Parolini PhD Thesis - core.ac.uk fileParolini PhD Thesis - core.ac.uk
Tsvaygboym PhD Thesis 2007 - BW
-
Upload
nanotech2masses -
Category
Education
-
view
215 -
download
6
Transcript of Tsvaygboym PhD Thesis 2007 - BW
RICE UNIVERSITY
Photochemical Studies of
Single-Walled Carbon Nanotube Ozonides and -Azoxy Ketones
by
Konstantin Tsvaygboym
A THESIS SUBMITTED
IN PARTIAL FULFILLMENT OF THE
REQUIREMENTS FOR THE DEGREE
Doctor of Philosophy
APPROVED, THESIS COMMITTEE:
Paul S. Engel,
Professor of Chemistry
W. Edward Billups,
Professor of Chemistry
Michael R. Diehl,
Assistant Professor of Bioengineering
HOUSTON, TEXAS
APRIL 2007
Volume I of II
ABSTRACT
Photochemical Studies of
Single-Walled Carbon Nanotube Ozonides and -Azoxy Ketones
by
Konstantin Tsvaygboym
This thesis contributes to two disparate problems in chemistry: studying properties of
carbon nanotube ozonides and products of their decomposition and determining behavior
of -azoxy radicals.
This work demonstrates that interaction of ozone with single-walled carbon
nanotubes (SWNT) results in formation of 1,2,3-trioxolanes (SWNTO3). Their formation
rate was found to be on the order of subseconds at room temperature for diluted SWNT -
1% aqueous SDS suspensions. SWNTO3 decayed to SWNT epoxides (SWNTO) with
release of molecular oxygen. Gas evolution measurements performed on dry ozonated
SWNT showed oxygen release to follow a simple exponential rise with rates
approximately 1.5 – 2 min-1
at r. t. The lifetime of SWNTO3, with a dissociation
activation energy of approximately 0.7 eV, depends on temperature and SWNT type. At
room temperature, it is less than two minutes for small-diameter SWNTs suspended in
water. Ozonides exhibited extreme quenching of SWNT fluorescence and substantial
bleaching of NIR absorption. The maximum number of 1,2,3-trioxolanes forming on the
surface of SWNT at any given time was found to be less than 4% of the theoretical value,
indicating a saturation point. Reaction of ozonated nanotubes with excess ozone is
limited by the SWNTO3 decomposition rate. Thinner tubes exhibited faster ozonide
decay rates resulting in greater oxidation levels over time in excess of ozone. Ozonation
with small quantities of ozone did not result in a D-band increase in the Raman spectra,
both for solid and liquid state experiments, though substantial decrease of the G band was
observed. IR absorbance kinetics of SWNT films revealed exponential intensity drift over
time with rates close to those in fluorescence and NIR absorbance techniques. Ozonated
SWNTs were found to abstract electrons from amines and thiols, thus resulting in
covalent attachment of nucleophiles to the sidewall.
The azoxy functional group greatly stabilizes an attached carbon-centered radical,
but the chemistry of such -azoxy radicals is unclear. This work reports that generation
of -azoxy radicals by irradiation of -azoxy ketones PhCO-C(Me)2-N=N(O)-R causes
ketone rearrangement to azoester compounds PhCOO-C(Me)2-N=N-R. This study
proposes a mechanism for this rearrangement.
Acknowledgments
I am grateful to my advisor, Prof. Paul S. Engel for allowing me to work on an
exciting, cutting edge project revolving around carbon nanotube ozonides. I have been
honored to work with a number of faculty, post docs, graduate and undergraduate
students, who immensely deepened my understanding of scientific principles and fostered
my teaching skills. There is no doubt some of them will become leading figures in
science, technology and business.
I would like to thank friends and relatives who were very supportive throughout my
graduate studies. Your help and advice are much appreciated.
Table of Contents
Volume I
Title Page i
Abstract iii
Acknowledgments v
Table of Contents vi
List of Symbols and Abbreviations ix
Part I
Chapter 1. Spectral and physical characteristics of reference SWNT samples 2
Introduction 3
References and Notes 12
Chapter 2. Carbon nanotube ozonides: formation rates, oxygen evolution,
decomposition rates and activation energies, determination of
saturation limits and a comparison of spectral changes in
fluorescence and UV-Vis-NIR absorption
13
Introduction, Results and Conclusions 14
Experimental Part 82
References and Notes 88
Chapter 3. Influence of SWNT ozonation on D and G bands in Raman spectra 91
Introduction, Results and Conclusions 92
vii
Experimental Part 106
References and Notes 107
Chapter 4. IR studies of SWNT ozonides and of products of their reactions
with different classes of compounds
109
Introduction, Results and Conclusions 110
Experimental Part 134
References and Notes 137
Chapter 5. Reaction of ozonated SWNT with electron rich nucleophiles
(amines, thiols and other)
139
Introduction, Results and Conclusions 140
Experimental Part 170
References and Notes 174
Chapter 6. Trapping reactive centers on SWNTOn with electron rich
nucleophiles (amines, thiols)
177
Introduction, Results and Conclusions 178
Experimental Part 181
References and Notes 182
Chapter 7. Reactions between ozonated SWNT and different classes of
compounds studied by X-ray photoelectron spectroscopy
183
Introduction, Results and Conclusions 184
Experimental Part 197
References and Notes 201
viii
Part II
Chapter 1. Photorearrangement of -Azoxy Ketones and Triplet Sensitization
of Azoxy Compounds
203
Introduction, Results and Conclusions 204
Experimental Part 221
References and Notes 227
Volume II
Appendix A Mathematics for regression analysis of fluorescence and NIR
absorbance data
235
Appendix B Supporting Information for Part I, Chapter 5. 1H NMR spectrum 251
Appendix C Supporting Information for Part I, Chapter 7. XPS spectra for
reactions of ozonated SWNT with different classes of compounds
253
Appendix D Supporting Information for Part II, Chapter 1. Calculated isotropic
Fermi contact couplings, computed structures, ESR, UV and NMR
spectra
323
ix
List of Symbols and Abbreviations
a.u. absorbance units
abs absorbance
ATR FT-IR attenuated total reflectance Fourier transform infrared
C60 fullerene C60
ca. Latin word for approximately
DTT dithiothreitol
ESCA electron spectroscopy for chemical analysis
em emission
ex excitation
HipCo high pressure carbon monoxide method
HOMO highest occupied molecular orbital
Imax maximum intensity
I/Imax normalized value(s)
Imax/I quenching factor, a degree of quenching, inverted normalized value(s)
absmax
local absorption maximum (spectral)
emmax
local emission maximum (spectral)
LUMO lowest unoccupied molecular orbital
NIR near IR
(n,m) carbon nanotube indices
O3 ozone
PM3 parametric method No. 3
x
PTFE polytetrafluoroethylene
r2 coefficient of determination, same as correlation coefficient
RBM radial breathing mode
SDS sodium dodecyl sulfate
SDBS sodium dodecyl benzyl sulfonate
SWNT single-walled nanotube
SWNTO3 product(s) of ozonation of single-walled carbon nanotube
lifetime
TMPD N,N,N’,N’-tetramethyl-p-phenylenediamine
uL microliter(s)
Wurster reagent N,N,N’,N’-tetramethyl-p-phenylenediamine (same as TMPD)
XPS X-ray photoelectron spectroscopy (same as ESCA)
Part I
Chapter 1
Spectral and physical characteristics of reference SWNT samples
3
1.1. Introduction
Single walled carbon nanotubes (SWNTs), a graphene sheet rolled up into a tubular
shape, may turn out to be a promising material for electronics, field emission, heat
transfer, sensing, material reinforcement, imaging, medicinal and other applications.1-3
Research in the area of carbon nanotubes increased significantly in the last several years
and is highly competitive, partly due to possible commercialization of their unique
properties. This chapter provides a brief introduction to key aspects of the spectroscopic
measurements of single walled carbon nanotubes (SWNT) discussed throughout this
thesis. Spectroscopic changes of SWNT after functionalization may not have the same
behavior as would be expected for a small molecule. An interesting example of this can
be found in Chapter 4 discussing IR absorption changes of SWNT over time after
ozonation. Chapter 1 contains an interconversion table of wavelengths and wavenumbers
that will be of use in Chapter 3, describing the Raman measurements performed on
aqueous SWNT suspensions as well as for discussion of IR results. Also, SWNT
fluorescence spectra obtained with different excitation sources are shown deconvoluted.
A brief table summarizes how much each tube contributes to the observed fluorescence
intensity. Other aspects like nomenclature and 3D structure are discussed as well. The
following section provides UV and NIR absorption spectra and talks about work with
different batches from the HipCo reactor (Rice University).
1.2. Near-IR fluorescence spectra
Two lasers, 660 nm and 785 nm were used for excitation of single-walled carbon
nanotubes (SWNT), the former one utilized for the majority of the spectra presented.
Wavenumber and wavelength scales are used interchangeably in this work. Table 1
4
shows the relation of the two scales. Specific Raman shifts from the 669.9 nm excitation
source are also included.
Table 1. Interconversion of wavelengths and wavenumbers for Visible, NIR and IR
regions. Raman shifts from 669.9 nm excitation source are provided.
Range , nm , cm-1
Shift, cm-1
Visible 669.9 14928 0
700.0 14286
733.5 13633 1294 (D)
749.5 13342 1585 (G)
785.0 12739
811.0 12330 2597 (G’) 830.0 12048
NIR 900.0 11111
1000 10000
1100 9091
1200 8333
1300 7692
1400 7143
1429 7000
IR 1500 6667
1600 6250
2000 5000
2500 4000
3333 3000
5000 2000
8333 1200
9091 1100
10000 1000
11111 900
12500 800
Aqueous SWNT-SDS suspensions are known to fluoresce when excited with suitable
lasers. Spectra obtained after excitation with 660 and 785 nm lasers are shown in Figure
1. The spectra were deconvoluted and peaks of interest assigned (n,m) numbers according
to published data.4
5
Norm
aliz
ed F
luore
scence
0.00
0.25
0.50
0.75
1.00
ex = 660 nm8,3
7,5
10,2
7,6
9,5 10,3
8,7
11,1
Optical frequency (cm-1
)
75008500950010500
Norm
aliz
ed F
luore
scence
0.00
0.25
0.50
8,3 7,510,2
11,310,5
8,7
6,57,6
9,7
ex = 785 nm
Figure 1. Fluorescence of aqueous SWNT-SDS suspensions. Tubes of interest are
marked with (n,m) numbers. The same (n,m) tube is shown with the same color and
symbol on both graphs. Top: excited with 660 nm. Bottom: excited with 785 nm laser.
Fluorescence changes in spectra obtained with em 660 nm were examined at four
distinct wavelengths: 954, 1027, 1125 and 1250 nm. The major contributors to
fluorescence intensity at each wavelength are summarized in Table 2.
6
Table 2. Major contributors to fluorescence intensity at four distinct wavelengths.*
em, nm , cm-1
(n,m) type of
major
contributors
tube diameter,
nm
% of total
emission at em
955.6 10465 8,3 0.782 95.4
6,5 0.757 1.9
1027.6 9731 7,5 0.829 85.0
10,2 0.884 5.3
8,1 0.678 4.3
1124.6 8892 7,6 0.895 78.9
8,4 0.840 8.3
9,2 0.806 3.7
9,4 0.916 3.4
1250.1 8000 9,5 0.976 39.8
10,3 0.936 30.3
11,1 0.916 12.0
8,7 1.032 6.2
10,5 1.050 3.8
8,6 0.966 3.0
* Excitation source ex
max 660 nm.
Minor contributors were excluded from Table 2 for clarity. Tube (8,3) contributed
95% of peak intensity at 954 nm, as deduced from spectrum deconvolution.5
Analogously, 85 % of peak intensity at 1027 nm was from tube (7,5). Tube (7,6) gave
only 79% of peak intensity at 1125 nm. The peak at 1250 nm was from a combination of
tubes, none contributing more than 40 % of total intensity.
Assignment of numbers (n,m) for carbon nanotubes is summarized in Figure 2.
7
Basis vectors
1,0 2,0 3,0 4,0 5,0 6,0 7,0 8,0 9,0 10,0 11,0 12,0
1,1 2,1 3,1 4,1 5,1 6,1 7,1 8,1 9,1 10,1 11,1 12,1
2,2 3,2 4,2 5,2 6,2 7,2 8,2 9,2 10,2 11,2
3,3 4,3 5,3 6,3 7,3 9,3 11,3
4,4 5,4 6,4 7,4 8,4 9,4 10,4
5,5 6,5 8,5 10,5
0,0
6,6 8,6 9,6 10,6
7,7 8,7 9,7
Armchair
ZigzagChiral
angle
13,1
12,2
12,3
11,4
11,5
7,7
10,6
13,0
8,7 9,7 10,7
8,8 9,8
8,3
7,5
7,6
9,5
10,38,3n = 8
m = 3
Roll-up vector
Basis vectors
1,0 2,0 3,0 4,0 5,0 6,0 7,0 8,0 9,0 10,0 11,0 12,0
1,1 2,1 3,1 4,1 5,1 6,1 7,1 8,1 9,1 10,1 11,1 12,1
2,2 3,2 4,2 5,2 6,2 7,2 8,2 9,2 10,2 11,2
3,3 4,3 5,3 6,3 7,3 9,3 11,3
4,4 5,4 6,4 7,4 8,4 9,4 10,4
5,5 6,5 8,5 10,5
0,0
6,6 8,6 9,6 10,6
7,7 8,7 9,7
Armchair
ZigzagChiral
angle
13,1
12,2
12,3
11,4
11,5
7,7
10,6
13,0
8,7 9,7 10,7
8,8 9,8
8,3
7,5
7,6
9,5
10,38,3n = 8
m = 3
Basis vectors
1,0 2,0 3,0 4,0 5,0 6,0 7,0 8,0 9,0 10,0 11,0 12,0
1,1 2,1 3,1 4,1 5,1 6,1 7,1 8,1 9,1 10,1 11,1 12,1
2,2 3,2 4,2 5,2 6,2 7,2 8,2 9,2 10,2 11,2
3,3 4,3 5,3 6,3 7,3 9,3 11,3
4,4 5,4 6,4 7,4 8,4 9,4 10,4
5,5 6,5 8,5 10,5
0,0
6,6 8,6 9,6 10,6
7,7 8,7 9,7
Armchair
ZigzagChiral
angle
13,1
12,2
12,3
11,4
11,5
7,7
10,6
13,0
8,7 9,7 10,7
8,8 9,8
8,3
7,5
7,6
9,5
10,38,3n = 8
m = 3
8,3n = 8
m = 3
Roll-up vector
Figure 2. Construction of a nanotube from a graphene sheet. Numbers n and m determine
the final position of a roll-up vector. Rolling sheet to superimpose hexagons (0,0) and
(8,3) will result in tube (8,3) with roll-up vector being perpendicular to tube direction.
Tubes of interest are emphasized with thick hexagons.
The physical structures of tubes of interest are shown in Figure 3.
(8,3) (7,5) (7,6) (9,5) (10,3)
Figure 3. Tubes (n,m) with the highest fluorescence intensity in HipCo samples for 661
nm excitation source. Each tube is shown in two projections (top and bottom).
8
It is important to note that there is no linear relationship between (n,m) tubes’
relative concentrations and their emission intensities for any given ex . This is because
SWNT fluorescence intensity is dependent on the wavelength of incident light. For
example, tubes (8,3), (7,5) and (7,6) with the highest emission intensity in the ex 660 nm
spectrum (Figure 1) are only a small fraction of a bulk sample (Figure 4).
6,4
9,1
8,3
6,5
7,3
7,5
8,1
10,2
9,48,4
7,6
9,2
12,1
8,6
11,3
9,5
10,3
10,5
11,1
8,7
14,0
13,2
9,7
12,411,4
12,2
10,6 11,6
9,8
15,1
14,3
10,8
13,512,5
13,3
14,1
11,7 12,7
10,9 11,9
15,2
16,0
14,4
5,0 7,0 8,0 10,0 11,0 13,0
5,1 6,1
6,2 7,2
5,3
5,4
12,8
11,10
13,6
armchair
zigzag
Figure 4. Distribution of (n,m) species in HipCo SWNT sample calculated from emission
spectra with ex 660 and 785 nm.6 Thickness of a hexagon is linearly proportional to tube
abundance in the sample.
Relative abundances of tubes were estimated by recording two separate emission
spectra with ex 660 and 785 nm. The knowledge of (n,m) tube abundance is of great
importance for absorption studies where measurements are performed on a bulk sample.
For example, if the bulk sample has two types of species, A and B, which transform over
time, independently of each other, into species A’ and B’ with corresponding rates c and
9
d, an overall absorbance can be expressed with a first order equation
dtct ebeatAbs )(
where a and b are Arrhenius prefactors derived from tube abundances. Typical HipCo
SWNT samples are estimated to have over forty different semiconducting tubes and
about fifteen metallic tubes. This means that observed absorbance can be affected by as
many as fifty five different species in a sample. Knowing relative abundances of specific
(n,m) tubes may help interpret absorbance kinetics.
Since metallic tubes do not fluoresce, their number is only an estimate. Studies of
SWNT radial breathing modes (RBM) in Raman spectra served as a basic for the relation
of abundances of metallic and semiconducting tubes.
Discussion of the mathematics behind (n,m) tube relative abundance calculations,
based on fluorescence emission spectra, is beyond the scope of this work and is not
included.5
1.3. UV-Visible and Near-IR absorption spectra
UV-Vis absorption spectrum for SWNT (HipCo, batch 162.4, Rice University) is
provided below:
10
Wavelength (nm)
300 400 500 600 700
Ab
so
rban
ce
(a
.u.)
0.2
0.4
0.6
Figure 5. UV-Vis absorption spectrum of aqueous SWNT – SDS suspension.
Absorption peaks in the area 450-550 nm are commonly assigned to metallic tubes.
Peaks in the area 650-750 nm are commonly assigned to semiconducting tubes.
NIR absorption of SWNT is thought to be caused by a conjugated network of double
bonds. It is not clear if the conjugated acene system in SWNT can be considered truly
aromatic. Hückel molecular orbital (HMO) theory states planarity as one of the most
important prerequisites of aromaticity. Ozonation of SWNT sidewall results in significant
decrease of NIR absorption. NIR absorption spectrum of pristine SWNT is provided
below.
11
Wavelength (nm)
900 1000 1100 1200 1300
Absorb
ance (
a.u
.)
0.15
0.20
0.25
Figure 6. NIR absorption spectrum of aqueous SWNT-SDS suspension.
Note the difference in the vertical scale for the above two spectra
1.4. Other spectra
Other reference spectra of SWNT, like IR, solid and liquid Raman and ESCA will be
introduced throughout the text.
1.5. Properties of different batches of SWNT
Different batches of SWNT from the HipCo process (Rice University) were used in
this work. All batches had similar or identical spectroscopic properties. Batch 153.3 was
used for fluorescence studies of the reaction between 2-methoxyethylamine and ozonated
SWNT. Batches 162.4 and 162.8 were used for IR studies. Batch 161.1 was used for UV,
liquid Raman and fluorescence studies. The majority of SWNT samples in this work were
used as synthesized, without purification. Unless otherwise noted, tubes were pristine.
SWNT – 1 wt. % aq. SDS suspension was prepared by a standard procedure outlined in
12
the experimental part. SWNT bundles, carbonaceous matter, metal catalyst and other
impurities are thought to be removed from the final SWNT – SDS suspension. Unless
otherwise noted, all SWNT – SDS samples used in this work were prepared by the same
procedure. Typically a large stock of SWNT – SDS suspension was prepared and used
for a great number of experiments.
1.6. References and Notes
1. Baughman, R. H.; Zakhidov, A. A.; de Heer, W. A., Carbon nanotubes - the route
toward applications. Science 2002, 297, (5582), 787-792.
2. Avouris, P., Molecular electronics with carbon nanotubes. Accounts of Chemical
Research 2002, 35, (12), 1026-1034.
3. Calvert, P., Nanotube composites - A recipe for strength. Nature 1999, 399, (6733),
210-211.
4. Weisman, R. B.; Bachilo, S. M., Dependence of optical transition energies on
structure for single-walled carbon nanotubes in aqueous suspension: An empirical
Kataura plot. Nano Letters 2003, 3, (9), 1235-1238.
5. Deconvolution performed with software package that accompanied NS1
NanoSpectralyzer (Applied NanoFluorescence LLC.).
6. Applied NanoFluorescence LLC http://www.appliednanofluorescence.com/.
13
Chapter 2
Carbon nanotube ozonides: formation rates, oxygen evolution,
decomposition rates and activation energies, determination of
saturation limits and a comparison of spectral changes in fluorescence
and UV-Vis-NIR absorption
14
2.1. Introduction
A number of publications have been dedicated specifically to ozonation of carbon
nanotubes. Recently, Chen1, 2
reported that 9 wt. % O3 in O2 bubbled through SWNT
suspension in perfluoropolyether (PFPE) at r. t. for periods ranging from 1 to 8 hours,
followed by a 30 minute purge with oxygen, resulted in SWNT shortening. Simmons et
al. 3 studied ozonation as a possible tool to selectively decrease conductivity of SWNT on
a microfabricated chip upon UV/ozone exposure. Samples exposed for one hour at r. t.
were shown to form characteristic carbonyl and ether bonds (XPS data), and SWNT
electrical resistance increased. The provided Raman spectra show D and G bands at
different times. After ten minutes of UV/ozone exposure, the G band decreased ca. five
times, but the D band did not change. The authors concluded that sidewall oxidation by
ozone and molecular oxygen resulted in - conjugated network disruption. Banerjee et
al.4-6
conducted a series of studies on ozonation of carbon nanotubes. The author noted
that Raman spectra of carbon nanotubes are strongly resonance enhanced, and as a result
signals from the functionalizing moieties are rarely seen in Raman spectra.4, 7
In a
different study, SWNT sidewall was ozonated (ca. ~10% O3 in O2) in a methanolic
suspension (100 mg in 150 mL) at -78 C for one hour and reacted with “cleaving”
reagents (either sodium borohydride or dimethyl sulfide).5 The authors assumed
formation of ozonides, by an analogy with alkenes, pointing out that C60O3 has been
reported in the literature.8 The “cleaving” step was introduced to alter relative distribution
of products (ethers, carbonyls and esters). The authors concluded that SWNT ozonation
15
could be used as a nondestructive method of introducing oxygenated functionalities
directly onto the sidewall.
In another study6 Banerjee et al. demonstrated that after solution phase ozonolysis of
SWNT (ethanolic suspensions, 2 hours), Raman peaks corresponding to smaller diameter
tubes were relatively diminished in intensity when compared to the profile of larger
diameter tubes. The author found no chiral selectivity (i.e. dependence on tube “twist,”
Figure 3, Chapter 1) and concluded that tube curvature and -orbital misalignment are the
main reasons for the observed selectivity. A theoretical study providing activation
energies for a reaction of ozone with SWNT has been reported.9 Cai et al.
10 reported
ozonation of SWNT and their assembly on top of oligo(phenylene ethynylene) self-
assembled monolayers. Oxidation produced oxygenated functional groups like carboxylic
acids, esters and quinone moieties. Depending on the degree of ozonation, the electrical
resistance was found 20 to 2000 times higher than that of pristine SWNT. Oxidation was
performed on a dry “bucky” paper with UV/O3 generator in ambient air for 25 minutes to
5 hours. Ozonated SWNT absorption in the IR region was shown to stop changing after 3
hours of ozonation. An IR peak at 1580 cm-1
was assigned to the stretching mode
(C=C) of double bonds in the nanotube backbone near functionalized carbon atoms.11
Ogrin et al.12
estimated an approximate molecular formula of SWNT ozonated for 3
hours to be C6O, i.e. every third double bond had an epoxide. None of the mentioned
publications focuses on SWNT ozonides kinetics.
A number of articles have been published on ozonation of fullerenes, a short analog
of SWNT, and their properties.13-19
Chibante and Heymann determined products of
16
ozonation of C60 in toluene solution included structures C60On, with n ranging from 1 to
6, and insoluble tan-colored precipitates.20
Bulgakov et al.21
found that epoxides C60On
(n = 1 – 6) are accumulated within the first three minutes of continuous ozonation.
Further ozone/oxygen mixture bubbling resulted in formation of ketone and ester
functional groups. Heymann et al.8 found that at 23 C ozonide C60O3 had a lifetime ca.
22 minutes in toluene, 330 minutes in a dry state and 770 min in octane.
Razumovskii et al.18, 19
reported ozonide formation rates for C60O3 (8.8 104 M
-1s
-1
at 0 C) and C70O3 (5 104 M
-1s
-1 at 22 C) in CCl4 solvent. The authors found that the
reactions obeyed a bimolecular rate law. The reactivity of C60 with ozone decreased ca.
90 times after the formation of C60O3. A similar tendency was found for C70, where the
formation of the first ozonide was 6 – 8 times faster than the subsequent ones. Fullerene
C70 was shown to uptake only 12 molecules of ozone within the first 16 minutes of
continuous O3/O2 gaseous mixture bubbling. The authors concluded that the formation of
the ozonide exerts an electronegative inductive effect on the adjacent network of
conjugated double bonds, similar to ozonation of divinylbenzene.22
Kinetics of SWNT ozonides have not been published to date. Among the reasons,
there are: different production methods resulting in different (n,m) types of SWNT in a
batch sample, a presence of a large number of different tubes in each SWNT sample,
poor solubility of SWNT in solvents, the need for efficient purification from the metal
catalyst, different purification techniques affect differently chemical and physical
properties of SWNT. Measuring kinetics on SWNT is a challenge. This chapter will
describe some interesting research findings discovered while attempting to study kinetics
17
of SWNT ozonides. Topics like deoxygenation of SWNT ozonides, NIR fluorescence
quenching degree, influence of high and low load of ozone on SWNTO3 decomposition
rates, proposed electronic transitions in SWNT and SWNTO3, decomposition rate
dependence on tube diameter, saturation limits in excess ozone, comparison of NIR
fluorescence and NIR absorption kinetics, establishing an average decomposition rate by
UV, structural changes and decomposition activation energies will be discussed.
2.2 Results and Discussion
A set of experiments was designed to measure the amount of oxygen evolving from
the surface of ozonated SWNT. No such study has been reported to date, even though a
number of articles on SWNT ozonation have been published.
Results of the experiment are summarized in Figure 1.
18
Time (min)
0 2 4 6 8 10
Pre
ssure
(m
To
rr)
0
25
50
75
100
Pressure (P1)
Pressure (P21
)
Pressure (P22
)
NIR
Absorb
an
ce (
a.u
.)
1.44
1.46
1.48
1.50
1.52
NIR Abs (A)
Pre
ssure
(m
To
rr)
0
25
50
75
100
P1
P21
P22
A
Figure 1. (Top) Pressure change at r. t. due to oxygen release from 2 () and 4 mg (
and ) of ozonated dry SWNT-coated glass (upper three curves) and corresponding
system-leak references (lower two curves, and ). (Bottom) NIR Absorbance
recovery of ozonated SWNT in solid form monitored at 1450 nm and r. t. (upper curve,
) and pressure change at r. t. due to oxygen release from 2 () and 4 mg ( and ) of
ozonated SWNT in solid state (lower three curves) after system leak correction. Curves
and were measured after the first and the second ozonation of the same sample
correspondingly.
Slurry of 2 or 4 mg of SWNT (as noted in Figure 1) in benzene (ca. 10 mL) was
added to the reaction vessel and was kept rotating until all the solvent was evaporated.
19
Such circular motion resulted in a thin SWNT film along the entire reaction vessel. A
vacuum line was degassed overnight, then the vessel was cooled to 5 C and 10 mL of
O3/O2 gaseous mixture (ca. 3 v/v % ozone23
) was injected to the bottom of the cylinder,
the cap closed and the vessel was left at atmospheric pressure for one minute. The valve
on the vessel was opened to the vacuum system and the vessel was evacuated for 1.5 min,
after which the pump was cut off and data were acquired. Degassing for one and half
minutes was found sufficient to bring the vacuum in the entire system to below 1 mTorr.
Time t = 0 min in Figures 1 and 2 indicates the point when the pump was cut off from the
system.
Time (min)
0 2 4 6 8 10
Pre
ssu
re (
mT
orr
)
0
25
50
75
100
Pressure (P1)
Pressure (P21
)
Pressure (P22
)
P1
P21
P22
A
NIR
Ab
so
rba
nce
(a
.u.)
1.44
1.46
1.48
1.50
1.52
NIR Abs (A)
Figure 2. Regression curves for NIR Absorbance at abs 1450 nm and for pressure
changes after SWNT ozonation in a dry state. Curves P1 () and P21 () correspond to
first ozonation of 2 and 4 mg of SWNT respectively. Curve P22 () was measured after
the sequential ozonation of 4 mg sample.
Cutting off the vacuum pump was followed by removing the ice bath and warming
the reaction vessel to r. t. with a water bath. Data points were collected until the observed
20
deoxygenation rate decreased to below the system leak rate value (ca. 0.5 mTorr/min).
The second sample (4 mg SWNT) was ozonated two times with approximately one hour
interval between oxidations.
The highest amount of oxygen evolved after gaseous ozonation of solid SWNT was
estimated as 0.72 umol within a 20 min time period at room temperature. This
corresponds to 0.2% of carbon atoms (or to 0.1% of double bonds) of SWNT (4 mg)
oxidized with ozone, assuming that all carbon soot was indeed SWNT or had a fullerene-
like structure. Weighing error of SWNT could bring an error into the calculated value. It
is possible that the number of double bonds reacted with O3 was higher, though it would
still be significantly less than the 3 – 4 %, estimated in UV studies at 260 nm. (NIR
absorbance estimation was at ca. 4 – 5 %). A possible explanation for such a low yield of
oxygen is SWNT bundling, which physically prevented large surface areas of SWNT
from reacting with gaseous ozone.
Ozonation of SWNT flakes resulted in their immediate burning. Ice bath cooling and
SWNT deposited along the glass wall of the reaction vessel were found necessary to
prevent this highly exothermic reaction from overheating.
NIR absorption was fitted with formula F1, while the pressure curves were fitted
with the 5-parameter two exponential rise formula F2 (formula selection discussed in
Appendix A):
minmin1
y
ceaey
yy
n
bt
bt
final
(F1)
21
)1()1(0n
bt
bt eceayy (F2)
Regression results are summarized in Table 1 below.
Table 1. Regression results for pressure changes and for NIR Absorption at
em 1450 nm after SWNT ozonation.*
Data Set Oxygen gas release NIR Absorption
Parameter F2, 4 mg (P21) F2, 4 mg (P22) F2, 2 mg (P1) F1
b, min-1
2.07 1.45 1.88 1.540
n 9.63 11.45 9 14.02
r2 0.9991 0.9987 0.9995 0.9995
ymin - - - 0.0000
* Ozonation of SWNT film deposited on a glass surface. The formula number and SWNT
amount used for the experiment are written at the head of each column. Rates b are
expressed in [min-1
]. Active constraints used in analysis were n > 9 (2 mg SWNT
pressure curve) and ymin > 0 (NIR absorption).
Points at time zero were excluded from regression because those were acquired at
5 C; all subsequent points were acquired at or near r. t. Constraints n > 9 and ymin > 0
were introduced to generate a better fit to the experimental data. Limiting n to greater
than nine was needed to better describe the term n
bt
ce , a “slow” component, for
pressure curve P1. Parameter n describes how many times the slow component is slower
than the fast one.
Approximately the same amount of ozone (O3/O2 gaseous mixture) was injected into
the reaction vessel in each experiment. The first time ozonation (P21) yielded a slightly
higher rate than the subsequent one (P22). All rates were comparable to those observed by
NIR fluorescence recovery, indicating that decomposition of a single ozonide is likely to
increase fluorescence intensity. This result means that the smallest section of SWNT
22
needed for a tube to fluorescence can be loaded with no more than one or two ozonides
on its surface, at least in an aqueous suspension.
Fluorescence studies demonstrated that 1,2,3-trioxolanes on the surface of SWNT
prevented the tube from emitting in the NIR region. If the “minimal” section of SWNT
needed for fluorescence carried several ozonides, all of them would have to decompose
before this section would gain its ability to fluoresce. If that were the case, then true
ozonide decay rates would be several times greater than those observed by fluorescence.
Observation of similar rates in vacuum deoxygenation of SWNTO3 and in fluorescence
techniques implies that decomposition of nearly every ozonide results in a fluorescence
increase.
It was found difficult to quench SWNT fluorescence completely. The highest
quenching degree (Imax/I) was less than 1000 times and tubes were shown to quickly
recover from that state. Quenching 1000 times means that 0.1% of previously emitting
“sections” of SWNT continued to fluoresce. Full fluorescence quenching was not
observed. A study was performed to investigate the fluorescence quenching degree
(Imax/I) as a function of the volume of injected O3/O2 gaseous mixture (ca. 3 v/v %
ozone). After excluding the most extreme points (i.e. the lowest intensity point after
ozonation), even with large amounts of ozone, such as 2 mL of O3/O2 gaseous mixture,
fluorescence could not be quenched more than 140 times (Figure 3).
23
O3 / O2 mixture volume (mL)
0.0 0.5 1.0 1.5 2.0
Flu
ore
sce
nce Q
ue
nchin
g (
I max/I
)
0
20
40
60
80
100
954 nm
1027 nm
1125 nm
1251 nm
Figure 3. Dependence of fluorescence quenching degree (Imax/I) on the amount of O3/O2
gaseous mixture (ca. 3 v/v % ozone) injected.
