid=1&groupid=10).

ll

"J TIJ'I'I

 

Figure 3: HR TEM Image of a typical Multi-wall CNT showing individual layers

 

p. -
055 25.56‘939
o'c‘o'o". *0

“2

Figure 4: Definition of the Chiral (Ch), Translational ('T) and Symmetry (R) Vectors

Vectors OA and OB define the chiral vector (Ch) and the translational vector (T),

respectively. R is the symmetry vector. The Figure corresponds to Ch=(4,2), T=(4,-5)
and R=(l, -l). The rectangle OABB’ defines the unit cell of the nanotube. (Taken from
R. Saito, G. Dresselhaus and MS. Dresselhaus, “Physical Properties of CNTs”, Imperial
College Press, 1998)

13

 

       
    

    

.I C. O
.... -.
’° 9:":

  

 
 

   
 

‘2‘? ° 5.

t. .o -
.0 0“. .6 .0

 

   

Figure 5: The Different Types of Chiral Structures of Carbon Nanotubes

(a) Armchair nanotube (carbon atoms are perpendicular to the nanotube axis) (b) Zigzag
nanotube (c) chiral nanotube (Taken from R. Saito, G. Dresselhaus and MS. Dresselhaus,
“Physical Properties of Carbon Nanotubes”, Imperial College Press, 1998)

 

K2

 

Figure 6: Relationship between Wave Vectors in a Carbon Nanotube. (Taken
from R. Saito, G. Dresselhaus and MS. Dresselhaus, “Physical Properties of
Carbon Nanotubes”, Imperial College Press, 1998)

15

 

Figure 7: Indices vs. Conductivity for Carbon N anotubes.
(Taken from R. Saito, G. Dresselhaus and MS. Dresselhaus, “Physical Properties
of Carbon Nanotubes”, Imperial College Press, 1998)

16

Mechanical Progrties of CNTs

The Young’s modulus of a material defines the stiffness or strength of a material. The
Young’s modulus of single-wall CNTs was reported to be as high as a few TPa using
innovative in-situ methods in TEM. Multi-wall CNTs have a high young’s modulus, or
elastic modulus, which is the ratio of stress over strain, of several Gigapascals (GPa) in
the direction of the tube axis. The TEM method [15] of measuring Young’s modulus of
CNTs utilizes thermal energy to cause the nanotubes to vibrate. More specifically, the
amplitude of the thermal oscillations is measured using TEM by utilizing several images

taken using a CCD camera. Then, treating the nanotube as a cylindrical cantilever, the
vibration energy W, is related to the effective spring constant k by W = gkug. Then

since the effective spring constant is a function of the Young’s modulus, one can simply
solve for the Young’s modulus. The Young’s modulus is a function of a CNT’s diameter

and helicity.

Another measure of strength of a material is the bending modulus. The bending modulus
is a measure of how much a material can be bent before it breaks. In another set of
experiments in the TEM, by using a frequency variable oscillating voltage supply, the

resonant frequency was determined. This is because resonance is selective based on the
nanotube’s dimensions. Then the resonant frequency, fl, of the CNT was used to

determine the bending modulus because it is dependent on the Bending modulus, EB, and

is given by:

 

 

.2 D2+ 2E
,B_I_\/( 01);; (7)

.:__'_
f' 87rL2 p

17

where D is the outer tube diameter, D] is the inner diameter and L is the length of the
nanotube. These parameters can be determined from a TEM image. The other
parameters are constants. The density of carbon is given by p, [31:].875 and B2=4.694,

for the first and second harmonics.

After a bit of rearranging, one finds that the Bending modulus, EB, as a function of the

resonant frequency and other parameters is given by the following equation:

= fr2647r2L4p

 

15,3 (8)

Hence, the once the resonant frequency, fi, of the CNT is found, one can determine the

bending modulus.

A third measure of strength is the strain energy. The strain energy, E0 of the single-wall
CNT, when considered as a graphene sheet, is given by:

3
z rzETdf
" 6d,

 

(9)

d, is the diameter of the nanotube, T is the length of the CNT per 1D unit cell in the
direction of the nanotube axis and E0 is the Young’s or Elastic modulus of the sheet.

Carbon is very unique in that it allows for hybridization of its a and 1t orbitals. The great
strength of CNTs in the direction of the nanotube axis comes from first, the strength of

the double bond between the carbon atoms due to the o bonding, 1t bonding and lastly,

l8

weak interlayer interactions. The strain energy experienced by a CNT increases with
decreasing diameter of the CNT. CNTs are very flexible in the direction perpendicular to

their axis.

CNTs also have the ability to store elastic energy, which is the potential energy stored
when a material is deformed. The first contribution to elastic energy in CNTs is due to

the o electrons. The second contribution comes from it electrons.

CNTs, in comparison with trans-polyacetylene, experience only a limited Peierl’s
distortion. Peierl’s distortion refers to the tendency of a double-single-bond combination
to freeze into a series of quantum wells and barriers rather than hybridize into a
delocalized electronic state, as shown in Figure 7. Peierl’s distortion causes a metal-
insulator transition, which combines a mechanical effect with an electronic outcome.
Peierl’s distortion has been a serious limitation in attempts to manufacture carbon-based
molecular electronics, such as polyacetylene-based molecular wires. However, it does
not look to be a serious issue in CNTs. In CNTs, the Peierl’s distortion is applicable only
to metals and hence to only one third of the CNTs, which are metallic in nature. Energy
analyses indicate that, though the metallic energy bands are unstable under the Peierl’s
distortion, the induced energy gap is very small, < 10'3 eV [13]. A distortion of 10'3 eV
would not be significant in a 0.5-] eV bandgap measurement, and even if observed could
be due to other causes. No serious Peierl’s distortion problems have been reported in the

literature to date.

19

Thus, CNTs have many exciting potential applications due to their excellent mechanical
properties, namely high Young’s modulus, high aspect ratio, and considerable strength
and flexibility. Multi-wall CNTs are already being used as Scanning Probe Microscopy
tips [16] and could possibly be used as linear bearings [17], and coaxial conducting wires

or capacitors. The formation of CNTs into space elevators [18] has also been theorized.

These excellent electronic and mechanical properties of CNTs have spurred research into
other potential applications such as a battery [19-20], a flow sensor and a beam source for
TEM [21 ~23]. Additionally, their high aspect ratio inspired research into their potential
use as field—emission displays, resulting recently in a highly compact CNT-based x-ray

device for portable medical applications [24].

20

(a)

(b)

 

Figure 8: PEIRELS DISTORTION (a) TRANS-POLYACETYLENE experiences a large
Peierl’s distortion while (b) Benzene is almost completely delocalized. CNTs exhibit only

a small Peierl’s distortion.

2]

Theoretical Growth Mechanism of CNTs
The axial growth mechanism (length-addition) of CNTs had been under debate for many
years. The two most widely debated growth models for the growth of CNTs were the

closed-growth and the open growth model. The closed growth model assumes low

growth temperature (~1100°C) and that the nanotube is always capped. According to this

model, growth occurs through the absorption of a C2 dimer near a pentagon in the cap of
the nanotube. In the open growth model, a high temperature is assumed, and growth is
also achieved by the addition of a C2 dimer. Capping to an open end occurs eventually

due to the formation of a heptagon-pentagon defect pair, as illustrated in Figure 8. These

heptagon-pentagon pair defects are also known as Stone-Wales defects.

Some support for the open growth method comes from the direct observation of open-end
CNTs [25]. Another piece of evidence which supports this model is the observation of
CNTs inside larger diameter nanotubes. CNT encapsulated C60 fullerenes (“bucky
balls”) have also been produced by pulsed laser vaporization, the majority of which
involve 1.4 nm diameter CNTs, possibly due to the favorable van der Waals separation of
0.3 nm [26]. There has been no experimental validation of the closed growth model. For
both of these carbon addition mechanisms, a metallic catalyst need not be present. A
catalyst was not used for the multi-wall CNTs from the NASA Goddard Space Flight

Center.

Other proposed growth models included the polyyne ring nucleus model [27] and the

Liquid-Metal-Solid (LMS) model [28]. To date, these methods rely on the presence of a

22

metallic catalyst. In the polyyne ring nucleus model, growth occurs when small
monocyclic rings (C10-C60) join to form polycyclic rings and convert to either fullerenes
or back to monocyclic rings in the presence of cobalt (Co) in combination with one of
three other catalysts: sulfur (S), bismuth (Bi) or lead (Pb). In the presence of cobalt, the
other catalysts sulfur, bismuth or lead tend to increase the yield of single-wall CNTs and

modify the diameter distribution [29]

The LMS model has been verified but only applies to the specific case where a catalyst
was deposited on a substrate and growth was conducted using plasma. In the LMS
model, essentially the carbon is absorbed into the catalyst and then upon saturation the
carbon is extruded in the form of a nanofiber or nanotube. We have evidence that
suggests that the carbon components of the CNT-Si nanowire heterostructures produced
by the experiments conducted at the University of Michigan were grown according to an

LMS model.

23

 

Figure 9: Formation Mechanism of End Caps in Carbon Nanotubes

Show are the formation of a disclination in carbon nanotubes via the a) addition of a
pentagon, b) removal of shaded hexagons, c) addition of a heptagon, highlighted in red
and (1) addition of two pentagons and a heptagon. Arrows are drawn to allow for easier
identification (Taken from MS. Dresselhaus, G. Dresselhaus, A. Jorio et al, “Raman
spectroscopy on isolated single wall carbon nanotube”, Carbon Vol. 40, pp. 2043-2061
(2002)

24

CHAPTER 2
ANALYSIS METHODS
The CNT-silicon nanowire heterojunctions and the CNTs pre- and post-irradiation were
investigated using four methods: (1) Micro-Raman spectroscopy and Surface—Enhanced
Raman spectroscopy, (2) scanning electron microscopy (SEM), (3) transmission electron
microscopy (TEM) and high-resolution transmission electron microscopy, both with
selected area diffraction, and (4) atomic force microscopy (AFM). These four methods

will now be discussed in further detail.

Raman Spectroscopy/Surface Ephanced Raman Spectroscopy

Raman spectroscopy reveals the bonding structure and the degree of purity of a volume
of a material equal to the spot size times the penetration depth of the laser beam. There
are two methods used in Raman spectroscopy: Anti-Stokes and Stokes. For this project,
the Stoke’s method was used, i.e., a laser beam is incident on a material, induces lattice
vibrations, or phonons, at certain frequencies, and causes the photon from the laser beam
to lose energy. This loss of energy to the sample causes a shift in the wavelength of the
photon. This shift is then detected by the monochromator. This results in peaks at
characteristic shifts away from the original wavelength of the beam. Sharp peaks indicate
a highly ordered material, whereas broad peaks are due to amorphous material. This
allows us to differentiate between different morphologies of carbon, such as SWCNTs

and MW CNTs.