Figure 3 demonstrates that injection of 0.3 mL of O3/O2 gaseous mixture decreased
fluorescence intensity of tube (8,3) with emmax
954 nm approximately 6 times. In
percent values it means that only 17% of all emitting “sections” were contributing to
fluorescence. One would expect that increasing the ozone load by 20 % could nearly
completely extinguish fluorescence from the tube (8,3). Interestingly, injection of 0.5 mL
of O3/O2 mixture quenched fluorescence only 16 times, i.e. 6% of SWNT was still
emitting. Further increase of the ozone load to 1.0 mL quenched emission only 41 times,
with 2.4% of emitters still left to be quenched. To conclude, increasing ozone load from
0.3 mL to 1.0 mL, i.e. by 330%, could not extinguish the remaining 17% of emitting
sections of SWNT. This observation meant that tubes are getting oxidized with ozone in
bands and not randomly.
24
Changes in SWNT fluorescence after oxidation with ozone
Wavelength (nm)
950 1050 1150 1250 1350
No
rma
lize
d F
luo
resce
nce
In
tensity
0.0
0.2
0.4
0.6
0.8
1.0
Before O3
1 min
3 min
9 min
ex = 660 nm
Figure 4. Addition of aqueous solution of ozone (50 uL, Abs (260 nm, 1 cm) = 1.25 a.u.)
to 0.5 mL SWNT-SDS aq. suspension. Used 660 nm laser for excitation. Fluorescence
emission quenching was followed by a slow recovery. Spectra recorded before, 1, 3 and 9
min after ozonation.
The overlaid spectra in Figure 4 show SWNT fluorescence change over time after
ozonation. The spectrum of pristine SWNT is provided for comparison.
SWNT oxidation was accomplished by an addition of a small volume of water
saturated with ozone. It was desired to prepare a saturated solution of ozone, thus
decreasing the volume of ozonated water needed for oxidation. A dilution of SWNT-SDS
suspension was a concern, since dilution could result in SWNT agglomeration, thus
leading to lower fluorescence intensity. In general, bubbling O3/O2 gaseous mixture
through the solution was of a greater benefit, since in that case there was no need to
25
worry about sample dilution. While dilution with 1% SDS decreases SWNT fluorescence
intensity, no comprehensive study was performed in this work to estimate the influence
of dilution on fluorescence. Aqueous SDS solution was not used for ozone accumulation
primarily because this surfactant is a known catalyst for conversion of ozone into
molecular oxygen.
For ozonation in the solution phase, fluorescence quenched to a different extent (20
to 100 times) had fairly close recovery rates as shown in logarithmic scale in Figure 5B.
26
I /
I ma
x
0.0
0.2
0.4
0.6
0.8
1.0
Time (min)
3 6 9 12 15
I ma
x / I
(lo
g s
cale
)
1
2
4
10
20
40
100 B
A
Figure 5. SWNT emission change at 1247 nm after ozonation. Used 661 nm laser for
excitation. Every fifth experimental point is shown with a symbol. (A) Change of
normalized fluorescence intensity (I/Imax) with addition of different amounts of ozone.
The higher level of ozone resulted in a lower fluorescence intensity (line ). The lowest
amount of ozone gave the highest intensity (line ). (B) Fluorescence quenching factor
(Imax/I) shown in logarithmic scale. Different oxidation degrees gave close decomposition
rates.
27
Time (min)
0 5 10 15
Invert
ed
No
rmaliz
ed F
luo
resce
nce
, I m
ax /
I
0
50
100
150
954 nm
1026 nm
1123 nm
1250 nm
A
B
C
D
0 1 2
5
10
15
20
A
Figure 6. Influence of ozone load on fluorescence quenching at different emmax
.
Inverted normalized fluorescence (Imax/I) at four distinct wavelengths is shown. Four
independent experiments (A-D) are shifted along the time axis for clarity. A zoom-in for
experiment A is provided in the upper left corner. Wider tubes ( em 1250 nm) were
quenched more at low ozone loads (A-B). With higher loads all tubes were quenched to
the same degree (C-D).
Large loads of ozone, typically above 1 mL of O3/O2 gaseous mixture (ca. 3 v/v %
O3), injected into 1 mL of SWNT – SDS suspension resulted in slower decay rates. Rates
obtained from samples with quenching degree (Imax/I) below 200 were reproducible.
Rates obtained from higher levels of ozonation were difficult to reproduce even with a
thermostated cuvette. The common problem was the curve deviation from simple
exponential decay.
Tubes emitting at longer wavelength, emmax
= 1250, typically with wider diameters,
were quenched to a higher degree within Imax/I range 20 to 130 (Figures 6A and B).
Higher ozonation loads resulted in all tubes getting quenched to the same degree (Figure
6C and D).
28
Time (min)
0 2 4 6 8 10
Inve
rte
d N
orm
aliz
ed
Flu
ore
sce
nce
0
25
50
75
100
2 4 6
5
9
13 A
B
C
D
em = 1026 nm
less O3
more O3
H
H
L
L
Figure 7. Regression fit for inverted normalized fluorescence at 1026 nm (formula F4).
Each regression curve represents an independent experiment and is shown with a solid
thin line. Samples were ozonated to a different extent; curves A and B correspond to a
low ozone load, while C and D to a high load. Arrows labeled L and H point to curve
deviation caused by slowly decaying ozonides. Curves B-D were shifted along the time
axis for clarity.
Emission kinetics at emmax
= 1026 nm for different ozone loads were fitted with
regression curves; the highest two points (after ozonation) on inverted normalized
fluorescence data sets were excluded from regression analysis. As described in Appendix
A equilibration periods should be excluded from ozonide decay regression.
Overall decomposition rates were found to be lower with higher ozone loads. Two
possible explanations for such a phenomenon are a) ozonides formed are either lateral or
longitudinal to tube axis (Figure 27), or b) closely situated ozonides affect decomposition
of nearby ozonides.
29
The ozonide decay rates for curves A – D (Figure 7) were calculated with the
following formula:
n
bt
bt
final ceaeyy (F4)
Regression results for Figure 7 are summarized in Table 2.
Table 2. Regression results calculated with formula F4 for inverted normalized
fluorescence data recorded at 1026 nm emission wavelength.*
Curves
Parameter A, n > 10 B, n > 10 C, n > 8 C, n > 10 D, n > 8 D, n > 10
a 28.45 58.43 102.4 195.1
b, min-1
1.96 1.59 1.27 1.24 1.36 1.32
yfinal 1.71 1.48 1.10 1.10
c 1.18 2.24 4.48 6.26
n 10 10.00 8.00 8.00
r2 0.9997 0.9997 0.9977 0.9954
* Curve one-letter symbol and a lower boundary for variable n are written at the head of
each column. Constraints used for regression were: 0 < a < 1000; 0 < b < 100; 0 < c < a;
n < 100; yfinal > 1.05. Rates b are in min-1
.
For regression purposes, 'tails' on inverted data sets were truncated to increase the
weight of points related to a fast decay. Rates were calculated with 5-parameter formula
F4. The formula has fast and slow exponential terms, the slow one being n times slower
than the fast one. (For details on mathematics behind regression see Appendix A.)
Parameter n was kept greater than 10 for low ozone load curves, since formation of
slowly decaying ozonides was minimal; n was set to be greater than 8 for high load
curves, since there was a greater number of slowly decaying ozonides. Decreasing n
value to less than 8 would increase influence of the slower component on regression
curve.
30
The higher load of ozone resulted in oxidation of sites with slower decay rates. Sites
that required higher activation energy for oxidation resulted in formation of more stable
ozonides, contributing to a slower component. In other words, double bonds that were
harder to oxidize gave slower 1,2,3-trioxolane decay rates.
Curves with lower ozone load were fitted well with n > 10 (i.e. small “slow”
component). Regression curves for higher ozone load had difficulty fitting to
experimental points and n value constraint was brought down to n > 8. Even such
adjustment did not help regression curve to fit D data set (Figure 7 D), Experiment D had
the highest ozonation degree. Arrows H and L point to deviation of experimental points
from regression line. (Rates for curves C and D were also calculated with n >10
constraint; see Table 2)
The main purpose of the introduction of the slow exponential component was to
improve correlation between normalized and inverted normalized data sets. Appendix A
explains this issue in great detail. Normalized experimental curves were found to have
slowly rising tails and required an introduction of a slow component. The assumption was
made that the slow component should be n times slower than the fast one.
Rates calculated for lower loads of ozone, curves A and B, were 1.96 and 1.59
accordingly. Rates for higher loads of ozone were 1.27 and 1.36 for curves C and D
accordingly. Curve D could not be fitted as well as other three curves. Higher degree of
ozonation resulted in lower coefficients of determination, r2
(SigmaPlot® software
package, used for regression analysis, defines r2 as a coefficient of determination).
31
SWNT oxidation with solvated ozone. Influence of ozone load on NIR absorption
and fluorescence.
Time (sec)
0 200 400 600 800 1000 1200
No
rma
lize
d I
nte
nsity
0.0
0.2
0.4
0.6
0.8
1.0
No
rma
lize
d I
nte
nsity
0.0
0.2
0.4
0.6
0.8
1.0
954 nm
1026 nm
1123 nm
1250 nm
F1
F2
A1
A2
Figure 8. Regression analysis of normalized absorbance (A1 and A2) and fluorescence
(F1 and F2) intensities of ozonated SWNT at four distinct wavelengths. SWNT sample
was oxidized with solvated ozone. Data points for absorption and fluorescence were
acquired sequentially with 1 sec delay. Points not used in regression are depicted with
dotted lines. Regression curves are shown with solid lines. Legends are the same for the
top and the bottom plots. Used 661 nm excitation source for fluorescence measurements.
Top: high load of trioxolanes, Bottom: low load of trioxolanes.
32
Curves in Figure 8 show NIR absorption and fluorescence change with introduction
of ozone into system. Intensities dropped down and then slowly recovered to sub initial
values. Formula used for regression on NIR absorption had six parameters (F1); formula
used for normalized fluorescence data sets had five parameters (F2).
minmin1
y
ceaey
yy
n
bt
bt
final
(F1)
n
bt
bt
final ceaey
y1
(F2)
In this particular experiment, water saturated with ozone was used instead of
bubbling gaseous ozone. Absorption points after reagent addition were adjusted to
compensate for dilution. Calculated ozonide decomposition rates are summarized in
Table 3.
33
Table 3. SWNT oxidation with solvated ozone. Regression results calculated with
formulas F1 and F2 for normalized NIR absorption and fluorescence data recorded at
four distinct emission wavelengths.*
Fluorescence NIR absorption
em, nm 954 1026 1123 1250 954 1026 1123 1250
High load F1 A1
b, s -1
0.0225 0.0146 0.0100 0.0061 0.0187 0.0110 0.0076 0.0058
b, min -1
1.35 0.88 0.60 0.37 1.12 0.66 0.46 0.35
n 10.57 10.00 10.00 10.00 20.00 20.00 20.00 20.00
r2 0.9995 0.9996 0.9996 0.9993 0.9652 0.9834 0.9964 0.9974
Low load F2 A2
b, s -1
0.0480 0.0428 0.0276 0.0132 - - 0.0302 0.0158
b, min -1
2.88 2.57 1.66 0.79 - - 1.81 0.95
n 15.00 15.00 12.62 10.00 - - 20.00 20.00
r2 0.9673 0.9895 0.9959 0.9985 - - 0.9822 0.9949
* Emission or absorption wavelength is written at the head of each column. Constraints
used for regression analysis were: 0 < a < 1000; 0 < b < 100; 0 < c < a; for fluorescence
10 < n < 15; for absorption n < 20; yfinal > 1.05. Abbreviations: r2 – coefficient of
determination, n – determines how many times slow exponential term is slower than the
fast one, b – 1,2,3-trioxolane decomposition rate.
Constraint 10 < n < 15 used in fluorescence regression was needed to prevent very
low n values, leading to a greater influence of the term n
bt
ce . Such reduction of n value
led to meaningless rates b, and it was imperative to keep 'slow' component as a small
contributor to the overall intensity change. Constraint n < 20 was set for NIR absorption
regression curve. With no upper constraint for parameter n, regression on NIR absorption
data set was attempting to set abnormally large n values for nearly straight 'tails'. Greater
n value resulted in a slower second term. When n values are abnormally high, regression
results in converting the curve into a straight line, which is not the case.
34
Fluorescence rates for low ozone load were found to be at least two times faster than
those for a high load of ozone. The same was true for NIR absorption rates.
Fluorescence and absorption rates obtained from the same sample were found to be
close, but not equal. For high load of ozone, fluorescence rates were slightly higher than
those for NIR absorption. For low load of ozone fluorescence rates were slightly lower.
Observation of a close relationship between fluorescence and NIR absorption growth
rates led to the following diagram of transition states (Figure 9).
En
erg
y
-4
-3
-2
-1
0
1
2
3
4
v1
v2
c1
c2
conduction
valence
v1
v2
c1
c2
Density of Electronic States
v1
v2
c1
c2
A B C
ozonideE
11 E22
E11
E11
E22
fluorescence abs
ozonide
NIR abs
Figure 9. Schematic density of electronic states for pristine and ozonated SWNTs. Thick
solid arrows depict optical excitation and emission transitions of interest; thin dashed
arrows denote nonradiative relaxation of the electron (in the conduction band) and the
hole (in the valence band) before emission. (A) Transitions of interest in pristine SWNT.
Diagram adopted from Science, 2002, 298, 2361-2366. (B) Transitions in ozonated
SWNT. Nonradiative relaxation c1 → ozonide → v1 is a major process and shown with
thick solid arrows. Fluorescence from c1 to v1 is a minor process and shown with a dotted
arrow. (C) NIR absorption of ozonated SWNT. Depletion of an electron density of band
v1 by ozonides resulted in a weaker NIR absorption v1 → c1 (v – valence band, c –
conduction band.)
35
SWNT oxidation with gaseous ozone (high load)
Time (sec)
0 200 400 600 800 1000 1200
Norm
aliz
ed Inte
nsity
0.0
0.3
0.6
0.9
954 nm
1026 nm
1123 nm
1250 nm LH
L
H
L
A
F
Figure 10. Regression analysis of normalized NIR absorption (A) and fluorescence (F)
intensities of ozonated SWNT at four distinct wavelengths. SWNT - SDS suspension was
bubbled with O3/O2 gaseous mixture (ca. 3 v/v % ozone). Data points for absorption and
fluorescence were acquired sequentially with 1 sec delay. Points not used in regression
are depicted with dotted lines. Regression curves are shown with solid lines. Each
wavelength is marked with an individual symbol. Labels L and H denote curve wobbling
above and below regression line. Used 661 nm excitation source for fluorescence
measurements. Upper curves: NIR absorption, Lower curves: fluorescence.
Regression results for Figure 10 above are summarized in Table 4.
36
Table 4. SWNT oxidation with gaseous ozone. Regression results were calculated with
formulas F1 for normalized NIR absorption and F2 for fluorescence data recorded at four
distinct emission wavelengths*
High load Fluorescence NIR absorption
em, nm 954 1026 1123 1250 954 1026 1123 1250
b, s -1
0.0211 0.0153 0.0113 0.0064 0.0180 0.0099 0.0079 0.0051
b, min -1
1.27 0.92 0.68 0.38 1.08 0.59 0.47 0.31
n 10.00 10.00 8.00 8.00 20.00 20.00 20.00 20.00
r2 0.9995 0.9998 0.9998 0.9979 0.9824 0.9918 0.9986 0.9987
* Emission or absorption wavelength is written at the head of each column. Constraints
used for regression were: 0 < a < 1000; 0 < b < 100; 0 < c < a; for fluorescence 8 < n <
10; for absorption n < 20; yfinal > 1.05. Abbreviations: r2 – coefficient of determination, n
– determines how many times slow exponential term is slower than the fast one, b –
1,2,3-trioxolane decomposition rate.
Formation rates for SWNT ozonides
Formation of SWNT ozonide is schematically shown in Scheme 1 below
Scheme 1
OO
O
SWNT SWNTO3
O3
Formation rates of 1,2,3-trioxolanes were measured at 25 C by monitoring
absorption at ozone absmax
= 260 nm (Figure 11).
37
Time (sec)
0 1 2 3 4 5
Ab
so
rba
nce
x 1
00
0
(a.u
.)
0
1
2
b1
b2
b3
Figure 11. Absorption kinetics at 25 ºC monitored at 260 nm after bubbling O3/O2
gaseous mixture through SWNT-SDS suspension. Absorption decrease represents ozone
consumption and 1,2,3-trioxolane formation rates. Curves are shifted along the vertical
axis to bring regression lines to approximate zero with t → . Curves are: ozonation of
4x diluted SWNT (line ), ozonation of 8x diluted SWNT (line ), and ozone bubbled
through preliminary ozonated 4x diluted SWNT (line ).
Each curve in Figure 11 represents a separate experiment. The same amount of gas
(0.5 mL) with the same concentration of ozone (ca. 0.5 v/v %) were used for all
injections. SDS was found to be unreactive with small amounts of ozone and its influence
on absorption was below detection limits. There is always a possibility that small percent
of impurities from SDS can affect absorption change, thus leading to misinterpretation of
kinetics results. This is thought not to be the case in the above-mentioned experiments
(Figure 11), because pre-ozonated SWNT gave a comparable ozone
consumption/trioxolanes formation rate. Additionally, all samples were bleached to levels
below original absorbance, indicating that some double bonds were no longer existing.
Based on these facts, it is believed that the measured kinetics curves are from chemical
38
reaction between ozone and SWNT and not from some unknown impurity. Dilution 4 and
8 times of stock SWNT – SDS suspension with 1 wt. % aq. SDS was necessary to acquire
sufficient data points for regression.
1,2,3-Trioxolane formation rates are summarized in Table 5.
Table 5. SWNT ozonide formation rates.
Curve b, s –1
r2 ccarbon, mg/L cdouble bond, mmol/L
4x diluted SWNT 1.44 0.8936 1.44 0.06
8x diluted SWNT 0.63 0.8498 0.72 0.03
4x dil. prelim. ozntd SWNT* 0.52 0.9475 1.44 0.06
* - Preliminary ozonated tubes were heated to 40 C for ca. 4 hours before their use in this
experiment; b – rate.
Rate for 8 times diluted SWNT – SDS suspension was found to be at least two times
slower than the one for 4 times diluted sample. Concentration of SWNT for 4 and 8 times
diluted suspensions are summarized in Table 5. Total concentration of double bonds in 4
times diluted sample was calculated 0.06 mmol/L, yielding the reaction rate constant
2.4·104 M
-1s
-1, which is of the same order of magnitude as the formation rate constants of
C60O3 (8.8 104 M
-1s
-1 at 0 C) and C70O3 (5 10
4 M
-1s
-1 at 22 C) in CCl4 solvent.
18, 19
The rate constant for 4 times diluted SWNT suspension is five orders of magnitude
slower than the diffusion rate constant 109 M
-1s
-1 for small organic molecules in hexane.
It is likely that water viscosity, solvation of ozone with water molecules, SDS
hydrophobic shell around SWNT, large molecular weight of the tubes and tubular
structure with large aspect ratio, all contributed to the rate decrease.
39
Establishing of a saturation limit with different amounts of ozone
Absorption changes after injections of different amounts of ozone into diluted
SWNT – SDS suspension are shown in Figure 12. SWNT - SDS suspension was diluted
eight times to decrease reaction rate between SWNT and ozone. Ozone concentrations
were approximately 0.5%, 0.6%, 0.75% and 1% in air stream (curves , , ,
accordingly). Ozone concentration was manipulated by dilution with air. A half milliliter
of O3 - air gaseous mixture was injected into 1 mL of SWNT – SDS suspension in each
case. All measurements were performed in a thermostated cuvette at 25.0 C.
Time (sec)
0 20 40 60 80 100 120
Absorb
an
ce x100 (
a.u
.)
7.6
7.8
8.0
8.2
8.4
3x
4x
5x
6x
O3
/ O2 mixture
dilution with air:
d8
saturation with ozonides
ozone consumption
Figure 12. Dependence of SWNT absorption on the amount of injected ozone. Each
curve represents a separate experiment. Kinetics curves were monitored at 260 nm and
25.0 C. Symbols are labeled with O3/O2 mixture dilution degrees. The difference
between initial and final absorbance is marked with d8.
Curves were shifted along the time axis to set injection point to zero seconds. Three
of the four curves were multiplied by the corresponding coefficients to bring the initial
40
absorbance to the same level (difference in absorbance before adjustment was very small;
initial absorbance ranged between 0.0737 and 0.0761 a.u.)
Spikes at 0 seconds are due to needle insertion and shown with dotted lines.
Exponential decrease of absorbance right after the injection represents ozone
consumption and 1,2,3-trioxolane formation.
Higher concentrations of ozone resulted in identical downward step (value d8 in
Figure 12), indicating that tubes got saturated with ozonides. Exponential decay is
schematically divided into two sections: saturation of SWNT with ozonides (left) and
ozone consumption by „emptied‟ sections of SWNT (right). Value d8 equaled to 0.0017
a.u. was found approximately 9 times smaller than downward step d in non-diluted
SWNT suspension (Figure 13). Addition of 0.5 mL of 6x diluted O3/O2 gaseous mixture
was not sufficient to saturate SWNT with 1,2,3-trioxolanes. Rate of reaction between
SWNT and the least concentrated ozone/air gas mixture (curve ) was calculated to be
b = 0.63 s-1
, corresponding to lifetime = 1.6 sec (see Figure 12 above for details.)
41
Influence of multiple ozone injections on SWNT saturation. Oxidation with 4 min
intervals
Time (min)
10 20 30
Ab
so
rba
nce
(a
.u.)
0.54
0.57
0.60
b6 b7b8
b5
b4
b3
b2
b1
d
Figure 13. Injections of 1.5 mL of O3/O2 gaseous mixture into 1.5 mL SWNT - SDS
suspension with 4 min time intervals. Suspension absorbance was monitored at 260 nm
and r. t. Upward spikes are due to needle insertion and are emphasized with arrows.
Difference in absorbance before and after the first injection is marked with letter d.
Difference in absorbance between before and after the first ozone injection was
d = 0.0154 a.u.(ca. 3%). This value is approximately 9 times larger than that for 8x
diluted SWNT suspension. Rates noted as b1 through b8 in Figure 13 are summarized in
Table 6.
Increase in rate b and decrease in the absolute value of step d (marked on Figure 13)
are believed to be associated with SWNT saturation. Experimental points in Figure 13
above, shown with circles, form simple exponential decay curves right after each
injection of ozone. This decay represents reaction of freely floating ozone with SWNT
and conversion of ozone to oxygen by collision with SDS and water molecules.
42
Table 6. Regression results for single exponential decay of absorbance after multiple
ozone injections into SWNT – SDS suspension.
Injection cycle*
1st 2
nd 3
rd 4
th 5
th 6
th 7
th 8
th
b, min-1
16.60 10.68 7.51 4.95 3.55 3.23 2.91 2.93
, sec 3.6 5.6 8.0 12.1 16.9 18.6 20.6 20.5
r2 0.9917 0.9971 0.9987 0.9987 0.9988 0.9994 0.9992 0.9988
* - Regression was performed with a single exponential decay formula. Variables: b –
decay rate, - lifetime, r 2
– coefficient of determination.
SWNT oxidation with different amounts of ozone. Estimation of a saturation limit
by NIR fluorescence.
Wavelength (nm)
950 1050 1150 1250 1350
Flu
ore
sce
nce
In
tensity
(nW
/nm
)
0.00
0.05
0.10
0.15
before O3
0.5 mL
1 mL
2 mL
3 mL
4 mL
5 mL
6 mL
8 mL
10 mL
1265
0.03
0.05
Figure 14. Fluorescence spectra after bubbling specific amounts of ozone through
SWNT-SDS suspensions. Each curve represents a separate experiment. One percent
aqueous SDS solution ozonated with 10 mL of O3/O2 gaseous mixture (ca. 3 v/v %
ozone) served as a baseline for all experiment. Samples were ozonated 3 days before
fluorescence acquisition. The 785 nm laser was used for excitation. Arrows point to the
saturation level. A zoom-in shows that saturation was reached with 3 mL of O3/O2
gaseous mixture (ca. 3 v/v % ozone; curve ). Tube (6,5) with emmax
977 nm was the
most difficult to oxidize.
43
SWNT suspension aliquots were bubbled with different amounts of O3/O2 gaseous
mixture (0.5 – 10 mL) and spectra overlaid (Figure 14). Gas was injected slowly in each
case. To avoid misinterpretations, spectra were recorded three days after ozonation,
which was plenty of time for ozonide decomposition and SWNT structural
rearrangements. One percent aqueous sodium dodecyl sulfate solution bubbled with
10 mL of O3/O2 gaseous mixture (ca. 3 v/v % ozone) served as a baseline for all curves in
Figure 14. Spectra overlay demonstrated that SWNT got saturated with ozone at 3 mL
O3/O2 gaseous mixture load (ca. 3 v/v % ozone). Curves for 8 and 10 mL have less
intense fluorescence than all other ones. It may be concluded that during slow O3/O2
gaseous mixture injection some of the ozonides have decomposed, thus allowing for a
greater amount of ozone to react with SWNT. Separate UV studies demonstrated that
after SWNT got saturated with 1,2,3-trioxolanes, which occurred at or below 3 mL load,
excess of ozone dissolved in aqueous media (Figure 13). This in turn provided an
additional supply of ozone for subsequent oxidation. Because collision of ozone with
SDS molecules leads to its conversion to oxygen (ozone decay rate in 1% aq. SDS is
about 0.43 min-1
at r. t.), slow addition of 8 and 10 mL provided sufficient amount of
ozone to overcome deactivation by SDS, hydroxyl and water species. Typically, injection
of 10 mL of O3/O2 gaseous mixture required more than a minute to complete.
Notably, tube (6,5) with emmax
977 nm (diameter = 0.76 nm) had the least ozonation
degree, i.e. was the most difficult to oxidize. This tube has a very small twist when
compared to other tubes in the experiment.
3D Structures of tubes of interest are shown in Figure 15.
44
(8,3) (6,5) (10,2) (11,3) (10,5)
Figure 15. Tubes (n,m) with well separated emission peaks on spectrum for 785 nm
excitation source. Each tube is shown in two projections (top and bottom). Tube number
is indicated below each pair. Tubes are drawn not to scale.
The physical and optical properties of tubes that were well separated in emission
spectrum with 785 nm excitation are summarized in Table 7.
Table 7. Summary of physical and optical properties of tubes with well separated peaks
in emission spectrum*
Tube (n,m) emmax, nm , cm
-1 diameter, nm
8,3 954 10486 0.78
6,5 977 10234 0.76
10,2 1056 9468 0.88
11,3 1201 8327 1.01
10,5 1253 7982 1.05
*- Used 785 nm excitation source.
The only tube that had difficulty getting ozonated with 10 mL of O3/O2 gaseous
mixture (ca. 3 v/v % ozone) was tube (6,5). It has the smallest “twist” out of all well
defined tubes in emission spectrum (Figure 14). It may be concluded that the tube “twist”
increases double bond reactivity with ozone. Tube (6,5) was also estimated to be the most
abundant in utilized HipCo sample (see Chapter 1 for tube abundance distribution). All
45
tubes had substantial difference in diameter and no conclusion could be made with regard
to dependence of ozonation degree on the tube diameter at 10 mL O3/O2 gaseous mixture
load (ca. 3 v/v % ozone). Particularly, tubes (8,3) and (6,5) had very close diameter, but
different oxidation degree, as evidenced by peaks at 954 and 977 nm (Figure 14).
SWNT oxidation with different amounts of ozone. Estimation of a saturation limit
by NIR absorption.
NIR absorption was measured on samples discussed above. For experimental details
see Figure 14 above and accompanying notes.
Wavelength (nm)
950 1050 1150 1250
Absorb
an
ce (
a.u
.)
0.15
0.20
0.25
before O3
0.5 mL
1 mL
2 mL
3 mL
4 mL
5 mL
6 mL
8 mL
10 mL
Figure 16. NIR absorption spectra after bubbling specific amounts of O3/O2 through
SWNT-SDS suspensions. Each curve represents a separate experiment. One percent
aqueous SDS ozonated with 10 mL of O3/O2 served as a baseline for all experiment.
Samples were ozonated 3 days prior to NIR absorption acquisition. Arrows point to the
saturation level, which was reached at 2 mL of O3/O2 gaseous mixture (ca. 3 v/v %
ozone; curve ). The area near 977 nm was the most difficult to oxidize and had the
least percent decrease.
46
Analogously to fluorescence spectra, the area near 977 nm had the least decrease in
absorbance as compared to percent values for all tubes. Saturation point was reached with
2 mL O3/O2 gaseous mixture (ca. 3 v/v % ozone; pointed with arrows in Figure 16). To
avoid misinterpretations, NIR absorption spectra were recorded three days after
ozonation.
Noisiness of spectra starting from 2 mL is presumed to be associated with production
of a number of nonequivalent sections of tubes. Notably, there are no peak shifts between
2 and 10 mL of O3/O2 gaseous mixture (ca. 3 v/v % ozone). This means that ozonation is
following a specific pattern, rather than a random one. An increase in the number of
peaks would be expected for random ozonation of SWNT. Despite significant decrease in
absorption (compare curves and ), the number of peaks and their absmax
were
preserved, thus indicating an ordered oxidation. Higher loads of ozone (within 10 mL
O3/O2 gaseous mixture) are thought to produce a greater number of „sections‟ of SWNT
ozonated with the same pattern.
SWNT oxidation with different amounts of ozone. Estimation of a saturation limit
by UV-Vis absorption.
UV-Vis spectrum of 1% aq. SDS was found to be unchanged in a region 235 - 800
nm after bubbling with 10 mL of O3/O2 gaseous mixture. Aqueous SDS solution purged
with 10 mL of O3/O2 gaseous mixture served as a baseline for all spectra in Figure 17.
47
Wavelength (nm)
250 350 450 550 650 750
Absorb
ance (
a.u
.)
0.2
0.4
0.6
245 285
0.55
0.59
0.63 before O3
0.5 mL
1 mL
2 mL
3 mL
4 mL
5 mL
6 mL
8 mL
10 mL
730
0.19
0.21
Figure 17. UV-Vis absorption spectra after bubbling specific amounts of ozone through
SWNT-SDS suspensions. Each curve represents a separate experiment. One percent
aqueous SDS ozonated with 10 mL of O3 served as a baseline for all experiment. Samples
were ozonated 3 days before UV-Vis absorption acquisition. Arrows point to the
saturation level, which was reached at or below 2 mL of O3/O2 gaseous mixture (ca. 3 v/v
% ozone; curve ).
SWNT spectra had smooth transition from 0 mL (curve ) spectrum to 10 mL one
(curve , Figure 17).
SWNT ozonation for a specific period of time. Influence of ‘saturation’ and
1,2,3-trioxolane decomposition rates on overall sidewall oxidation as monitored by
NIR fluorescence.
Ozonation of SWNT-SDS suspensions was conducted for specific periods of time,
ranging from 30 sec to 30 min. One percent aq. solution of sodium dodecyl sulfate,
bubbled with ozone for specified periods of time served as a baseline for each curve (i.e.
each curve had its own baseline). Interesting spectral changes were observed and are
summarized in Figure 18.
48
Wavelength (nm)
950 1050 1150 1250 1350
Flu
ore
sce
nce
In
ten
sity (
nW
/nm
)
0.00
0.04
0.08
0.12
0.16before O3
0.5 m
1 m
2 m
4 m
5 m
10 m
30 m
Figure 18. Influence of ozonation on SWNT fluorescence spectra. Each curve represents
a separate experiment. Symbol legends denote ozone bubbling times in minutes. Spectra
were recorded 3 days after ozonation. Ozone was bubbled through samples at room
temperature. The 785 nm laser was used for excitation. Arrows point to fluorescence
curves after 0.5 and 5 min of continuous bubbling of O3/O2 gaseous mixture (ca. 3 v/v %
ozone) through SWNT-SDS suspension. Curves for 5, 10 and 30 min and before O3 are
shown with thick lines.
Two sets of pairs of arrows in Figure 18 demonstrate location of curves after 30 sec
and 5 min of continuous bubbling of O3/O2 gaseous mixture. The gas flow rate was
approximately 26 mL/min. This means 0.5 and 5 min bubbling correspond to 13 and 130
mL of O3/O2 gaseous mixture.
The upper left arrow points to tube (6,5) which was found to be fairly robust to
ozonation with 10 mL of O3/O2 gaseous mixture (ca. 3 v/v % ozone). In terms of percent
values, intensities of all tubes except (6,5) were substantially bleached within 30 sec of
continuous bubbling. Notably, tube (8,3), with roughly the same diameter as (6,5), was
bleached more than (6,5) after 30 sec ozonation. (Tube properties are summarized in
49
Table 7). Interestingly, after 5 min of bubbling of O3/O2 gaseous mixture, tube (6,5) is
nearly gone, while the ones emitting at longer wavelengths are still present. Comparing
fluorescence intensities at 977 and 1255 nm for curves and in Figure 18, one may
believe that tube (6,5) had a high degree of aromaticity, or conjugation of double bonds,
thus making it difficult to oxidize. Once that conjugation was disrupted, oxidation of this
particular tube was comparable to those with similar diameters.
Fluorescence of SWNT with shorter emmax
was much weaker than those with longer
emission wavelengths after 2 minutes of continuous bubbling of O3/O2 gaseous mixture
(curve ). It is believed that 1,2,3-trioxolanes formed on tubes emitting in a range 900-
1100 nm decayed faster then those emitting at longer wavelengths (as evidenced by
comparison 30 sec and 2 min curves, and ).
After saturation limit of trioxolanes on the surface of SWNT was reached (curve )
tube (8,3) with emmax
954 nm decreased a lot more than those emitting at longer
wavelengths. It was concluded that tube diameter is one of the major factors affecting
rates of trioxolane decomposition, thus allowing for more ozonides to be formed on the
SWNT surface.
Faster decay rates lead to a greater number of ozonides formed, resulting in a higher
turnover within the same period of time. After 5 min of continuous bubbling fluorescence
at 954 nm is negligible, but at 1150 nm and longer wavelengths is only a little less than in
the spectrum for 30 seconds bubbling.
50
SWNT ozonation for a specific period of time. Influence of ‘saturation’ and 1,2,3-
trioxolane decomposition rates on overall sidewall oxidation as monitored by NIR
absorption.
NIR absorption was measured on samples that were bubbled with O3/O2 gaseous
mixture (ca. 3 v/v % ozone) for specific periods of time, as discussed in the section
above. In the same manner, aq. SDS solutions bubbled with ozone for noted periods of
time served as a baselines in each case (i.e. each curve had its own baseline).
Wavelength (nm)
950 1050 1150 1250
Absorb
ance (
a.u
.)
0.05
0.10
0.15
0.20
0.25
before O3
0.5 m
1 m
2 m
4 m
5 m
10 m
30 m
1 hr
Figure 19. Influence of ozonation on SWNT NIR absorption spectra. Each curve
represents a separate experiment. Symbol legends denote ozone bubbling times in
minutes. Spectra were recorded 3 days after ozonation. Ozone was bubbled through
samples at room temperature. Arrows point to curves after 0.5 and 10 min of continuous
bubbling of O3/O2 gaseous mixture (ca. 3 v/v % ozone) through SWNT-SDS suspension.
Bleaching of NIR absorption of ozonated SWNT was less informative when
compared to fluorescence spectra. Saturation of SWNT with trioxolanes was reached at
or before 30 sec of continuous bubbling of O3/O2 gaseous mixture (ca. 3 v/v % ozone).
51
Initial bleaching at shorter wavelengths was less than that at longer ones. As evidenced
from comparison of the curves before and after 30 sec of ozone bubbling (curves and
), the change in absorbance at 977 nm was only a fraction of that at 1123 nm or 1250
nm.
After 10 min of continuous bubbling, all characteristic peaks below 1100 nm
disappeared. This means that faster decomposition rates of 1,2,3-trioxolanes of thinner
tubes resulted their destruction, while thicker ones were still maintaining their
characteristic absorption. The peaks maxima did not shift after 0.5 min of ozonation and
the curve stayed smooth (curve ). This means that NIR absorption peaks in Figure 19
above are for a small section of a tube. Bubbling for one minute resulted in a very „noisy‟
curve. This meant that after one minute of continuous bubbling damage to tubes was so
great that it finally manifested itself in NIR spectra. It is still not clear how large these
sections were.
SWNT ozonation for a specific period of time. Influence of ‘saturation’ and
1,2,3-trioxolane decomposition rates on overall sidewall oxidation as monitored by
UV-Vis absorption.
UV-Vis absorption was measured on samples that were bubbled with O3/O2 gaseous
mixture (ca. 3 v/v % ozone) for specific periods of time, as discussed in two previous
sections. In the same manner, aq. SDS solutions bubbled with ozone for a specific period
of time served as a baseline for each experiment (each curve had its own baseline).
Below are overlaid 1% aq. SDS spectra after bubbling for specified periods of time.
With regard to baseline subtractions, SDS absorption increased in the range 250 – 330
52
nm to acceptable levels, i.e. it was sufficiently low. At wavelengths below 250 nm it
increased greatly and that region will not be discussed in this work.
Wavelength (nm)
300 400 500 600 700
Ab
so
rba
nce
(a
.u.)
0.2
0.3
0.4
0.5
0.6 before O3
0.5 m
1 m
2 m
3 m
4 m
5 m
10 m
30 m
1 hr
Figure 20. UV-Vis spectra of 1% aq. SDS bubbled with O3/O2 gaseous mixture (ca. 3 v/v
% ozone) for noted periods of time.
SDS was found to be sufficiently robust to oxidation with ozone and could be used
as baseline reference for wavelengths above 250 nm.
Below are overlaid UV-Vis spectra of SWNT ozonated for noted periods of time
(Figure 21). SWNT spectrum obtained before ozonation is provided for reference.
53
Wavelength (nm)
300 400 500 600 700
Ab
so
rba
nce
(a
.u.)
0.1
0.2
0.3
0.4
0.5
0.6 before O3
0.5 min
1 m
2 m
3 m
4 m
5 m
10 m
30 m
1 hr
310 330
0.30
0.35
0.40
0.45
0.50
Figure 21. Influence of ozonation on UV-Vis absorption spectra of SWNT. Each curve
represents a separate experiment. Symbol legends denote ozone bubbling times in
minutes. Spectra were recorded 3 days after ozonation. Ozone was bubbled through
samples at room temperature. Arrows point to curves 3, 4 and 5 min of continuous
bubbling of O3/O2 gaseous mixture (ca. 3 v/v % ozone) through SWNT-SDS suspension.
SWNT were found to give essentially the same UV-Vis spectra for 3, 4 and 5 min of
continuous ozonation. Three minutes of bubbling are equivalent of a slow injection of
about 78 mL of O3/O2 gaseous mixture (ca. 3 v/v % ozone). It is not clear why there was
a „freeze‟ in spectral changes after 3 min. With small ozone loads, saturation point for
SWNT was found to be around 0.5 min of bubbling as described in the sections above.
SWNT spectra had smooth transition from 0 min spectrum to 30 min one.
54
Ozone consumption as monitored by absorption at 260 nm
Ozone consumption after the first injection was monitored near its absmax
= 260 nm
(Figure 22).
240 250 260 270 280
Absorb
an
ce (
a.u
.)
0.52
0.54
0.56
0.58
0.60
0.62before O3
13.7 sec
27.5 s
41.2 s
55.0 s
68.7 s
Wavelength (nm)
240 250 260 270 280
0.00
0.02
0.04
Figure 22. UV spectrum change with ozone consumption. Left: spectra as recorded.
Right: spectra after subtraction of pristine SWNT spectrum (). Symbols are the same
for both graphs and denote times after the injection of O3/O2 gaseous mixture. One
percent aq. SDS served as a baseline. Curves , and are shown with thick lines.
One and half milliliter of O3/O2 gaseous mixture (ca. 3 v/v % ozone) was injected
into 1.5 mL SWNT – SDS suspension. Injection was done during 2 second interval
between curves and in Figure 22 (left side).
Decrease of ozone absorption was several times greater than that of SWNT (compare
curves , and in Figure 22). SWNT bleaching is the difference between absorbance
before and after SWNT ozonation. In absolute values SWNT bleaching at 260 nm was
0.0028 a.u. between curves and . Ozone absorbance change was approximately
0.047 a.u., or 17 times greater than SWNT bleaching. Sodium dodecyl sulfate, a
55
surfactant, also contributed to ozone decomposition (ozone decay rate in 1% aq. SDS is
about 0.43 min-1
at r. t.).
It was established that ozonide decomposition results in increase of SWNT
absorbance. At 260 nm this rise is barely noticeable. Effect on absorbance is more
pronounced with longer wavelengths. Typically, absorption curve plotted against time
would have a fast rising beginning and a slowly changing tail. Initial change in
absorbance is attributed to 1,2,3-trioxolanes decomposition. Tail part within the first 30
min after ozonation is attributed to a combination of structural rearrangements of SWNT
and decomposition of slowly decaying ozonides (section ii in Figure 23).
Ozonide decomposition was studied by UV-Vis at 25 and 40 C. To avoid possible
chemical transformations, the longer wavelength, 735 nm, was chosen for monitoring. A
time period of five hours at 25 C was not sufficient to reach the maximum absorbance.
Temperature was increased to 40 C and the experiment was repeated (Figure 23).
56
Time (min)0 100 200 300 400
Ab
so
rba
nce
at
73
5 n
m (
a.u
.)
0.17
0.18
0.19
0.20
0.21
5 6 7
0.17
0.18
0.19
0.20
0.21
50 150 250 350
0.198
0.199
0.200
0.201
i ii iii
Figure 23. SWNT absorbance change at 40 C after ozonation. The point of injection is
emphasized with an arrow. An upward spike at the time of injection is due to a needle
insertion. 2.5 mL of O3/O2 gaseous mixture was bubbled through 1 mL of SWNT
suspension over a period of 20 sec. A fast changing beginning and a slowly changing
„tail‟ are zoomed for clarity. An approximate place of absorbance maximum after the
injection is marked with a dotted vertical line. Plot is schematically divided into: (i) – fast
O3 decay, (ii) – slow O3 decay and rearrangements, (iii) – rearrangements.
Heating for 6 hours at 40 C was needed to bring absorbance at 735 nm to its
maximum value. It is presumed that the slow change is associated with SWNT structural
changes and migration of double bonds. Absorbance at 735 nm after recovery was 4%
lower than that recorded before ozonation. This result is in good agreement with SWNT
absorption decrease measured in UV region. Ozonation of SWNT resulted in 3 – 4 %
absorbance bleaching at 260 nm.
57
Establishing an average rate of ozone consumption
Time (min)
0 5 10 15 20 25
Absorb
an
ce (
a.u
.)
0.0
0.2
0.4
0.6
0.8 Abs (H2O/O3)
Abs (diluted H2O/O3)
b1
b2
SDS
SWNT - SDS
Figure 24. Estimation of an average rate of ozone consumption by SWNT. Absorption
was monitored at 260 nm and 25 C. Curves and represent two separate
experiments. A suspension of SWNT was added to an aqueous solution of ozone (curve
). The second experiment included an injection of an equivalent volume of an aqueous
SDS solution. The points of injection are marked with arrows. Expected absorbance level
for curve after dilution is marked with a horizontal dash line. Experimental points used
for regression analysis are shown with thick lines. Experimental points not used in
regression are shown with dotted lines. Regression lines are depicted with thin black
lines.
Figure 24 demonstrates a difference between rates of ozone decay in presence (curve
) and absence of SWNT (curve ). Two separate experiments were designed to
compare rates of ozone consumption obtained by UV and NIR fluorescence
measurements. The first experiment included injection of 700 uL of SWNT-1% SDS aq.
suspension into 800 uL of an aqueous solution of ozone. The initial absorbance of ozone
solution in the first experiment was above 0.8 a.u. Injection of SWNT suspension is
marked with an arrow (Figure 24, curve ). Dashed lines on the same figure are shown
58
to visually demonstrate the predicted absorbance after dilution. The selected points for
regression analysis are shown with thick solid lines. Resulted regression curves are
shown with thin black lines. Experiment was repeated with an injection of 700 uL of 1%
aq. SDS solution only (curve ). In the case of the second experiment initial ozone
absorbance was only 0.65 a.u.; this was sufficient for the purpose of the experiment.
Preparation of an ozone solution for the second experiment with the absorbance of 0.8
a.u. was not necessary. Sections of the curves before injection have gradual slopes,
common for slow ozone decay in water. Regression results are summarized in Table 8
below.
An upper boundary for absorbance after reagent injection was calculated in
accordance with the following formula: ofbid AAAkAA
where dA – absorbance after dilution, iA – absorbance at the time of injection, bA – SWNT
bleaching at saturation point ( bA = 0.008 a.u.; cf. Figure 13), fA – final absorbance, k – dilution
coefficient, k = 0.533, oA – absorption of consumed ozone.
Estimation of the upper boundary was needed to determine which points must not be
included in regression analysis. Due to rapidly changing exponential decay immediately
after ozonation, there was no clear division between the effects of dilution and
absorbance decay. The decrease of ozone absorption due to the formation of ozonides is
thought to be several times larger than SWNT bleaching (see Figure 22), but the exact
number is difficult to estimate.
59
Table 8. Regression results for single exponential decay of absorbance after injections of
SWNT – 1 wt. % aq. SDS suspension and of 1 wt. % aq. SDS solution into aqueous
solutions of ozone.
b, min-1
, min r2
dA fA
SWNT-SDS (b1) 1.65 0.61 0.9978 0.6981 – oA 0.265
SDS (b2) 0.29 3.45 0.99997 0.3602 – oA 0.015
b diff 1.36 0.74
* - Regression was performed with a single exponential decay formula. Variables: b –
decay rate, - lifetime, r 2
– coefficient of determination, dA – absorbance after dilution,
fA – final absorbance, oA – absorption of consumed ozone.
An amount of ozone in aqueous solution with absorbance 0.8 a.u. exceeds many
times what could react with SWNT at any given time. Bleaching of SWNT, i.e. decrease
of absorbance due to reaction with ozone, was estimated to be near 0.008 a.u. based on
the data obtained in previous studies (this chapter, cf. Figure 13). The final absorbance of
SWNT – SDS suspension after full consumption of ozone was found to be fA = 0.265
(Table 8). The final absorbance fA of SWNT-SDS suspension would be higher if water
had no oxidizer in it. This implies that only 0.008/0.265 = 3 % of a total number of
double bonds were consumed in excess of O3 (i.e. within ca. 3 min after the injection,
curve in Figure 24). The saturation point was reached within seconds after injection
(cf. formation kinetics in Figure 11). A number of consumed double bonds is likely to be
higher than 3%, but it is still a very small number. Vacuum deoxygenation of SWNTO3
gave an estimate of 0.2 % of double bonds were consumed by ozone within 1 min of
reaction at 5 C in solid state (this chapter, Figure 1). Fast decay of curve (Figure 24)
60
is mainly due to ozone consumption by SWNT or decomposition by collision with SDS,
and not because of SWNT bleaching (Figure 22).
An assumption was made that about the same number of collisions occurred between
molecules of ozone and SDS in both experiments, thus making it possible to extract the
rate of ozone consumption by SWNT via subtraction of rate b2 from b1. The difference is
b diff = 1.36 min-1
, which is somewhere in between the rates obtained for fluorescence
measurements at four distinct wavelengths: 954, 1026, 1123 and 1250 nm.
It is important to keep in mind that the decomposition of 1,2,3-trioxolanes is a rate
determining step in the formation of subsequent or “new” ozonides after the previous
ones underwent decomposition. This is always true when the amount of ozone in the
system exceeds the saturation limit. SWNT samples saturated with 1,2,3-trioxolanes were
found to have 3 to 4 % of the total number of double bonds converted to ozonides. The
decomposition rates obtained from fluorescence studies were employed in the
interpretation of UV studies. Particularly, the fluorescence recovery rate was found to
increase exponentially with a decrease in tube diameter (plot is not shown). Thus, the
faster decay rates were observed for thinner tubes.
Metallic tubes are believed to react with ozone in the same fashion the
semiconducting tubes do, though they do not fluoresce. The overall rate b diff = 1.36 min-1
indicates that metallic tubes may have similar distribution of tube diameters.
More in-depth investigation is needed to study differences between UV results
recorded at 260 nm and fluorescence data to make a better estimation of how many and
what types of metallic tubes are present in the sample. As noted earlier, metallic tubes do
61
not fluoresce, and their quantitative analysis needs significant improvements. Currently,
there is no fast and efficient way to estimate distribution of metallic tubes in a sample.
An influence of the injection period on SWNT saturation with 1,2,3-trioxolanes
Time (min)
2 10 18 26 34
0.54
0.57
0.60
0.63
16 min
Absorb
ance a
t 260 n
m (
a.u
.)
0.54
0.57
0.60
0.63
8 min
0.54
0.57
0.60
0.63
4 min
b4
b8
b16
d4
d8
d16
Interval:
3rd
3rd
3rd
Figure 25. An influence of the injection time interval on SWNT saturation with 1,2,3-
trioxolanes. Three curves (top, middle and bottom) are three separate experiments.
Reactions were monitored by absorption at 260 nm and 25 C. Experimental points are
shown with circles. Upward spikes are due to needle insertion into cuvette at the time of
injection of O3/O2 gaseous mixture. First two injection points are emphasized with thin
black arrows. All three experiments had multiple injections with intervals: 4 min (top,
), 8 min (middle, ) and 16 min (bottom, ).
62
The same amount of ozone was injected into SWNT – SDS samples with different
time intervals. Results are summarized in Figure 25. SWNT suspension was subjected to
multiple ozonation with different time intervals. Ozone consumption rates were
calculated for an injection at or near 34 min (Figure 25).
Nine injections of 1.5 mL of O3/O2 gaseous mixture into 1 mL of SWNT – SDS
suspension with 4 min intervals resulted in a small upward step (marked as d4;
d4 = 0.0002 a.u.) and ozone consumption rate b4 = 3.42 min-1
(Figure 25). Extending time
interval to 8 min gave a larger downward step, d8 = 0.0033 a.u., and faster ozone
consumption rate, b8 = 3.95 min-1
(Table 9). A sixteen minute interval further increased
the rate, b16 = 4.63 min-1
, and increased the step to d16 = 0.0068 a.u. These results indicate
that neither four nor eight minutes are sufficient to fully decompose all ozonides.
Additional experiments would be of benefit to determine if 16 min is long enough to
decompose all trioxolanes. Separate vacuum deoxygenation studies indicated that oxygen
from ozonide decomposition can evolve for 20 minutes before deoxygenation amount per
minute falls below system leak threshold. The downward step d18 after the third injection
(curve ) was larger than the third steps on two other curves ( and ).
63
Table 9. Results summary for multiple injections of O3/O2 gaseous mixture into SWNT –
SDS suspensions with different time intervals*
3rd
injection at 34 min
Interval, min d, a.u. d, a.u. b, min-1
, min r2
4 0.0061 (0.0002) 3.42 0.29 0.9991
8 0.0040 0.0033 3.95 0.25 0.9988
16 0.0068 0.0068 4.63 0.22 0.9969
* - Regression was performed with a single exponential decay formula. Variables: d –
SWNT bleaching step, b – decay rate, - lifetime, r 2
– coefficient of determination.
Saturation of SWNT with 1,2,3-trioxolanes by reaction with solvated ozone
Time (min)
0.0 0.5 1.0 1.5
Ab
so
rba
nce
at
26
0 n
m (
a.u
.)
0.4
0.5
0.6
s
s
a3
a2
a1
d1
d2
Figure 26. Additions of water saturated with ozone to SWNT – SDS suspension.
Absorbance was monitored at 260 nm. Points of injection are emphasized with arrows.
Upward spikes caused by mixture stirring are marked with letter s. An expected
absorbance change due to dilution is shown with levels d1 and d2 (dotted lines) for
comparison.
The first addition of 0.2 mL of water saturated with ozone (absorbance ca. 0.85 a.u.)
to SWNT suspension (1 mL) resulted in previously seen ozone consumption „tail‟,
64
indicating that amount of ozone exceeded SWNT saturation limit. Observed absorbance
after the first injection was a2 = 0.4811, which is 0.0162 a.u. less than the expected
absorbance d1 when accounted for dilution. In terms of percent values, SWNT
absorbance was bleached by 1 - a2/d1 = 3.3 % at 260 nm. This is a very small number of
double bonds. It is believed that electron withdrawing by 1,2,3-trioxolanes has a
profound influence on tube ability to react with molecules of ozone. The second injection
of aqueous ozone gave even slower rate of ozone consumption, indicating tube saturation
with ozonides. An overall SWNT bleaching after the second injection was 1 - a3/d2 = 3.1
%. The error in these calculations may be significant, due to the fact that ozone slowly
reacts with SDS with an increase in absorbance at 260 nm. The more SWNTs are
saturated, the higher the percentage of SDS molecules will react with ozone. While the
error estimation is a subject for further discussion, the fact that only a small percent of
double bonds could be converted into trioxolanes is undeniable.
Structural changes of SWNT after ozonation
HipCo SWNTs were found to release a large amount of heat when reacted with
ozone at room temperature. Typically, such heating results in SWNT burning. Enthalpy
estimation for SWNT-O3 and SWNT and O3 with PM3 RHF method supports this
experimental observation (data not provided). Below are 3D pictures of semiconducting
tube (8,0) with three ozonides next to each other (Figure 27):
65
Figure 27. PM3 RHF optimized lateral (C) and longitudinal (A and B) trioxolanes on the
surface of semiconducting tube (8,0). Directions of ozonides are shown with dotted lines
and marked a, b and c for corresponding ozonides. Tube diameters of pristine and
ozonated sections are marked with labels s1 and s2 correspondingly. (Top) Diagonal
projection. (Bottom) Side projection. Optimization results are courtesy of S. Ghosh and S.
Bachilo.
An optimization of oxidized SWNT yielded tube widening in the area of ozonides
due several sp3 hybridized carbons. A longitudinal trioxolane, optimized with PM3
method for tube (10,0), had sp3 carbons significantly above the surface of SWNT
(marked with distance d2 in Figure 28). Carbon-carbon bond in ozonide was found longer
than what would be expected for a single C-C bond; a further optimization may be
needed.
Figure 28. An optimized structure of semiconducting tube (10,0) with a longitudinal
1,2,3-trioxolane on its surface. PM3 method was used for calculations. Shown are two
different projections. Optimization results are courtesy of S. Bachilo.
66
Ozonide decomposition will ultimately lead to a formation of an epoxide.
Semiconducting tube (8,0) with a lateral epoxide was optimized with PM3 RHF method
(Figure 29)
Figure 29. An optimized structure of semiconducting tube (8,0) with a lateral epoxide on
its surface. An optimization yielded elevation of sp3 carbons of epoxide above the surface
of the tube (distance d2). Used PM3 method for calculations. Shown are three different
projections. Optimization results are courtesy of S. Ghosh and S. Bachilo.
Previously performed optimization of ozonated fullerenes24
gave essentially the same
elevation of sp3 carbon atoms of an epoxide. Two fullerene structures, with „open‟ and
„closed‟ epoxides are shown of Figure 30. In both cases ratio d2/d3 was found to be
greater than 1, indicating fullerene distortion, or a “squeeze.”
67
A B C
D E F
Figure 30. PM3 optimized open (oxidoannulene) and closed (epoxide) structures of
C60O. d2 – height, d3 – width; (A-C) – an open form of C60O; (D-F) – a closed form of
C60O. Three different projections are shown for both structures. Optimization results are
courtesy of S. Bachilo.
Interaction of visible and NIR light with SWNT
SWNTs are black in color and seemed to absorb electromagnetic waves throughout
the entire electromagnetic spectrum. Spectra presented in this work covered UV, NIR, IR
areas with frequencies 41000 – 400 cm-1
. Pristine SWNTs had high absorbance at low
wavenumbers, below 1000 cm-1
. This absorbance is attributed to metallic SWNTs.
Ozonation of SWNTs resulted in irreversible damage to tube structure and disappearance
of low frequency transitions (IR studies, Chapter 4).
For tube to be black in color, an absorbing chromophore must have a substantial
number of conjugated rings in acene system. Figure 31 shows possible interactions of a
diode laser beam with semiconducting tubes (8,0) and (10,0).
68
(8,3)
(10,3)
Figure 31. Possible ways for interaction of light with SWNT. Tubes are shown from a
side. Possible chromophores are shown with a darker color; their direction is marked with
double arrows a1, a2 and a3. Top: semiconducting tube (8,3); Bottom: semiconducting
tube (10,3).
A profound work by Clar,25
who synthesized and studied UV-Vis spectra of
hundreds of polycyclic hydrocarbons, revealed that a significant number of conjugated
rings is needed to make compound absorb at wavelengths above 700 nm (Figure 32).
p absmax 737 nm
dark green-black
7.8,15.16-dibenzoterrylene
p absmax 603 nm
dark blue
bisanthene
p absmax 582 nm
violet - blue
pentacene
p absmax 693 nm
dark green
hexacene
Figure 32. Bathochromic effect for separate and conjugated acene systems. Provided are
absmax
for p-band transitions. Conjugation of acene systems increases p absmax
.
69
Bathochromic effect is a result of “dilution” of an aromatic sextet (shown with
circles). One sextet is shared among annealed rings. Clar observed strong bathochromic
effect for compounds with multiple conjugation of acene systems (dibenzoterrylene and
bisanthene in Figure 32). Observed NIR absorption and fluorescence peaks of SWNT are
likely from fairly small segments. Such an interpretation of SWNT absorption absmax
and emission emmax
wavelengths would explain why SWNT with up to 3 % of double
bonds being quenched by ozonides do not show peak shifts for absmax
or emmax
after
ozonation.
Activation energies for decomposition of SWNT trioxolanes
Activation energies Eact were estimated from trioxolane decomposition rates, b.
Rates were obtained by regression on inverted normalized fluorescence data sets. Curves
tails were truncated as needed to increase „weight‟ of initial points. A primary goal was to
estimate activation energies of trioxolane decay, a fast process. Activation energies for
slow processes were not studied. Oxidation reactions were conducted in a thermostated
cuvette with temperatures set to 15.4, 20.1, 25.0 and 29.9 C. Amount of ozone bubbled
through an aqueous SWNT suspension was sufficient to substantially decrease SWNT
fluorescence intensity. It was found that at higher temperatures more ozone had to be
injected to get the same oxidation degree. Ozone is less stable at higher temperatures,
thus the observed phenomenon is expected. Constituents of reaction media, such as
sodium hydroxide and sodium dodecyl sulfate are known for their ability to speed up
ozone decomposition, and are expected to increase ozone decay rates due to a greater
70
number of collisions between ozone molecules and hydroxyls or dodecyl sulfate anions at
higher temperatures. Due to the nature of the experimental setup, air above an aqueous
SWNT suspension was purged right after bubbling of O3/O2 gaseous mixture (ca. 3 v/v %
ozone). This step was needed to avoid penetration of gaseous ozone accumulated above
the liquid back into suspension, thus decreasing an observed ozonide decay rate.
Surfactant sodium dodecyl sulfate formed foam after bubbling O3/O2 gaseous mixture
and helped in retaining unreacted ozone right above the aqueous media. Purging of air
above the liquid was primarily to destroy foam bubbles, thus releasing entrapped ozone.
Rates obtained from regression were plotted against reciprocal temperatures to
obtain decay activation energies (Figure 33).
1 / T (103 K
-1)
3.30 3.35 3.40 3.45
Ln
(b
)
-5.5
-5.0
-4.5
-4.0
-3.5
-3.0 954 nm
1026 nm
1123 nm
1250 nm
Figure 33. Arrhenius plot. SWNTO3 decomposition rates (b, s-1
) were measured at
different temperatures for 954, 1026, 1123 and 1250 nm emission wavelengths.
Many factors contributed to possible errors in trioxolane decomposition rates. Larger
error was observed for longer emission wavelengths. While 95% of emission at emmax
71
954 nm is coming from tube (8,3), fluorescence at emmax
1250 nm is a combination of
signals from tubes, none contributing more than 40% of observed intensity. Assuming
that ozonide decay of different tubes will be affected differently with temperature
changes, larger error was expected.
Generally, experiments were repeated many times with different loads of ozone to
obtain the most accurate rates. The most common problem observed with rates
measurements was having either too much ozone or too little ozone. Fluorescence of tube
(8,3) with emmax
= 954 nm was quenched less efficiently than that of tubes with longer
emission wavelengths, e.g. (9,5) and (10,3).
Too little ozone resulted in insufficient number of exponential decay points for tube
(8,3) with emmax
954 nm. Too much ozone resulted in a decay curve distortion, or
waviness, mainly for wider tubes, e.g. (9,5) with emmax
~ 1244 nm. Getting the right
amount was a challenge.
Obtained ozonide lifetimes and activation energies are summarized in Table 10.
72
Table 10. Activation energies and ozonide lifetimes for tubes emitting at 954, 1026, 1123
and 1250 nm wavelengths.
Lifetime , sec
T, C 954 nm 1026 nm 1123 nm 1250 nm
15.4 85 94 137 221
20.1 47 54 67 92
25.0 31 38 50 75
29.9 22 27 35 48
Eact (kJ/mol) 68.9 61.6 66.8 72.1
Eact (eV) 0.71 0.64 0.69 0.75
r2 0.9907 0.9886 0.9618 0.9289
* - Abbreviations are: Eact - activation energy for ozonide decay; - ozonide lifetime,
which is an inverse value of decomposition rate b, i.e. = 1/b; r 2
– coefficient of
determination.
SWNT fluorescence sensitivity to 1,2,3-trioxolanes on its surface
Fluorescence of tubes with larger diameters, mainly (9,5), (10,3) and (11,1),
represented by curve in Figure 34, was quenched four times stronger than that of tube
(8,3) (curve ). Such a difference in fluorescence quenching degrees is thought to be due
to varied sensitivity of tubes to similar number of ozonides on their surface.
73
Time (min)
0 10 20 30 40 50
Inve
rte
d N
orm
aliz
ed
Flu
ore
sce
nce
( I
ma
x /
I )
0
10
20
30
40
No
rma
lize
d F
luo
resce
nce
0.0
0.2
0.4
0.6
0.8
1.0
954 nm
1027 nm
1125 nm
1251 nm
A
B
Figure 34. Formation and decomposition kinetics of SWNT ozonides at four distinct
wavelengths. Ozone addition at ca. 2 min led to a sharp decline in fluorescence intensity
followed by partial recovery over time. (A) Normalized fluorescence, (B) inverted
normalized fluorescence (Imax/I). Symbols denote wavelengths at which kinetics were
monitored. The 660 nm laser was used for excitation.
Tubes of interest and their diameters are summarized in Table 11.
74
Table 11. Major contributors to fluorescence intensity at four distinct wavelengths*
em, nm , cm-1
(n,m) type of
major
contributors
tube diameter,
nm
% of total
emission at em
955.6 10465 8,3 0.782 95.4
6,5 0.757 1.9
1027.6 9731 7,5 0.829 85.0
10,2 0.884 5.3
8,1 0.678 4.3
1124.6 8892 7,6 0.895 78.9
8,4 0.840 8.3
9,2 0.806 3.7
9,4 0.916 3.4
1250.1 8000 9,5 0.976 39.8
10,3 0.936 30.3
11,1 0.916 12.0
8,7 1.032 6.2
10,5 1.050 3.8
8,6 0.966 3.0
* Excitation source ex
max 660 nm.
Tube surface area is l times greater than its diameter, where l – is a tube length.
Evidently, surface area increase is no more than 20% for tubes with larger diameters
( emmax
~ 1250 nm). All tubes listed in Table 11 have a chiral angle, meaning they are
“twisted”. 3D structures of major contributors to fluorescence intensity are shown below
(drawn not to scale).
(8,3) (7,5) (7,6) (9,5) (10,3)
75
The tube curvature is playing a key role in ozonide decomposition rates, though there
is no clear understanding why fluorescence quenching of ozonated tubes is four times
greater for (9,5) and (10,3) when compared to (8,3). One possible explanation of such
difference between quenching degrees (Imax/I) is a varied tube sensitivity to presence of
1,2,3-trioxolanes on its surface. While it is possible to suggest that the reaction rate
between ozone and tube (9,5) may be faster than that for (8,3), normalized fluorescence
intensity of tube (9,5) fifty three minutes after ozonation was ca. 10% lower than that for
tube (8,3) (Figure 34). The 10% difference in normalized intensities could be because of
a wider diameter of tube (9,5).
76
Wavelength (nm)
950 1050 1150 1250
Norm
aliz
ed F
luore
scence I
nte
nsity
0.0
0.2
0.4
0.6
0.8
1.0
Flu
ore
sce
nce Inte
nsity (
nW
/nm
)
0.0
0.5
1.0
1.5
2.0
53 min
2 min
before O3
A
B
951 954
0.9
1.0
Figure 35. An influence of SWNT ozonation on emmax
locations. The reaction was
monitored at r. t. Plots are: (A) Fluorescence spectra as measured, (B) normalized
fluorescence (I/Imax). Curves are: – pristine SWNT suspended in 1 wt. % aqueous
SDS, – SWNT two minutes after bubbling 2 mL of O3/O2 gaseous mixture (ca. 3 v/v
% of ozone), – spectrum 53 min after ozonation. Legends are the same for the top and
the bottom plots.
77
A very small blue shift was observed for tube (8,3) with emmax
~ 954 nm right after
ozonation (curve ). This phenomenon is thought to be due to a depletion of an electron
density of the -cloud. This shift disappeared over time (compare curves and , see
zoom-in in Figure 35B). Fluorescence of tube (8,3) recovered to about 70 % level of its
original intensity 53 min after ozonation. Regression performed on the curve (curve ,
Figure 35A) estimated that final recovery level would not get higher than 75 % of an
initial intensity. It is not clear why fluorescence peak maxima are so insensitive to
sidewall modification. It is reasonable to expect emmax
peaks to shift for functionalized
tubes, though it is not observed. At the same time, a variety of surfactants (CTAB, SDBS,
SDS, Brij 700) were shown to affect peak maxima.26
From these observations a
conclusion can be made, that fluorescence from modified sections of SWNT does not
recover and that an observed recovery is likely coming from an increase of an electron
density on SWNT after ozonides‟ decay.
PM3-level optimization of SWNT tube with several 1,2,3-trioxolanes on its surface
gave a “squeezed” tube, with tube getting wider at the place of functionalization (Figure
27). Similar results were obtained for ozonated fullerenes. Such distortion affects tube
dimensions, and thus can affect its emmax
. Tube distortion and disrupted conjugated
-system are likely to be two main reasons for lack of fluorescence from functionalized
sections of SWNT.
Overall, fluorescence peaks maxima were found to be very insensitive to side wall
modification with 1,2,3-trioxolanes, i.e. there was no significant shift observed. It is
believed that fluorescence after ozonation is coming from pristine sections of tubes only.
78
Fluorescence recovery is likely due to an increase of an electron density of a -cloud
after ozonides decomposition to SWNT epoxides and molecular oxygen.8 Fluorescence
from functionalized areas is thought to be irreversibly lost. Special treatment, like tube
baking at high temperature, would be needed to restore original SWNT structure.