25

The Micro-Raman system used for this project is shown in Figure 10. It consists of an
Argon-ion laser from Coherent, an optical microscope, an Electron Tubes 9124B photo-
multiplier tube and a Jarrell-Ash 25-100 series double Czerny-Tumer monochromator set
up in back-scattering configuration. The Raman system and the software for controlling
the Raman system were designed and built by Joerg Mossbrucker for his Master’s thesis
project [30]. The wavelength at which the CNTs were examined was 514.5 nm. The
Raman spectra were deconvolved with Lorentzian peak amplitudes using PeakFit®
software. For the Raman investigations the photomultiplier tube was air-cooled, and the
room lights were turned off, to reduce noise. The samples were placed on the microscope
stage and the laser beam was focused to a spot size of approximately 30 microns using

the 60x objective.

For the Tubes@Rice single-wall nanotubes, at the 514.5 nm wavelength the two strongest
tangential modes occur at the Raman shifts of 1567 cm'] and 1591 cm']. The peaks occur

at the same Raman shifts at the 488 nm wavelength. However, the peaks have a smaller

full width at half maximum (FW HM) and are sharper. The D—band or “disorder peak” for

CNTs and planar graphite occurs at 1350 cm.1 and is a broad peak. For silicon, at the

Raman peak occurs at 523.286 cm'] and does not overlap with the carbon peaks in any

experiments where both might be present. Multi-wall peaks have 3 Raman shift at about

1580 cm".

The radial breathing modes of CNTs occur at 248/dt cm'], where d, is the diameter of the

nanotube [13]. This spectral information was unfortunately available to us due to the

26

overlap of the high brightness laser line for the 514.5 and 488 nm wavelengths. An
attempt to analyze the breathing modes using the HeNe laser was also unsuccessful. 633
nm (red light) has a stronger coupling to carbon sp2 bonds than either 514.5 nm or 488
nm wavelengths. However we found that our PMT lacked the requisite sensitivity in the

red range.

Surface-Enhanced Raman Spectroscopy employs the use of metallic nano-particles
(colloidal silver or gold) to enhance the response of the sample to the laser excitation.
The enhancement is primarily due to large local electromagnetic fields which come from
resonant optical excitation of surface plasmons (packets of electrons) [31] That is,
energy transfer between photons and the photons which occurs only when the wavelength
where the quantum energy carried by the photons exactly equals the quantum energy

level of the plasmons.

27

 

Buckey Paper. 488 nm, 7124/00

3500 .,_

,
O
O
O
(V)

2500 i, -

O
O
O

500

(we) MSILM

1000

28

 

Ramon ShIIt (cm ")

 

Figure 10: Raw Raman Spectra Taken at 488 nm Showing Tangential Modes for Tubes@Rice

 

mlflOl’ 2

 

1 r— Ar-ion Laser
I___

entrance slit
_ lens
I 7/1/ H / lens
4 I 5?

 

 

 

 

 

 

0')

 

 

 

 

 

 

 

 

 

 

 

 

 

 

S 3
O
3
o 3
3'
3 1
3 .
3 .
-— o mirror /

beam splitter

 

sample

\

 

 

 

 

 

 

 

—

Figure 1]: Schematic of Raman System in Backseattering Configuration: Numbers 1-6
represent the order in which the light traverses the beam path

29

Scanning Electron Microscopy

Scanning electron microscopy (SEM) allows surface analysis of materials at the micron
to nanometer scale. For the scanning electron microscope used in this project, the
electron beam is generated by an electron gun, guided in a raster pattern down the column
by two scanning coils in the objective lens and the electrons hit the atoms in the sample,
resulting in various types of scattered electrons. They are secondary, back-scattered, and
Auger electrons. Secondary electrons are low energy electrons excited from 100

nanometers or less within the sample, back—scattered electrons are hi gh-energy electrons
which are reflected by an angle greater the 90°. These scattered electrons are then

detected by an Everhart-Thomley detector. The front of the detector is a Faraday cage.
Secondary electrons, once inside the Faraday cage, are then attracted to the scintillator
which is the second component of the detector, across which a 12 kV voltage is applied.
The Scanning Electron Microscope was useful for determining the distribution of the

CNTs in the Si-CNT heterostructure sample.

The Scanning Electron Microscopes used for this project were the CamScan 44FE Field
Emission Scanning Electron Microscope and the Hitachi S-47001] Field Emission
Scanning Electron Microscope. A field-emission scanning electron microscope differs
from a regular scanning electron microscope in that emission from a sharp-pointed
emitter occurs through application of a high electric field. Generally field—emission
scanning electron microscopes allow for higher resolution and induce less damage to a

sample.

30

High-Resolution Transmission Electron Microscopy/Transmission Electron

Max
In TEM, an electron beam is generated by an electron gun. For the Hitachi H-800

(E=200 KeV, 7t=.00251 nm), the electron gun is thermionic, i.e., the electron beam is
produced by the resistive (thermionic) heating of a tungsten filament. The electrons are
sent down a column, hit atoms in the sample, whereby the electrons are scattered in a
variety of ways (diffraction, reflection, and backscattering). Forward scattered electrons
are the primary source of signals used in the TEM. High-Resolution transmission
electron microscopy (HR-TEM) is used to analyze the atomic layer structure of nanoscale
materials. HR-TEM is therefore valuable in the study of CNTs because it reveals their
structure in atomic scale, their crystallinity and various mechanical properties.
Transmission electron microscopy (TEM) provides structural information with resolution
equal to that of Atomic force microscopy (AFM). HR-TEM has higher resolution than

either AFM or TEM to atomic plane spacings.

Contrast in a TEM/HR-TEM is in three forms: amplitude, diffraction and phase contrast.
Amplitude contrast occurs when using a small objective aperture, excluding all electrons
except those of either the transmitted electron beam or the diffracted beam. Diffraction
contrast occurs when you have a two beam condition, so that you have only the main
beam and one strongly diffracted beam. Phase contrast occurs due to differences in the
phases of the electron waves scattered from a sample. A phase contrast image is taken
with a large objective aperture. Regions of the sample with higher mass density or which

are strongly biased will scatter more electrons, and therefore appear dark in the image

31

[32]. How much information in reciprocal space is transferred to the image is dependent
on H(u), the contrast transfer function. H(u) then is the product of the aperture function,

envelope function and the aberration function. The phase distortion function is x(u).

x(u)= gCS/l3u4 + II/lAfu2 , where C, is the spherical aberration coefficient, A is the

wavelength of the electron, 4f is the defocus value, and u is the magnitude of the unit
vectors in reciprocal space. x(u) is valuable because it allows one to determine the effect

of spherical aberration, acceleration voltage and defocus value on the resolution.

In the case of the CNT-Si nanowire heterostructures, the amplitude contrast allowed
ready differentiation between the silicon, iron and carbon components. The accuracy of
the measurement of CNTs using TEM varies with the diameter of the tube. It has been
reported that in general, based on simulations of TEM images, for CNTs with diameter
larger than 1.0 nm in diameter, the estimate of the tube diameter agrees to within 10% of
the value, whereas for those sub nanometer in diameter, the difference could be as large

as 30% [32].

The CNTs were prepared for TEM and HR—TEM by ultrasonicating in ethanol, and then
pipetting onto 3 mm diameter mesh copper grids covered with holey carbon film. The
holey carbon film was prepared according to the procedure outlined in Appendix A. The
primary benefits of holey carbon film are that it is stable under the electron beam,
prevents charging of the material being investigated and supports more material than the

copper grid alone.

32

Calculation of diffraction patterns for CNTs

Selected area diffraction was an important tool for our analyses. The diffraction pattern
in general reveals the crystallinity and the identity of a material because the diffraction
patterns for materials differ significantly for amorphous, polycrystalline and crystalline
materials. Amorphous materials produce rings for the diffraction pattern, due to the
random orientation of the atoms, polycrystalline materials produce speckled rings, and
crystalline materials produce spots. If the crystal does contain defects, then streaks are
produced. One can calculate the relevant parameters of interest, such as the lattice

parameter and symmetry of a crystalline material using the diffraction pattern.

As previously mentioned, the brillouin zone of the CNT is comprised of rectilinear
segments inside the hexagon of carbon atoms. The reciprocal lattice vectors, b1 and b;

for the CNT, are

The various lattice spacings, d, for different orientations of a CNT or any crystal, can be
found via the simple equation Rdsz, where R is the distance between diffraction spots
or rings on a TEM diffraction pattern, L is the camera length, and I. is the wavelength of

the electron. For CNTs, HR-TEM reveals whether they are multi-wall or single-wall.

The expected diffraction pattern for CNTs of different indices can be calculated by using

the following equation given in a previous report [33]:

33

A(k) = f(k)Zexplik - RjJE 2A....(k)
j n,m

(11)
A(k) = f(k)z kZ — 27r[%+ 2H!" (klr)exp[i[n[y/k — (0+ 1;] + kzzfl
P

n,m

where f( k) is the carbon structure factor of the electron or X-ray, Rj are the atomic

positions along the helix, (n,m) are two integers representing the indices of the CNT,

_ k
k = (kx,ky,kz) is the wave vector transfer, kt = ,jk? +k§ , Wk 2 tan-{El}, 1,. is the
z

nth order cylindrical Bessel function (r,(p, z) are cylindrical coordinates of the “initial”

atom of the helix of radius r, pitch P and atomic repeat distance p in the z-direction.

The one disadvantage posed in using TEM and HR-TEM is that although they allow for
highly detailed investigation of the structure of a material, due to the minute amount of
the sample examined, it does not allow one to determine the uniformity of properties of

the majority of a material.

At_o_r_nic Force Microsw

Atomic force micrOSCOpy (AFM) is a form of scanning probe microscopy (SPM), which
provides structural information with tens of nanometers scale resolution. In AFM, a
sharp tip made of SiN or silicon, which is attached to a cantilever, is brought close
enough to the sample so that the atomic orbitals of the tip interact with those of the

sample. The tip is raster scanned across the sample by a piezoelectric scanner. A laser

34

beam is reflected off the back of the cantilever which holds the tip and detected by a four-
cell photodiode. Therefore, when there is a change in the position of the tip, there is a
change in the strength of the signal being detected by the photodiode quadrants. The
topographical image is formed via the stored vertical position of the piezoelectric scanner
at each (x,y) point which was required to maintain either a constant height or a constant
deflection or a constant oscillation amplitude near the cantilever resonant frequency
(Tapping Mode, discussed below). Constant values are maintained through the use of a

feedback 100p.

There are two modes of AFM: Tapping ModeTM and contact mode. In Tapping Modem,

a micro machined silicon tip at the end of a silicon “diving board” cantilever is used. The
diving board is oscillated near its natural resonant frequency by means of a small external
signal. When the tip nears the sample surface, interactions with the sample induce small
changes in the oscillation frequency, the rms amplitude of which is detected by the
photodetector, and plotted as (2: x,y). The phase difference between the natural resonant
and interaction frequencies can also be plotted (Ad): x,y) and this information is
particularly sensitive to material composition differences, and also to edges. This work
was performed in Tapping Mode most of the time. In contact mode, a micro-machined
SiN tip on the back of a triangular non-oscillating cantilever is used. Variations in the
choice of the spring constant k of the cantilever allow flexibility in examining different
surfaces, where k is the same as that in Hooke’s law, F =kx. The height, or the

deflection, of the tip with respect to the sample surface is kept constant.

35

AFM, because of its high-resolution capability, provides detailed information as to the
surface structure of individual nanotubes as well as the presence of defects which go
through to the surface. The phase image reveals differences in the composition of a
sample. Atomic Force microscopy is useful for structural and compositional analysis of

the CNT and carbon nanofiber surfaces.

Scanning probe microscopy was conducted using the Nanoscope IIIa Multimode SPM

System from Veeco Metrology Group/Digital Instruments, operated in ambient air.

36

CHAPTER 3

SYNTHESIS AND SAMPLE PREPARATION METHODS

The CNTs examined for this project were synthesized by three different methods. One
set of CNTS were commercially produced by Tubes@Rice and Carbon Nanotechnologies
via the HiPco method (high pressure carbon monoxide (CO) decomposition) and
produced exclusively single—wall CNTS. Another set of CNTs was produced at the
NASA Goddard Space Flight Center using the carbon arc method in a helium atmosphere
without a catalyst. This method produced exclusively multi-wall CNTS with well-defined
walls composed of single graphene layers. The third method involved growth of
MWCNTs as part of a MWCNT-Silicon nanowire heterostructure using plasma
deposition in an inductively coupled plasma (ICP) Reactor. Two other hollow core
carbon fibers with Well-defined wall structures were also investigated for comparison
with the CNTS. These were: Electrospun carbon nanofibers and vapor grown carbon

fibers. All synthesis conditions will now be briefly described.

Production of Single-Wall Nanotubes

Single-wall CNTs, from the Rice University spin-off companies Tube@Rice, now called
Carbon Nanotechnologies [http:/l cnanotech.com/pages/about/4-1_background.html],

were produced via the HiPco process and suspended in toluene. In the HiPco process,
flowing CO is mixed with Fe(CO)5 through a heated reactor. The reaction produces iron
clusters, which then serve as catalyst particles for CNT growth. More specifically the

solid carbon is produced via the following reaction:

CO +co —5‘1—> C(s) + C02 (10)

37

The CNTs are then suspended in a toluene solution. Before analysis by our group they
were purified and formed into “Bucky Paper” according to the procedure reported by
Rinzler et a1 [34]. In this process the Tubes@Rice suspended in toluene were first
refluxed in 2-3 M nitric acid for 45 hours. After the reflux, the CNTs were placed in a
container and allowed to settle and the brownish-yellowish supernatant liquid on the top
of the container was decanted. This was then followed by suspending the CNTS in
deionized water via ultrasonication, allowing the CNTs to settle and then decanting the
supernatant liquid and repeating the previous steps two or three times until the liquid is
dark throughout. Finally, the solution was dispersed in NaOH via ultrasonication and

then vacuum filtered through a membrane and allowed to dry, forming the “bucky paper”

Mtion of Mplti-wall Carbomnotufibes at the NASA Goddard Space Flight
93mg

To produce the multi-wall CNTS at the NASA Goddard Space Flight Center, a graphite
rod was welded at a steady rate with a Helium pressure of 50 psi and a current of 19
Amps. The catalyst requirement was relaxed. The relationship between amount and
quality of growth and experimental parameters including pulse width, current amplitude
and helium pressure, has been under development. The specimens utilized in the
radiation experiments were also subjected to a unique (patent pending) filtering process,
which resulted in high MW CNT yields. Figure 32 shows examples of NASA GSFC

MW CNTS, indicating good defect free walls between 5-20 layers in thickness.

38

Productiop of Mplti-wpll Carbon Nanotppe-Si Nanowire Heterostructures

 

Carbon nanostructures of approximately micron-scale lengths were grown in an
inductively coupled plasma system, using a 20-nanometer layer of iron catalyst on a p-
type (100) silicon wafer in methane-hydrogen-argon plasma. The carbon nanostructures
were shown to be multi—wall CNTs and hollow core vapor grown carbon nanofibers.
Simultaneously, tapered silicon nanowires were etched that retained the (100) orientation
and crystallinity of the original silicon substrate. Increasing reactor times resulted in a
progressive etch of the silicon nanowires, but no change in the approximately micron—
scale lengths of the CNTS and nanofibers. Junctions were observed between the silicon

and carbon nanostructures.

The simultaneous growth/etching was achieved in an inductively coupled plasma system.
Inductively coupled plasmas (ICP) have high radial uniformity and low pressure/low
temperature operating characteristics [42] which make them attractive for large-scale
applications. The inductively coupled plasma system was based on a modified Gaseous
Electronics Conference (GEC) Reference cell [43]. The original GEC reactor had two
parallel electrodes, 2.54 cm apart. The bottom electrode was biased with a 13.56 MHz
radio frequency (RF) power supply at 50 Watts. In the modified cell, the top electrode
was replaced by a 5-tum coil that was also powered by a 13.56 MHz RF power supply at
475 Watts. The pressure of the gas mixture was 100 mTorr with 87% Argon, and 6.5%
methane and 6.5% hydrogen. Argon was used to sustain the plasma in the inductively

coupled mode. Further details are reported in reference [44].

39

Samples were prepared by forming a 10-20 nm thick Fe layer on a p-type (100) silicon
substrate, using laser ablation [45-46]. Individual depositions were carried out over

time periods of 1 hour 0 minutes, 3 hours 10 minutes, and 5 hours 23 minutes.

This work was performed as part of a collaborative effort with the University of
Michigan using the University of Michigan Gaseous Electronic conference (GEC)
Reference Cell, with its unique Spatially Resolved Optical Emission Spectrometer
(SROES). Nanoscale heterojunctions between silicon and other nanotube/nanowire
materials may have important applications in nanoscale electronic devices, and as
interfaces between new nano-materials and conventional silicon electronics. Formation
of such nano-heterostructures by means of a common catalyst and a Liquid-Metal-Solid
(LMS) growth mechanism for both parts of the heterostructure have been recently

reported [35-41].

Production of Electrospun Carbon Nanofibers

The pre-irradiation carbon nanofibers were electrospun from polymer monomers. The
basic components of a typical electrospinning set—up are shown in Figure 33. The ES-
CNF’s in the present experiment were about 200 nm in diameter, but our collaborative

group [60] has also spun fibers down to about 25 nm in diameter.
Electrospinning involves the application of electrostatic force between polymer

monomers in solution kept in a narrow bore pipette or a syringe (capillary), and a metal

electrode kept at a suitable distance [47]. A charged drop of monomer + solution is

40

formed at the tip of the capillary, sometimes by tilting the syringe. With an increase in
electric potential, the charged drop of polymer solution formed at the tip of the capillary
is deformed into a cone, know as “Taylor’s cone”. 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. While in
transit, most of the solvent molecules evaporate away and the different polymer strands in
the jet separate out due to mutual repulsion (“splaying”). When these reach the collecting
electrode the diameters of the self—assembled fibers are in the range of a few micrometers
to nanometers. A basic electrospinning set-up and examples of electrospun carbon
nanofibers are shown in Figure 33. The electrospun carbon nanofiber samples from
MSU were provided by Professor Melvin S. Schindler, Department of Biochemistry and

Molecular Biology, Michigan State University.

Production of Vapor Grown Carbon Fiberp

Well—graphitized vapor grown carbon fibers were obtained from G. Tibbett’s research
group at General Motors Research and Development Center in Warren Michigan. These
were produced by carbon vapor feedstock thermal decomposition, flowed over a metallic

catalyst bed, as described in Reference [48]

41

CHAPTER 4

NEW HETEROSTRUCTURE FORMATION MECHANISM
The formation of silicon and carbon hetero-nanostructures in an inductively coupled
plasma system by a simultaneous growth/etching mechanism was reported by our group
in Reference [42]. Multi-wall CNTs were grown during one, three and five hour
depositions while tapered silicon nanowires were progressively etched. The carbon and
silicon nanostructures and the interfaces between them were studied by electron
microscopies and micro Raman spectroscopies. This method has potential for large-scale
controlled production of nano heterostructures without the requirement of a common

catalyst.

Scanning electron microscopy (SEM), transmission electron microscopy (TEM) with
selected area diffraction (SAD), high-resolution transmission electron microscopy (HR-
TEM) with Energy Dispersive x-ray Spectroscopy (EDS), and micro-Raman
Spectroscopy and Surface Enhanced micro Raman Spectroscopy (SERS) were used to
characterize the resulting carbon and silicon nanostructures. The SEM was performed
using 3 Cam Scan 44FE. The TEM was performed using a Hitachi H-800 Transmission
Electron Microscope operated at 200 kV and the HR-TEM was performed using a JEOL
JEM 2010F operated at 200 kV, at Michigan State University. A JEOL 4000 EX operated
at 400 kV at the University of Michigan was also used for some of the HR-TEM analysis.
Samples were prepared for TEM/HR-TEM by wetting the surface of the sample with
ethanol and gently scraping, followed by dispersing the ethanol and sample via a pipette

onto holey carbon film TEM grids.

42

The micro Raman spectroscope consisted of a 0.75 m double monochrometer coupled
with an optical microscope, operated in backscattering configuration, with an argon ion
laser as the excitation source. 514.5 nm and 488 nm excitation wavelengths were both
used in these investigations. SERS spectroscopy was performed as described in Ref.

[48], using 20 nm of sputtered gold on a glass substrate.

Scanning Electron Microscgpy Resplps
For the three deposition time periods, SEM images taken at 0° tilt showed a uniform

sparse coverage of fibrous nanostructures on a dotted background. A typical example
from the five-hour sample is shown in Figure 12. The coverage appeared to be uniform
over roughly 4 cm2 sample areas. Upon tilting to 40°, it became apparent that the dotted
backgrounds were tips of tapered nanostructures perpendicular to the surface. Interfaces

between the fibrous and tapered nanostructures were observed by SEM, as shown in

Figure 12 (a) and (b).

TEM and Selected Area Diffraction Results and Micro-Raman Results
TEM analysis indicated that the fibrous nanostructures observed in the surface plane had

hollow cores and micron-scale lengths, as shown in Figures 13 (a)-(c). Selected area
diffraction (SAD) patterns of these nanostructures revealed spot splitting of about 30°,

consistent with that expected for nanotubes [50]. An example of a typical diffraction
pattern is shown in Figure 14. Micro-Raman and SERS spectroscopies of the 5-hour

sample also indicated the presence of well-graphitized carbon by relatively sharp peaks at

43

about 1580 cm" with 8-10 cm" full width half maximum (FWHM). A typical peak is

shown in Figure 15.