Work with ozone: gaseous vs. aqueous ozone and parameters that affect decay rates
observed by fluorescence
Ozone diffusion through water was found to be somewhat inefficient. In a separate
experiment a slow addition of an equal amount of water to an aqueous solution of ozone,
with Abs 260 ~ 1, did not yield absorbance decrease to an estimated dilution level even
within several minutes after addition. Mechanical stirring was required to halve the
absorbance value. The same problem was encountered when water saturated with ozone
was added to SWNT-SDS aqueous suspension. Seemingly large quantities of ozone were
quenching SWNT fluorescence inefficiently, even with fast injections. Speed of
injections was found to affect the effectiveness of mixing. Stirring with a spatula was
shown to increase SWNT oxidation degree. Bubbling of known volumes of gaseous
ozone through SWNT suspension gave a much better reproducibility of fluorescence
quenching degrees (Imax/I) when compared to addition of aqueous ozone. The main
advantage of using gaseous O3/O2 mixture was thorough suspension stirring with gas
bubbles. On the flip side, gaseous ozone was found entrapped in foam bubbles above the
aqueous suspension after injections and required an additional purge of space above the
liquid to eliminate ozone penetration from the gaseous phase back into suspension.
79
Sodium dodecyl sulfate, a surfactant for SWNT, is known for its ability to foam. Ozone
has a long decay rate in the air. Purging of air above the liquid was primarily to destroy
foam bubbles, thus releasing entrapped ozone. Penetration of gaseous ozone back into the
liquid phase affected negatively rates observed by fluorescence, meaning actual ozonide
decay rates are faster then the ones observed. An experiment was performed by bubbling
the same amounts of ozone through two 1 wt. % aqueous SDS suspensions. Air above the
liquid in a second cuvette was purged and decay rates were measured by absorbance
change at 260 nm. The purged cuvette had a slightly faster ozone decay rate, 0.47 vs.
0.43 min-1
.
Decay rates of SWNT 1,2,3-trioxolanes measured with fluorescence technique are
also dependant on solution acidity and surfactant aggregation at low temperature.
Decreasing temperature to below 15 C may result in surfactant precipitation. Acidity per
se does not seem to affect 1,2,3-trioxolane decomposition, but it has a strong influence on
SWNT fluorescence. Increase in acidity would give a weaker emission signal. Keeping
suspension pH in the range 8 -9 was necessary.
To summarize, injection of gaseous ozone was found to be the most efficient way to
oxidize SWNT. Injections of gaseous ozone should be followed by a brief stirring with a
spatula and a thorough purge of the gaseous phase above the SWNT suspension. It is
desirable to have the same injection speed in all runs. Running experiments in a
thermostated cuvette was found to be necessary for obtaining reproducible kinetics
results. These guidelines were followed in obtaining 1,2,3-trioxolane decay rates.
Suspension acidity should be kept in a range pH 8 – 9.
80
Fluorescence intensity drift
SWNT fluorescence was found to be fluctuating. Such fluctuations are thought to be
dependant on a number of parameters. Two of the most important factors were
suspension temperature and the length of laser pulses. Even small changes in the
temperature were shown to affect fluorescence intensity of SWNT suspended in 1 % aq.
SDS solutions. Such fluctuation was less noticeable in sodium dodecyl benzyl sulfonate
(SDBS) surfactant. An attempt to use SWNT suspended in SDBS was not successful, as
it was found that ozone does not quench SWNT fluorescence efficiently in presence of
SDBS. SDBS itself is reacting with ozone, thus reducing amount of ozone available for
reaction with SWNT. The laser pulse length was found to affect fluorescence fluctuation.
The least drifts were obtained with > 10 sec intervals between 500 msec pulses.
Decreasing duty cycle to below ten seconds resulted in random intensity drifts, both
upward and downward relative to the initial fluorescence. The majority of experiments in
this work were repeated numerous times to ensure that the observed kinetics were
reproducible and not affected by fluorescence drifts.
Ozone loads
Abnormally high loads of ozone, typically greater than 2 mL of O3/O2 gaseous
mixture (ca. 3 v/v % ozone) were found to distort single exponential decay curves.
Particularly, three types of decay were observed with high loads of ozone: fast ozonide
decay, slow ozonide decay and some rearrangement processes that did not result in
oxygen release. All three processes were found to follow exponential decays. The fast
and slow ozonide decays could be regressed with 5-parameter two exponential decay
81
formulas. Oxygen release was observed during these two processes. The very slow
rearrangement process was always observed, but its rate was not studied. Particularly,
random fluorescence drifts were overshadowing slowly changing fluorescence intensity,
thus introducing a large error.
2.3. Conclusions
An interaction of ozone with single-walled carbon nanotubes (SWNT) resulted in the
formation of 1,2,3-trioxolanes (SWNTO3). Obtained formation rate was 2.4·104 M
-1s
-1 for
SWNT – 1% aq SDS suspension, which is of the same order of magnitude as the
formation rates reported for carbon tetrachloride solutions of C60O3 and C70O3. SWNTO3
decayed to SWNT oxides (SWNTO) with release of molecular oxygen. A vacuum
deoxygenation technique performed on dry ozonated SWNT showed oxygen release to
follow simple exponential rise with rates ca. 1.5 – 2 min-1
at r. t. The lifetime of
SWNTO3, was shown to depend on temperature and SWNT type, and at room
temperature was less than two minutes for small-diameter SWNTs suspended in water.
Ozonides exhibited an extreme quenching of SWNT fluorescence and a substantial
bleaching of NIR absorption. The maximum number of 1,2,3-trioxolanes forming on the
surface of SWNT at any given time was found to be less than 4% of the theoretical value,
indicating a saturation point. Reaction of ozonated nanotubes with excess ozone was
limited by the SWNTO3 decomposition rates. Thinner tubes exhibited faster ozonide
decay rates resulting in greater oxidation levels over time in excess of ozone.
82
2.4. Experimental Part
Setup for measurement of oxygen evolution from SWNT ozonated in solid state
A set of experiments was designed to measure the amount of oxygen evolving from
the surface of ozonated SWNT. A vacuum line, shown in Figure 36 was equipped with an
electronic pressure gauge.
Figure 36. (Left) A vacuum line for measurement of oxygen release. A pressure gauge is
in the left top corner; a valve to cut off an oil diffusion pump is at the lower left. (Right)
A reaction vessel with a cap at the top and a valve on a side.
The entire system with a reaction vessel attached was vacuumed for at least one day
before each experiment. Typically, such vacuuming resulted in a background leak of
0.5 mTorr/min. This was a sufficiently low leak for the purposes of measurements.
Ozone collection
An example of a syringe used for ozone collection is shown in Figure 37. Typically,
samples were collected for 1 minute in a 3 mL disposable syringe with the O3/O2 gaseous
mixture flow set to 25 mL per minute (ca. 3 v/v % ozone).
83
Figure 37. (A) Teflon tubing coming out of a corona discharge ozonator and a
homemade assembly for syringe attachment. (B) 3 mL disposable syringe outer core is
attached to an assembly for ozone collection.
It was found necessary to connect a syringe plunger to the outer core immediately
after it was taken off of the ozonation line. This was done to reduce air flows above the
syringe and made the amount of ozone reproducible with each collection.
Experimental determination of the amount of oxygen evolved from ozonated
SWNT.
The following procedure describes determination of the volumes of the reaction
vessel and the vacuum line. Calculations of the amount of evolved oxygen are also
provided.
The vacuum line was vented to the air to get 803 mTorr in the entire system. The
reaction vessel was cut off and the rest of the system evacuated. The pump was cut off
and the reaction vessel opened; gas from the vessel spread through the system and caused
pressure to change to 197 mTorr, indicating the ratio of volumes k = 803/197 = 4.08. The
reaction vessel was taken off the vacuum line, filled with water and its volume
determined, Vol (vessel) = 34.0 mL, from which volume of the system was determined
84
Vol (system) = 139 mL. Calculation of the number of moles of oxygen evolved after
ozonation was determined with the following equations:
obs
obsobs
stp
stpstp
T
VP
T
VP can be rewritten as
obsstp
stpobs
obsstpTP
TPVV
Parameters are:
Tstp = 273.15 K
Tobs = 293 K
Pstp = 760 Torr (1 atm)
Vm = 22.414 L/mol (at 0 C and 1 atm)
][1071.1760293
15.273139.0 4 TorrP
PLV obs
obsstp
The molar amount of oxygen was estimated with the following equation
414.222
stp
m
stp
O
V
V
V
The results are summarized in Table 12:
85
Table 12. Conversion of oxygen gas pressure into amount
O2), umol Vstp, uL Pobs, mTorr
0.23 5.1 30
0.30 6.8 40
0.38 8.6 50
0.45 10 60
0.54 12 70
0.62 14 80
0.67 15 90
0.72 16 95
Equipment
Near-IR fluorescence and absorbance in the range 900-1350 nm were recorded on a
NS1 Nanospectralyzer (Applied NanoFluorescence LLC, Houston, TX). Built-in lasers
660 and 785 nm were employed for excitation as noted in the text. 1% aq. sodium
dodecyl sulfate (Aldrich) solution served as a reference. Software that came with the NS1
Nanospectralyzer was used for spectra deconvolution. UV-Vis absorbance in a range
250-900 nm was recorded on a Cary 4E UV-Vis Spectrophotometer with 1 cm quartz
cuvette. Unless otherwise noted, 1 cm quartz cuvettes were used for UV-Vis and NIR
absorption measurements. Oxygen gas measurements were performed with a 275 Mini-
Convectron (Granville-Phillips Co.) pressure gauge.
Preparation of a surfactant coated SWNT aqueous suspension
Single-walled carbon nanotubes (60 mg, SWNT, HipCo, batch 161.1, raw,
unpurified, Rice University) were debundled in 1 wt. % aqueous solution (200 mL) of
sodium dodecylsulfate (Aldrich) with a bath sonicator (FS 14, Fisher Scientific) for
4 hours. Further dispersion was performed by applying intense ultrasonic agitation
(7 watt) with a tip sonicator (Microson XL2000) for 15 min. The sample was ultra-
86
centrifuged (Sorvall, Discovery 100SE) for 4 hours at 151514.2 g to obtain a clear, dark
grey decant, which was separated from the precipitate right after the centrifugation. The
SWNT concentration in the decant was approx 23 mg/L estimated by visible absorption.
(Absorbance values measured at 632 and 763 nm were divided by coefficients 0.036 and
0.043 L/mg respectively to calculate the SWNT concentration (mg/L).)27
The resulting
suspension was diluted four times with 1% aq. SDS solution and used for aqueous
experiments. An oxygen flow meter (Puritan-Bennett Corp.) was used at flow rates 1/16
and 1/32 L/min as noted.
Ozonation procedures
Ozone was generated by passing oxygen (industrial grade, Matheson Tri-Gas) with a
flow rate 1/32 L/min at r. t. through a high frequency corona discharge ozonator
(GE60FM, Yanco Industries Ltd, www.ozoneservices.com) set to a maximum output
(power level 10). The ozonator was idled before use for at least 7 min to reach its
maximum output as instructed by manufacturer. Three different ozonation procedures
used in this work are described below.
Direct ozonation of aqueous SWNT-SDS suspension. A gaseous mixture of O3/O2
(ca. 3 v/v % ozone) was bubbled through an aqueous SWNT suspension in a test tube for
a desired period of time.
Gaseous ozone/oxygen injection with a syringe. Ozone was collected in a plastic
syringe for 1 min (3 mL volume; cf. Figure 37B) or 2 min (5-10 mL volume), capped
with a plunger and the desired volume injected into a cuvette with SWNT suspension
followed by brief stirring with a spatula and purging of the air above the liquid.
87
Dry ozonation. The oxygen flow rate was kept at 1/16 L/min. A SWNT film was
ozonated for a desired period of time at r. t. in a small chamber by holding 1/16” ID
Teflon tubing above the dry SWNT film. Alternatively, a specific volume of O3/O2
gaseous mixture (ca. 3 v/v % ozone) could be squeezed from a syringe pointing the gas
stream at the SWNT film.
Oxygen evolution measurements
A corresponding amount of SWNT (2 or 4 mg, SWNT, HipCo, batch 162.8, raw,
unpurified, Rice University) was debundled in benzene (10 mL) with a bath sonicator
(Fisher Scientific, FS 60) and a slurry added to a cylindrical reaction vessel. The vessel
was tilted horizontally and rotated until all solvent was evaporated. Thin Teflon tubing
with a constant flow of nitrogen gas was inserted into the vessel to speed up the
evaporation process. The resulting SWNT-coated glass vessel had an evenly distributed
layer of nanotubes. The vessel was attached to a vacuum line and the entire system
evacuated for a day. When the system leak decreased to below 0.5 mTorr/min following
pump cut off, the vessel was cooled with an ice bath, an oxygen/ozone gaseous mixture
injected (10 mL, ca. 3 v/v % ozone), the vessel capped; waited for 1 min and then the
entire system was degassed for 2.5 min, then the pump was cut off and the pressure
monitored for 20 min. For the second ozonation, the system was thoroughly degassed,
reaction vessel cooled, the O2/O3 gaseous mixture (10 mL) was injected and the above
procedure followed.
88
References and Notes
1. Chen, Z. Y.; Hauge, R. H.; Smalley, R. E., Ozonolysis of functionalized single-
walled carbon nanotubes. Journal of Nanoscience and Nanotechnology 2006, 6,
(7), 1935-1938.
2. Chen, Z. Y.; Ziegler, K. J.; Shaver, J.; Hauge, R. H.; Smalley, R. E., Cutting of
single-walled carbon nanotubes by ozonolysis. Journal of Physical Chemistry B
2006, 110, (24), 11624-11627.
3. Simmons, J. M.; Nichols, B. M.; Baker, S. E.; Marcus, M. S.; Castellini, O. M.;
Lee, C. S.; Hamers, R. J.; Eriksson, M. A., Effect of ozone oxidation on single-
walled carbon nanotubes. Journal of Physical Chemistry B 2006, 110, (14), 7113-
7118.
4. Banerjee, S.; Hemraj-Benny, T.; Balasubramanian, M.; Fischer, D. A.; Misewich,
J. A.; Wong, S. S., Ozonized single-walled carbon nanotubes investigated using
NEXAFS spectroscopy. Chemical Communications 2004, (7), 772-773.
5. Banerjee, S.; Wong, S. S., Rational sidewall functionalization and purification of
single-walled carbon nanotubes by solution-phase ozonolysis. Journal of Physical
Chemistry B 2002, 106, (47), 12144-12151.
6. Banerjee, S.; Wong, S. S., Demonstration of diameter-selective reactivity in the
sidewall ozonation of SWNTs by resonance Raman spectroscopy. Nano Letters
2004, 4, (8), 1445-1450.
7. Dresselhaus, M. S.; Dresselhaus, G.; Jorio, A.; Souza, A. G.; Pimenta, M. A.;
Saito, R., Single nanotube Raman spectroscopy. Accounts of Chemical Research
2002, 35, (12), 1070-1078.
8. Heymann, D.; Bachilo, S. M.; Weisman, R. B.; Cataldo, F.; Fokkens, R. H.;
Nibbering, N. M. M.; Vis, R. D.; Chibante, L. P. F., C60O3, a fullerene ozonide:
Synthesis end dissociation to C60O and O2. Journal of the American Chemical
Society 2000, 122, (46), 11473-11479.
9. Liu, L. V.; Tian, W. Q.; Wang, Y. A., Ozonization at the vacancy defect site of
the single-walled carbon nanotube. Journal of Physical Chemistry B 2006, 110,
(26), 13037-13044.
10. Cai, L. T.; Bahr, J. L.; Yao, Y. X.; Tour, J. M., Ozonation of single-walled carbon
nanotubes and their assemblies on rigid self-assembled monolayers. Chemistry of
Materials 2002, 14, (10), 4235-4241.
11. Mawhinney, D. B.; Naumenko, V.; Kuznetsova, A.; Yates, J. T.; Liu, J.; Smalley,
R. E., Infrared spectral evidence for the etching of carbon nanotubes: Ozone
89
oxidation at 298 K. Journal of the American Chemical Society 2000, 122, (10),
2383-2384.
12. Ogrin, D.; Chattopadhyay, J.; Sadana, A. K.; Billups, W. E.; Barron, A. R.,
Epoxidation and deoxygenation of single-walled carbon nanotubes:
Quantification of epoxide defects. Journal of the American Chemical Society
2006, 128, (35), 11322-11323.
13. Cataldo, F., Polymeric fullerene oxide (fullerene ozopolymers) produced by
prolonged ozonation of C60 and C70 fullerenes. Carbon 2002, 40, (9), 1457-1467.
14. Cataldo, F.; Heymann, D., A study of polymeric products formed by C60 and C70
fullerene ozonation. Polymer Degradation and Stability 2000, 70, (2), 237-243.
15. Churilov, G. N.; Isakova, V. G.; Weisman, R. B.; Bulina, N. V.; Bachilo, S. M.;
Cybulski, D.; Glushchenko, G. A.; Vnukova, N. G., Synthesis of fullerene
derivatives. Physics of the Solid State 2002, 44, (4), 601-602.
16. Deng, J. P.; Mou, C. Y.; Han, C. C., Oxidation of fullerenes by ozone. Fullerene
Science and Technology 1997, 5, (5), 1033-1044.
17. Heymann, D.; Weisman, R. B., Fullerene oxides and ozonides. Comptes Rendus
Chimie 2006, 9, (7-8), 1107-1116.
18. Razumovskii, S. D.; Bulgakov, P. G.; Ponomareva, Y. G.; Budtov, V. P., Kinetics
and stoichiometry of the reaction between ozone and C70 fullerene in CCl4.
Kinetics and Catalysis 2006, 47, (3), 347-350.
19. Razumovskii, S. D.; Bulgakov, R. G.; Nevyadovskii, E. Y., Kinetics and
stoichiometry of the reaction of ozone with fullerene C60 in a CCl4 solution.
Kinetics and Catalysis 2003, 44, (2), 229-232.
20. Chibante, L. P. F.; Heymann, D., On the Geochemistry of Fullerenes - Stability of
C60 in Ambient Air and the Role of Ozone. Geochimica Et Cosmochimica Acta
1993, 57, (8), 1879-1881.
21. Bulgakov, R. G.; Nevyadovskii, E. Y.; Belyaeva, A. S.; Golikova, M. T.;
Ushakova, Z. I.; Ponomareva, Y. G.; Dzhemilev, U. M.; Razumovskii, S. D.;
Valyamova, F. G., Water-soluble polyketones and esters as the main stable
products of ozonolysis of fullerene C-60 solutions. Russian Chemical Bulletin
2004, 53, (1), 148-159.
22. Razumovskii, S. D.; Zaikov, G. E., Ozone and its reactions with organic
compounds. ed.; Elsevier Science Publishers: New York, NY, 1984; Vol. I, p.
403.
90
23. Concentration of ozone in oxygen stream was provided by the manufacturer of the
ozonator.
24. Weisman, R. B.; Heymann, D.; Bachilo, S. M., Synthesis and characterization of
the "missing" oxide of C60: [5,6]-open C60O. Journal of the American Chemical
Society 2001, 123, (39), 9720-9721.
25. Clar, E., Polycyclic Hydrocarbons. Academic Press: New York, 1964; Vol. 1, p.
487.
26. Moore, V. C.; Strano, M. S.; Haroz, E. H.; Hauge, R. H.; Smalley, R. E.; Schmidt,
J.; Talmon, Y., Individually suspended single-walled carbon nanotubes in various
surfactants. Nano Letters 2003, 3, (10), 1379-1382.
27. Moore, V. C. Single walled carbon nanotubes: Suspension in aqueous/surfactant
media and chirality controlled synthesis on surfaces. Rice University, Houston,
2005.
91
Chapter 3
Influence of SWNT ozonation on D and G bands in Raman spectra
92
3.1. Introduction
Studying disorder-induced D and tangential G modes of graphite crystallites in
Raman spectra can be tracked back to the pioneering work of Tuinstra and Koenig1
published in 1970. Kastner et al.2 compared Raman and infrared spectra of carbon
nanotubes to those of highly oriented pyrolytic graphite (HOPG). Raman active E2g in-
plane stretching mode, often designated as the G mode, was found at 1574 cm-1
for
carbon nanotubes. A disorder induced D-band was found at ca. 1350 cm.-1
The IR
spectrum of HOPG had lines at 868 (out-of-plane mode, A2u) and 1588 cm-1
(in-plane
mode, E1u), while that of carbon nanotube had the same A2u peak and a much broader
asymmetrical E1u peak at 1575 cm.-1
The IR spectrum of carbon nanotubes presented in
the article does not resemble the ones observed for pristine HipCo tubes in this work. The
peaks A2u and E1u were observed for ozonated SWNT samples, but not for pristine ones
(this work). Pimenta et al.3 studied disorder-induced D and D’ Raman features, as well as
the G’-band (the overtone of the D-band which is always observed in defect-free
samples). The authors determined G band (appearing near 1582 cm-1
in graphite) as a
doubly degenerate phonon mode (E2g symmetry) at the first Brillouin zone center that is
Raman active for sp2 carbon network. Pimenta noted that integrated intensity ratio ID/IG
for the D band and G band is widely used for characterizing the defect quantity in
graphitic materials. Expanding the work of Tuinstra and Koenig,1, 4
Cancado et al.5
provided the following formula for estimation of in-plane crystallite size La:
93
1
410104.2)(G
Dlasera
I
InmL , where crystallite size La and excitation
wavelength are expressed in nanometers. Parameters ID and IG are integrated intensities
of the G and disorder-induced D bands in Raman spectra. Thomsen and Reich6 linked
excitation energy dependence of the D mode to a double resonant process. Ferrari et al.7
examined relation of D, G and G’ bands in graphene, and found the dependence of G’
band shape and location on the number of graphene layers. A recent summary of IR and
Raman spectral changes of purified vs. pristine SWNT has been reported.8 An analysis of
Raman spectra of SWNT from different sources was published by Hennrich et al.9
It appears that the influence of small quantities of ozone on SWNT Raman features
has not been reported to date. This chapter will give a brief summary of the influence of
ozonation of aqueous and solid samples on Raman disorder (D) and tangential (G) modes
in SWNT. As will be demonstrated, small loads of ozone resulted in a significant
decrease of G band and had little or no influence on the disorder band.
94
3.2. Results and discussion
Wavelength (nm)
710 730 750 770 790 810
Counts
, x10
3
5
10
15
20
749.5
5 min
10 m
17 m
24 m
32 m
37 m
46 m
D
G
G'
Figure 1. Raman spectra acquired over time on an aqueous SWNT – SDS suspension
after bubbling O3/O2 gaseous mixture (ca. 3 v/v % ozone) for 3 min. The sample was
excited with 669.9 nm laser source. Consult Chapter 1 for conversion of wavelengths into
wavenumbers. Symbols are labeled with approximate times from the beginning of
ozonation.
Several Raman spectra of an aqueous SWNT – SDS suspension were measured over
time after its ozonation. Gas bubbles evolving during ozonides decomposition were found
to be interfering with Raman measurements causing random, not reproducible peaks in
the acquired spectra. An attempt was made to focus on SWNT spectral changes after
major portion of ozonides has already decomposed, thus decreasing chances of
misinterpreting the data. The first spectrum in Figure 1 was recorded ca. 2 minutes after
bubbling of gaseous ozone through a SWNT – SDS suspension.
As seen in Figure 1, the G-peak increased over time by 33% from its 5 min value.
Within the same timeframe the basis function of D-band did not change. A change in
95
fluorescence background between 5 and 10 min spectra is noticeable. Small random
peaks due to light interference with oxygen bubbles are seen in the region 710 – 725 nm
(curve ).
It is believed that multiple processes contribute to an increase in the intensity of the
G- band, two major ones being ozonide decay (a fast process) and structural
rearrangement of epoxides to oxidoannulenes (a slow process). By analogy with
fullerenes,10
an oxidation of SWNT is likely to yield multiple functional groups on its
surface. It was concluded from IR studies of ozonated SWNT films (see Chapter 4) that
epoxides, aryl ethers, -diketones, anhydrides and esters are likely to be dominant
functional groups on the SWNT surface after ozonide decay. G-band intensity growth in
Figure 1 is thought to be due to SWNT structural changes leading to a greater number of
Raman active double bonds.
Multiple injections of O3/O2 gaseous mixture (ca. 3 v/v % ozone) into a SWNT –
SDS suspension led to absorbance bleaching. Absorbance at 260 nm fell by
approximately 10% after eight injections with 4 min intervals between each (Chapter 2,
Figure 13). Raman spectra were acquired before and 15 hrs after ozonation. As was
demonstrated in the fluorescence studies (cf. Chapter 2 for details), extended time period
was needed for tubes to recover. As shown in Figure 2, G-peak recovered only to 62 %
level of its original intensity. A slight increase of background is thought to be due to
fluorescence of certain oxidized sections of SWNT. The basis function and the shape of
the D-band did not change after ozonation (Figure 2).
96
Wavelength (nm)
710 730 750 770 790 810
Coun
ts, x10
3
10
20
30
40before O
3
15 hr after
730 750
Cou
nts
, x1
03
3
6
D
G
G'
D
G
Figure 2. Raman spectra of SWNT-SDS aqueous suspension before and 15 hrs after
ozonation. Sample was excited with 669.9 nm laser source. Eight O3/O2 gaseous mixture
injections (ca. 3 v/v % ozone), 1.5 mL each, were made with 4 min intervals and
suspension kept at r. t. for 15 hours.
Though the D/G peak intensity ratio changed, this result does not support the
common point of view that D band intensity should increase with a higher degree of side
wall functionalization.
For a vibration to be Raman active, it needs a change in polarizability and no change
in a dipole moment. Carbon atoms participating in C-O bonds of SWNT epoxides and
oxidoannulenes can not be considered as truly sp3 hybridized, since bond angles do not
match those in a diamond. Furthermore, if oxidoannulene rearranges to diaryl ether,
carbon atoms become sp2 hybridized. Introduction of epoxides, oxidoannulenes,
anhydrides, esters and -diketones is likely to increase the intensity of asymmetric
stretches, contributing little or nothing to stretches with zero dipole. Theoretically
97
peroxides, epoxides and aryl ethers should have Raman active stretches in the region
1000-1200 cm-1
, near D-band, but experimentally no D-band growth could be observed
for an aqueous suspension of ozonated SWNT.
The ambiguity surrounding D-band growth during side wall modification led to a
series of UV measurements. As mentioned above, eight 1.5 mL injections of O3/O2
gaseous mixture (ca. 3 v/v % ozone) were bubbled through 1.5 mL of SWNT - SDS
aqueous suspension with 4 min intervals. The reaction of SWNT with ozone was
monitored at absmax
(O3) = 260 nm (Figure 3). Each injection led to an initial absorbance
increase followed by a fast exponential decay. The amount of ozone that could be
dissolved in 1% SDS solution was found to be limited. An absorbance jump in the
presence of SWNT was less than 0.1 a. u. An increase in the upward jump amplitude
(first to fifth injections, dotted line in Figure 3) and longer times needed to consume free
floating ozone was interpreted as a consequence of SWNT saturation with 1,2,3-
trioxolanes, also known as primary ozonides. A four minute time period at r. t. was found
to be insufficient to decompose all ozonides. This interpretation explains why downward
“step,” or SWNT bleaching, is getting smaller with each subsequent ozonation.
98
Time (min)
10 20 30
Ab
so
rba
nce
(a
.u.)
0.54
0.57
0.60
Figure 3. Absorbance of SWNT – SDS aqueous suspension at 260 nm and room
temperature. Eight injections of O3/O2 gaseous mixture (ca. 3 v/v % ozone) were made
with 4 min intervals (emphasized with arrows). Solid line represents absorbance change
due to needle insertion. Points above 0.62 a.u. are extraneous and were excluded from the
graph. A diagonal dotted line marks an increase of absorbance right after each ozone
injection.
The resulting suspension was kept at r.t. for 15 hrs and then the Raman spectrum was
recorded (Figure 2).
Oxidation of SWNT with different amounts of ozone as monitored by Raman.
Injection of specific volumes of ozone (0.5 – 10 mL) led to a decrease of the G-peak
intensity with no change in the D band (Figure 4).
A functionalization of SWNT side wall with 1,2,3-trioxolanes was accompanied by a
broad increase of background fluorescence intensity. The G-peak dropped approximately
26% from its initial value after an injection of 4 mL of O3/O2 gaseous mixture (ca. 3 v/v
% ozone) and reached a saturation point. An injection of 5 or 6 mL of ozone brought the
99
G peak to the same intensity level. Higher loads of ozone (8 and 10 mL) resulted in
increased background fluorescence (curves and in Figure 4).
Wavelength (nm)
710 730 750 770 790 810
Counts
, x10
4
1
2
3
4
749.5
Cou
nts
, x1
04
2.8
3.2
3.6
before O3
0.5 mL
1 mL
2 mL
3 mL
4 mL
5 mL
6 mL
8 mL
10 mL
734
Cou
nts
, x1
04
0.4
0.8
D
G
G'
D G
Figure 4. The influence of different volumes of ozone on G peak intensity. Specific
amounts of gas were injected into separate aliquots of aq. SWNT - SDS suspension. All
samples were heated to 40 C for 30 min before measuring Raman spectra. The 669.9 nm
laser was used for excitation.
To better understand how many 1,2,3-trioxolanes can form on SWNT at any given
time, aliquots of SWNT suspension were bubbled with ozone oxygen mixture (OOM) for
specific periods of time. The first bubbling time was set to 30 seconds, which would be
the equivalent of injecting 13 mL of O3/O2 gaseous mixture. Other times were 1, 2, 3, 4,
5, 10, 30 min and 1 hour. Bubbling O3/O2 gaseous mixture for one hour resulted in a
complete disappearance of D and G bands and a substantial increase in fluorescence. The
sample changed color from light grey to light brown. This spectrum will not be presented
here. Thirty minute ozonation gave a very strong fluorescence and small D and G peaks,
100
as expected. When corrected for fluorescence background, the basis function under D-
band preserved its size and shape (curve in Figure 5).
Wavelength (nm)
710 730 750 770 790 810
Counts
, x10
4
0
1
2
3
4before O3
30 s
1 m
2 m
3 m
4 m
5 m
10 m
30 m
D
G
G'
Figure 5. Bubbling O3/O2 gaseous mixture (ca. 3 v/v % ozone) through SWNT – SDS
suspension for noted periods of time. Spectra are shown as measured and overlaid for
comparison. The 669.9 nm laser was used for excitation. Each curve represents a separate
experiment. All samples were heated to 40 C for 30 min before measuring Raman
spectra.
A decrease in G band intensity was not accompanied by D band increase after
ozonation (Figure 5). D band was found unchanged within the first ten minutes of
bubbling O3/O2 gaseous mixture (curves and ).
Fluorescence of functionalized SWNT was found to be fairly strong. Interestingly,
the Raman spectrum of a phenylsulfonated SWNT aqueous suspension in 1% SDS (not
shown) resembled the one ozonated for 30 min (curve in Figure 5).
To simplify a visual comparison of D and G peaks for different curves, plots were
shifted along vertical axis to make D-band (734 nm) at the same intensity level as in
101
SWNT sample before ozonation. This is done to compensate for unequal fluorescence
background in all samples.
Wavelength (nm)
710 730 750 770 790 810
Counts
, x10
4
0
1
2
3
4
before O3
30 s
1 m
2 m
3 m
4 m
5 m
10 m
30 m
749.5D
GG
G'
Figure 6. Bubbling O3/O2 gaseous mixture (ca. 3 v/v % ozone) through SWNT – SDS
suspension for noted periods of time. Raman spectra were shifted along vertical axis to
make D-band (734 nm) at the same intensity level as in SWNT sample before ozonation.
The 669.9 nm laser was used for excitation. Each curve represents a separate experiment.
All samples were heated to 40 C for 30 min before measuring Raman spectra.
Figure 6 demonstrates that the basis function for G band stopped decreasing at or
before 2 min of bubbling of O3/O2 gaseous mixture (ca. 3 v/v % ozone), indicating a
saturation point. Two minutes of bubbling is equivalent to a slow injection of 50 mL of
O3/O2 gaseous mixture. It is clearly seen in Figure 6 that ozonation for the first 10 min
does not affect the D band. After an initial fast decrease of G band (curves , and )
curves for 2 through 5 min of continuous ozonation (, , and ) gave essentially the
same intensity of G band. Longer ozonation times resulted in further decrease of G band
(curves and ).
102
The influence of sidewall modification on Raman spectra. A comparison between
different reactions.
A comparison was made between Raman spectra of pristine, dodecylated,
phenylated, phenylsulfonated and ozonated SWNT. Both liquid (dispersed tubes) and
solid state (bundled tubes) spectra were acquired, showing close similarity. Water
suspensions of the above mentioned samples of SWNT were prepared in 1 aq. % SDS by
bath sonication. Raman spectra of suspensions were acquired with 669.9 nm laser source
and had slightly broader peaks (spectra not shown). Fluorescence from samples was
much stronger in liquid state and therefore only solid state spectra will be discussed.
Wavenumber (cm-1
)
500 1000 1500 2000 2500 3000
Co
un
ts
x1
04
0
2
4
6
8
10
12Pristine SWNT
dodecyl
Ph
C6H
4-SO
3H
epoxide
Modified SWNT:
G
G'D
1295 1595
1
2
3
4
5
D
G
Figure 7. Raman spectra of pristine and functionalized SWNT overlaid for comparison.
Solid SWNT samples were excited with 785 nm source. All intensities are shown in
counts, as measured. SWNT samples are: – pristine, – dodecylated, –
phenylated, – phenylsulphonated, – ozonated. Dodecylated and phenylsulfonated
SWNT samples are courtesy of Feng Liang.
103
Intensities of spectra in Figure 7 are shown in counts, as measured. Commonly used
normalization of different spectra at the G-peak frequency will not be used in this work
for reasons discussed below. Utilization of ratio of intensities of D and G bands (D/G
ratio) will be used instead to describe a degree of SWNT functionalization. There is no
doubt that D band changes in its shape and intensity after SWNT functionalization, but
quantifying those changes cannot be easily derived from D/G ratios. Comparison of
spectra in Figure 7 clearly demonstrates the complexity of the quantification problem.