TEM analysis of the tapered nanostructures is shown in Figure 16. The lengths of these
nanostructures varied systematically as a function of reactor time. Approximate average
lengths 0.4 microns, 0.5 microns, and 0.9 microns were observed for the 1-hour, 3—hour,
and 5-hour samples. The average base width for the 1-hour and 3—hour samples was about
100 nm, while the average base width for the 5—hour sample was about 150 nm. Typical

examples of the progressively etched silicon nanowires are shown in Figure 17.

Selected area diffraction (SAD), which included tilting to multiple zone axes, showed
that these tapered structures were single crystal silicon nanostructures with the [100]
orientation along the long axis. The orientation analysis is shown in Figure 18. The

structures were tapered silicon nanowires, not hollow core silicon nanotubes.

TEM images revealed dark contrast dots of about 5-20 nm at the tips or along the sides of
most of the tapered silicon nanowires, as shown in Figure 19. These dark contrast dots

were subsequently shown by HR-TEM to be nanocrystalline iron.

High-Resolution TEM and EDS Results

HR-TEM was used to assess details of the silicon, carbon and iron nanostructures. Two
size scales of fibrous carbon nanostructures were observed. Based on inner and outer

diameters, these were identified as multi-wall CNTS and vapor grown carbon nanofibers,

44

as shown in Figure 20. Background from the carbon holey film limited further resolution
of the wall structures. The dark contrast dots were shown by HR-TEM to be crystalline
with well-defined atomic spacings consistent with bcc iron. EDS spectra provided
further confirmation that these dark contrast features contained predominantly iron. The
HR-TEM and EDS results are shown in Figures 21 (a)—(d). Iron nano-dots were
sometimes observed at the tips of the carbon nanostructures as well as at the tips of the

tapered silicon nanowires, also as shown in Figure 20.

Details of a heterojunction between a tapered silicon nanowire tip and a multi-wall CNT
are shown in Figures 22 (a)-(c). The heterojunction appears to be about 4-6 mm wide.
An EDS spectrum, shown in Figure 22 (b), provided further support for the identification

of distinct silicon and carbon component parts.

Two types of nanostructures were observed in the plane of, and perpendicular to, the
sample surfaces for 1 hour 0 minutes, 3 hours 10 minutes, and 5 hours 23 rrrinutes reactor
times. One was fibrous and micron—scale in length for all reactor times. The other was
tapered, solid, and with an aspect ratio which changed as a function of reactor time.
Analysis by TEM/SAD, HR-TEM/EDS, and micro-Raman/SERS indicated that the

material compositions of these nanostructures were, respectively, carbon and silicon.

The carbon nanostructures were shown by TEM and HR-TEM to be multi-wall CNTs
and hollow core vapor grown carbon nanofibers. The iron catalyst layer was shown by

TEM and HR-TEM to be broken up into 10’s of nanometer-sized droplets by the action

45

of the plasma. Iron dots with diameters between 5-20 mm were observed at the tips and
along the sides of most of the tapered silicon nanowires. They were also observed at the
ends of some of the CNTs and nanofibers. The observed multi-wall CNTs and carbon
nanofibers would be consistent with growth via a Liquid-Metal-Solid growth mechanism

in the presence of the iron catalyst droplets.

The silicon nanowires were shown by TEM and SAD to be solid tapered nano-wires,

with the [100] orientation in the sample normal direction, and with progressive length
variations as a function of reactor time. This results is consistent with formation by an
etching mechanism of the (100) silicon substrate. Similar etchings of silicon [51-52] and
other semiconductor materials [52-53] have been observed in argon-methane plasmas.
However, in our experiments, the etching of the silicon nanostructures was
simultaneously accompanied by LMS—mechanism growth of the carbon nanostructures.
As the iron catalyst layer was broken up into 10’s of nanometer—sized islands by the
action of the plasma, it may have served as a nano-mask for the etching of silicon, as well

as the LMS nucleation site for the growth of the multi-wall CNTS.

SEM images indicated that a nano-heterostructure junction was formed between the
carbon nanostructures and the silicon nanowires at the pointed tips of the silicon. The
formation of such junctions was observed for all three reactor times. The sample
preparation method used in the present analysis reduced the probability of finding an
intact heterojunction, but one intact specimen was located and analyzed by HR-TEM.

The multi-wall CNT-silicon nanowire heterojunction appears to be about 4—6 nm wide,

46

which is a truly nanoscale heterojunction. Further experiments are needed to determine
how many junctions had a similar structure, and if this type of connection is found within

the macroscopic junctions observed by SEM.

The conditions for simultaneous LMS growth of multi—wall CNTs and etching of high
aSpect ratio silicon nanowires can be achieved in an Inductively Coupled Plasma reactor
system. ICP reactors are the current state of the art in the semiconductor fabrication
industry for large-scale silicon etching, due to high radial unifomrity and low
pressure/low temperature operating characteristics. Our results indicate that the
incorporation of nano-materials produced by an LMS growth mechanism within a
growth/etching fabrication cycle may also be possible. This may in turn lead to practical
large-scale manufacturing of nano-components of multiple materials, using an existing

reactor tool for both the conventional silicon and the new nano- electronic materials.

47

29:3. So:-m 05 Eat Ba mousé 05% BE. .8353 =u ecu :30? can wagon—305w:
3:58 we omfio>cu smote: Damned a @8865 :5 oo 3 coo—S momma: 2mm

«Ema—m m2 E 9:53 03m eo¢< custom BEES—«Z
6920 553 8 5:3 co ae< 2E co E. ..o a :95. 8mg: 23 “2 .2sz

48

 

 

 

 

 

858.585: @288 05 Ea
265$ 05 5238 228:3 8365 .z: 09» ”32> on: By new z: 00 ”32> a8 95 a coo—S momma: Sum
883 m ao¢< moeaosbmohzom 25052 5-.qu m0 mEmm @3685 ”mg oSwE

scram gonna

 

49

douse—Em 28 30:0: ecu magma": :EEE 303:.an9“ 836E 33an
597m 3 new 59:“ 3v 55:-— AS 05 Eot $589585: £55m 05 mo momma: 2m...

«Emma ...—E E mesa: m use .m .

S: cm

_ coca. @258 3583.5 .5330 .8 momma: 2m: ”3 SEE

 

50

EE comma .588 2: Eat 5 23026 coatsm one 05 556 89c on 38 mwct
Boise/x .8385: .8330 no 058:2: mEEEm 8% wEBonm Eugen 5:856
D<m 60:8 3:25:52 3v use. 3:30 235% Q<m E3 owns: Eoc Emtm 3
«SEE mm: E nooseoi

33252 :ofiaU a Co 5225 8:855 wEncomwoboU ES 8me Emu. ”w. 833m

 

51

._-Eo owfl “a 7:8
oTw 80% 22% 5:3 33m 92m E3332 3 338.585: 598 Busfimfim :03 mo 5:258 05
2365 29:3 .52; 2: Bo: Ammmmv 388.58% SEEM 2&2 nougncm 88:6 93 SEEM 822

85825 BNEFEEO

4.85 wccaofifi .3203 m 8% «8.35 mm: E 3an 3%? 88m 8.8mm =aEaM-8o_2 ”3 829m

p-83 “Em 525mm

ca: x2; co. xmc; 00— x8.— co_xmm._ 02 “am.— coo
. 4 _ ‘ _ ‘ 4 1 _

 

l

ooor

com?

I

oovr

('n'e) Ausuam

 

l

oomw

 

 

ooow

 

coon

52

.03an 50:6 2: 60$ 3 owns: o>ona 2F
.ouatsm 2: 9 336593 8:80 203 8588: REES 05 @2865 3925 Em;
8.32585: cog—V. c2093 2: mo mung: Emh u: 2sz

 

53

.moEEmm 52% 3 was .50;-m3v .52.-— Qv BE: .8 5:25.,“ a ma bozmmuhmoa
not? $53385: “.2093 05 mo mfiwfi: 05 35 8365 woman: Em;
«Ema—m mum E 8:03 n g was as: N at .50: _ AS Jana hoes moans: Emfi.
85825052

82% be £223 8mm Ea mfimaoq mEhamEoU momma: Emh ”w. 235

 

54

.38 wac— ofi wcofi couficoto 8o: 2: 5:5 :82: Embo
BwEW go 3:865 23:3 30% 3.22585: 3.592 2: So: momma: 558:6 Q<m
8.95% :53 m EB":

8025 ocESmbU 225 358250an 2: 35 35:5 5:855 35. @8020 no. 2&5

 

55

3:325: 82% 3.093 2: .8 82m 2: mac? new was 05 a 306 68:50 MED
Sea 32:50 {am mo >553 mo cosaomto> ”om oSwE

 

56

ESE 5E8 c2 532 05 E :32: 3 8585.585:
.598 05 we as 2: 3 8:0on moESuEOm 203 flow 35 so: .EUV £0535:
E530 :32w 59; van CZV 89:85: .588 £3.22: 8865 woman: 23-x:
«Emu—m 33:00 b03835

E 2:0: m “can 32605 ESE :8th can #20 .«o Emflnmm gm oSwE

 

57

Spectrum 2

 

Figure 22 (a)—(b): HR-TEM Images of Silicon Components of CNT—Si
nanowire Heterojunctions and Iron Nanodots

(a) The dark contrast dots showed atomic spacings consistent with
nanocrystalline iron. (b) EDS spectra were used to further confirm the
identification of the iron nanodots (Spectrum 1) on the silicon nanowires
(Spectrum 2).

58

2 E2839 $255: :0: 2: ho 5585202 05 CELEB 52:3 9 tom: 33 mam
$25.82 so: .8 Epsooam mam ”on mm EsmE

 

 

('0
Lo.
,-
O

 

:0

£0

 

 

 

 

 

 

_ €28on

 

 

 

59

.8 82.50an 8:383 .525 2: he cozaocccoE 2: 85:8 Esta 8 new: 33 mam
3:26.52 .525 he E2825 mam AB mm oSwE

 

 

 

N €28on

 

 

 

 

 

 

 

 

 

 

 

 

 

6O

 

Figure 23 (a)—(b): HR—TEM IMAGES an Intact CNT-Si—Nanowire
Heterojunction (a) HR-TEM image of a heterojunction between a MWCNT
and a tapered silicon nanowire tip (b) Lower magnification view of the
MWCNT-silicon nanowire heterojunction.

6]

cosmocscofl 50:3 2: comma 8 com: 33 mam
cozocsaohzom o.__Bo:mZ-_m-._.ZU 885 as ho E8825 mam x8 mm 8:?”—

 

 

 

 

 

 

 

 

o« m o n m m c n N w o
va Dbhbl—bthPhbbb—bbhbhbbhnhbbprbPLD>hbrbih>~>>Dhb>btrhpbb>>h>b>Pbl>>hh>>bhpbb1bbPFF‘>L>>>LI}D’>
so . on c on 3 —
.w :0 O
:0
:EEQoum
o

 

 

 

 

62

CHAPTER 5

HEAVY-ION INTERACTIONS IN NEW N ANOSCALE MATERIALS

Introduction

To reiterate, space exploration is entering a new era, with the advent of new micro space
probes, and the increasing use of miniaturized equipment. The United States space
agency, NASA, is planning to deploy in space electronic devices and enhanced
composites constructed from the newly developed nanotube and nanowire components
[55]. Heavy ion irradiation is a significant component of the Low Earth Orbit (LEO)
space radiation environment, where an increasing number of space platforms and
satellites will operate. Heavy ions are known to induce failures in current state of the art
components and electronic devices, through uncontrolled ion implantation, which alters
the channel conductance [56], and through lattice disruption mechanisms [57-58]. These
problems may be exacerbated in microprobes, which have less radiation shielding. In
addition, state of the art conventional transistors are increasingly sub-micron in

dimension and more susceptible to radiation damage.

Experimental Procedure

The radiation experiments were performed at the National Superconducting Cyclotron
Laboratory (N SCL) at Michigan State University, whose available beam energies well
match the energy spectra of abundant charged particles in space radiation environments.
To simulate radiation doses from hi gh-Z ions, the samples were irradiated using the

Krypton-86 standard primary beam at 142A MeV/nucleon. As there are no published

63

guidelines for beam exposure times in nanoscale materials, these experiments were
conducted with doses and exposure times comparable to those that can be expected to
damage or destroy current state—of-the art silicon circuits. Failures under dose in a diverse
range of silicon devices and technologies indicate that initial damage thresholds are at

about 10 Gy [58]. Thresholds for radiation-hardened devices are at about 1000 Gy.

Almost all devices will fail before doses reach 105 Gy.

Single wall CNTS (SW-CNT), multi-wall CNTS (MW-CNT), tapered silicon nanowires
(T -SiNW), electrospun carbon nanofibers (ES-CNF), and vapor-grown graphitic carbon
fibers (V G-CF) were irradiated. The choice of samples allowed us to investigate heavy
ion interactions with multiple carbon wall structures: graphene single crystal, i.e., SW-
CNT, nested graphene single crystal, i.e., MW-CNT, highly ordered organic networks,
i.e., ES-CNF, and stacked graphite crystal platelets, i.e., VG-CF. Investigation of the
silicon nanowires will allow comparisons between the behaviors of carbon versus silicon
at the nanoscale. The SWCNTs were in the form of a buckypaper to maximize the

volume available for heavy-ion interaction.

Sample Preparation

The NASA CNTS and the “Tubes@Rice” were ultra-sonicated in ethanol, placed in
quartz capillary tubes and the ethanol was then allowed to evaporate, leaving the samples
(except the carbon nanofibers and the CNT—Si nanowire heterostructures on the walls of
the tube. In the case of the CNT-Si nanowire heterostructures, the wafers on which they

were grown were simply placed in the quartz tubes after being cleaved to a size that

allowed them to fit into the tubes. The electro—spun carbon nanofibers were placed in the

tubes in their original form.

100 Gy, 1,000 Gy and 10,000 Gy were set as the target doses for our experiments. The
individual doses varied somewhat as a function of mass, sample type (density), and
packing volume. Fifteen samples were placed in quartz tubes arranged as 6/5 tubes per
row wide (close packing) and 2 rows deep in a rectangular holder. Each individual quartz
tube had an outer diameter of 3 m, an inner diameter of 2 mm, and a total length of 70
mm, of which about 20 mm at the top, containing the samples, was exposed to the beam.
The quartz tube array was placed in front of the beam port. Dose calculations included
the effects of the quartz sidewalls in the first and second rows, the zirconium foil vacuum
window and the small air gap between the window and the sample array. The beam was
transported to the NSCL N4 vault end station by dipole and quadrupole magnets, and
focused there to an area having dimensions of 2—cm (height) by 2-cm (width), as
measured from known dimensional markings on the beam-viewing scintillator placed at
the beam line exit vacuum window. The beam area was sufficient to irradiate the
complete extent of the samples. The upstream current was measured in the extraction
channel of the K1200 cyclotron, using a remotely operated beam probe. The exit and
upstream current measurements were ratioed to access the beam factor. For each
subsequent experiment, the upstream current was measured both before and after the run.
From the Faraday cup measurement and subsequent probe measurements, a stable current

of 0.3 particle-picoamperes (ppA) or 1.87 x 10° particles per second, was estimated as the

65

beam current at the exit point during the experiments. Therefore, the particle flux was

4.68 x 103 particles per second per m2.

Pre-Irradiation Calculations

Prior to the irradiation of the samples, calculations were performed to estimate the
irradiation times necessary to achieve the desired doses of 100, 1,000, and 10,000 Gys
First, the necessary dose was determined. This was done as follows:

1) The mass of the carbon was calculated. The density of the carbon was estimated
by the value for graphitic carbon, 2.24 g/cm3. The density was then multiplied by
estimated volume of the carbon, 2.84 E-02 cm3 (thhickness=.05 cm, height=1
cm).

2) The energy loss was determined after a series of calculations using a specially
developed software package at the N SCL. Hence, the dose per part Gy/ion, which
is the energy loss in the carbon divided by the mass of the carbon determined to

be 7.35 E+06 per part Gy/ion

Secondly, the total number of ions needed to achieve the desired doses was determined
by dividing the desired doses by the dose per part Gy/ion. This gave the total number of
ions needed to achieve 100, 1,000 and 10,000 Gy. Finally, to determine the necessary
amount of time needed to achieve the closes, the total number of ions necessary to achieve
the desired doses was divided by the beam current, which was .3 particle pico Amps

(ppA) =1 .8 E+06 particles per second

66

 

The results are tabulated below in Table 1

Table 1. Calculations Performed to Determine Irradiation Times

 

 

Dose per part Dose Time at .3 ppA
86 Kr Ions/sec

Gy/ion (Gy) (Sec) (Min) (Hr)
142 MeV 7.35 E-06 100 1.36 E+O7 1.13 E +00 1.89 E-02 3.15 E04

 

1,000 1.36E+08 1.13E+00 1.89E—01 3.15E-03

 

10,000 1.36E+09 1.13E+00 1.89 E+00 3.15E-02

 

 

 

 

 

 

 

 

 

Post-Irradiation Calculations

Detailed calculations were made in order to verify the exact dose that the single wall
CNT film was exposed to, seen in Table 2. The approximate area of the carbon film was
calculated and was carefully weighed. The parts per second were then calculated to
determine how many ions were reaching the sample and the required time in order to
achieve the desired dose.

Table 2. Calculations Performed to Determine Actual Doses.

 

 

 

 

Sample {5:801} Area Sample Mass (g) Ions Time (sec) Sample Dose (Gy)
1 16.92 0.01 3 7.92E+04 900 3.42E+02
2 16.92 0.005 7 .92E+04 90 8.89E+01
3 16.92 0.0099 7.92E+04 15 7.49E+00

 

 

 

 

 

 

 

 

Radiation Exgriment Results

Pre versus post irradiated specimens were investigated for structural and chemical
alterations. SEM, FE SEM and AFM were used to investigate changes to the external
structure. Transmission electron microscopy and high—resolution transmission electron

microscopy were used to investigate changes in the internal structure. Selected area

67

diffraction was used to detect any departure from crystallinity. Micro Raman
spectroscopy/Surface enhanced Raman Spectroscopy and was used to investigate the

molecular bonding present and to monitor the appearance of any defect induced peaks.

The results are presented in decreasing order of damage effects, to show the damage
mechanisms present in the VGCF and ESCNF samples under heavy ion irradiation, and

their absence in the irradiated SWCNT and MW CNT samples.

Radiation Damage Effects in the VGCF Samples

Pre-irradiation characterization of the VG-CF was obtained by SEM, TEM and micro
Raman spectroscopy and indicated well graphitized hollow core stacked platelet [56]
fibers with fairly uniform 150 nm diameters of about 150-200 nm, as shown in

Figure 23(a). By 90 seconds (1,000 Gy) irradiation, substantial damage was observed
over most of the sample area, as shown in Figure 23 (b). The fibers become vaporized

into disordered or amorphous carbon with very few fibers remaining.

Sample damage was so severe that insufficient amounts of the 15 and 900 second
samples remained for analysis, after extraction from the quartz tubes. In the future, a
different sample preparation technique using thin fiat quartz plates will be employed.
Also, it will be known in advance that this type of sample can be reduced to a very fine
powder which is very difficult to handle. Figure 24 shows a possible damage mechanism
associated with this type of radiation damage. The radiation beam introduces an

interstitial, an excited 86Kr ion, which induces a vacancy in the plane of the graphite.

68

This interstitial also creates a vacancy in the graphite. There is then a subsequent
contraction in the in-plane (a-plane) direction and expansion in the out of plane (c-plane)

direction, resulting in the disintegration of the fiber.

Radiation Damage Effects in the ESCNF Samples
The electrospun carbon nanofibers (ES CNF) produced at Michigan State University

using a homebuilt system used poly e-caprolactone as the polymer monomer; the ES
CNF fibers produced at the Donaldson company (industrial partner of Professor Melvin
S. Schindler) used a propriety mixture. Tapping Mode Atomic Force Microscopy of the
post-irradiation electrospun carbon nanofibers indicated damage spots (10’s of nms),
which may be due to intense local heating which caused “melting”. FE SEMs, shown in
Figure 32, were used to confirm that the results of the AFM analyses were truly

representative.

TEM analysis also showed progressive damage with increasing radiation, as shown in
Figure 35. We noticed the appearance of dark contrast areas, which seemed to be
increasingly prominent by 900 sec. Dark contrast areas may indicate changes in

molecular structure, corroborating the hypothesis that local “melting” did occur.

This interpretation would be consistent with a known radiation damage mechanism in
polymer materials [65]. Intense heating could be caused by scission of the side chains or
by transfer of phonons down a carbon chain until an end site is reached. Further damage

could have been generated through formation and release of volatile oxygen compounds

69

such as CO and C02. The nature and degree of cross-linking and hence the availability

of side chains as a function of tube diameter is currently under investigation, and will be

investigated in a further series of radiation experiments.

Radiation Results for the SWCNT SW

A continuous film of single wall CNTS (bucky paper) was studied to maximize the
possibility of an interaction. The following results are shown in Figures 27 (3)-(b).
Tapping Mode Atomic Force Microscopy studies of the surface morphology did not
indicate any obvious damage at any radiation dose. Wide area analysis by FE SEM (20

microns x 20 microns) also did not indicate any obvious morphological changes.