There is a striking difference in intensities of peaks of common aromatic compounds
versus SWNT in Raman spectra. Spectra in Figure 7 were acquired with an attenuated
laser on a single scan. Recording a spectrum of picric acid with the same conditions gave
much weaker intensities. The highest peak on a spectrum of picric acid (not shown) had
approximately the same height as the D band in pristine SWNT. The high intensity of the
G peak, ascribed to sp2-hybridized carbon atoms, is thought to be due to resonance of
conjugated double bonds.
A disruption of conjugation and introduction of electron withdrawing groups during
oxidation with ozone led to a substantial decrease of the G peak intensity and, at the same
time, gave almost no change of D band (curve ). Some broadening of the D peak is
associated with formation of epoxides and other functional groups. Banerjee11
in his
studies of ozonated SWNT samples encountered exactly the same problem with weak
increase in D band. He proposed that Raman cross sections for sp2 and sp
3 hybridized
carbons are dramatically different.
Clearly, conversion of an sp2-hybridized carbon to an sp
3 carbon after ozonation did
not result in increase of D band intensity (Figure 7). This result can be explained by
104
distinguishing sp3 carbon atoms of SWNT framework formed by an attack of a carbon-
based reactive center (e.g. phenyl radical) and other reactive species (e.g. ozone). An
electron withdrawing character of an attacking species is likely to decrease the intensity
of symmetrical stretches (Raman active) and increase the intensity of asymmetrical
stretches (IR active) of sp3 carbon atom on SWNT. Thus, an introduction of epoxides
gave no intense symmetric stretches. IR studies of ozonated samples demonstrated that
some epoxides can rearrange to other functional groups or be further oxidized with
ozone. Particularly, aryl ethers, -diketones, lactones and anhydrides are thought to be
forming in addition to epoxides. Conversion of epoxides to other functional groups is
another reason why there was no substantial increase of intensity at or near D band.
Recording Raman spectra right after ozonation of an aqueous suspension of SWNT
with a small amount of ozone (~2 mL O3/O2 gaseous mixture, ca. 3 v/v % ozone) gave no
increase of D band, a clear indication that formed ozonides and epoxides do not produce
intense symmetric stretches. In fact, no additional peaks were observed in Raman, even
though the presence of ozonides within first several minutes was proved by changes in
UV, IR, fluorescence and by measuring an amount of evolved oxygen. Raman spectra
show a change in G peak intensity over time, i.e. G peak increased with ozonide
decomposition, but no additional peaks attributable to symmetric stretches of epoxides or
ozonides could be found.
Phenylated () and phenylsulfonated () SWNT samples gave different intensities
of D band. There are three parameters that affect D band intensity: a) quantity of SWNT
in the beam of excitation source, b) degree of functionalization, and c) nature of a group
105
covalently attached to SWNT framework. Disorder, or D band, of phenylsulfonated
SWNT was found to have lower intensity than that in phenylated sample. Due to the
nature of Raman microscope measurement, it is very difficult to have exactly the same
amount of a SWNT in the laser spot. While there is a possibility that the amount of
SWNT-(C6H4-SO3H)n sample in a laser spot was less than that in phenylated sample, the
presence of electron withdrawing group on the ring is likely to decrease intensity of
symmetric stretches of sp3 carbon atoms on SWNT. Both phenylated and
phenylsulfonated samples were prepared with roughly the same amount of SWNT per
studied surface area (for solid state Raman).
Intensity of D band in dodecylated sample (curve ) is thought to be a composite of
symmetric vibrations of sp3 carbons in dodecyl chains and in SWNT. There is no clear
explanation why symmetric stretches of methyl and methylene groups, expected to
manifest themselves at 2850-2950 cm-1
are not present in the spectrum (curve ). Tight
packing of SWNT bundles could have influenced the symmetry of CH3 and CH2
stretches. It should be noted that fluorescence of a dodecylated sample was much higher
than of any other sample in Figure 7. Ideally, fluorescence background should be
subtracted for more accurate comparison of basis functions of D and G peaks.
3.3. Conclusions
The major change in Raman spectra of SWNT after functionalization is a decrease of
G peak. The intensity change is directly proportional to the degree of disruption of
conjugated - system. The D peak was shown to increase, but the growth was only a
fraction of the intensity lost at the G peak. The presence of an electron withdrawing
106
group next to sp3 carbon on SWNT is likely to decrease intensity of sp
3-carbon
symmetric stretches, thus resulting in minimal or no growth of D band in functionalized
samples. Formation of sp3 carbon centers on SWNT during reaction with ozone gave no
additional peaks (for ozonides and for epoxides) and did not result in D band growth. The
shape of the D band changed slightly, likely due to appearance of different functional
groups on SWNT surface. In-plane crystallite size (La) decreased approximately 1.8 times
after two minutes of bubbling O3/O2 gaseous mixture (ca. 3 v/v % ozone). Qualitatively,
phenylsulfonated SWNT were found to have lower D band intensity when compared to
phenylated SWNT. Published cases of D band growth values based on D/G ratios are less
representative of D band growth and more representative of G band decrease. While there
is no doubt that the D band increases, at least in cases with phenylated and dodecylated
samples, this increase is minor when compared to G band decrease. In discussion of
factors influencing D band growth, Raman data of dodecylated SWNT samples should be
treated with caution, since dodecyl chains themselves have a peak at 1300 cm-1
, near the
D band, thus causing it to go to a higher intensity level.
3.4. Experimental Part
SWNT - SDS aqueous suspension was prepared by a standard protocol (see
Experimental Part for Chapter 2). UV-Vis absorption in a range 250 - 900 nm was
recorded on a Cary 4E UV-Vis Spectrophotometer with 1 cm quartz cuvette. Solution
Raman spectra were recorded on a Jobin Yvon Spex Fluorolog with a 4 x 4 mm quartz
cuvette and an external 669.9 nm diode laser. Solid state Raman spectra were recorded on
107
Renishaw Raman Microscope; each spectrum was obtained in one pass with 785 nm
excitation source.
3.5. References and Notes
1. Tuinstra, F.; Koenig, J. L., Raman Spectrum of Graphite. Journal of Chemical
Physics 1970, 53, (3), 1126.
2. Kastner, J.; Pichler, T.; Kuzmany, H.; Curran, S.; Blau, W.; Weldon, D. N.;
Delamesiere, M.; Draper, S.; Zandbergen, H., Resonance Raman and Infrared-
Spectroscopy of Carbon Nanotubues. Chemical Physics Letters 1994, 221, (1-2),
53-58.
3. Pimenta, M. A.; Dresselhaus, G.; Dresselhaus, M. S.; Cancado, L. G.; Jorio, A.;
Saito, R., Studying disorder in graphite-based systems by Raman spectroscopy.
Physical Chemistry Chemical Physics 2007, 9, (11), 1276-1291.
4. Tuinstra, F.; Koenig, J. L., Characterization of Graphite Fiber Surfaces with
Raman Spectroscopy. Journal of Composite Materials 1970, 4, 492.
5. Cancado, L. G.; Takai, K.; Enoki, T.; Endo, M.; Kim, Y. A.; Mizusaki, H.; Jorio,
A.; Coelho, L. N.; Magalhaes-Paniago, R.; Pimenta, M. A., General equation for
the determination of the crystallite size La of nanographite by Raman
spectroscopy. Applied Physics Letters 2006, 88, (16), 163106.
6. Thomsen, C.; Reich, S., Double resonant Raman scattering in graphite. Physical
Review Letters 2000, 85, (24), 5214-5217.
7. Ferrari, A. C.; Meyer, J. C.; Scardaci, V.; Casiraghi, C.; Lazzeri, M.; Mauri, F.;
Piscanec, S.; Jiang, D.; Novoselov, K. S.; Roth, S.; Geim, A. K., Raman spectrum
of graphene and graphene layers. Physical Review Letters 2006, 97, (18), 187401.
8. Kim, U. J.; Furtado, C. A.; Liu, X. M.; Chen, G. G.; Eklund, P. C., Raman and IR
spectroscopy of chemically processed single-walled carbon nanotubes. Journal of
the American Chemical Society 2005, 127, (44), 15437-15445.
9. Hennrich, F.; Krupke, R.; Lebedkin, S.; Arnold, K.; Fischer, R.; Resasco, D. E.;
Kappes, M., Raman spectroscopy of individual single-walled carbon nanotubes
from various sources. Journal of Physical Chemistry B 2005, 109, (21), 10567-
10573.
108
10. Weisman, R. B.; Heymann, D.; Bachilo, S. M., Synthesis and characterization of
the "missing" oxide of C60: [5,6]-open C60O. Journal of the American Chemical
Society 2001, 123, (39), 9720-9721.
11. Banerjee, S.; Hemraj-Benny, T.; Balasubramanian, M.; Fischer, D. A.; Misewich,
J. A.; Wong, S. S., Ozonized single-walled carbon nanotubes investigated using
NEXAFS spectroscopy. Chemical Communications 2004, (7), 772-773.
109
Chapter 4
IR studies of SWNT ozonides and of products of their reactions with
different classes of compounds
110
4.1. Introduction
Ozonation of alkenes is known to proceed at very low temperatures with formation
of primary ozonides.1 Formed 1,2,3-trioxolanes are typically not stable above – 100 C.
1-4
Hull et al.2 determined that the majority of 1,2,3-trioxolanes studied rearranged to 1,2,4-
trioxolanes at temperatures above – 100 C. Kohlmiller and Andrews1 and Samuni and
Haas5 assigned the following IR active bands to primary ozonide of ethylene: 846 cm
-1
( sym O-O-O stretch), 927 cm-1
(C-O stretch) and 983 cm-1
(C-O stretch). Hull et al.2
found characteristic bands for primary ozonides of different alkenes to be in 850-1050
cm-1
region. Andrews and Kohlmiller6 compared strong IR bands of 1,2,3-trioxolanes of
propene, trans-2-butene, 2-methylpropene, tetramethylethylene and trans-
diisopropylethylene. The C-O stretches for all ozonides were found in the region 850-
1000 cm.-1
Contrary to small molecule alkenes, ozonides of fullerene (C60O3) 7 and of carbon
nanotubes (SWNT(O3)n) (this work, Chapter 2) were found to be significantly more
stable. Heymann et al.7 found that at 23 C ozonide C60O3 has a lifetime ca. 22 minutes in
toluene. Lifetimes of SWNT(O3)n, depending on tube type, were found in the range 0.5 –
2 minutes at 20 C (this work, Chapter 2). Thus, it is reasonable to expect characteristic
IR active stretches of primary ozonides of C60 and SWNT at room temperature. This
chapter will discuss the influence of oxidation with ozone/oxygen gaseous mixture on IR
spectra of carbon nanotubes (SWNTs). A brief summary of published data on ozonation
of SWNT is discussed in the introduction of Chapter 2. The majority of articles discussed
had IR spectra of final products with carboxylic, ester, quinone, and other functional
111
moieties on SWNT, and none provided any spectral examination of SWNT ozonides. To
date, IR verification of the existence of SWNT ozonides has not been published. Kinetics
determined in this chapter will be compared to those discussed for fluorescence and UV
techniques (Chapter 2).
4.2. Discussion and results
An extensive IR study has been accomplished here in search of peaks that could be
assigned to 1,2,3-trioxolanes or products of their decomposition, like epoxides and
oxidoannulenes.8 For the majority of spectra presented in this section, constant purging of
the IR chamber with nitrogen was done to decrease the intensity of water and carbon
dioxide peaks.
The SWNT film was formed on a BaF2 window by adding several drops of SWNT
slurry freshly dispersed in benzene. Dry SWNT film was ozonated for 30 seconds with a
stream of O3/O2 gaseous mixture (ca. 1.5 v/v % ozone) and the absorption IR spectrum
measured (Figure 1).
No peaks attributable to ozonides or epoxides could be found. Slow absorbance drift
over time was observed. Peaks for ozonides and epoxides were expected in the area
800 – 1200 cm-1
. Kamaras et al.9 and Hu et al.
10 attributed absorbance depletion at
frequencies below 1000 cm-1
(see Figure 1) to covalent modification of metallic SWNT.
112
Wavenumber (cm-1
)
1500300045006000
Ab
so
rba
nce
(a
.u.)
0.15
0.20
0.25Before O3
2 min
5.4 min
15.8 min
Figure 1. Absorbance IR of SWNT film before and after purging with O3/O2 gaseous
mixture (ca. 1.5 v/v % ozone) for 30 sec. Curves are: - before ozonation, - two
minutes after ozonation, – 5.4 min after O3, - 15.8 min after O3.
The experiment was repeated with fullerene C60, since its ozonides have been
reported in the literature.7, 8
The T1u vibrations of C60 at 1182 and 1429 cm-1
observed in
this work (Figure 2) are in good agreement with published data.11
Fullerene ozonide
(C60O) lifetime was reported ~ 22 min at 23 C 8 and deemed long enough to be
detected by IR. It was reasonable to expect to see asymmetric stretches imparted by the
1,2,3-trioxolane moiety (Figure 2).
113
Wavenumber (cm-1)
1500300045006000
Ab
so
rban
ce
(a
.u.)
0.04
0.08
0.12
before O3
2 m
5.4 m
2 m
4.8 m
after 1st O3
after 2nd
O3
Figure 2. Absorbance IR of ozonated C60 film on BaF2 window before and after purging
with O3/O2 gaseous mixture (ca. 1.5 v/v % ozone) for 20 sec. Curves are: - before
ozonation, - two minutes after 1st ozonation, – 5.4 min after 1
st O3, - 2 min after
2nd
O3, - 4.8 min after 2nd
O3. Sharp peaks are 1182 and 1429 cm.-1
A small background shift was observed at frequencies above 4500 cm-1
after each
ozonation; no peaks attributable to ozonides could be found.
An attempt was made to run IR of SWNT buckypaper without IR windows. Though
such paper had high absorption below 1500 cm-1
, its transmission increased greatly right
after ozonation. Spectrum acquisition was performed on buckypaper stretched across a
4 x 6 mm opening as shown in the pictures below.
114
Figure 3. Buckypaper stretched across a 4 x 6 mm opening for windowless IR
measurement. (A-B) SWNT film attached to adhesive tape, (C-D) IR holder with 4 x 6
openning, (E-F) film inside the holder.
Buckypaper was flushed with O3/O2 gaseous mixture (ca. 1.5 v/v % ozone) for
30 sec and spectra recorded (Figure 4).
Wavenumber (cm-1
)
150030004500
Abso
rba
nce
(a.u
.)
1.3
1.7
2.1
2.5
2.9
before
2 m
10.4 m
2 m
10.8 m
2 m
45.1 m
2 m
11.8 m
after 1st O
3
after 2nd
O3
after 3rd
O3
after 4th O
3
Figure 4. Absorbance IR of SWNT buckypaper after four 30 sec purging with O3/O2
gaseous mixture (ca. 1.5 v/v % ozone). Symbols denote times after the beginning of each
ozonation. (Windowless IR.)
No peaks below 1500 cm-1
attributable to ozonides could be found. Spectra were
fairly noisy at frequencies below 1050 cm-1
. Absorbance drift over time was observed as
115
in previous experiments. Two of the most prominent peaks in the spectra, near 1220 and
1570 cm-1
, are likely to be from asymmetric stretches of single (C-C-EWG) and double
(C=C-EWG) bonds on SWNT surface located in the vicinity of electron-withdrawing
groups (EWG). Drifts of peaks’ maxima towards higher frequencies with subsequent
ozonations are due to an increase of the number of electron withdrawing groups on the
SWNT surface.
Raman and IR spectra overlay shows the proximity of D and G bands and two major
peaks in the IR spectrum of ozonated SWNT:
Absorb
ance (
a.u
.) (
IR)
0.48
0.52
0.56
0.60IR after O3
D
G
G'
Wavenumber (cm-1
)
0 500 1000 1500 2000 2500
Counts
x10
5 (
Ram
an)
0.0
0.4
0.8
1.2
Raman before O3
ab
Figure 5. Overlay of absorbance IR spectrum of ozonated SWNT with Raman spectrum
of pristine SWNT.
The experiment was repeated with a longer ozonation period to bring the absorbance
at 1000 cm-1
to below 1.5 a.u. (Figure 6).
116
Wavenumber (cm-1
)
1000200030004000
Ab
so
rban
ce
(a
.u.)
0.75
1.50
2.25
3.00
before
6 m
18.3 m
6 m
46.1 m
6 m
50.3 m
6 m
29.4 m
after 1st O3
after 2nd
O3
after 3rd
O3
after 4th O3
Figure 6. Absorbance IR of SWNT buckypaper after four 5 min purging with O3/O2
gaseous mixture (ca. 1.5 v/v % ozone). Symbols denote times after the beginning of each
ozonation. (Windowless IR.)
The signal to noise ratio increased dramatically for area below 1500 cm-1
(compare
Figures 4 and 6), but no peaks attributable to ozonides could be found. Absorbance drift
was substantial after the first 5 min of purging with O3/O2 gaseous mixture (compare
curves and ).
117
Wavenumber (cm-1
)
1000200030004000
Absorb
ance (
a.u
.)
0.6
0.8
1.0
1.2
after 4th
O3
17
65
34
70
12
22
15
72
13
71
11
44
Figure 7. Absorbance IR of SWNT buckypaper after fourth 5 min purging with O3/O2
gaseous mixture (ca. 1.5 v/v % ozone).
Additional peaks found in SWNT spectra after multiple oxidations were 3470, 1765,
1371 and 1144 cm-1
(Figure 7). The peak at 1765 cm-1
could be attributed to 1,2-
diketones, anhydrides, unsaturated esters or a combination of thereof; 1371 cm-1
to
epoxides and 1144 cm-1
to diaryl ethers. Absorbance kinetics were measured and are
presented below. Found peaks’ maxima were close to those observed by Cai et al.12
118
10 20
Absorb
ance (
a.u
.)
1.4
1.5
1.6
1.7
1.8
1.9
Time (min)
10 20 30 40 50
1.2
1.3
1.4
1.5
1.6
1.7
10 20 30 40 50
1.0
1.1
1.2
1.3
1.4
1.5 Abs at 1315 cm-1
Abs at 1625 cm-1
Abs at 2250 cm-1
A
A B
A B
after 3rd
O3after 2nd
O3after 1st O3
Figure 8. Absorbance IR kinetics of SWNT buckypaper after the first three 5 min
purging with O3/O2 gaseous mixture (ca. 1.5 v/v % ozone). For comparison purposes,
absorbance and time axes have the same scale on all three graphs. Each symbol is marked
with a distinct wavenumber at which kinetics was monitored. Schematic diagram: A –
ozonide decay, B – structural rearrangements. Left: after 1st O3, Middle: after 2
nd O3,
Right: after 3rd
O3.
Absorbance changes within the first ten minutes were observed after the first three
five-minute ozonations. The absorbance change is believed to be associated with ozonide
decay. An upward movement, i.e. absorbance increase, could be caused by an increase of
electron density on SWNT after ozonide decay.
119
Lack of success with finding characteristic ozonide peaks prompted the use of a KBr
window in place of BaF2. Pictures of SWNT film on KBr are shown below.
Figure 9. SWNT film on KBr window (Fisher Scientific) is shown from different angles.
Solid state Raman was recorded on KBr window (shown in Figure 9) before and
after ozonation (Figure 10; ozonation was monitored by IR). The intensity of D band was
found unchanged after extensive oxidation with ozone. The 785 nm laser was used for
excitation.
Wavenumber (cm-1
)
500 1000 1500 2000 2500
Counts
x10
5
0.0
0.4
0.8
1.2
before O3
after O3
D
G
G'
1292 1591
Counts
x10
3
1.7
7.0
D
G
Figure 10. Raman spectra of SWNT film on KBr window before and after several rounds
of ozonation. Zoom-in shows D band did not change. G and G’ bands were bleached
substantially. The 785 nm laser was used for excitation.
Raman spectra shown in Figure 10 correspond to curves and in Figure 11.
120
Wavenumber (cm-1
)
200040006000
Absorb
ance (
a.u
.)
0.2
0.4
0.6
0.8
1.0
before
1 m
8.9 m
1.5 m
23.6 m
1.5 m
19.5 m
1.5 m
9.3 m
1.5 m
5.8 m
after 1st O3
after 6th O3
after 7th O3
after 8th O3
after 10th O3
Figure 11. Absorbance IR spectra of SWNT film on KBr window after ten rounds of
ozonation. Oxidations 1 through 5 were done with 1 mL of O3/O2 gaseous mixture (ca.
1.5 v/v % ozone). Ozonations 6 through 10 were performed by blowing O3/O2 gaseous
mixture (ca. 1.5 v/v % ozone) onto SWNT film for 30 sec. Some spectra are not shown to
avoid clutter. Curve is the spectrum of a pristine SWNT film.
Spectra on KBr were found to swing around a pivoting point near 5000 cm-1
.
Absorbance drift over time was larger than what was observed for BaF2 and windowless
IR experiments. Compare curves and in Figure 11. An initial population of low
frequency vibrations (1000 cm-1
) and a depopulation of electronic transitions (7000 cm-1
)
drifted over time to repopulate electronic transitions. Such a swing is believed to be
associated with the charge transfer from ozonated SWNT to KBr or vice versa.
Below are zoomed-in spectra demonstrating a swing motion of the spectra.
121
Wavenumber (cm-1
)
1500300045006000
Absorb
ance (
a.u
.)
0.45
0.60
1 m
15.6 m
1 m after 5th O3
after 4th O3
Absorb
ance (
a.u
.)
0.50
0.65
1 m
11.2 m
1 m after 3rd
O3
after 2nd
O3
A
B
Figure 12. Absorbance IR spectra of SWNT film on a KBr window between 2nd
and 5th
rounds of ozonation. Oxidations were done with 1 mL of O3/O2 gaseous mixture (ca. 1.5
v/v % ozone). Legends to symbols denote times after the beginning of the corresponding
ozonation. (A) Spectra after 2nd
, before 3rd
and after 3rd
ozonation, (B) spectra after 4th
,
before 5th
and after 5th
O3. See note to Figure 11 for more details.
A depopulation of NIR absorbance was observed for SWNT – SDS aqueous
suspensions right after ozonation (Chapter 2, Figure 8). A similar tendency was observed
for a solid SWNT film near the 7000 cm-1
region in Figure 12A. Ozonation resulted in a
122
depopulation of electronic transitions in NIR, with a slow recovery over time (Figure
12A, curves and ). Swing motion was accompanied by an overall bleaching (Figure
12B).
A fast scanning (2 scan averaging) on BaF2 window was performed to measure the
kinetics right after an ozone injection. The O3/O2 gaseous mixture (ca. 3 v/v % ozone)
was injected directly into IR chamber with the needle pointed towards SWNT film. The
chamber was under a constant nitrogen flush. The first point was acquired 10 sec after the
injection.
Time (min)
0 3 6 9 12
Absorb
ance a
t 2250 c
m-1
(a.u
.)
0.64
0.65
A B
Figure 13. Absorbance IR kinetics of SWNT film on BaF2 window at 2250 cm-1
.
Oxidation was done with 5 mL of O3/O2 gaseous mixture (ca. 1.5 v/v of ozone). A – fast
decay; B – slow rearrangement.
A five parameter two-exponential decay formula was used for the regression. A
decay rate b was calculated to be 2.6 min-1
, which corresponds to ~ 0.4 min. Kinetics
123
measurement were repeated on a KBr window and a comparable rate was obtained
(Figure 4).
Time (min)
0 5 10 15 20
Absorb
ance a
t 2250 c
m-1
(a.u
.)
0.76
0.77
0.78 after 1st O
3
Figure 14. Absorbance IR kinetics of SWNT film on KBr window at 2250 cm-1
.
Oxidation done with 3 mL of O3/O2 gaseous mixture (ca. 1.5 v/v of ozone).
The decay rate obtained from regression was 2.1 min-1
, which corresponds to
~ 0.48 min.
The following conclusions can be made for IR spectra of ozonated SWNT. No
characteristic peaks could be found in the area typical for 1,2,3-trioxolanes or peroxides.
Commonly throughout all experiments, absorbance drift was observed. In the case with
KBr window used as a support, the change resembled a swing with a pivoting point in the
range 4000-5000 cm-1
. In the case with BaF2 absorbance decreased over time at
2250 cm.-1
In the case with windowless IR, the change led to an increase of absorbance in
the range 600 – 4000 cm-1
during the first ten minutes. Regression yielded decay rates
2.1 and 2.6 min-1
for KBr and BaF2 experiments. Ozonide peaks could not be found for
124
C60O3 either. There is no clear understanding why ozonides do not have characteristic
C-O bands in IR in the area 850 – 1050 cm-1
. The absorbance drift is thought to be
associated with ozonide decay. Interestingly, for films on KBr and BaF2 windows
absorbance decreased over time at 2250 cm-1
; for windowless IR on buckypaper the
absorbance increased during the first ten minutes. IR window may be involved in a
charge transfer to or from ozonated SWNT.
Influence of amines and solvents on ozonated SWNT
Ozonation of SWNT film with O3/O2 gaseous mixture to a moderate degree yields a
characteristic wavy curve (Figure 15, curve ). This curve was used as a reference for
reactions of ozonated SWNT with amines.
Wavenumber (cm-1
)
1500300045006000
Ab
so
rba
nce
(a
.u.)
0.1
0.2
0.3
0.4
before
after 1st O
3
after 2nd
O3
after 3rd
O3
after 4th
O3
f
10001500
a
c
b
b'
ed
d'
Figure 15. Absorbance IR spectra of SWNT film on BaF2 window ozonated four times.
Peak maxima values (cm-1
): a = 1760, b = 1550, b' = 1559, c = 1369, d = 1190, d' = 1206,
e = 1163, f = 3495.
125
A comparison was made between dry ozonated SWNT (Figure 15) and SWNT
ozonated in suspension. Three solvents used for this comparison were methanol, ethanol
and trifluoroethanol. Dilute suspensions of SWNT in solvents of interest were bath
sonicated for 1 hour, O3/O2 gaseous mixture (ca. 3 v/v % ozone) was bubbled through
suspensions for 1 min at r. t. and IR spectra recorded 1 hour after ozonation. Spectra were
overlaid for comparison and shown below (Figure 16).
Wavenumber (cm-1
)
1500300045006000
Absorb
ance (
a.u
.)
MeOH
EtOH
CF3CH
2OH
Figure 16. A solvent shielding effect on SWNT ozonation degree. Bubbling O3/O2
gaseous mixture (ca. 3 v/v % ozone) for 1 min at r. t. through SWNT suspension in
methanol (), ethanol () and trifluoroethanol () yielded spectra identical to those
obtained after dry ozonation. Spectra were recorded 1 hour after ozonation as a dry
SWNT film on BaF2 window.
All three solvents were found unreactive with ozonated SWNT. No IR active peaks
attributable to solvents’ residues were seen on spectra. Oxidation in ethanol resulted in
the lowest ozonation degree, while trifluoroethanol gave the highest oxidation degree. It
was concluded that each solvent possesses a unique shielding ability, preventing ozone
126
from reacting with SWNT. An alternative explanation would be a different degree of
ozone solubilization in each solvent. Chen et al.13
reported that perfluoroethers have a
good solubility of ozone.
Below is an example of a reagent, acetic acid, which does not react with ozonated
SWNT.
Wavenumber (cm-1
)
1500300045006000
Absorb
ance (
a.u
.)
0.00
0.15
0.30
0.45
0.60
before
after
reference
AcOH
Figure 17. Addition of acetic acid to ozonated SWNT film on BaF2 window did not
affect SWNT curve. Curves are: – SWNT film before O3, – reference, mixed
SWNT film with acetic acid for 1 min then dried with a heat gun, –SWNT film purged
with O3/O2 gaseous mixture (ca. 1.5 v/v % ozone), – ATR converted spectrum of an
authentic sample of acetic acid for comparison. Curves and were measured on the
same film. All curves except are absorbance IR spectra.
A few drops of n-butyl amine were added to SWNT plate treated first with ozone and
then with acetic acid. The BaF2 plate was dried with a heat gun and an IR spectrum
obtained (Figure 18). The arrows schematically show repopulation of electronic
transitions at 7000 cm-1
and depopulation of vibrational ones at 1000 cm-1
. A spectrum of
127
an authentic sample of n-butyl amine (curve ) is provided for comparison. It is believed
that the first step in the reaction between amine and SWNT is an electron transfer from
electron rich amine to ozonated SWNT. It is not clear why ozonated SWNT plays a role
of an oxidizer. As demonstrated by XPS (Chapter 7), such a reaction does not take place
between pristine SWNT and amines at r. t.
Wavenumber (cm-1
)
1500300045006000
Absorb
an
ce (
a.u
.)
0.00
0.15
0.30
0.45
0.60
before
after O3 & AcOH
added n-BuNH2
n-BuNH2
15003000
c
de
a b
c
ab
de
Figure 18. A reaction of ozonated SWNT film with n-butyl amine. A repopulation of
electronic transitions and a depopulation of vibrational ones are emphasized with arrows.
Curves are: – SWNT film before O3, – an ozonated SWNT film treated with acetic
acid, –added n-BuNH2 to SWNT treated with O3 and AcOH, reacted for 1 min and
dried with a heat gun, – ATR converted spectrum of an authentic sample of n-butyl
amine for comparison. Curves , and were measured on the same film. All curves
except are absorbance IR spectra.
128
SWNT films, ozonated for 30 seconds, were treated with n-butylamine forty minutes
and two days after ozonation. In both cases amines were shown to react with SWNT
(Figure 19). The sample with a forty minute delay was dried on a vacuum pump before
IR measurements. Sample with a two-day delay was dried with a heat gun.
Wavelength (cm-1
)
1500300045006000
Ab
so
rba
nce
(a
.u.)
0.00
0.15
0.30
after O3
2 day + amine
40 min + amine
n-BuNH2
Figure 19. A reaction of ozonated SWNT film with n-butyl amine 40 min and 2 days
after ozonation. Repopulation of electronic transitions and depopulation of vibrational
ones is emphasized with arrows. Curves are: – SWNT film after O3, – ozonated
SWNT film treated with n-butylamine 2 days after O3, –added n-BuNH2 to SWNT
forty minutes after O3, – an ATR converted spectrum of an authentic sample of
n-butylamine is overlaid for comparison. Curve was scaled down by a factor of 0.87 to
compensate for the higher load of SWNT on IR window. Curves and were
measured on the same film. All curves except are absorbance IR spectra.
129
Secondary and tertiary amines were also tested. The resultung spectra are shown
below. Triethyl amine was shown to react with SWNT, which further supports the
likelihood of an electron transfer between ozonated SWNT and amine.
Wavenumber (cm-1
)
1500300045006000
Ab
so
rba
nce
(a
.u.)
0.00
0.15
0.30
0.45
0.60
before
after
reference
Et3N
15003000
de
abc
a
de
Figure 20. A reaction of ozonated SWNT film with triethyl amine. Curves are: –
SWNT film before O3, – ozonated SWNT film treated with amine, – added amine
to pristine SWNT, reacted for 1 min and dried with a heat gun, – an ATR converted
spectrum of an authentic sample of triethyl amine is overlaid for comparison. Curves
and were measured on the same film. All curves except are absorbance IR spectra.
Reaction between triethyl amine and SWNT does not seem to be as effective as
between primary amines and ozonated SWNT. The product is thought to be an
ammonium salt. The first step is an electron transfer from amino group to ozonated
SWNT; the second step is thought to be a covalent attachment of radical Et3N+
to the
surface of SWNT. Similar types of reactions had been studied by Isobe et al. on a
fullerene C60 substrate.14
The author proposed for a reaction of fullerene C60 with
dialkylamine a formation of a long lived pair of aminium radical R2NH+ and C60
–
130
radical anion. Such pair existed only in the absence of oxygen. When the mixture was
exposed to the air, two radicals reacted immediately. The author observed a trace of
hydrogen peroxide, which was formed during oxygen reduction by an aminium radical.
Wavenumber (cm-1
)
1500300045006000
Absorb
ance (
a.u
.)
0.00
0.15
0.30
0.45
0.60
before
after
reference
Et2NH
Figure 21. A reaction of ozonated SWNT film with diethyl amine. Curves are: –
SWNT film before O3, – ozonated SWNT film treated with amine, – reference,
added diethyl amine to SWNT, reacted for 1 min and dried with a heat gun, – ATR
converted spectrum of an authentic sample of diethyl amine is overlaid for comparison.
Curves and were measured on the same film. All curves except are absorbance
IR spectra.
Reaction of ozonated SWNT with diethyl amine was essentially the same as with
n-BuNH2 and Et3N.
131
Reaction between 2-methoxyethyl amine and ozonated SWNT is shown below.
Wavenumber (cm-1
)
1500300045006000
Absorb
ance (
a.u
.)
0.00
0.15
0.30
0.45
0.60
before
after
reference
MeOCH2CH
2NH
2
ba
c
15003000
ba
c
de
f
de
f
Figure 22. A reaction of ozonated SWNT film with 2-methoxyethyl amine. Curves are:
– SWNT film before O3, – ozonated SWNT film treated with amine, – reference,
added amine to SWNT, reacted for 1 min and dried with a heat gun, – ATR converted
spectrum of an authentic sample of amine is overlaid for comparison. Curves and
were measured on the same film. All curves except are absorbance IR spectra.
Enrichment of amines and amides on the surface of SWNT (IR monitoring)
An enrichment of n-butyl amine on the surface of SWNT was successfully realized
by cycling ozonation with amine addition. The experiment was performed on the same
BaF2 plate. SWNT film ozonation was alternated with amine addition. Butyl amine, used
in the experiment, was chosen for its low boiling point. A heat gun was used to remove
traces of unreacted amine from the surface of SWNT. Distinct amide, C(=O)O, OH and
NH stretches were observed in the IR spectra (Figure 23).