Micro-Raman is a volumetric measure over an even wider area. For our Raman system

the 60x objective used in these experiments, and the spot size was 30 microns in
diameter. The Micro-Raman spectroscopy of the tangential mode region, 1500-1700 cm-
1, clearly indicated the double peak at about 1560 and 1590 cm'] , that corresponds to the
well-studied signature of mixed metallic and semiconducting single wall CNTS [14]. In
the 900 second (10,000 Gy) Raman spectrum, a slight peak at about 1350 cm.1 may

indicate the presence of some disordered carbon. The conclusion was the SWCNTs did
not show any obvious radiation damage, when investigate by AFM, FE SEM, and Micro-

Raman Spectroscopy

70

 

Radiation Results for the MWCNT Sarpplgg

FE SEMs of the multi-wall carbon nanotubes also showed consistent radiation resiliency
throughout all radiation doses. In the pre irradiated sample, the nanotubes were difficult
to find and seemed to be buried in the disordered carbon bulk. The MWCNT show with
increasing cleanness in the 15 sec, 90 sec and 900 sec FE SEM image. It is possible that
the heavy ion irradiation vaporized some of the amorphous carbon content of the

samples, leaving the MW CNT intact and exposed.

One sample of the Multi wall CNTS was studied via High-Resolution transmission
electron microscopy. The 90 second (1,000 Gy) irradiated sample was showed no visible
structural damage as seen in Figure 26. Further HR TEM of all of the irradiated MW CNT

samples is ongoing, and further Micro Raman spectroscopy is also ongoing.

Summary of Heavy Ion Irradiation Results

 

Morphological and molecular bonding characterizations of single wall buckypaper CNT
films and multi-wall CNT samples indicated little damage for 100, 1,000 and 10,000 Gy
86-Krypton radiation doses at 142 AMeV/nucleon. By comparison, vapor grown carbon
fibers showed substantial disruption after 90 seconds heavy ion irradiation (1,000 Gy),
and total loss by 900 seconds. Electrospun carbon nanofibers showed damage at all
radiation doses, with increasing damage at increasing doses, and extensive damage by
900 seconds. The damage pattern for electrospun carbon nanofibers appears to be
different than that for vapor grown carbon fibers. Known damage radiation mechanisms

for graphitic vs. polymeric materials are consistent with the observed results for vapor

71

grown carbon fibers and electrospun carbon nanofibers. However: neither damage

mechanism was explicitly observed in irradiated single and multi—wall carbon nanotubes.
These results indicate that CNT textiles and components may have good heavy ion

radiation resilience that would make them appropriate for deployment in space

applications.

72

 

 

mucooom cm as omega dosage 3v aoufivmbmoum A3 .3 2mm .8: 02 Son“ E 53:36 Rona noummustméa 23.
Hopi 8330 8580 “23> .«o momaEH 2mm #N 0.5me

SC

00999

 

73

 

Figure 25: Damage Mechanisms in graphite due to radiation (Taken from [66]:
TD. Burchell, “Radiation Effects in Graphite and Carbon-Based Materials”, MRS
Bulletin Special Issue: Materials Performance in a Radiation Environment, Vol.
22, pp. 29-35 (1997))

74

 

lebettc Well-Gmhltlzod Flbora, 9I1912000

('n'e) Ausuarm

a a a a

8

V

75

 

E

90+3008'l
90+3081.‘ L
90+3091'l
90+30W.’ L
90+30ZL' l
90+300[ L
90+3089' L

' 90+3099‘ l

90+30179'L
90610391
904510091
90+3089'l
90+3099’L
90+30179' l
90+EOZS' l

, 90+3009'L

90+30817' l
90+30917' L
904‘301717' I
90+EIOZI7' l
90+30017' L
90+308‘E' L

’ 90+3099'l

90+30VE L

- ' 9090?: L

90+3008'l
9061083 I
90+EIOQZ' L
904301731

3 90+aozu
~ 90+300u

Raman Shm (cm “)

 

Figure 26: Raman Spectra of well-graphitized fibers prior to irradiation showing a strong peak at 1580 cm-1

 

0 Data Type Height 5 um 0 Data Type Height 5 um
Z—range 1.80 um Z—range 729 nm

(a) (b)

      

. 5 um 0 Data Type Height 5 um
Data Type Height Z-range 1.97 um
Z-range 1.34 um (d)

(C)

Figure 27: AFM Height Images of Electrospun Carbon Nanofibers Electrospun carbon
nanofibers (precursor monomer poly(e-caprolactone)). The diameter is about 200 nm.
Tapping Mode AFM of (a) pre-irradiation (b) radiation damage at 15 seconds (100
Gray), (c) radiation damage at 90 seconds (1,000 Gray), (d) radiation damage at 900
seconds (10,000 Gray)

76

(a)

 

C > Data Type Phase 2.50 um
Z-range 86.1 deg

Figure 28 (a): AFM PHASE Image of Tubes@Rice Prior to Irradiation
Continuous film of single-wall carbon nanotubes, individual diameter about 1.4
nm. Tapping Mode AFM images all resembled (a) indicating no observable
morphological damage.

77

(b)

 

Buckey Paper. 488 nm, 7I24I00

3500

3000

2500 A

O O

C) C)

8 :9
('n'e) Ausuaaul

78

1000

 

500

' 90+39w

90+308' l
90+38L' l
904*391' L
908‘“: l
90+30L' l
90+389' I
90+399' L
90989 I
90+3L 9' L
90+389‘ I.
90+399' I
90+379' I
904-319 I
90+36V L

90+3W' l
90+BZV l
90+368' I
90+BLS‘ L
90+3V€ l
90+3Z8' l
90+308' l
90+EIZ' 1

Raman snm (cm ")

i, ;'90+39z1

90+BZZ‘ I
90+BOZ' l

 

Figure 28 (b): RAMAN SPECTRA OF TUBES @RICE PRIOR TO IRRADIATION GRAY.

Micro Raman spectroscopy of the pre-radiated sample

(c)

 

Bucky Paper (900 s) on 81 water, 514.5 nm,4l7l2003

1800 .

1600

 

1400

90+369‘l
90+389'L
90+399‘l.
90+399'L
90"389'1
90+3Z9'l
90+309'l
90+369' l
90+319‘L
908991
9091791
90+399'l.
90319 L
9043091
90+3817'l.
scary I
90+ESV'L
90+Ell717'l
90+EZV' 1
90431171
90*368'1
90+38S‘l
90+399' L
. 90+398‘l
90+3€8l
90+3ZE‘1
9043081

0

O O O O O O
O O O C) O O
N O 00 (O V N
r- <—

(‘n'al Ausuaaui

79

Raman sum (cm 4)

 

Figure 29: Raman Spectra of Tubes@Rice After a 900 seconds.

Micro Raman spectroscopy of irradiated sample after 900 seconds (340 Gy) also indicated

little damage

 

Figure 30: HR-TEM lrnages of Multi—wall CNTs after 90 seconds.
Both images show no radiation damage after 90 seconds

80

 

.
,
a“"

Figure 31: FE SEM Images of Tubes @Rice prior to irradiation.
[Images taken by Benjamin Jacobs]

81

a)

C)

 

Figure 32: FE SEM Images of MWCNTs grown at the NASA Goddard Space
Flight Center taken after irradiation.

a) 15 seconds b) 90 seconds and c) 900 seconds of irradiation

[Images taken by Benjamin Jacobs]

82

 

Figure 33: HR-TEM Images of NASA Goddard Space Flight Center Multi-wall
CNTS Showing Distinct Wall Layers

83

Home-built electrospinning system:

 

 

 

 

 

 

 

 

 

 

Polymer \
Monomers .mo M . .
in e 3
Suspension l Plate
f Power \I\
Supply f ~15 kV

 

Self-Assembly Technique
Diameter ~ 200 - 25 nm

Figure 34: Electrospinning Set-up

84

.3 .
a:

We.

4.....

(C)

 

1V6

HR-TEM Images of Electrospun Carbon Nanofibers Showing Progress

.
c

Figure 35

Damage from 0- 900 seconds (a) Pre-irradiation (b) 90 seconds (c) 900 seconds

85

CHAPTER 6

DISCUSSION
The CNTS from the NASA Goddard Space Flight Center were not affected up to the
highest radiation dose over 900 sec. These CNTs were shown by HRTEM and Raman to
be high quality multi-wall CNTs with diameters ranging from 5-10 nm. They were
produced via the carbon arc discharge method in a helium atmosphere at the NASA
Goddard Space Flight Center. Single-wall CNTs from Carbon Nanotechnologies
(T ubes@Rice) were also not affected up to the highest radiation dose over 900 sec. By
contrast, two other hollow core carbon structures/nanostructures, vapor grown carbon
fibers (V GCF) and electrospun carbon nanofibers (ESCNF), showed considerable
damage by the highest radiation dose and significant damage even at the lower doses.
The radiation damage mechanism appeared to be different for the VGCF and ESCNF,
being possibly amorphitization in one case and local melting in the other. Thus, while
these two different damage mechanisms were possible outcomes of our radiation
experiments, neither seemed to apply to single and multi wall carbon nanotubes up to the
radiation doses studied. These results indicate that CNT textiles and components may
have good heavy ion radiation resilience that would make them appropriate for

deployment in space applications.

Heavy ion radiation, although a significant component and a serious problem for low
earth high declination orbits, is not the only type of radiation found in space. For
example, another important feature in LEO is the South Atlantic Anomaly (SAA), which

is an area of exceptionally high proton density, and placed over much of South America

86

and the South Atlantic Ocean. Currently those satellites that travel through the SAA
experience higher rates of single event effects (SEE) than those that do not. In
laboratory experiments on earth, it has been noted that purified single-wall CNTs
(particularly isolated single-wall CNTs) do not appear to be stable under the electron
beam of the electron microscope [63—64]. Multi-wall CNTS were also reported by Bursill
et al to be unstable under the electron beam of the transmission electron microscope [64].
Electron radiation (gamma radiation) is also a type of ionizing radiation found in space.
These reports underscore the need for total ionizing dose (T ID) experiments on nanotube
components and circuits. Our group currently has research efforts underway for TID
experiments on nanotube components and nanotube circuits as part of a collaborative
effort of Michigan State University with both the NASA Goddard Space Flight Center

and the NASA Jet Propulsion Laboratory.

Transmrt

Transport studies are the major emphasis of the thesis project of Benjamin W. Jacobs,
who will be continuing this research. Heavy-ion radiation and total dose effects on CNT
and Gallium Nitride (GaN) nanowire-based nanocircuits will both be examined.
Transport evaluation studies will be conducted with scanning tunneling microscopy, two-
point, and four-point probe measurements. Some of these studies will also investigate the
effect of thermal variation on transport, using the world-class cryogenic facilities at the

NASA Goddard Space Flight Center (micro to milli Kelvin).

87

Gold and Palladium contacts will both be investigated. The goals of this part of the

research are to carefully investigate the nature of the Schottky-barrier at the nanocontacts.

NASA has a well-established generic need for radiation tolerant technologies for flight
electronics, spanning from spacecraft systems to scientific instruments. Current state-of-
the-art radiation-hard microelectronics cannot be used without heavy shielding in
environments such as the Jovian or in proximity of a Radioisotope Thermal Generator
(RTG). This impacts heavily the mass budget of a mission. A potential solution to this
problem can be sought in using nanometer—sized devices, whose small active volume
makes them inherently radiation tolerant. During the transport studies, the electrical
characteristics will be monitored for changes as a function of the absorbed dose. The
results will be analyzed to (I) extract the essential characteristics of the technology, (2)
evaluate its radiation tolerance, and (3) assess the feasibility of nanotechnology for

infusion in NASA missions.

88

APPENDD( A

Preparation Procedure for TEM Gn'ds With Holey Carbon Film

1 Dissolve about 0.4% formvar or butvar in chloroform (make about 100ml), add

about 1% glycerol and shake until you have an emulsion.

2. Decant into a tall 25 ml beaker and dunk a clean glass slide into the solution. Remove

from the solution at a slow but constant rate. Allow it to flash in a vertical position.

3. Hold the dried slide in a saturated steam plume until its well beaded with condensate

but the condensate isn’t sheeting.

4 Scrape the edges of the slide with a plastic bottle cap to break the film free.

Lit

Float the film off the slide by gently immersing the slide in a dish of DI water. It

should appear to silver to gray in reflected light.

6 Make an array of grids on the film.

7 Lay a piece of paper (typing paper or a PostIt note paper) over the grids and allow the

paper to wet.

8. Gently lift the paper off the water surface and allow to dry, grid side up.

89

9. When dry, carbon coat.

10. Place the carbon coated grids, still on the lifting paper in a glass petri dish into which

you’ve placed a pad of filter paper completely soaked in chloroform. Cover and

allow the petri dish to stand under a hood until the chloroform flashes off.

90

APPENDD( B
Irradiation Procedure using the Coupled Cyclotron Facility at the National

Superconducting Cyclotron Laboratory

Samples were irradiated using a primary beam of 86Kr at 142A MeV. When having 142

MeV per nucleon, 86Kr has a total energy of 12.212 GeV. The beam was transported to

the NSCL’s N4 vault end station by dipole and quadrupole magnets, and focused there to
an area having dimensions of 2-cm (height) by 2—cm (width), as measured from known
dimensional markings on the beam-viewing scintillator placed at the beam line exit
vacuum window. The vacuum window was a zirconium foil, having a diameter of about 7

cm and a thickness of about 0.02159 mm.

The beam current was initially measured at the exit point using a Faraday cup. A beam

stop was pressed into a long copper pipe for accurate electron current collection. The
beam stop was a solid cylinder of Hevimeta, an alloy of tungsten, nickel, and copper. Its

diameter was 5.08 cm and its length was 5.093 cm. Insulating rings were placed around
the Hevimet-copper-pipe assembly, and the assembly was placed in the beam line,

forming a Faraday cup for current integration. For reference, the rangeb in Hevimet of an

ion of 86Kr having 12,212 MeV is about 1.12 mm.

The upstream current was measured in the extraction channel of the K1200
cyclotron, using a remotely operated beam probe. The exit and upstream current
measurements were ratioed to assess a beam transmission factor. For each subsequent

experiment, the upstream current was measured both before and after the run. From this,

91

a stable current of 0.3 particle-picoamperes (ppA) was estimated as the beam current at

the exit point during the experiments.

a HD18, by Mi—Tech Metals, Inc., 1340 N. Senate Ave., Indianapolis, IN 46202; HD18
is 95% tungsten, 3.5% Ni, and 1.5% copper, and has a density of 18 g/cm3.

b Ranges and energy losses reported in our paper were calculated using the computer
code SRIM-2003: J. F. Ziegler and J. P. Biersack, SRIM code, http://www.srim.org/

(2003).

92

BIBILIOGRAPHY

[l] N. Bingefors, T. Ekelof, C. Eriksson, G. Mork, M. Paulsson, and A Sjolund,
“Radiation Hardness Tests with a Demonstrator Preamplifier Circuit Manufactured in
Silicon on Sapphire (SOS) VLSI Technology”, Nuclear Instruments and Methods A, Vol.
316 (2-3) pp. 349-363 (1992)

[2] S. J. Mathew, G. Niu, S.D. Clark, JD. Cressler, MJ. Palmer, and W.B. Dubbelday,
“Radiation-Induced Back-channel Leakage in SiGe CMOS on Silicon-on-Sapphire (SOS)
Technology”, IEEE Transactions on Nuclear Science, Vol. 46 (6) pp. 1848-1853 (1999)

[3] CF. Edwards, W. Redman-White, M. Bracey, B.M. Tenbroek, M.S.L. Lee and MJ.
Uren, “A Multibit 2A Modulator in Floating-Body SOS/S01 CMOS for Extreme
Radiation Environments”, IEEE Journal of Solid-State Circuits, Vol. 34 (7) pp. 937—948
(1999)

[4] K.P.H. Lui and FA. Hegmann, “Ultrafast Carrier Relaxation in Radiation-Damaged
Silicon on Sapphire Studied by Optical-Pump-Terahertz-Probe Experiments”, Applied
Physics Letters, Vol. 78 (22) pp. 3478-3480 (2001)

[5] J. Handon and H. Veloric, “Radiation Hardened Silicon-on-Sapphire”, Electronic
Engineering, Vol. 56 (685) pp. 61—62 (1984)

[6] F1. Cui, 2.]. Chen, J. Ma, G.R. Xia and Y. Zhai, “Atomistic Simulation of Radiation
Damage to Carbon Nanotube”, Physics Letters A, Vol. 295, pp. 55-59 (2002)

[7] V.V. Klimov and V.S. Letokhov, “Hard X-radiation Emitted by a Positron Moving in
a Carbon Nanotube”, Physica Scripta, 56 (5) pp. 480-486 (1997)

[8] S. lijima, “Helical Microtubules of Graphitic Carbon”, Nature, Vol. 354, pp. 56-59
(1991)

[9] R0. Avakian, LA. Gevorgian, K.A. Ispin’an and R.K. Ispirian, “Spontaneous and
Stimulated Radiation of Particles in Crystalline and Nanotube Undulators”, Nuclear
Instruments and Methods B, Vol. 173, pp. 112-120 (2001)

[10]V.V. Klimov, and V.S. Letokhov, “Hard-directional X-radiation Emitted by a
Charged Particle Moving in a Carbon Nanotube”, Physics Letters A, Vol. 222, (6) pp.
422—428 (1996)

[11] V.V. Klimov and V.S. Letokhov, “Monochromatic Gamma-radiation Emitted by a

Relativistic Electron Moving in a Carbon Nanotube”, Physics Letters A, Vol. 226 (3-4)
pp. 244-252 (1997)

93

[12] D. Grimm, R.B. Muniz, and A. Latgé, “From Straight Carbon Nanotubes to Y-
shaped Junctions and Rings”, Physical Review B, Vol. 68 (19) pp. 193407-19310 (2003)

[13] R. Saito, G. Dresselhaus, and M. Dresselhaus, Physical Properties of Carbon
Nanotubes, by Imperial College Press, London (1998)

[14] SJ. Tans, A.R.M. Verschueren, and C. Dekker, “Room-temperature Transistor
Based on a Single Carbon Nanotube”, Nature, Vol. 393, pp. 49-52 (1998)

[15] M.M. J. Treacy, T.W. Ebbesen, and J .M. Gibson, “Exceptionally High Young’s
Modulus Observed for Individual Carbon Nanotubes”, Nature, Vol. 381, pp. 678-680
(1996)

[16] AT Woolley, C. Guillemette, C.L. Cheung, D.E. Housman and CL. Lieber, Nature
Biotechnology, “Direct Haplotyping of Kilobase-Size DNA Using Carbon Nanotube
Probes”, Vol. 18 pp. 760-779 (2000)

[17] J. Cumings and A. Zettl, “Low Friction Nanoscale Linear Bearing Realized From
Multiwall Carbon Nanotubes”, Science, Vol. 289, pp. 602-604 (2000)

[18] BC. Edwards, “Design and Deployment of a Space Elevator”, Acta Astonautica,
Vol. 47 (10) pp. 735-744 (2000)

[19] Q. Wang and J .K. Johnson, “Optimization of Carbon Nanotube Arrays for Hydrogen
Adsorption”, Journal of Physical Chemistry B, Vol. 103, pp. 4809-4813 (1999)

[20] X. Qin, X.P. Gao, H.Liu, H.T. Yuan, D.Y. Yan, W.L. Gong, and D.Y. Song,
“Electrochemical Hydrogen Storage of Multi-Wall Carbon Nanotubes”, Electrochemical
and Solid-State Letters, Vol. 3 (12) pp. 532-535 (2000)

[21] AG. Rinzler, J.H. Hafner, P. Nikolaev, L. Lou, S.G. Kim, D. Tomanek, P.
Nordlander, D.T. Colbert, and RE. Smalley, “Unraveling Nanotubes: Field Emission
From an Atomic Wire”, Science, Vol. 269 (5230) pp. 1550-1553 (1995)

[22] W.A. de Heer, A. Chatelain, and D. Ugarte, “A Carbon Nanotube Field Emission
Electron Source”, Science Vol. 270 (5239) pp. 1179-1180 (1995)

[23] PG. Collins and A. Zettl, “A Simple and Robust Electron Beam Source From
Carbon Nanotubes”, Applied Physics Letters, Vol. 69 (13) pp. 1969-1971 (1996)

[24] Y. Cheng Y and O. Zhou, “Electron field emission from carbon nanotubes”,
Comptes Rendus Physique, Vol. 4, pp. 1021-1033 (2003); 0. Zhou, S. Chang, Y. Cheng
Y and J. Zhang, “Potential applications of nanotechnology in radiation oncology”,
Medical Physics, Vol. 30, pp. 1414-1414 (2003)

94

[25] S. lijima, “Growth of Carbon Nanotubes”, Materials Science and Engineering, Vol.
B19. PP. 172-180 (1993)

[26] B.W. Smith, M. Monthioux, and DE. Luzzi, “Carbon Nanotube Encapsulated
Fullerenes: A Unique Class of Hybrid Materials”, Chemical Physics Letters, Vol. 315,
pp. 31-36 (1999)

[27] C.-H. Kiang and W.A. Goddard HI, “Polyyne Ring Nucleus Growth Model for
Single-Layer Carbon Nanotubes”, Physical Review Letters, Vol. 76 (14) pp. 2515-2518
(1996)