132
Wavenumber (cm-1
)
1500300045006000
Ab
so
rba
nce
(a
.u.)
0.1
0.2
0.3
0.4
before
after 1st O3
added amine (1st cycle)
after 2nd O3
added amine (2nd cycle)
after 3rd O3
added amine (3rd cycle)
after 4th O3
Figure 23. Cycled reaction of ozonated SWNT film with n-butyl amine. Curves are:
– SWNT film before O3, – ozonated SWNT film, – ozonated SWNT film treated
with amine (1st cycle), – after second ozonation, - added amine (2
nd cycle),
– after third ozonation, - added amine (3rd
cycle), - after forth ozonation. Water
spikes in the area 1600 cm-1
were removed manually to improve clarity. All curves are
absorbance IR spectra.
SWNTs were debundled in ethanol by bath sonication, and then added as a slurry to
the surface of BaF2 window. The spectrum of pristine SWNT was measured and the film
was subjected to a stream of O3/O2 gaseous mixture (ca. 1.5 v/v % ozone) for 1 min at r.
t. The IR spectrum of ozonated SWNT was recorded and several microliters of n-butyl
amine added to the surface of BaF2. The plate was kept under the cover for 1 min and
then dried on vacuum pump for 3 min. Ozonation, IR measurements and reactions with
amine were repeated for three more cycles (Figure 23). A zoom-in of the region 4000-
700 cm-1
is shown in Figure 24.
133
Wavenumber (cm-1
)15003000
Absorb
ance (
a.u
.)
0.10
0.15
0.20
0.25
ab
c
d
e
f g
h
i
j
Figure 24. Cycled reaction of ozonated SWNT film with n-butyl amine. Curves are:
– SWNT film before O3, – ozonated SWNT film, – ozonated SWNT film treated
with amine (1st cycle), – after second ozonation, - added amine (2
nd cycle),
– after third ozonation, - added amine (3rd
cycle), - after forth ozonation. Water
spikes in the area 1600 cm-1
were removed manually to improve clarity. All curves are
absorbance IR spectra.
Peak data for Figure 24 are summarized in Table 1 below.
Table 1. IR peaks assignment for amine/amide enriched SWNT film
Abbreviation Wavenumber Peak assignment
a 3300 N-H stretch; amide or amine;
b 2982- 2829 C-H stretches (CH2, CH3)
c 1760 1,2-diketones, anhydrides, unsaturated esters
d 1672 amide C=O stretch
e 1573 C=C and N-H bending from amides/amines
f 1451 CH2
g 1375 C-N stretch (amide); CH2 scissoring; epoxides
h 1196 CH2 scissoring
i 1118 C-N stretch in amine
j 683 N –H wagging
134
The observed increase in peak intensities is thought to be directly linked to the
amount of amine or amide attached to the surface of SWNT.
4.3. Conclusions
IR monitoring of dry ozonated SWNT film at 2250 cm-1
revealed exponential
absorbance change over time. The rate of the change was found to be close to the rates
obtained with fluorescence and NIR absorbance techniques for ozonated SWNT. For
experiments performed on KBr window, ozonated SWNT demonstrated an initial
bleaching of absorbance in a wide range, above 5000 cm-1
. The change was repeatable
with each injection. Bleaching at frequencies greater than 5000 cm-1
was accompanied by
an absorbance increase at frequencies lower than 4000 cm-1
.
Ozonated SWNT films were shown to react with amines of different structure
(primary, secondary and tertiary). Reaction with amines was independent of time period
after ozonation. SWNT ozonated two days prior to addition of amine were shown to
undergo the same reaction as the ones that were ozonated right before amine addition.
An enrichment of amines/amides on the surface of SWNT was successfully realized
by cycling ozonation, addition of neat amine and vacuum drying.
4.4. Experimental Part
Ozonation procedures
Ozone was generated by passing oxygen (industrial grade, Matheson Tri-Gas) with a
flow rate 1/16 L/min at r.t. through a high frequency corona discharge ozonator
135
(GE60FM, Yanco Industries Ltd, www.ozoneservices.com) set to the maximum output
(power level 10). The ozonator was idled before use for at least 7 min to reach the
maximum output as instructed by the manufacturer. Two different ozonation procedures
used in this chapter are described below.
Continuous ozonation of SWNT suspensions in a solvent of interest. A gaseous
mixture of O3/O2 (ca. 3 v/v % ozone) was bubbled through SWNT suspension in a test
tube for the desired period of time.
Gaseous ozone/oxygen injection with a syringe. Ozone was collected in a plastic
syringe for 1 min (3 mL volume) or 2 min (5-10 mL volume), capped with a plunger and
the desired volume released by blowing on top of SWNT film.
Continuous ozonation of dry SWNT films. The oxygen flow rate was kept at 1/16
L/min. Unless otherwise noted, a SWNT film was ozonated for 1 min at r. t. in a small
chamber by holding 1/16” ID Teflon tubing above the dry SWNT film.
Routine IR spectra measurements
Bath sonicated SWNTs (HipCo Lab, batch 162.8 or 161.1, Rice University) in
benzene were used for IR measurements. The Fourier Transform Infrared Spectrometer
(JASCO FT/IR-660 Plus) chamber was flushed with nitrogen unless otherwise noted. All
measurements were performed on dry films at or near r. t.; no special temperature
monitoring was made inside the chamber during IR measurements. Experiments had
either attenuated total reflectance (ATR) or absorbance setup which is noted on each
figure; KBr and BaF2 windows were used for absorbance IR measurements.
Some of the spectra of carbon nanotube films, or buckypapers, were recorded
without IR windows. A special film holder with 3 x 5 mm rectangle opening was
136
constructed to hold the buckypaper. Films were prepared by a dropwise addition of a
concentrated SWNT suspension in benzene onto the surface of a filter paper followed by
careful SWNT film removal. Paper twisting and turning was required to visually detach
nanotube film from the filter paper before film removal. Extreme care should be observed
to ensure nanotube film integrity.
Some spectra have a carbon dioxide peak present right after ozonation. This is
caused by an introduction of air into the chamber during IR holder handling. Over time, a
constant nitrogen flush removed carbon dioxide from the chamber. Nearly all FT/IR
spectra were acquired with constant nitrogen gas purge to reduce intensity of carbon
dioxide and water peaks.
IR kinetic measurements
The spectrometer was purged with nitrogen before and during measurements. For
kinetic studies 3 mL of O3/O2 gaseous mixture (ca. 3 v/v % ozone) was injected directly
into the IR chamber with a needle pointed towards SWNT film on BaF2 or KBr window.
137
4.5. References and Notes
1. Kohlmiller, C. K.; Andrews, L., Infrared-Spectrum of the Primary Ozonide of
Ethylene in Solid Xenon. Journal of the American Chemical Society 1981, 103,
(10), 2578-2583.
2. Hull, L. A.; Heicklen, J.; Hisatsun.Ic, Low-Temperature Infrared Studies of
Simple Alkene-Ozone Reactions. Journal of the American Chemical Society
1972, 94, (14), 4856-4864.
3. Hisatsune, I. C.; Kolopajlo, L. H.; Heicklen, J., Low-Temperature IR Studies of
Some Chloroethylene-Ozone Reactions. Journal of the American Chemical
Society 1977, 99, (11), 3704-3708.
4. Mile, B.; Morris, G. W.; Alcock, W. G., Infrared-Spectra and Kinetics of
Decomposition of Primary Ozonides in the Liquid-Phase at Low-Temperatures.
Journal of the Chemical Society-Perkin Transactions 2 1979, (12), 1644-1652.
5. Samuni, U.; Haas, Y., An ab-initio study of the normal modes of the primary and
secondary ozonides of ethylene. Spectrochimica Acta Part a-Molecular and
Biomolecular Spectroscopy 1996, 52, (11), 1479-1492.
6. Andrews, L.; Kohlmiller, C. K., Infrared-Spectra and Photochemistry of the
Primary and Secondary Ozonides of Propene, Trans-2-Butene, and
Methylpropene in Solid Argon. Journal of Physical Chemistry 1982, 86, (23),
4548-4557.
7. Heymann, D.; Bachilo, S. M.; Weisman, R. B.; Cataldo, F.; Fokkens, R. H.;
Nibbering, N. M. M.; Vis, R. D.; Chibante, L. P. F., C60O3, a fullerene ozonide:
Synthesis end dissociation to C60O and O-2. Journal of the American Chemical
Society 2000, 122, (46), 11473-11479.
8. Weisman, R. B.; Heymann, D.; Bachilo, S. M., Synthesis and characterization of
the "missing" oxide of C-60: [5,6]-open C60O. Journal of the American Chemical
Society 2001, 123, (39), 9720-9721.
9. Kamaras, K.; Itkis, M. E.; Hu, H.; Zhao, B.; Haddon, R. C., Covalent bond
formation to a carbon nanotube metal. Science 2003, 301, (5639), 1501-1501.
10. Hu, H.; Zhao, B.; Hamon, M. A.; Kamaras, K.; Itkis, M. E.; Haddon, R. C.,
Sidewall functionalization of single-walled carbon nanotubes by addition of
dichlorocarbene. Journal of the American Chemical Society 2003, 125, (48),
14893-14900.
138
11. Chase, B.; Herron, N.; Holler, E., Vibrational Spectroscopy of C-60 and C-70
Temperature-Dependent Studies. Journal of Physical Chemistry 1992, 96, (11),
4262-4266.
12. Cai, L. T.; Bahr, J. L.; Yao, Y. X.; Tour, J. M., Ozonation of single-walled carbon
nanotubes and their assemblies on rigid self-assembled monolayers. Chemistry of
Materials 2002, 14, (10), 4235-4241.
13. Chen, Z. Y.; Ziegler, K. J.; Shaver, J.; Hauge, R. H.; Smalley, R. E., Cutting of
single-walled carbon nanotubes by ozonolysis. Journal of Physical Chemistry B
2006, 110, (24), 11624-11627.
14. Isobe, H.; Tanaka, T.; Nakanishi, W.; Lemiegre, L.; Nakamura, E., Regioselective
oxygenative tetraamination of 60 fullerene. Fullerene-mediated reduction of
molecular oxygen by amine via ground state single electron transfer in dimethyl
sulfoxide. Journal of Organic Chemistry 2005, 70, (12), 4826-4832.
139
Chapter 5
Reaction of ozonated SWNT with electron rich nucleophiles (amines,
thiols and other)
140
5.1. Introduction
Reactions of ozonated SWNT with electron rich nucleophiles like amines and thiols
have not been discussed in the scientific literature yet. This chapter will expand on the
subject of reacting ozonated SWNT and C60 with amines and thiols. Studies will be
supported by SWNT fluorescence, UV absorbance, XPS, NMR and visual color changes.
The greatest advantage of this methodology is the speed of the reaction. Reaction of
amines with ozonated C60 was on the order of subseconds for a diluted fullerene solution
in toluene. Reaction of amines with SWNT was found to be slower, with a rate ca.
2 min-1
for a diluted SWNT suspension in ethanol. The presented experiments were
conducted in different solvents (water, ethanol and toluene), at r. t. and required only
minutes for completion. Particularly, an experiment with N,N,N’,N’- tetramethyl–p-
phenylenediamine (TMPD) was shown to reach completion within ca. 10 min after
mixing of ozonated SWNT and TMPD. (This strategy has been successfully extended
onto a covalent attachment of amino acids to SWNT sidewall and is discussed in chapters
6 and 7. For IR monitoring of reactions between ozonated SWNT and amines see Chapter
4.)
Closely related work performed on fullerene C60 with four peroxide moieties
(–OOBut), covalently attached to it, has been studied and published very recently.
1
Particularly, primary and secondary amines were shown to get covalently attached to the
fullerene framework with peroxides still intact. Reactions were reported to proceed fast.
Peroxides significantly decreased the activation energy for reaction between C60(OOBu-
t)4(O)1 and amines. While the reader may find it unique that amines were not oxidized by
141
peroxy-t-butyl moieties on C60, researchers who conducted this study were able to obtain
X-ray structures of some of the compounds.1
Reactions of amines with underivatized and derivatized fullerenes have been studied
and reported in peer reviewed journals for more than a decade.1-9
A typical reaction of
C60 with amine requires very long times (days to weeks at r. t.) or heating.8
Methodologies developed in this work and the one published by Hu et al.1 have a great
advantage of being fast.
It should be noted that the lifetime of SWNT ozonides (1,2,3-trioxolanes) was
determined in this work to be in the range 20 – 200 sec at r. t. (see Chapter 2). In view of
this result, the term “ozonated SWNT” can mean both SWNT with ozonides still present
on its surface and SWNT with all ozonides decomposed. Several experiments in this
work were specifically designed to demonstrate that amines can react with ozonated
SWNT in the absence of ozonides (Chapter 4, Figure 19).
The only readily available example of reaction of ozonated SWNT with electron rich
nucleophile is the reaction of SWNTO3 with dimethyl sulfide (DMS) at -78 C.10
This
easily oxidizable reagent was added to SWNT ozonated for an extended period of time.
The author was assuming an analogy between the reduction of secondary ozonides of
small molecules and the reduction of primary ozonides on the surface of SWNT. The
goal was to preserve ketone and aldehyde groups on the surface of SWNT, while
converting dimethyl sulfide into dimethyl sulfoxide. Ozonide decomposition rates
obtained in this work (Chapter 2) clearly indicated that at temperature – 78 C, 1,2,3-
trioxolanes should be very stable. Also, formation of a secondary ozonide in place of a
142
primary one is not likely for rigid SWNT structure. Article did not mention any attempt
to covalently attach dimethyl sulfide to the surface of SWNT. It is likely that DMS
reacted not only with 1,2,3-trioxolanes, but also with SWNT sidewall. Provided XPS
spectra did not cover S 2s or S 2p regions, thus it is not clear if any sulfur was present on
the surface of SWNT. While conversion of organic sulfides and phosphines to
corresponding oxides in reactions with 1,2,4-trioxolanes is known,11-13
it is not clear what
will be the products of reduction of 1,2,3-trioxolanes formed on rigid SWNT framework.
Formation of 1,2,4-trioxolanes by the Criegee mechanism,12
as it happens with small
organic molecules, seems unlikely.
With regard to SWNT thiolation, Lim et al.14
reported the conversion of terminal
carboxylic moieties of 200 nm long SWNT to –CH2SH by a series of reduction,
chlorination (SOCl2) and thiolation. Nakamura et al.15
reported a sidewall attachment of
sulfur-containing functionalities by irradiation of aliphatic disulfides in presence of
SWNT. The reaction was conducted with a low pressure mercury lamp (60W; > 200
nm) for 4 hours at r. t. Liu et al.16
reported an attachment of a thiol through an amide
linkage. Another example of an indirect attachment of amines includes the work of Peng
et al.17
who reported SWNT sidewall functionalization through the reaction with succinic
or glutaric acid acyl peroxides in o-dichlorobenzene at 80-90 C. Subsequent treatment
with thionyl chloride and amines yielded amide functionalized carbon nanotubes.
The most practical functionalization of SWNT with amines can be accomplished
through the reaction of fluorinated SWNT (SWNTF).18
Interestingly, the IR spectrum of
SWNTF reported by Stevens et al.18
is nearly identical to those demonstrated in this work
143
for ozonated SWNT (e.g. Figure 4 in Chapter 4). A typical procedure requires heating of
fluoronanotubes in excess of amine (used as a solvent) to 100-170 C for 4 hours.
Holzinger et al.19
reported preparation of alkoxycarbonylaziridino-SWNTs by
reaction of nanotubes with azidoformates at 160 C in 1,1,2,2-tetrachloroethane.
Derivatized tubes had a much better solubility in DMSO than pristine ones, thus allowing
for their separation.
A number of derivatized fullerenes with indirectly attached amines, i.e. nitrogens are
not attached to C60 framework, have been reported.20, 21
This topic is of less interest and
will not be discussed in this chapter.
144
5.2. Results and discussion
The degree of fluorescence recovery of ozonated SWNT with and without addition
of electron rich nucleophiles (amines, thiols and other)
Side wall functionalization of SWNT with ozonides has been demonstrated in this
work by various techniques: oxygen evolution, fluorescence bleaching, UV absorbance
bleaching, NIR absorbance bleaching, Raman G-band bleaching, IR absorbance increase
over a wide range of vibrational frequencies, typically below 4000 cm-1
, and an increased
percent of oxygen atoms on SWNT surface as determined by XPS.
0.1
0.3
0.5
0.7
0.9
1.1
1.3
1.5
1.7
1.9
2.1
2.3
2.5
10001100
1200
1300
1400
1500
1600
02
46
8
Flu
ore
scence Inte
nsity (
nW
/nm
)
Wavelength (nm)
Time (min)
0.20.40.6
0.81.01.21.3
1.51.7
1.9
Figure 1. Formation of SWNT ozonides and their reaction with 2-methoxyethylamine
monitored by NIR fluorescence at r. t. Ozone addition at ca. 30 sec led to a sharp decline
in fluorescence intensity. An addition of amine at ca. 1.5 min resulted in a fast recovery.
The 660 nm laser was used for excitation.
A full fluorescence recovery was observed after an addition of 3-methoxyethylamine
to ozonated SWNT. 3D kinetics at different wavelengths is shown in Figure 1. As
145
demonstrated later in this chapter, intercalation of amines into SDS quasi-liquid shell can
increase SWNT fluorescence intensity, though not much. Observed fluorescence increase
to the level above the preozonated one is thought to be due to: a) reaction of ozonated
SWNT with amines and b) amine intercalation. SWNT fluorescence is thought to be
strongly dependent on electron density of a -cloud formed by a conjugated system of
double bonds. An initial injection of 2 mL of O3/O2 gaseous mixture (ca. 1.5 v/v %
ozone) resulted in substantial fluorescence intensity depletion. A time period of 1 min
was allotted to ensure there is no more free ozone in aq. suspension. Amine (6 uL,
9 umol, 13 v/v % in water) addition to 1 mL of SWNT – SDS suspension led to a fast
fluorescence intensity rise.
Below is a graph demonstrating the influence of SWNT sidewall modification with
1,2,3-trioxolanes and a subsequent reaction with 2-methoxyethylamine on emmax
locations (Figure 2).
Several nanometers blue shift of the largest peak, near 953 nm, is likely due to tube
- cloud depletion (Figure 2B). Addition of amine red shifted emmax
of tube (8,3) near
953 nm. It is not clear if the red shift is caused by amine covalent attachment or by an
SDS shell intercalation or both. As shown later in this chapter, at lower ozone loads there
was no blue shift observed for ozonated SWNT. Thus small emmax
blue shifts were
observed only at extreme conditions, such as a high load of ozone. XPS and IR studies
confirmed that no chemical modification occurred upon mixing pristine SWNT with
amine at normal pressure and room temperature (see Appendix C for XPS spectra; see
curve in Figure 21 in Chapter 4 for an IR example). Thus, an intercalation of SDS
146
shell with amine molecules brought the final fluorescence intensity to a level higher than
that in the initial SWNT sample.
Wavelength (nm)
950 1050 1150 1250 1350
No
rma
lize
d F
luo
resce
nce
0.0
0.2
0.4
0.6
0.8
1.0
Flu
ore
sce
nce
In
ten
sity (
nW
/nm
)
0.0
0.5
1.0
1.5
2.0
2.5
A
B
951 954 957 960
0.8
0.9
1.0
SWNT(O3)n + amine
pristine SWNT
SWNT(O3)n
Figure 2. Influence of SWNT modification on em
max location. (A) Fluorescence spectra
as measured, (B) normalized fluorescence and zoom-in for tube (8,3) with
emmax
= 954 nm. Curves are: – pristine SWNT suspended in 1 wt. % aqueous SDS,
– SWNT after bubbling 2 mL of O3/O2 gaseous mixture (ca. 3 v/v % of ozone), – a
spectrum recorded 10 min after addition of 2-methoxyethylamine to ozonated SWNT.
The 660 nm laser was used for excitation.
147
Generally, SWNT fluorescence recovery after ozonation has a much slower rate than
in the case with electron rich nucleophiles added (Figure 3).
0.1
0.3
0.5
0.7
0.9
1.1
1.3
1.5
1.7
1.9
2.1
10001100
1200
1300
1400
1500
1600
05
10
15
20
Flu
ore
scence Inte
nsity (
nW
/nm
)
Wavelength (nm)
Time (min)
0.20.40.6
0.81.01.21.3
1.51.7
1.9
Figure 3. Formation and decomposition kinetics of SWNT ozonides at r. t. Ozone
addition at ca. 2 min led to a sharp decline in fluorescence intensity followed by a partial
recovery over time. The 660 nm laser was used for excitation. A small amount of ozone
was used for an oxidation (0.17 mL O3/O2 gaseous mixture, ca. 3 v/v % ozone).
Fluorescence of tubes emitting at shorter wavelengths was quenched less and
recovered faster when compared to tubes at longer wavelengths. There was a much
slower recovery of ozonated tubes when amine was not added (compare to Figure 1).
148
Reaction of ozonated SWNT with N,N,N’,N’-tetramethyl-p-phenylenediamine.
Ozonated SWNTs were reacted with N,N,N’,N’- tetramethyl-p-phenylenediamine
(TMPD), also known as Wurster reagent, to investigate the radical nature of the first step
of an electron transfer from amine to oxidized SWNT. Reaction between amines and
ozonated SWNT was demonstrated to proceed even two days after ozonation (Chapter 4),
indicating that oxidation with ozone converted SWNT into some form of a mild oxidizer.
TMPD (5.1 umol) was added to a very dilute suspension of SWNT in ethanol right after,
10, 20, 40 and 60 min after ozonation, reacted for 15 min and UV-Vis spectra obtained
(Figure 4).
Wavelength (nm)
400 500 600
Absorb
an
ce (
a.u
.)
0.5
2.0
3.5
565 615
2.5
3.5
no O3
0 min
10 m
20 m
40 m
60 m
Figure 4. UV-Visible spectra of reaction of N,N,N’,N’- tetramethyl-p-phenylenediamine
with ozonated SWNT. The same amount of Wurster reagent was added to all samples.
Additions were made right after, 10, 20, 40 and 60 min after ozonation. Non-ozonated
SWNT in ethanol, containing Wurster reagent, served as a reference sample (curve ).
UV spectra were recorded approximately 15 min after the addition of Wurster reagent.
149
All ozonated suspensions gained a deep purple color regardless of when Wurster
reagent was added. Reaction mixtures were collected after kinetics measurements and
photographed (Figure 5).
a b c d e f
Figure 5. Reaction of N,N,N’,N’- tetramethyl-p-phenylenediamine with ozonated SWNT.
Samples b-f gained a deep purple color due to a production of TMPD+ radicals. The
same amount of Wurster reagent was added to all samples. Black suspended flakes seen
on all pictures are SWNT bundles. (a) A reference sample, an ethanolic mixture of
pristine SWNT and TMPD, (b) TMPD added right after ozonation, (c) TMPD added 10
min after O3, (d) TMPD added 20 min after O3, (e) TMPD added 40 min after O3, (f)
TMPD added 60 min after O3.
Black flakes seen in Figure 5 are SWNT bundles. All samples were prepared from a
single stock suspension of SWNT in ethanol. Analogously, Wurster reagent was prepared
as a stock solution and aliquots were drawn for each experiment. SWNT flakes in ethanol
tended to precipitate over time and care was taken to prevent disturbance of samples
during spectra acquisition. As expected, addition of Wurster reagent right after and 10
min after ozonation gave higher degrees of its conversion into radical species TMPD+.
Addition of reagent 20, 40 and 60 min after ozonation gave essentially the same level of
consumption of TMPD (see Figure 4).
Absorbance change was monitored at 565 nm, near the local maximum (Figure 6).
Kinetics acquisition started approximately 10 sec after addition of the reagent. The delay
was used for mixing TMPD with SWNT flakes. TMPD+ production kinetics for samples
150
with reagent added right after, 40 and 60 min after ozonation were overlaid for
comparison and shown in Figure 6.
Time (min)
0 2 4 6 8 10 12 14 16
Absorb
an
ce (
a.u
.)
1
2
30 min
40 m
60 m
Figure 6. Absorbance rise at 565 nm with TMPD+ radical formation by oxidized SWNT.
Wurster reagent was added to ozonated SWNT right after, 40 and 60 min after SWNT
ozonation.
All three curves (, and in Figure 6) were fitted with 3 – parameter
exponential rise formula (F1) with excellent r2 values.
)1(0
bteayy (F1)
Table 1. Regression parameters for TMPD+ radical formation by oxidized SWNT.
Curve y0 a b, min-1
r2
0 min () 0.6793 2.9385 0.5077 0.9998
40 min () 0.3692 3.1732 0.4649 0.9996
60 min () 0.3043 3.1059 0.5196 0.9998
Close TMPD+ radical production rates and r
2 ~ 1 indicate that TMPD was in large
excess and reaction was first order with respect to the number of „reactive‟ centers on
151
SWNT. Obtained rate constants b corresponded to TMPDconsumed lifetime ~ 2 min at r. t.
in the presence of ozonated SWNT. The dependence of concentration of SWNT or
TMPD on TMPD+ radical production rate was not studied. A greater amount of TMPD
+
radicals produced with no “waiting” time is attributed to an additional reaction between
1,2,3-trioxolanes and TMPD. An appearance of a strong purple color is associated with
an electron transfer from TMPD to oxidizing species. Though pH affects the coloration of
TMPD solution, it does not impart such an intense purple color. Pictures of TMPD
coloration in EtOH/H2O (1:1 v/v) at different pH values are shown in Figure 7 below.
a b c d e f
Figure 7. TMPD coloration at different pH. An acidity of a TMPD solution in
water/ethanol mixture (1:1 v/v) was adjusted with HCl or NaOH. pH values were (a) 0.3,
(b) 1.8, (c) 3.6, (d) 8.7, (e) 13.1, (f) 13.9.
As seen in Figure 7, protonated TMPD at pH 3.6 (sample c) had a light crimson
color. No purple color could be obtained by adjusting pH. Scheme 1 below summarizes
observed reactions between TMPD and ozonated SWNT.
152
Scheme 1
NN
pristine SWNT
EtOH
no reaction
colorless
Wursterreagent
OO
O
epoxide
1,2,3-trioxolane
EtOH
O
NN
solvdeep purple
O
+
solv
or
otherproducts+
A reasonable question to ask: if solution turns purple, does TMPD+ radical get
attached to SWNT surface? Could it be that radical produced is dissolved by the media
and not attached to SWNT surface?
To answer on this question, XPS spectra were acquired on samples (Teflon® coated
with SWNT) dipped into a solution of TMPD in ethanol, followed by washing with
ethanol (2 times) and water. The first sample was subjected to a stream of O3/O2 gaseous
mixture (ca. 1.5 v/v % ozone) for 1 min before dipping it into TMPD solution (2.8 mg in
500 uL of ethanol). A washing step was designed to remove traces of unreacted TMPD.
As seen in Figure 8, ozonated SWNT film got a small number of TMPD molecules
attached. A delocalization of an unpaired electron in TMPD+
and solvation with ethanol
molecules are thought to be the major reasons for turning solution color to a deep purple.
It is reasonable to expect a greater percent of radicals, generated from aliphatic amines,
getting attached to SWNT surface. Aliphatic amines do not have electron delocalization
seen in TMPD and are thought to be more reactive.
153
Binding Energy (eV)
02004006008001000
Co
un
ts /
se
c
x1
04
1
2
3
4
Co
un
ts /
se
c
x1
04
1
2
3
4
5
- O
KLL
- O
1s
- N
1s
- F
e 2
p3/2
- O
1s
- F
e 2
p3/2
- C
1s
SWNT + Wurster reagent
Ozonated SWNT + Wurster reagent
- C
1s
Figure 8. XPS spectra of SWNT reacted with an ethanolic solution of TMPD. (Top) a
product from reaction with pristine SWNT, a reference, (bottom) a product from reaction
with ozonated SWNT.
Fluorescence studies of reaction between dithiothreitol (DTT) and ozonated SWNT
Reaction between a thiol and ozonated SWNT has been studied by fluorescence
(Figure 9).
SH
OH
OH
HS
dithiothreitol
154
Time (sec)
0 200 400 600
Norm
aliz
ed F
luore
scence
0.0
0.3
0.6
0.9
954 nm
1026 nm
1123 nm
1250 nm
Figure 9. An addition of dithiothreitol to ozonated SWNT – SDS suspension. The sample
was excited with 661 nm laser source. Fluorescence of tube (8,3) at 954 nm recovered at
once to a 90% level of its initial intensity. Other peaks returned to lower levels. The
timing of DTT addition is emphasized with an arrow.
Fluorescence decreased at least 200 times upon reaction with ozone at all four
emission wavelengths. Partial ozonide decay brought fluorescence intensities up to 1 % at
954 nm and 2 % at 1250 nm of initial levels after which dithiothreitol was injected. An
immediate fluorescence recovery followed thiol addition.
Different tubes recovered to different levels. As observed in other experiments, tubes
with emmax
at shorter wavelengths recovered to higher percent levels. Dithiothreitol was
added 2.5 minutes after injection of ozone/oxygen gaseous mixture. UV studies
demonstrated (Chapter 2) that all freely floating ozone should be gone within less than 2
min at r. t. A fluorescence rise was another indication of the absence of ozone. It is
believed there was no more unreacted ozone in aqueous SWNT suspension at the time of
a dithiol injection.
155
Alkyl thiols, analogously to aliphatic amines, are thought to serve as electron donors
during the first step of interaction with ozonated SWNT. An increase of electron density
on SWNT is supported by the evidence of increased NIR absorbance of SWNT at
frequencies above 7000 cm-1
upon reaction with amines (IR studies, Chapter 4). A greater
number of electrons in a -cloud causes SWNT to absorb stronger in the NIR.
Noticeably, recovery of fluorescence after thiol addition was permanent and showed
no signs of equilibration after chemical addition (compare to Figure 10). The jump was
abrupt and did not change over time. There are three possible explanations for this
phenomenon. Ozonide decay by thiol is the most obvious one. The second one is a
covalent attachment of thiol to SWNT. A third possibility is that hydrophobic interaction
of SWNT with thiol led to an accumulation of thiols on the surface of tubes, in the same
manner as with 2-methoxyethylamine. All explanations are likely to be valid, though the
third one would not result in such a dramatic intensity change.
Injections of water-miscible 2-methoxyethylamine into SWNT-SDS aqueous
suspensions were shown to yield fluorescence increase (Figure 10). It is reasonable to
assume that thiols would result in a similar fluorescence intensity increase.
156
Time (min)0 10 20
Norm
aliz
ed F
luore
scence
1.0
1.2
1.4
954 nm
1026 nm
1123 nm
1250 nm
Figure 10. Fluorescence intensity change of pristine SWNT at four distinct wavelengths
after several additions of 2-methoxyethylamine. Injections are emphasized with arrows.
The 661 nm laser was used for excitation.
Larger diameter tubes, contributing to emission at 1250 nm, had no fluorescence
increase with second through fourth additions of amine. Spikes right after additions were
due to misbalance and are not meaningful. In contrast, thinner tubes, emitting at 954 nm,
had fluorescence increase after all four injections.
SDS shells around nanotubes are thought to be in quasi liquid state. The above
experiment demonstrated that: a) an increased number of amine molecules at the surface
of SWNT causes stronger emission and b) SDS shell around thinner tubes ( emmax
954
nm) could not be saturated with amine molecules as it was in the case with thicker ones
( emmax
1250 nm). It is concluded that SDS shells around thicker tubes are more rigid,
more hydrophobic and could be harder to penetrate for a heavily solvated reagent like 2-
methoxyethylamine.
157
Spikes right after chemical additions are thought to be due to an increased
concentration of amine at the surface of SWNT. System equilibration results in expulsion
of amine molecules from quasi-liquid shells of sodium dodecyl sulfate. Lack of an
“equilibration” period in Figure 9 suggests that thiol could get covalently attached to the
SWNT surface.
Wavelength (nm)
950 1050 1150 1250
Flu
uo
resce
nce
In
ten
sity (
nW
/nm
)
0.00
0.01
0.02
0.03
0.04
0.05
951 955 959N
orm
aliz
ed
0.7
0.8
0.9
1.0Pristine SWNT
1st
2nd
3rd
4th
amine addition
Figure 11. Dilution of SDS shell with 2-methoxyethylamine and its influence on SWNT
emission. The 660 nm laser was used for excitation. Equal amounts of amine were
injected each time. Main plot: fluorescence spectra as measured, zoom-in: normalized
fluorescence near 954 nm peak. Curves are: – pristine SWNT suspended in 1 wt. %
aqueous SDS; Symbols , , , and correspond to spectra after 1st, 2
nd, 3
rd and 4
th
injections of amine.
As seen in Figure 11, each subsequent injection of amine (4 umol, 5 uL,
c = 6.7 v/v % in 1 wt. % aq. SDS) resulted in fluorescence rise at peak maxima near 954,
1026 and 1123 nm. No change was observed for 1250 nm. Rise was greater with shorter
wavelengths, indicating higher degree of penetration of quasi-liquid SDS shell. No peak
shifts were observed. This is in contrast to small shift at 954 nm observed for reaction of
158
amines with ozonated SWNT. In general, peak shifts of SWNT should be treated with
caution, since different surfactants (SDS, SDBS, CTAB, Brij 700 and other) were shown
to affect peak maxima location.22
Dithiothreitol solubility in water is significantly lower than that of SDS;
15.4 g/L for dithiol vs. 200 g/L for sulfate. The argument could be made that DTT
preferentially intercalated in between SDS chains on the surface of SWNT. It is worth
noting that such high solubility of SDS is achieved through formation of micelles. In this
work 1 wt. % SDS (10mM) aqueous suspensions were used for all fluorescence and
liquid Raman measurements. Critical micelle concentration for SDS is about 8mM in
water.23
Encapsulation of DTT molecules within SDS micelles is yet another possibility.