[28] H. Kanzow and A. Ding, “Formation Mechanism of Single-Wall Carbon Nanotubes
on Liquid-Metal Particles”, Physical Review B, Vol. 60 (15) pp. 11180-11186 (1999)

[29] C-H. Kiang , W.A. Goddard II], R. Beyers and D.S. Bethune, “Carbon Nanotubes
With Single-Layer Walls”, Carbon Vol. 33 (7) pp. 903-914 (1995)

[30] J. Mossbrucker, “Measurement and Characterization of the Raman Spectrum of
Diamond Films”, MS. Thesis, Michigan State University (1994)

[31] M.I. Stockman, V.M. Shalaev, M. Moskovits, R. Botet, and T.F. George, “Enhanced
Raman scattering by fractal clusters: Scale-invariant theory” Physical Review B, Vol. 46,
pp. 2821 -2830

[32] C. Qin and L. —M. Peng, “Measurement Accuracy of the Diameter of a Carbon
Nanotube from TEM Images”, Physical Review B, Vol. 65, pp. 155341 (2002)

[33] A.A. Lucas, V. Bruyninckx, and P. Lambin, “Calculating the Diffraction of
Electrons or X-rays by Carbon Nanotubes”, Europhysics Letters, Vol. 35 (5) pp. 355-360
(1996)

[34] A.G. Rinzler, J. Liu, H. Dai, P. Nikolaev, C.B. Huffman, F.J. Macias-Rodriguez, P.J.
Boul, A.H. Lu, D. Heymann, D.T. Colbert, R.S. Lee, J .E. Fisher, A.M. Rao, P.C. Eklund
and RE Smalley, “Large-Scale Purification of Single-Wall Carbon Nanotubes: Process,
Product, and Characterization”, Applied Physics A, Vol. 67, pp. 29-37 (1998)

[35] MS. Gudiksen, L.J. Lauhon, J. Wang, D.C. Smith, and CM. Lieber, “Growth of
Nanowire Superlattice Structures for Nanoscale Photonics and Electronics”, Nature, Vol.
415, pp.6l7-620 (2002)

[36] CM. Lieber, “Nanowire Superlattices”, Nano Letters, Vol. 2, pp. 81-82 (2002)

[37] J. Hu, M. Ouyang, P. Yang, and C. M. Lieber, “Controlled Growth and Electrical

Properties of Heterojunctions of Carbon Nanotubes and Silicon Nanowires”, Nature, Vol.
399, PP. 48-51 (1999)

95

[38] Y. Wu, R. Fen, and P. Yang, “Block-by-Block Growth of Single-Crystalline Si/SiGe
Superlattice Nanowires”, Nano Letters, Vol. 2, pp. 83-86 (2002)

[39] MT. Bjork, B.J. Ohlsson, T. Sass, A.I. Persson, C. Thelander, M.H. Magnusson, K.
Deppert, L.R. Wallenberg, and L. Samuelson, “One-Dimensional Heterostructures in
Semiconductor Nanowhiskers”, Applied Physics Letters, Vol. 80, pp. 1058-1060 (2002)

[40] MT. Bjork, B.J. Ohlsson, T. Sass, A.I. Persson, C. Thelander, M.H. Magnusson, K.
Deppert, L.R. Wallenberg, and L. Samuelson, “One-Dimensional Steeplechase for
Electrons Realized”, Nano Letters, Vol. 2, pp. 87-89 (2002)

[41] YR Zhang, Y.H. Tang, Y. Zhang, C.S. Lee, I. Bello, and ST. Lee, “Deposition of
Carbon Nanotubes on Si Nanowires by Chemical Vapor Deposition”, Chemical Physics
Letters, Vol. 330, pp. 48-52 (2000)

[42] G. A. Hebner, “Spatially resolved Excited State Densities and Neutral and Ion
Temperatures in Inductively Coupled Argon Plasmas”, Journal of Applied Physics, Vol.
80, Pp. 2624-2636 (1996)

[43] P. Hargis Jr., K. Greenberg, P. Miller, J. Gerado, J. Torczynski, M. Riley, G.
Hebner, J. Roberts, J. Olthoff, J. Whetstone, R. Van Brunt, M. Sobelewski, H. Anderson,
M. Splichal, J. Mock, P. Bletzinger, A. Butterbaugh, M. Brake, M. Passow, J. Pender, A.
Lujan, M. Elta, D. Graves, H. Sawin, M. Kushner, J. Verdeyen, R. Horwath, and T.
Turner, “The GEC Reference Cell: A Defined Parallel-Plate RF—System for Experimental
and Theoretical Studies of Plasma Processing Discharges”, Review of Scientific
Instruments, Vol. 65, pp. 140-154 (1994)

[44] C]. Collard, “Growth of CNT in an Inductively Coupled Plasma”, Chapter 4, PhD.
Thesis, University of Michigan (2002)

[45] B. Qi, R.M. Gilgenbach, Y.Y. Lau, M.D. Johnston, 1. Lian, L.M. Wang, G.L. Doll,
and A. Lazarides, “Ablation Plasma Ion Implantation Experiments: Measurement of Fe
Implantation Into Si”, Applied Physics Letters, 78 pp. 3785-3787 (2001)

[46] B. Qi, Y.Y. Lau, and RM. Gilgenbach, “Extraction of Ions From the Matrix Sheath
in Ablation-Plasma Ion Implantation”, Applied Physics Letters, 78 pp. 706-708 (2001)

[47] B. Sundaray, V. Subramanian, T.S. Natarajan, R.-Z. Xiang, C.-C. Chang, and W-S.
Fann, “Electrospinning of continuous aligned polymer fibers”, Applied Phys. letters,
Vol. 84, pp. 1 222-1224 (2004)

[48] G. Tibbetts, “Applications Research on Vapor Grown Carbon Fibers”, Science and

Application of Nanotubes, pp. 35-51, , Eds. D. Tomanek and R. Enbody, Kluwer
Academic Press, 2000, New York

96

[49] P. Corio, S.D.M. Brown, A. Marucci, M.A. Pimenta, K. Kneipp, G. Dresselhaus, and
MS Dresselhaus, “Surface-Enhanced Resonant Raman Spectroscopy of Single-Wall
Carbon Nanotubes Adsorbed on Silver and Gold Surfaces”, Physical Review B, Vol. 61,
pp. 13202-13211 (2000)

[50] L. C. Qin, T. Ichihashi and S. lijima, “On the Helicity of Carbon Nanotubes”,
Ultramicroscopy, Vol. 67, pp. 181-189 (1996)

[51] GS. Oerhlein, J.F. Rembetski and EH. Payne, “Study of Sidewall Passivation and
Microscopic Silicon Roughness Phenomena in Chlorine-based Reactive Ion Etching of
Silicon Trenches”, Journal of Vacuum Science and Technology B, Vol. 8, pp.1199-1211
(1990)

[52] El. Givargizov, A.N. Stepanova, L.N. Obolenskaya, E.S. Mashkova, V.A.
Molchanov, M.E. Givargizov, and I.W. Rangelow, “Whisker Probes”, Ultramicrosc0py,
Vol.. 82, pp. 57-61 (2000)

[53] PS. Nam, L.M. Ferreira, T.Y. Lee, and K.N. Tu, “Study of Grass Formation in
GaAs Backside Via Etching Using Inductively Coupled Plasma System”, Journal of
Vacuum Science and Technology B, Vol. 18, pp. 2780-2784 (2000)

[54] J .E. Parrneter, R.J. Shul, A.J. Howard, and RA. Miller, “Treatment of InP Surfaces
in Radio Frequency H-2 and H-2JCH4/Ar Plasmas: In Situ Compositional Analysis, Etch
Rates, and Surface Roughness”, Journal of Vacuum Science and Technology B, Vol. 14,

pp. 3563-3574 (1996)

[55] NJ. Pinto, A.T Johnson, A.G. MacDiarmid, C.H. Mueller, N. Theofylaktos, D.C.
Robinson, and FA. Miranda, “Electrospun Polyaniline/Polyethylene Oxide Nanofiber
Field-Effect Transistor”, Applied Physics Letters, Vol. 83 (20) pp. 4244-4246 (2003)

[56] J .P. Bibn'ng and F. Rocard, “Ion-Irnplantation Phenomena in Space”, Radiation
Effects and Defects in Solids, Vol. 65 (1-4) pp. 159-165 (1982); M. Maurette, “Ion-
Irnplantation Effects in Space”, Nuclear Instruments and Methods, Vol. 132, pp. 579-586
(1976)

[57] C. Ruehlicke, MA Briere, and D. Schneider, “AFM Studies of a New Type of
Radiation Defect on Mica Surfaces, Caused by Highly-Charge Ion Impact”, Nuclear
Instruments and Methods in Physics Research B, Vol. 99 (1-4) 528-531 (1995)

[58] C. Ascheron, J .P. Biersack, P. Goppeltlanger, and J. Erxmeyer, “Influence of
Ionization Processes on Radiation Defect Formation”, Nuclear Instruments and Materials

in Physics Research B, Vol. 80 (1) pp. 3—6 (1993)

[59] “Handbook of Accelerator Physics and Engineering”, edited by Alexander Wu Chao
and Maury Ti gner, World Scientific, River Edge, NJ (1999)

97

[60] Schindler group, Department of Biochemistry and Molecular Biology

[61] D.M. Fleetwood, P.S. Winokur, and RE. Dodd, “An Overview of Radiation Effects
on Electronics in the Telecommunications Environment”, Microelectronics Reliability
Vol. 40, pp. 17-26 (2000)

[62] Z.L. Wang, P. Poncharal, and W.A. de Heer, “N anomeasurements of Individual
Carbon Nanotubes by In Situ TEM”, Pure and Applied Chemistry, Vol. 72 (1-2) pp. 209-
219 (2000)

[63] Z.L. Wang, P. Poncharal, W.A. de Heer, Z.R. Dai, and Z.W. Pan, “Mechanical and
Electrostatic Properties of Carbon Nanotubes and Nanowires”, Materials Science and
Engineering C, Vol. 16 (1-2) pp. 3-10 (2001)

[64] LA. Bursill, J .-L. Peng, and X.-D. Fan, “Cross-Sectional High-Resolution
Transmission Electron Microscopy Study of the Structures of Carbon Nanotubes”,
Philosophical Magazine A, Vol. 71 (5) pp. 1161-1176 (1995)

[65] D.A. Evans and M.A. Crook, “Irradiation of Plastics: Damage and Gas Evolution”,
MRS Bulletin Special Issue: Materials Performance in a Radiation Environment, Vol. 22,
pp. 36—41 (1997)

[66] TD. Burchell, “Radiation Effects in Graphite and Carbon-Based Materials”, MRS

Bulletin Special Issue: Materials Performance in 3 Radiation Environment, Vol. 22, pp.
29-35 (1997)

98

   

1[111111le111111111111111]!