Due to a great number of variables in the system, fluorescence intensity change alone is
not a proof of covalent attachment. An independent analytical method such as X-ray
Photoelectron Spectroscopy (XPS) is needed to establish elemental composition of atoms
on the surface of SWNT.
In conclusion, the fluorescence technique was useful in demonstrating how different
chemicals affect ozonide stability, but it provided no actual proof of covalent attachment
of such reagents to SWNT sidewall. Amines and thiols were shown to decompose
ozonides on the SWNT surface, thus affecting SWNT fluorescence intensity.
Fluorescence studies of reaction between 2-methoxyethylamine and ozonated SWNT
at different times after ozonation
The influence of amines on fluorescence of ozonated SWNT at different time
intervals after oxidation was investigated.
159
Time (min)
0 10 20 30
No
rma
lize
d F
luo
resce
nce
0.0
0.3
0.6
0.9
Imax
L1L2
a bc
d
e
em = 1250 nm
No
rma
lize
d F
luo
resce
nce
0.0
0.3
0.6
0.9
46 umol
35 umol
23 umol
4 umol
4 umol
4 umol
em = 1026 nmI
max
L1L2
a b
c
d
e
Figure 12. Addition of 2-methoxyethylamine to ozonated SWNT – SDS suspension.
Each curve represents a separate experiment and is marked with a symbol. All curves are
overlaid for comparison purposes. All reactions are run at 23.1 C. Symbol legends are
the same for the top and the bottom graphs. Imax is a SWNT fluorescence level before
bubbling ozone. Letters: (a) a fluorescence drop with an injection of 1 mL of O3/O2
gaseous mixture (ca. 3 v/v % ozone), (b-d) addition of 4 umol of amine at different times,
(e) an addition of 23, 35 and 46 umol of amine at approx. the same time. Top: emission at
1026 nm. Bottom: emission at 1250 nm.
160
A determination of ozonide decomposition rates made it possible to distinguish
between reactions of amines with SWNT epoxides and SWNT ozonides. Based on
oxygen evolution studies, all ozonides are thought to be decomposed within 20 minutes at
r. t. (Chapter 2, Figure 2). Normalized fluorescence spectra of oxidized SWNT treated
with different quantities of 2-methoxyethylamine and at different times are shown in
Figure 12.
2-Methoxyethylamine was considered a good choice for such studies, since it is
miscible with water and is not expected to have hydrophobic interactions, or adhesion to
SWNT surface. Two heteroatoms on a molecule greatly facilitate its solvation with water
dipoles.
Levels L1 and L2 in Figure 12 show schematically the dependence of the
fluorescence recovery on the amount of amine added. Larger amounts of amine resulted
in a slightly higher level of fluorescence (compare L1 to L2 in Figure 12) and faster
reaction rates as judged by curve slopes right after amine addition. Injections of 23 to 46
umol of amine gave steeper slopes than those after 4 umol. Addition of 4 umol of a
nucleophile at different times resulted in the same recovery levels (schematic line L1).
This result means that the amount of amine was in a large excess to the number of 1,2,3-
trioxolanes formed on the surface of SWNT after ozone bubbling.
In accordance with the Le Chatelier principle, an increased number of amine
molecules in the suspension resulted in a higher intercalation level of SDS shell with
amines, leading to steeper fluorescence recovery slopes (Figure 12).
Fluorescence jumped upward after addition of amine 24.3 min after ozonation (d of
Figure 12). It is thought that jump is associated with a) covalent attachment of amines to
161
ozonated SWNT and b) intercalation of SDS shell with freely floating amines. As seen in
Figure 10, addition of the same amount of amine to non oxidized SWNT gave
fluorescence increase by only 0.15 (at 1026 nm) and 0.05 (at 1250 nm). Separate X-ray
Photoelectron Spectroscopy measurements further confirmed that SWNT surface is
modified with nitrogen containing groups (Chapter 7).
Molar ratio of the amount of ozonated double bonds to that of added amine
A change in absorbance at 260 nm is directly proportional to the number of 1,2,3-
trioxolanes formed. Epoxides and 1,2,3 trioxolanes are not expected to absorb at such a
short wavelength. Whether SWNT absorbance changes linearly with decrease in the
number of double bonds is not known. Let‟s assume that at low levels of SWNT
ozonation it is linear. Bubbling ~ 1 mL of O3/O2 gaseous mixture (ca. 3 v/v % ozone) was
used for a low degree ozonation. Resting on the above assumptions, the following
calculations were made. In a separate experiment SWNT absorbance was bleached less
than 4 % at 260 nm after bubbling 1.5 mL of O3/O2 gaseous mixture through suspension
(Chapter 2, Figure 3). An approximate number of double bonds reacted with ozone can
be estimated from this result. A SWNT-SDS suspension had a tube concentration of 5.75
ug of carbon/mL. Four per cent of this amount corresponds to 0.23 ug/mL, or 0.019
umol/mL. Dividing this number by two yields the amount of double bonds reacted with
ozone, 0.01 umol/mL. This is ca. 400 times less than the lowest amount of amine injected
(see Figure 12).
162
Interestingly, a reaction between amine and ozonated SWNT can occur even in the
absence of 1,2,3-trioxolanes. Amines were shown to react with SWNT two days after
ozonation (Chapter 4, Figure 19).
Possible reaction mechanisms between electron rich nucleophiles and ozonated
SWNT are summarized in Scheme 2.
163
Scheme 2
OO
OO
OO
OO O
Nu
epoxide
O
oxidoannulene
O O
Nu
1,2,3-trioxolane
Nu = AlkSH, ArSH, R3N, RNH2, R2NH, PR3, guanidine, etc.
Nu
Nu
Nu
SWNT=
OO
epoxide
RNH2
RNH2 (xs)
RNH2
Route A
Route B
RNH3
RNH
RNH O
NH
R
RNH3
RNH2
O
NH
R
H
OO
ONu
NuOno attachment ofNu to sidewall
Route C
Reactions between amines and ozonated fullerene C60Ox
Reactions observed between ozonated SWNT and amines prompted a series of
r. t. experiments with ozonated C60Ox. Thus, a purple solution of C60 in toluene was
ozonated briefly for 5 sec with a stream of O3/O2 gaseous mixture (ca. 3 v/v % ozone),
purged thoroughly with argon gas and amine was added. Purging with inert gas was done
to ensure absence of ozone in toluene. An orange color of ozonated fullerene immediately
changed to a dark brown, indicating the reaction proceeded fast at r. t. Primary and
164
tertiary amines were used to demonstrate the electron transfer nature of this reaction.
Triethylamine, in the same manner as n-butylamine, reacted with ozonated fullerenes to
produce a highly polar brownish product, which oiled out of toluene over time. Results of
these reactions are summarized in Scheme 3.
Scheme 3
C60
BuNH2
Et3N
BuNH2
Et3N
light purpleorange
no reaction
no reactiontoluene
toluene
NEt3
toluene
dark brown
NH2Bu
dark brown
C60 On + m
C60
OO
O
O
n
m
C60 On + m
k
k
r.t.
r.t.
r.t.
otherproducts+
otherproducts+
Reaction mixtures were photographed and shown in Figure 13.
Figure 13. A solution color comparison of C60 and derived products. Solutions are in
toluene. All reactions were run for 5 min at r. t. Labels are: (A) fullerene C60, (B)
fullerene ozonated for several seconds C60(O3)x and (C) reaction product of ozonated
fullerene and n-butyl amine C60Ox(NH2Bu)y
Reaction of ozonated fullerene with triethylamine produced the same dark brown
color as in reaction with n-butyl amine, indicating that electron transfer from nitrogen to
C60(O3)x is likely to be the first step. An analysis of product from reaction with
triethylamine by 1H NMR and IR indicated the presence of triethyl ammonium salt.
Characteristic shifts in 1H NMR (MeOH-d4) were at 3.18 and 1.28 ppm for ethyl groups,
165
as would be expected for an ammonium salt. An authentic sample of
tetraethylammonium chloride in
MeOH-d4 had chemical shifts 3.35 and 1.33 ppm, which are very close to the ones in
the fullerene derivative.
To demonstrate the ability of amines to react with ozonated fullerenes even in the
absence of 1,2,3-trioxolanes, n-butyl amine was added to ozonated C60 twenty minutes
after oxidation. The same color change from orange to dark brown occurred (Figure 14).
This implied that the color change is likely due to a rearrangement of conjugated double
bonds on fullerene.
Figure 14. A solution color comparison of C60 and derived products. Solutions A and B
are in toluene; sample C was dried under vacuum and redissolved in MeOH. All reactions
were run for 5 min at r. t. Labels are: (A) fullerene C60, (B) mixture of C60 and n-BuNH2
and (C) amine added to C60Ox twenty minutes after ozonation
Cross-linking of aminated fullerenes has been reported in literature, which could be
the case here as well.2 IR spectrum and peak assignments for structure C60Ox(NH2Bu)y
are provided below.
166
Wavenumber (cm-1
)
100020003000
Absorb
ance (
a.u
.)
0.1
0.2
0.3
C60
Ox(NH
zBu)
y
a
b
c
d
e
f
g
h i
j
k
Fugure 15. IR spectrum of C60Ox(NH2Bu)y
Table 2. IR peak assignments for structure C60Ox(NH2Bu)y:
Symbol Position Assignment
a 3025 N-H stretch in R2NH2+; may be combined with peak from C60Ox
b 2958 b-d CH3 and CH2 stretches
c 2930 ditto
d 2872 ditto
e 1711 C=O from C60Ox
f 1641 C=C
g 1586 N-H bend
h 1465 CH3 as bend & CH2 sym bend (scissoring)
i 1378 CH3 sym bend
j 1081 C-N stretch
k 736 N-H wagging
167
IR spectrum of C60Ox(NEt3)y and peak assignments are provided below.
Wavenumber (cm-1
)
100020003000
Absorb
ance (
a.u
.)
0.05
0.10
0.15
0.20C
60O
x(NEt
3)y · n MeOH-d4
a
b
c
d
e
fg
h i
j
k
Figure 16. IR spectrum C60Ox(NEt3)y
Table 3. IR peak assignments for structure C60Ox(NEt3)y:
Symbol Position Assignment
a 2000-
3500
unassigned; it could be from C60Ox or C60Ox(NEt3)y(MeOH)n
b 2984 unassigned
c 2955 c-d CH3 and CH2 stretches
d 2917 ditto
e 2850 ditto
f 1732 C=O from C60Ox
g 1614 unassigned
h 1455 CH3 as bend & CH2 sym bend (scissoring)
i 1390 CH3 sym bend
j 1096 C-N stretch
k 804 unassigned
IR spectra had multiple bands coming from amine residues. Band assignments are
listed in the corresponding tables under the IR spectra. Amines are believed to be
168
covalently attached to oxidized C60 species. Samples changed color from purple to deep
orange during ozonation, and then from orange to dark brown upon amine addition.
Amines chosen for reaction had low boiling points and were expected to be completely
evaporated during vacuum drying (10-3
Torr). Products slowly aggregated and oiled out
of toluene as dark brown – black liquid drops. It is likely that formed species were
charged and solvated by the excess of freely floating amine molecules. Toluene was
evaporated and product redissolved in methanol-d4, forming a dark brown solution. C60 is
extremely insoluble in solvents like methanol. Obtained products C60Ox(NH2Bu)y and
C60Ox(NEt3)y had excellent solubility in methanol. Ability to redissolve C60Ox(amine)y in
methanol is an indication of a highly polar product. Change in color from orange to dark
brown is thought to be due to electron transfer from amine to C60Ox species. Exact
structures of products were not established. Repetition of experiment with longer
ozonation period resulted in a better solubility of C60Ox(amine)y in methanol, indicating
an increased number of amine residues on a single C60 skeleton.
Benzene oxidation with ozone is known.24
Some IR peaks may be coming from non-
volatile species formed by toluene oxidation. The nature of the very broad peak at
2000-3500 cm-1
in reaction with triethylamine is not clear. It is thought that methanol-d4
could get involved in protonation of the reactive species.
13C NMR was not performed on a crude mixture. MALDI-TOF characterization of
crude products gave complex spectra for both triethyl and n-butyl amines. It is believed
that there were multiple amine residues on C60Ox. Multiple charges, expected for
C60Ox(NEt3)y species, further complicated MS peaks‟ assignment. Oxidation of several
169
double bonds on C60 is well documented in the literature.25
The number of amines on
each C60 skeleton was not established.
Though reaction between amines and C60O3 was clearly manifested by changes in
color, solubility in methanol, characteristic peaks in 1H NMR and IR, thorough study is
needed for unambiguous structure assignments. Particularly, separation of products on
HPLC and MS and NMR characterization of each product is needed.
5.3. Conclusions
Ozonated SWNT suspended in aqueous SDS solution were shown to react with
electron rich nucleophiles, like amines and thiols. While reaction between SWNT and
amines is known to proceed with a very slow rate at normal conditions, the technique
presented here allowed SWNT side wall modification within minutes. The key to
decreasing activation energy of the process was oxidation of SWNT with ozone, resulting
in conversion of SWNT to a mild oxidizer, capable of abstracting electrons from thio and
amino moieties. Formed radicals are thought to be of a high energy and reacted with
SWNT in a matter of minutes. Reaction of Wurster reagent with ozonated SWNT was
monitored by UV and the rate of TMPD+ radical production found to be ca. 0.5 min
-1 for
a very dilute ethanolic suspension of SWNT. Reactions of amines with ozonated
fullerene C60 demonstrated similar behavior. n-Butyl and triethyl amines reacted with
C60On within seconds with corresponding color change from deep orange to deep brown.
Formed products were found to be well soluble in methanol, thus C60 solubility in
methanol was greatly increased.
170
5.4. Experimental Part
Reaction of ozonated SWNT with N,N,N’,N’-tetramethyl-p-phenylenediamine
Ethanol was used as a solvent (200 proof, Aaper Alcohol and Chemical Co.). A
preliminary sonicated for 40 min concentrated stock suspension of carbon nanotubes in
ethanol (Rice University, HipCo raw tubes, unpurified, batch 162.8) was used for
reactions. All sonication was done with a bath sonicator (Fisher FS60). The Ozone
Services Inc. ozonator (model GE60/FM 500) with a power level set to the maximum and
oxygen gas flow to 1/16 L/min was used for ozone production. Ozonator was warmed up
for 7 min per manufacturer instruction to reach its maximum output. UV measurements
were performed on Cary 4E spectrophotometer with 1cm quartz cuvette. Ethanol
spectrum served as a baseline. Wurster reagent was prepared by dissolution of N,N,N’,N’-
tetramethyl-p-phenylenediamine (11.5 mg) in ethanol (2 mL).
A very dilute suspension of SWNT (< 0.1 mg) was prepared in 20 mL of ethanol and
bath sonicated for 30 sec. Finely dispersed suspension was shaken to “crash out” SWNT
back to carbon nanotube flakes, and then distributed 2.5 mL aliquots into separate glass
vials. Suspension was swirled before each aliquot was taken out of a stock solution.
Wurster reagent (150 uL, 5.2 umol) was added to the first sample. Second sample was
ozonated by bubbling O3/O2 gaseous mixture (ca. 3 v/v % ozone) through the suspension
for 1 min at r. t. Air was blown off the top of the vial with a stream of argon for about 20
sec, then argon was bubbled through ozonated solution for about 30 sec with speed
approx. 1/16 L/min, and blowing air off of the top of the vial with argon was repeated as
before. Wurster reagent (150 uL) was added to ozonated sample. Other samples were
171
analogously ozonated and purged with argon, but waited for 10, 40 and 60 min before
addition of Wurster reagent.
Reaction of ozonated SWNT with dithiothreitol
A suspension of SWNT (HipCo, batch 161.1, Rice University) in 1 wt. % aq. SDS
was prepared by a standard procedure. Suspension pH was adjusted with NaOH to pH 8.
SWNT concentration was estimated at 5.75 mg/L. The O3/O2 gaseous mixture (1 mL, ca.
3 v/v % ozone) was bubbled through 1 mL of SWNT – SDS suspension, waited for 2.5
min and added dithiothreitol (5 uL, a diluted solution in water). Reaction was monitored
at 22.2 C by fluorescence. The 661 nm laser was used for excitation.
Influence of amine on pristine SWNT fluorescence intensity
A suspension of SWNT (HipCo, batch 161.1, Rice University) in 1 wt. % aq. SDS
was prepared by a standard procedure. Suspension pH was adjusted with NaOH to pH 8.
SWNT concentration was estimated at 5.75 mg/L. Multiple additions of 2-methoxyethyl
amine (4 umol, 5 uL, 6.7 v/v % in 1 wt. % aq. SDS) to SWNT – SDS suspension at 23.1
C were monitored by fluorescence. The 661 nm laser was used for excitation.
Reaction of 2-methoxyethylamine with ozonated SWNT
SWNT-SDS suspension was prepared by a standard procedure from HipCo 161.1
raw tubes (see Chapter 2 for details). pH was adjusted to 8 with 0.1N aq. NaOH,
suspension shaken and bath sonicated for 1 min. Solutions of 2-methoxyethylamine
(0.77M and 2.3M) in 1 wt. % aq. SDS were used for amine additions. All reactions were
run at 23.1 C in a thermostated quartz cuvette with 1 mL of SWNT – SDS suspension.
Spectra were recorded on NanoSpectralyzer NS1 with 661 nm excitation source. Points
172
were acquired every 10 sec with 500 ms excitation pulses. The O3/O2 gaseous mixture
(1 mL, ca. 3 v/v % ozone) was bubbled through SWNT suspension. In three separate
experiments amine (4 umol) was added at times 1.7, 9.5 and 24.3 min. In another three
different experiments 23, 35 and 46 umol of amine were added at approx. 5.5 min.
Resulted emission curves were normalized and overlaid for comparison.
Reaction of amines with ozonated C60
Purple-colored aliquots of C60 in toluene (0.25 mL; conc. not available) were
bubbled for 5 sec with O3/O2 gaseous mixture (ca. 3 v/v % ozone) followed by thorough
purging with nitrogen and addition of amine (5 uL). Waited for 5 min, evaporated toluene
and dried on vacuum for 5 min. Residue was redissolved in MeOH-d4 and a few drops
were added to BaF2 window. Plate was dried in vacuum for 15 min and IR measured.
Chamber was purged with nitrogen to remove carbon dioxide and moisture. The
remaining portion of solution was used for 1H NMR and MS. MeOH-d4 served as a
reference. 1H NMR (MeOH-d4), C60Ox(NR3)y 3.18 (q, J = 7.2 Hz) and 1.28 (dt, J1 =
7.7 Hz; J1 = 3.3 Hz); 1H NMR of n-BuNH2 was poorly resolved and data is not provided.
Reaction of 2-methoxyethylamine with ozonated SWNT (3D kinetics experiment)
SWNT – 1 wt. % aq. S DS suspension was prepared by a standard procedure from
HipCo 153.3 raw tubes (details in experimental part of Chapter 2). pH was adjusted to 8
with 0.1N aq. NaOH, sample shaken and bath sonicated for 1 min. The O3/O2 gaseous
mixture (2 mL, ca. 3 v/v % ozone) was injected into 1.0 mL of SWNT – SDS suspension
(c ~ 5.75 mg/L; 0.48 umol carbon atoms/mL), waited for 1 min and amine was injected
173
(6 uL, 9 umol, 13 v/v % in water). Kinetics was monitored by fluorescence. The 660 nm
laser was used for excitation.
SWNT ozonation (3D kinetics experiment)
SWNT – 1 wt. % aq. S DS suspension was prepared by a standard procedure from
HipCo 153.3 raw tubes (details in experimental part of Chapter 2). pH was adjusted to 8
with 0.1N aq. NaOH, sample shaken and bath sonicated for 1 min. The O3/O2 gaseous
mixture (0.17 mL, ca. 3 v/v % ozone) was injected into 1.0 mL of SWNT – SDS
suspension (c ~ 5.75 mg /L; 0.48 umol carbon atoms/mL). Kinetics was monitored by
fluorescence. The 660 nm laser was used for excitation.
Procedure for Wurster reagent (for XPS measurements)
A Teflon plate covered with SWNT film was ozonated for 1 min with a stream of
O3/O2 gaseous mixture (ca. 3 v/v % ozone). Plate was dipped into the solution of TMPD
(2.7 mg in 500 uL EtOH) for 15 min, and then washed two times in EtOH (1 mL) and
one time in water (1 mL) for 3 min each. The sample was dried in vacuo before XPS
measurements.
The pH adjustments
Millipore Milli-Q water (18.2 MOhm/cm at 25 C), ethanol, 0.1 and 1N aq. NaOH
(Fisher Scientific), 0.1 and 1N aq. HCl (Fisher Scientific) were used to prepare TMPD
solutions with different pH. Measurements were performed with a digital pH meter
(Fisher Scientific, Dual Channel pH meter AR 50).
174
5.5. References and Notes
1. Hu, X. Q.; Jiang, Z. P.; Jia, Z. S.; Huang, S. H.; Yang, X. B.; Li, Y. L.; Gan, L.
B.; Zhang, S. W.; Zhu, D. B., Amination of 60 fullerene by ammonia and by
primary and secondary aliphatic amines - Preparation of amino C60 fullerene
peroxides. Chemistry-a European Journal 2007, 13, (4), 1129-1141.
2. Manolova, N.; Rashkov, I.; Beguin, F.; Vandamme, H., Amphiphilic Derivatives
of Fullerenes Formed by Polymer Modification. Journal of the Chemical Society-
Chemical Communications 1993, (23), 1725-1727.
3. Nakamura, E.; Isobe, H., Functionalized fullerenes in water. The first 10 years of
their chemistry, biology, and nanoscience. Accounts of Chemical Research 2003,
36, (11), 807-815.
4. Isobe, H.; Tanaka, T.; Nakanishi, W.; Lemiegre, L.; Nakamura, E., Regioselective
oxygenative tetraamination of C60 fullerene. Fullerene-mediated reduction of
molecular oxygen by amine via ground state single electron transfer in dimethyl
sulfoxide. Journal of Organic Chemistry 2005, 70, (12), 4826-4832.
5. Isobe, H.; Tomita, N.; Jinno, S.; Okayama, H.; Nakamura, E., Synthesis and
transfection capability of multi-functionalized fullerene polyamine. Chemistry
Letters 2001, (12), 1214-1215.
6. Isobe, H.; Ohbayashi, A.; Sawamura, M.; Nakamura, E., A cage with fullerene
end caps. Journal of the American Chemical Society 2000, 122, (11), 2669-2670.
7. Isobe, H.; Tomita, N.; Nakamura, E., One-step multiple addition of amine to C60
fullerene. - Synthesis of tetra(amino)fullerene epoxide under photochemical
aerobic conditions. Organic Letters 2000, 2, (23), 3663-3665.
8. Schick, G.; Kampe, K. D.; Hirsch, A., Reaction of C60 Fullerene with Morpholine
and Piperidine - Preferred 1,4-Additions and Fullerene Dimer Formation. Journal
of the Chemical Society-Chemical Communications 1995, (19), 2023-2024.
9. Isobe, H.; Nakanishi, W.; Tomita, N.; Jinno, S.; Okayama, H.; Nakamura, E.,
Gene delivery by aminofullerenes: Structural requirements for efficient
transfection. Chemistry-an Asian Journal 2006, 1, (1-2), 167-175.
10. Banerjee, S.; Wong, S. S., Rational sidewall functionalization and purification of
single-walled carbon nanotubes by solution-phase ozonolysis. Journal of Physical
Chemistry B 2002, 106, (47), 12144-12151.
11. Kuwabara, H.; Ushigoe, Y.; Nojima, M., Synthesis and reaction of cyano-
substituted 1,2,4-trioxolanes. Journal of the Chemical Society-Perkin
Transactions 1 1996, (9), 871-874.
175
12. Bailey, P. S., Ozonation in Organic Chemistry. Academic Press: New York,
1978; Vol. 1, p. 272.
13. Bailey, P. S., Ozonation in Organic Chemistry. Academic Press: New York,
1982; Vol. 2, p. 497.
14. Lim, J. K.; Yun, W. S.; Yoon, M. H.; Lee, S. K.; Kim, C. H.; Kim, K.; Kim, S.
K., Selective thiolation of single-walled carbon nanotubes. Synthetic Metals 2003,
139, (2), 521-527.
15. Nakamura, T.; Ohana, T.; Ishihara, M.; Tanaka, A.; Koga, Y., Sidewall
modification of single-walled carbon nanotubes with sulfur-containing
functionalities and gold nanoparticle attachment. Chemistry Letters 2006, 35, (7),
742-743.
16. Liu, Z. F.; Shen, Z. Y.; Zhu, T.; Hou, S. F.; Ying, L. Z.; Shi, Z. J.; Gu, Z. N.,
Organizing single-walled carbon nanotubes on gold using a wet chemical self-
assembling technique. Langmuir 2000, 16, (8), 3569-3573.
17. Peng, H. Q.; Alemany, L. B.; Margrave, J. L.; Khabashesku, V. N., Sidewall
carboxylic acid functionalization of single-walled carbon nanotubes. Journal of
the American Chemical Society 2003, 125, (49), 15174-15182.
18. Stevens, J. L.; Huang, A. Y.; Peng, H. Q.; Chiang, L. W.; Khabashesku, V. N.;
Margrave, J. L., Sidewall amino-functionalization of single-walled carbon
nanotubes through fluorination and subsequent reactions with terminal diamines.
Nano Letters 2003, 3, (3), 331-336.
19. Holzinger, M.; Vostrowsky, O.; Hirsch, A.; Hennrich, F.; Kappes, M.; Weiss, R.;
Jellen, F., Sidewall functionalization of carbon nanotubes. Angewandte Chemie-
International Edition 2001, 40, (21), 4002-4005.
20. Yang, J. H.; Wang, K.; Driver, J.; Barron, A. R., The use of fullerene substituted
phenylalanine amino acid as a passport for peptides through cell membranes.
Organic & Biomolecular Chemistry 2007, 5, (2), 260-266.
21. Rouse, J. G.; Yang, J. Z.; Barron, A. R.; Monteiro-Riviere, N. A., Fullerene-based
amino acid nanoparticle interactions with human epidermal keratinocytes.
Toxicology in Vitro 2006, 20, (8), 1313-1320.
22. Moore, V. C.; Strano, M. S.; Haroz, E. H.; Hauge, R. H.; Smalley, R. E.; Schmidt,
J.; Talmon, Y., Individually suspended single-walled carbon nanotubes in various
surfactants. Nano Letters 2003, 3, (10), 1379-1382.
176
23. Esposito, C.; Colicchio, P.; Facchiano, A.; Ragone, R., Effect of a weak
electrolyte on the critical micellar concentration of sodium dodecyl sulfate.
Journal of Colloid and Interface Science 1998, 200, (2), 310-312.
24. Nakagawa, T. W.; Andrews, L. J.; Keefer, R. M., The Kinetics of Ozonization of
Polyalkylbenzenes. Journal of the American Chemical Society 1960, 82, (2), 269-
276.
25. Manning, T. J.; Olsen, K.; Hardin, L.; Purcell, J.; Ayers, T. M.; Duncan, M. A.;
Phillips, D., Extensive ozonation of C60: Degradation or polymerization? Ozone-
Science & Engineering 2006, 28, (3), 177-180.
177
Chapter 6
Trapping reactive centers on SWNTOn with electron rich nucleophiles
(amines, thiols)
178
6.1. Introduction
This chapter examines the possibility of trapping reactive centers on SWNT after it
was subjected to ozonation. Successfully trapped oxidative areas on “SWNT” should
loose their ability to withdraw electrons from N,N,N’,N’-tetramethyl-p-
phenylenediamine, thus preventing purple coloration. Such trapping methodology for
ozonated SWNT has not been described in the literature.
6.2. Results and discussion
Reactive centers on ozonated carbon nanotubes (SWNTOn) were efficiently trapped
with reagents of choice, thus preventing N,N,N’,N’-tetramethyl-p-phenylenediamine
(TMPD) from getting oxidized. TMPD undergoes a characteristic color change from a
pale tan to a deep purple due to radical cation generation when reacted with oxidizers. A
number of reagents were tested in reaction with ozonated SWNT: guanidine,
dithiothreitol, 2-methoxyethylamine and sodium salts of arginine, -aminobutyric acid,
lysine and cysteine. All reagents were expected to give up an electron in reaction with
ozonated SWNT. Reaction was run against a reference sample, which was prepared by
exactly the same procedure but did not include an ozonation step.
For reaction to proceed, an amino group of a reagent must be in its free form. As
seen in Figure 1, having charged amino groups, guanidine·HCl and
-aminobutyric acid were not able to quench reactive centers on SWNTOn. Subsequent
additions of TMPD turned suspensions to purple. Without NaOH added, lysine was found
to quench SWNTOn reactive centers inefficiently. Addition of TMPD yielded a faint
purple color.
179
The following is a brief summary of the procedure. A diluted stock suspension of
SWNT (less than 2.0 mg/70 mL) was prepared by bath sonication and aliquots were
drawn from each sample. Stock solutions were prepared for each nucleophilic reagent to
ensure that an equal amount of reagent was used for both reaction and reference samples.
SWNT suspensions were ozonated for 1 min, thoroughly purged with argon, then a
reagent of interest was added and Wurster reagent added 15 min after. The results of
these reactions are summarized in Figure 1.
Cysteine was found to have a superb affinity for ozonated SWNT. Addition of this
reagent to ozonated SWNT formed grey flakes that precipitated over time (pointed with
an arrow in Figure 1). Formation of a cross-linked network where both amino and thio
groups participate in reaction is likely. Cysteine in the reference sample did not show
such behavior.
180
a b c d e
a guanidine·HCl + 1 eq NaOH
b guanidine·HCl + 1 eq NaOH (reference)
c guanidine·HCl
d guanidine·HCl (reference)
e TMPD
f g h i j
f GABA + 1 eq NaOH
g GABA + 1 eq NaOH (reference)
h GABA
i GABA (reference)
j TMPD
k l m n o
k Lysine + 1 eq NaOH
l Lysine + 1 eq NaOH (reference)
m Lysine
n Lysine (reference)
o TMPD
p q r s t
p Cysteine + 2 eq NaOH
q Cysteine + 2 eq NaOH (reference)
r Dithiothreitol
s Dithiothreitol (reference)
t TMPD
u v w x y
u MeOCH2CH2NH2
v MeOCH2CH2NH2 (reference)
w Arginine + 3.5 eq NaOH
x Arginine 3.5 eq NaOH (reference)
y TMPD
p q pz qz
p Cysteine + 2 eq NaOH
q Cysteine + 2 eq NaOH (reference)
pz Zoom-in of p
qz Zoom-in of q
Figure 1. Trapping reactive centers on SWNTOn with electron rich amines and thiols
181
6.3. Conclusion
Amines and amino acids salts were found to be effective traps of oxidized sections of
SWNT. Addition of TMPD resulted in no coloration. The first step in trapping is an
electron transfer from amine to ozonated SWNT with subsequent attachment of aminium
radical to SWNT surface. Guanidine hydrochloride, having no available lone pair, was
unable to prevent formation of TMPD˙+ radical.
6.4. Experimental Part
Wurster reagent was prepared by dissolution of TMPD (55 mg) in EtOH (10 mL,
200 proof). A stock suspension of SWNT (pristine SWNT, HipCo, batch 162.8, Rice
University) was prepared in ethanol with concentration less than 2 mg/70 mL. Occasional
swirling of SWNT stock suspension was performed to ensure equal amounts of carbon
nanotubes were withdrawn during sample preparation. Stock solutions were prepared for
each of the reagents to ensure an identical concentration of reagents in reference and
reaction samples (see Table 1). SWNT suspension (less than 0.1 mg; 2.5 mL EtOH) was
bubbled with O3/O2 gaseous mixture (ca. 3 v/v % ozone) for 1min at r. t., sample was
purged with argon, then a solution of a compound of interest was added and the mixture
swirled. After 15 min, Wurster reagent (150 uL) was added. For the corresponding
reference sample a reagent of interest (see conditions in Table 1) and Wurster reagent
(150 uL) were added to SWNT suspensions without ozonation or purging.
182
Table 1. Reagent concentrations for trapping of SWNT(O)n reactive centers*
Compounds
FW
weight eq. of
1N
NaOH other
H2O
vol
total
vol
mg mmol NaOH vol, uL solvent needed uL
Cysteine 121.16 5 0.0413 2 82.5 17.5 100.0
Lysine 146.19 5 0.0342 1 34.2 65.8 100.0
Lysine- no NaOH 5 0.0 100.0 100.0
Arginine-HCl 174.2 5 0.0287 3.5 100.4 0 100.0
GABA 103.1 5 0.0485 1 48.5 51.5 100.0
GABA- no NaOH 5 0.0 100.0 100.0
Guanidine-HCl 95.53 10 0.1047 1 104.7 95.3 200.0
Guanidine-HCl -
no NaOH 10 0.0 200.0 200.0
Dithiothreitol 154.25 10 0.0648 0 0.0 EtOH 0.0 200.0
MeOCH2CH2NH2 10ul 0 0.0 EtOH 0.0 200.0
* - Double the amounts to prepare stock solutions of each reagent.
6.5. References and Notes
There are no references for this chapter.
183
Chapter 7
Reactions between ozonated SWNT and different classes of compounds
studied by X-ray photoelectron spectroscopy
184
7.1. Introduction
A covalent attachment of amines to ozonated SWNT allows extension of this
methodology to biologically active molecules with thio and amino moieties. This chapter
examines conditions needed to improve covalent attachment of amino acids to SWNT
sidewall. A particular problem to be addressed is the determination of the amount of
sodium hydroxide needed for the reaction.
The successful attachment of a single amino acid would imply that a number of
amino acids, or peptides, could be attached to the surface of SWNT, fullerenes and
similar species. The simplicity and the speed of the reaction between ozonated SWNT or
ozonated C60 and amines make it a great tool for synthetic chemistry.
7.2. Results and Discussion
A great number of reagents containing heteroatoms like N, S, Cl, Br and F have
being tried in reactions with ozonated SWNT (see Appendix C). Amines and thiols were
shown to be the most effective ones. Nearly all essential amino acids were shown to react
with ozonated SWNT at r. t. within minutes. XPS gave nitrogen concentrations in a range
2 – 3 atomic % on the surface of SWNT for a majority of compounds.
To ensure that the observed peaks in XPS spectra are not the result of a hydrophobic
interaction or some unknown side reaction, all reactions were run against their own
references. Reference samples were prepared by exactly the same procedure, but did not
include an ozonation step. A special ‘washing’ step was developed for a complete
removal of an unreacted reagent from SWNT surface. Introduction of the washing step
was imperative for complete removal of nonvolatile reagents like amino acids. Judged by
185
the absence of characteristic peaks (N, S, F), a vast majority of reference XPS spectra had
no unreacted reagents present.
Reaction is believed to proceed through an electron transfer from an electron rich
nucleophile (amine, thiol) to an ozonated SWNT. A conversion of an amino group to its
free form, i.e. RNH3+ to RNH2, was necessary for reaction to take place. Ammonium
salts, RNH3+, were shown to have no reaction with ozonated SWNT within 15 min after
addition (Chapter 6).
A thin film of SWNT, a buckypaper, deposited onto a Teflon® 0.2 um filter by
filtration (Figure 1), was employed for ozonation and subsequent reactions with reagents
of choice.
Figure 1. Plates for XPS measurements. SWNT film was deposited onto 0.2 um
Polypropylene backed PTFE filter by filtration. Ruler is provided for reference.
With respect to XPS spectra, Teflon ® surface was found to be of a greater benefit
when compared to indium foil. Particularly, obtained spectra did not have intense peaks
from the substrate commonly seen on spectra with indium support. From a sample
preparation standpoint, it was important that all samples had the same thickness of
SWNT film on a substrate. Whatman 0.2 um Polypropylene backed PTFE membrane
filter (WTP type) was ideal for such purpose. Bath sonicated slurry of SWNT in ethanol
was filtered through PTFE filter to obtain a uniformly distributed SWNT layer on the
186
surface of a membrane. SWNT film was found to have reasonable conductivity and XPS
measurement (with neutralizer turned on) did not result in charge accumulation on
SWNT surface.
Figure 2. Plate holder for XPS measurements. Zoom-in shows a SWNT film at the
bottom of the well and a Teflon® ‘lid’. (A) 24-well holder for XPS plate handling, (B) a lid to
prevent SWNT samples from flying around due to static, (C) 4 x 6 mm SWNT film deposited
onto PTFE 0.2 um filter at the bottom of the well.
Teflon ‘plates’ with SWNT film on it had dimensions 4 x 6 mm, and were found to
be very sensitive to static. Figure 2 shows a plate holder used for samples’ collection and
transportation.
Typical XPS spectra for reference and ozonated samples are shown in Figure 3.
187
Binding Energy (eV)
02004006008001000
Co
un
ts /
se
c
x1
04
1
2
3
4
5
Co
un
ts /
se
c
x1
04
1
2
3
4
5
- O
KLL
- F
KLL
- F
1s
- O
1s
- N
1s
- C
1s
- F
e 2
p3
/2
- O
KLL
- O
1s
- F
e 2
p3
/2
- C
1s
392400408
SWNT + Guanidine
Ozonated SWNT + Guanidine
NH
H2N NH2
Figure 3. XPS spectra of SWNT reacted with an aqueous solution of guanidine. Top: a
product from reaction with pristine SWNT, a reference. Bottom: a product from reaction
with ozonated SWNT. A zoom-in in the right upper corner shows N 1s peak. A fluorine
peak comes from PTFE substrate.
The characteristic peak for nitrogen at ca. 400 eV was observed for amines and a
majority of amino acids used in this work. The spectrum obtained on a sample treated
with ozone and guanidine was estimated to have ca. 6 atomic % concentration of
nitrogen. This number is about 2 – 3 times higher than those seen for amines. Such an
increase is thought to be due to the three nitrogen atoms in guanidine molecule. Also,
guanidine does not have bulky substituents to hinder it from incident X – rays during
spectrum acquisition.
188
A typical XPS spectrum for a product of reaction of an amino acid and ozonated
SWNT is shown in Figure 4.
Binding Energy (eV)
02004006008001000
Co
un
ts /
se
c
x1
04
1
2
3
4
5
Co
un
ts /
se
c
x1
04
1
2
3
4
5
- O
KL
L
- F
KL
L
- F
1s
- O
1s
- N
1s
- C
1s
- F
e 2
p3
/2
- O
KL
L
- O
1s
- F
e 2
p3
/2
- C
1s
392400408
SWNT + Argininate, Na salt
Ozonated SWNT + Argininate, Na salt
- N
a 1
s
- N
a K
LL
O
ONa
NH2
NH
H2N
NH
Figure 4. XPS spectra of SWNT reacted with an aqueous solution of argininate. Top: a
product from reaction with pristine SWNT, a reference. Bottom: a product from reaction
with ozonated SWNT. A zoom-in in the right upper corner shows N 1s peak. A fluorine
peak comes from PTFE substrate. Atomic concentrations (%) for bottom spectrum are:
C1s 72.9, O1s 20.0, N1s 3.4, Na1s 2.6, Fe2p3/2 1.0.
Sodium peaks were commonly observed in XPS spectra of samples that underwent
ozonation and treatment with sodium salts of amino acids. The peak is likely from the
carboxylate of an amino acid. Other possibilities, like formation of carboxylates during
ozonation and sodium anion exchange during chemical treatment should not be excluded.
189
Carbon nanotube ozonides were shown to decompose with lifetime < 3 min at r. t.
(Chapter 2). The ability of SWNT to react with amines at different time intervals after
ozonation was studied by XPS. Spectra were recorded at 0 min, i.e. right after, 10, 20, 40
and 60 min after ozonation.
Binding Energy (eV)
300400500300400500
Co
un
ts /
se
c
x1
04
1
2
3
4
300400500
- O
1s
- N
1s
- C
1s
- O
1s
- N
1s
- C
1s
- O
1s
- C
1s
0 min 60 minReference
Figure 5. XPS spectra of SWNT reacted with 2-methoxyethylamine. Labels N 1s mark
nitrogen peaks. Left: a product from reaction with pristine SWNT, a reference. Middle:
amine added right after SWNT ozonation. Right: amine added 60 min after ozonation.
Spectra for reactions of 2-methoxyethylamine with ozonated SWNT right after and
60 min after ozonation are shown in Figure 5. A corresponding reference spectrum is
shown for comparison. All samples were found to have amine attached to its surface.
This means that the covalent attachment of amine to SWNT is independent of the
presence of ozonides. The decomposition of 1,2,3-trioxolanes resulted in the formation of
a carbon nanotube SWNTOn , exhibiting unique oxidative properties not seen in pristine
SWNT.
190
As a part of the experiment, amine/amide enrichment of the SWNT surface was
accomplished by cycling ozonation and reagent addition. The results are summarized in
Figure 6.
Binding Energy (eV)
300500300500
Co
un
ts /
se
c
x1
04
1
2
3
4
5
300500 300500 300500
- O
1s
- N
1s
- C
1s
- O
1s
- N
1s
- C
1s
- O
1s
- O
1s
- N
1s
- C
1s
- O
1s
- N
1s
- C
1s
- C
1s B EDA C
Figure 6. SWNT surface enrichment with 2-methoxyethylamine. Shown are XPS
spectra. Samples C-E went through cycling ozonation - reagent addition. Each section
(A-E) represents a separate experiment. Labels N 1s mark nitrogen peaks; (A) after
reaction with pristine SWNT, a reference, cN 0 at. %; (B) after reaction with ozonated
SWNT, cN 2.8 atomic %; (C) 2 cycles of ozonation – amine addition, cN 4.5 at. %; (D) 3
cycles, cN 6.0 at. %; (E) 4 cycles, cN 7.8 at. %.
Four cycles were made; each one included: SWNT plate (4 x 6 mm buckypaper on
Teflon® substrate) ozonation for 1 min with a stream of O3/O2 gaseous mixture (ca. 1.5
v/v % ozone), a reaction with neat amine (8 uL) for 1 min and a vacuum drying for 3
min. XPS gave ca. a 2 % increase in atomic concentration of nitrogen on the surface of
SWNT with each subsequent cycle. Similar amine/amide enrichment work was done with
IR measurement, results of which are discussed in Chapter 4. Conversion of amines to
191
amides is expected for multiple ozonation cycles, and was demonstrated by formation of
characteristic amide bonds at 1680 cm-1
(Chapter 4). The possibility of nitrogen peak
splitting into two nonequivalent ones was not studied by XPS. More in-depth work is
needed to determine possible conversion of amines to amides during subsequent
ozonation steps. Values of nitrogen atomic % on the surface of SWNTOn were obtained
with survey scans and will need to be refined in future work.
Reagents found to have good reactivity with ozonated SWNT (SWNTOn) are shown
in Figure 7 (see XPS spectra in Appendix C.)
192
NH
H2N NH2
HON
N
NN
NH2
NHNH
O
ONaO
OONa
N
N
O
ONa
NH2
NaO
O
ONa
O
NH2
O
ONa
NH2
NaO
O
O
ONaH2N
HN
N NH2
ONa
O
NH2
ONa
O
H2N
NH2
OH
O
H2N
ONa
O
NH2
S
H2N
O
ONa
O
NH2
ONa
O
H2N
O
NH2
O
ONa
NH2
NH
H2N
NH
Argininate, Na+
Glutaminate, Na+
Asparaginate, Na+
Methioninate, Na+
Lysinate, Na+
Histidinate, Na+
Glycinate, Na+
Glutamate, 2Na+
Aspartate, 2Na+
Alaninate, Na+
Lysine
Wurster reagent
Folic acid, 2 Na+ salt
Guanidine
NH2
OH O
OH Threoninate, Na+
NH2HO
ONa
O
Tyrosinate, Na+
O
ONaH2N GABA, Na+ salt
ONH2
2-methoxyethylamine
Aqueous ammonia
H2NNH2
H2NOH
NH2Isoamylamine
ethanolamine
ethylenediamine
NH3aq.
N
N
HO
SONa
O
OHEPES, Na+
Figure 7. Reagents found to be reactive with ozonated SWNT. All compounds gave
characteristic N 1 s peak in XPS spectra.
193
Compounds with more than one nitrogen, e.g. lysinate, argininate, guanidine,
ethylene diamine, gave stronger signals for N 1s peak. In a separate study, ethylene
diamine was shown to crosslink SWNT, leading to its inability to get debundled upon
bath sonication in ethanol. Such behavior is another indication of a covalent attachment
of reagents to the surface of SWNT.
Possible reaction mechanisms are proposed on Scheme 1. Ozonides are thought to
oxidize amines and thiols, yielding no covalent attachment to SWNT surface.
Scheme 1
OO
OO
OO
OO O
Nu
epoxide
O
oxidoannulene
O O
Nu
1,2,3-trioxolane
Nu = AlkSH, ArSH, R3N, RNH2, R2NH, PR3, guanidine, etc.
Nu
Nu
Nu
SWNT=
OO
epoxide
RNH2
RNH2 (xs)
RNH2
Route A
Route B
RNH3
RNH
RNH O
NH
R
RNH3
RNH2
O
NH
R
H
OO
ONu
NuOno attachment ofNu to sidewall
Route C
194
Reactions with epoxides and oxidoannulenes were shown by XPS and IR to yield
reagent attachment to the SWNT sidewall.
Samples that had poor reactivity with SWNTOn included: isoleucinate, valinate,
serinate, cysteinate, aniline, triethylamine, thiophenol and p-fluorothiophenol. The latter
two compounds produce stabilized radicals, either PhS+ and F-C6H4S
+ or PhSH
+ and F-
C6H4SH+ or a combination of thereof and were expected to have poor attachment. Both
thiols were shown to have attachment to corresponding reference samples, either through
a side reaction or a non-covalent interaction.
Hydrophobic interaction or inefficient ‘washing’ steps are possible. Valinate,
isoleucinate and triethylamine had poor attachment probably due to steric hindrance.
Aniline, having its nitrogen lone pair in conjugation with an aromatic ring was shown to
give a weak signal for N 1s peak, as expected. The acidity constant of PhNH3+ is pKa ~
4.6 (in water) which is much lower than that of amines (pKa > 8).
Samples with uncertain reactivity towards SWNTOn included: uracil, adenine and
2-mercaptopyridine. All compounds have nitrogen atoms with delocalized lone pairs.
HN
NH
O
O
SNH
N
NNH
N
NH2
H
N SH
pKa ~ 9
pKa ~ 3.5
uracil adenine 2-mercaptopyridine
delocalized lone pair
The nature of uracil attachment is not clear. While uracil can be deprotonated at
pH > 9.2, its protonation would require a very strong acid. Lone pairs of nitrogen atoms
of amide-like groups can be protonated at pH ~ 0. In turn, this means that this lone pair is
195
not ‘available’ for an electron transfer to SWNTOn. Poor reactivity with SWNTOn for all
three compounds was expected.
Samples that had no reactivity with SWNTOn included: urea, NaCN, NaBr, TsCl and
CF3CH2OH. Lack of reactivity between urea and SWNTOn seemed reasonable, taking
into account that lone pairs on urea nitrogen atoms are delocalized onto the adjacent
carbonyl. Reasonably good nucleophiles, like CN –
and Br –
, were shown to be unreactive
with SWNTOn, indicating that nucleophilic attack on SWNTOn may not be a
requirement for amine or thiol attachment observed in other experiments.
Trifluoroethanol did not react with ozonated SWNT as expected.
Samples that were shown to have hydrophobic interaction with SWNT included
cysteine and phenylalanine (see spectra in Appendix C). Cysteine (10 mg/200 uL) was
shown to crash out of aqueous solution onto the surface of SWNT film. Apparently, a
thiol group had a superb affinity for carbon nanotubes (Figure 8).
Figure 8. Comparison of SWNT films dipped into aqueous solutions of: (A-B) cysteine
sodium salt (10 mg/200 uL plus 2 eq. of NaOH) and (C-D) cysteine (10 mg/200 uL).
Top: ozonated SWNT. Bottom: reference SWNT
Development of cysteine or thiol based surfactants for SWNT solubilization may be
an interesting area to look into. XPS of samples C and D verified that precipitate is
cysteine. Phenylalanine was observed on both reference and reaction spectra. It is
believed that hydrophobic interaction kept this acid attached to SWNT surface.
196
7.3. Conclusions
A new methodology for SWNT sidewall functionalization has been developed for
amines and thiols and was extended to a class of amino acids. The main requirement for
amino acids to be reactive with ozonated SWNT was the presence of a sufficient amount
of sodium hydroxide to convert all free carboxylic groups to sodium salts, thus freeing up
amino groups for an electron transfer to ozonated SWNT.
Nearly all amino acids got attached to ozonated SWNT, while no attachment was
observed for non ozonated SWNT. A borderline case, where attachment could be either
covalent or hydrophobic or both, included phenylalanine. Cysteine was shown to have a
very good affinity for ozonated SWNT. It precipitated out of an aqueous solution forming
a thin film on the surface of SWNT. Usage of cysteine terminated surfactants for SWNT
encapsulation could be an area for further research. As expected, nucleophiles like CN –
and Br – were not able to get attached to SWNT surface, indicating that electron transfer
is the first step.
197
7.4. Experimental Part
General information
A sonication bath (Fisher Scientific, FS 60) was used for all bath sonications.
MiniVortex (Fisher Scientific) was used mainly for solubilization of amino acids. X-ray
Photoelectron Spectroscopy measurements were performed on PHI Quantera. XPS
samples were prepared either on indium foil (Aldrich) or on polypropylene backed PTFE
membrane filters (Whatman, 0.2 um, WTP type).
Millipore Milli-Q water (18.2 MOhm/cm at 25 C) was used for experiments with
amino acids. A pH meter (Fisher Scientific, Dual Channel pH meter AR 50) and 1N aq.
NaOH (Fisher Scientific) were used for pH adjustments.
All -amino acids used in this work were natural. Adenine, uracil and the majority of
essential amino acids were from US Biological; asparagine, NH4OH, glycine and urea
were obtained from Fisher; cysteine, arginine hydrochloride and guanosine were from
Aldrich; glutamic acid was from Acros; benzene from EMD Chemicals Inc.; ethyl
alcohol (200 proof) was from AAPER Alcohol & Chemical Co.; kimwipes ® were from
Kimberly-Clark.
SWNT plate preparation
SWNT (2-4 mg, SWNT, HipCo, batch 162.8, raw, unpurified, Rice University) were
debundled in benzene (10 mL) with a bath sonicator (Fisher Scientific, FS 60) until there
were no more solid particles present in the suspension, then concentrated to a slurry and
added dropwise to the surface of PFTE filter.
198
Procedure for Wurster reagent
A Teflon plate covered with SWNT film was ozonated for 1 min with a stream of
O3/O2 gaseous mixture (ca. 1.5 v/v % ozone). The plate was dipped into a solution of
TMPD (2.7 mg in 500 uL EtOH) for 15 min, and then washed two times in EtOH (1 mL)
and one time in water (1 mL) for 3 min each.
Procedure for enrichment of 2-methoxyethyl amine on the surface of ozonated
SWNT
A Teflon plate covered with SWNT film was ozonated for 1 min with a stream of
O3/O2 gaseous mixture (ca. 1.5 v/v % ozone). Neat amine (8 uL) was added, SWNT plate
was covered with a jar to prevent amine evaporation, reacted for 1 min, the SWNT plate
was dried under vacuum for 3 min and this cycle was repeated as many times as
necessary.
Procedure for reaction of NaBr, NaCN and CF3CH2OH with ozonated SWNT
A Teflon plate with deposited SWNT film was ozonated for 1 min with a stream of
O3/O2 gaseous mixture (ca. 1.5 v/v % ozone). Neat trifluoroethanol (8 uL) was added,
SWNT plate covered with a jar, reacted for 1 min and dried under vacuum for 3 min
before XPS measurements.
For reactions with salts, SWNT plates were dipped into dilute aqueous solutions of
NaBr and NaCN for 1 min, then dipped two times into deionized water (3 min each),
wiped off excess of water with Kimwipes ® and dried under vacuum for 3 min before
XPS measurements.
199
A general procedure
Deionized water (MilliQ) was used for all aqueous solutions. Solvents and amounts
of reagents are noted in Table 1. To help with solubilization, bath sonicator and
MiniVortex were used as needed. Prepared solutions were stored at 10 C before use.
All reactions and washes were performed in 1.5 mL microcentrifuge tubes
(Eppendorf vials). Pristine SWNTs (HipCo, batch 162.8, Rice University) were bath
sonicated in ethanol for 40 min, solvent was filtered out on 0.2 um PTFE filter
(polypropylene backed, Whatman, WTP type) to form an equally distributed SWNT film
on its surface. Filter was cut into 4 x 6 mm sections for use in chemical reactions.
A 4 x 6 mm section of SWNT film on Teflon ®
substrate was subjected to a steam of
O3/O2 gaseous mixture (ca. 1.5 v/v % ozone) for 1 min. The film was removed from the
ozonation chamber and submerged into a solution of interest (see Table 1 for details) for
3 min. Film was taken out of a reaction solution, excess of liquid was removed with
Kimwipes.® Unless otherwise noted in Table 1, washed reaction plate two times in
deionized water (1 mL) for three minutes each. Plates were dried under vacuum for 15
min before loading into XPS machine.
200
Table 1. Preparation of reagents for XPS experiment
Reagent FW mg mmol NaOH
eq
1N NaOH
uL
Other
solvent
H2O
uL
Total
uL
Alanine 89.09 5 0.0561 1 56.1 43.9 100
Cysteine 121.16 10 0.0825 2 165.1 30.0 200
Cysteine No NaOH 121.16 10 0.0825 0 0.0 200.0 200
Aspartic acid 133.1 5 0.0376 2 75.1 24.9 100
Glutamic acid 147.13 10 0.0680 2 135.9 0.0 136
Phenylalanine 165.19 5 0.0303 1 30.3 69.7 100
Glycine 75.07 5 0.0666 1 66.6 33.4 100
Histidine 155.16 5 0.0322 1 32.2 67.8 100
Isoleucine 131.17 5 0.0381 1 38.1 61.9 100
Lysine 146.19 5 0.0342 1 34.2 65.8 100
Lysine No NaOH 146.19 5 0.0342 0 0.0 100.0 100
Methionine 149.21 5 0.0335 1 33.5 66.5 100
Asparagine 132.12 5 0.0378 1 37.8 62.2 100
Proline 115.13 5 0.0434 1 43.4 56.6 100
Glutamine 146.15 5 0.0342 1 34.2 65.8 100
Arginine-HCl 174.2 5 0.0287 2 57.4 42.6 100
Serine 105.09 5 0.0476 2 95.2 4.8 100
Threonine 119.12 5 0.0420 1 42.0 58.0 100
Valine 117.15 5 0.0427 1 42.7 57.3 100
Tyrosine 181.19 5 0.0276 2 55.2 44.8 100
GABA 103.1 5 0.0485 1 48.5 51.5 100
urea 60 10 0.1667 0 0.0 100.0 100
uracil a 112 5 0.0446 0 100.0 40 uL
CH3CN
- 140
adenine a 135.1 5 0.0370 0 100.0 40 uL
CH3CN
- 140
NH4OH aq. solution b 35 20 uL - 0 0.0 H2O 80.0 100
Guanidine·HCl 95.53 10 0.1047 1 104.7 95.3 200
2-Mercaptopyridine c 111.17 10 0.0900 0 0.0 CH3CN 0.0 200
HEPES 238.3 10 0.0420 1 42.0 158.0 200
Folic acid d 441.4 10 0.0227 2 45.3 154.7 200
a First wash 40 : 100 uL (CH3CN : 1N aq. NaOH), then 2x water 1 mL;
b - Dilute 5x NH4OH
conc; dip teflon plate into this solution; c - First wash 1 mL CH3CN then 2x H2O;
d - First wash
200 uL of 1N NaOH, then 2x H2O (1 mL).
201
7.5. References and Notes
There are no references for this chapter.
Part II
203
Chapter 1
Photorearrangement of -Azoxy Ketones and Triplet Sensitization of
Azoxy Compounds
204
1.1. Abstract
Ph
O
NN
O
R' O NN
R'Ph
O
hv
a R' = t-Bu
b R' = Ph4 8
A recent review on the radical chemistry of the azoxy group1 revealed that the nature
of -azoxy radicals remains unclear. This chapter presents the generation of -azoxy
radicals under mild conditions by irradiation of -azoxy ketones 4a,b. These compounds
undergo -cleavage to yield radicals 5a,b, whose oxygen atom then recombines with
benzoyl radicals to produce presumed intermediate 15. Formal Claisen rearrangement
gives -benzoyloxyazo compounds 8a,b, which are themselves photolabile, leading to
both radical and ionic decomposition. The ESR spectrum of 5a was simulated to extract
the isotropic hyperfine splitting constants, which showed its resonance stabilization
energy to be exceptionally large. Azoxy compounds have been found for the first time to
be good quenchers of triplet excited acetophenone. The main sensitized photoreaction of
7Z in benzene is deoxygenation. The principal direct irradiation product of 4bZ and
model azoxyalkane 7Z is the E isomer, whose thermal reversion to Z is much faster than
that of previously studied analogues.
1.2. Introduction
Despite its occurrence in a number of biologically active molecules,2-6
the azoxy
functional group has received less attention than its lower oxidation state analogue, the
205
aliphatic azo group.7 The latter one is a widely used source of free radicals, but as pointed
out in a recent review,1 the radical chemistry of azoxy compounds is sparse.
The generation of -azoxy radical 2, also known as hydrazonyl oxide 3, has been
proposed in the literature, but no product attributable to this radical could be found.
Perester decomposition yielded -azoxy radical 1, which in turn underwent fragmentation
to ethylene with a rate constant below 2 × 105 s
-1 at 120 C (Scheme 1).
8 On the other
hand, bromination of azoxy compounds is known to occur at the distal carbon (i.e. away
from azoxy oxygen), presumably via -azoxy radicals.3, 9
The goal of the present work
was to generate -azoxy radicals under mild conditions in hopes of learning more about
their chemistry.
Scheme 1
NN
O
NN
O
H2C CH2
NN
O
31 2
1.3. Results
Norrish Type I photochemical cleavage of -azoxy ketone 4, a previously unknown
structural type, was chosen as the path to -azoxy radicals. Photolysis of some phenones
and benzyl ketones yields alkyl radicals in solution;10-14
hence, it is not unreasonable that
4a-c might also exhibit -cleavage (Scheme 2).
206
Scheme 2
R1
O
NN
O
R2hv
a: R1 = Ph, R2 = t-Bu
b: R1 = Ph, R2 = Ph
4
NN
O
R2
R1
O
c: R1 = PhCH2, R2 = t-Bu
5
Photochemistry of 4a. The extinction coefficients of 4a at two photochemically
useful wavelengths are included in Table 1, and the spectra can be found in Appendix D.
Table 1. UV Extinction Coefficients of Azoxy and Azo Compounds
compd (313 nm) (366 nm)
4a 128 16
4bZ 830 22
8a 4.7 26
8b 212 80
6 32 1.9
7Z 262 7.7
PhCOMe 41 6.7
Ph
O
NN
O
NN
O
Ph
O
NN
O
Ph
NN
O
Ph
313 = 32 313 = 262
313 ~ 40
313 = 830 313 = 128
313 ~ 40
7Z6
207
The bichromophoric molecule absorbs twice as strongly as the sum of its two
individual chromophores, represented by acetophenone and azoxy-tert-butane 6.15
Irradiation of 4a in benzene at 313 nm and 25 C caused clean rearrangement to an
unexpected product, azoester 8a (Scheme 3), whose extinction coefficients are included
in Table 1. The quantum yield for appearance of 8a was 0.02 at 23, 60, and 100 C, and
the reaction even proceeded at -78 C ( was established with DBH actinometry).
Scheme 3
Ph
O
NN
O
R' O NN
R'Ph
O
hv
a R' = t-Bu
b R' = Ph
4 8
The disappearance rate of 4a was not diminished by the inclusion of 0.1 M
biphenyl16
as quencher, but 0.1 M 1,3-cyclohexadiene, a quencher of much lower triplet
energy than biphenyl, decreased the conversion of 4a to 8a by 18.1%. Table 2
summarizes triplet state energies of acetone, acetophenone, biphenyl and
1,3-cyclohexadiene (CHD) used in this work.
Table 2. Triplet energies and lifetimes for excited acetone, acetophenone, Ph-Ph and
CHD.
Compd ET, kcal / mol Lifetime T, us Reference
acetone 80 17
PhCOMe 73.6 0.86 this work
Ph-Ph 65.5 16
CHD 54 1.3 18, 19
208
To determine whether -cleavage of the ketone moiety was on the pathway from 4a
to 8a, the benzoyl radicals were trapped with tert-butyl thiol.20
Irradiation of 4a at 313
nm with 0.16 M tert-butyl thiol afforded benzaldehyde in 25% yield, as determined by
NMR. A corresponding reduction in the amount of 8a was observed. No product from
trapping of 5a could be identified.
The photolysis of 4a was also investigated by ESR spectroscopy. Irradiation of 4a in
toluene-d8 at + 8 C with a full 500 W mercury arc lamp gave a weak but structured ESR
signal. Suspecting that the signal was due to radical 5a, this intermediate was generated
independently by irradiation of di-tert-butyl peroxide21, 22
containing
tert-butyl(ONN)azoxy-2-propane 9 at -78 C. A stronger signal ~ 60 G wide containing
about 37 lines was observed and was very similar in appearance to the one from 4a (See
ESR spectra in Appendix D).
Scheme 4
The GC trace of the irradiated solution showed several peaks, none of which were in
the right region for the C-C dimer of 5a.8 To support the assignment of the ESR signals to
5a, the experiment was repeated with 9 at -50 C. The superior resolution (see Appendix
D) allowed isotropic simulation (coefficient of determination r2 = 0.926) with the
program WinSim,23
which led to the following splitting constants: N1 13.75 G, N2 1.91
G, 3H 5.64 G, 3H 5.12 G. A second radical was present in the spectrum but since its
intensity was much lower than that of 5a, this minor species was not studied further.
209
Photochemistry of 4bZ. As shown in Table 1, 4bZ absorbs about three times more
strongly at 313 nm than the sum of its two chromophores, which are represented by
acetophenone and 7Z.24-26
Irradiation of 4bZ at 313 nm both at 25 and -78 C gave the
intensely yellow 8b, but the major reaction was Z E photoisomerization about the
azoxy double bond.27, 28
The E isomer (4bE) reverted to the Z isomer 4bZ with a half-life
of 14 min at 50 C. This behavior supports the structural assignment of 4bE.
Scheme 5
The chemical shift changes in toluene-d8 upon Z-E isomerization (4bZ Me2 1.716
ppm vs 4bE Me2 at 1.377 ppm) are similar to those of 7Z 7E (1.472 vs 0.943 ppm).
The magnitude of the upfield shift upon Z E isomerization is larger than that in the
methyl and ethyl analogues.28
Irradiation of 4bZ at ambient temperature and 313 nm in
acetone-d6 as solvent and triplet sensitizer29-32
exhibited the same rearrangement quantum
yield as in toluene-d8. Irradiation of 4bZ in toluene at -53 C in the ESR cavity led to a
weak signal about 50 G wide consisting of 13 peaks, which were much broader than
those from 4a. This spectrum was definitely not due to phenyldiazenyl radicals33
arising
from secondary photolysis of 8b. Instead, the similarity of the major nitrogen coupling to
that of 5a suggests that this radical is 5b.
210
Irradiation of 4c. This compound was designed to decarbonylate photochemically
but it proved to be quite photostable. Thus 4c upon irradiation with an unfiltered 500 W
mercury lamp gave neither 8c nor the E azoxy isomer but instead slowly decomposed to a
mixture of many products.
Triplet Sensitization of Model Azoxy Compounds. Because the
photorearrangement of 4bZ 4bE in acetone-d6 is the first reported triplet sensitized
reaction of an azoxy compound, we examined the azoxy group as a triplet quencher.
Acetophenone was chosen as the donor because of its high triplet energy (73.6
kcal/mol),17
its structural similarity to the ketone moiety of 4, and because it is known to
phosphoresce in solution.34
Azoxy-tert-butane 6 (0.012 M) and phenylazoxy-tert-butane
7Z (0.0021 M) were found to be strong quenchers of acetophenone emission intensity.
The quenching rate constant in isooctane was obtained by quenching the triplet lifetime
of acetophenone, giving kq values of 9.8 × 108 M
-1 s
-1 for 6 and 4.4 × 10
9 M
-1 s
-1 for 7Z.
Since 6 and 7Z proved to be good triplet quenchers, the triplet sensitized
photochemistry of 7Z was investigated. Two degassed and sealed NMR tubes were
prepared, one containing 0.0191 M 7Z in toluene-d8 and the other containing the same
concentration of 7Z plus 0.437 M acetophenone. The high concentration of acetophenone
was needed to make it the main light absorber because the extinction coefficient of 7Z
exceeded that of acetophenone at 366 nm (cf. Table 1). Irradiating both tubes in parallel
for 4 h at 366 nm led to the same products but in different amounts (cf. Table 3).
211
Table 3. Relative GC Peak Areas in the Direct and Sensitized Photolysis of 7Z at 366 nm
in toulene-d8
a Conditions: -5 C, 4 h, 62% of 7Z left unreacted as determined by an internal standard.
b Conditions: PhCOMe, -5 C, 4 h, 58% of 7Z left unreacted.
Whereas Z E isomerization was dominant under direct irradiation, this process is
actually negligible with acetophenone present. We calculate that 4.6% of the incident
light was absorbed directly by 7Z in the sensitized experiment, as compared to 16% in the
direct irradiation, where the overall absorbance was only 0.073. Since the contribution of
direct photolysis in the sensitized reaction was 4.6 / 16 = 29%, the bulk of the 7E seen in
the latter case arose by direct irradiation of 7Z.
The photoisomer 7E was thermally labile, exhibiting a half-life for reversion to 7Z of
2.3 h at 25 C. Due to steric repulsion, 7E is much more labile than its ethyl analogue,
which reverts completely to the Z isomer in 5 h at 110 C.28
On the other hand, 7E is
slightly more stable than 10Z, whose half-life for conversion of 10E is 1.36 h at 25 C.35
The set 7E, 10Z provides a rare opportunity to compare the thermal isomerization rate of
a cis azoalkane with that of its related azoxy compound.
Deoxygenation was the main process observed under triplet sensitization but the fate
of oxygen is unknown. The structure of the deoxygenated product, phenylazo-tert-butane
10Z,E, was proven by comparison with an authentic sample.36, 37
The sensitized
212
irradiation was repeated in C6D6 to rule out hydrogen / deuterium abstraction from
toluene-d8, but the outcome was the same.
Because direct irradiation of azoxy compounds can led to oxadiaziridines27, 38-40
as
well as Z E isomerization,27, 28
it was important to verify the structure of 7E.
Surprisingly, no 1H NOE was seen for the tert-butyl protons and the ortho H's of the
aromatic ring. This negative result prompted for an examination of the NOE of a model
compound 10Z, whose trans-cis photoisomerization is well-known.35
Since 10Z also
failed to exhibit NOE, experimentally measured 15
N chemical shifts were compared to
those calculated theoretically,41
as shown in Table 4. The proximal and distal nitrogens
were assigned unambiguously by 1H-
15N HMBC.
Table 4. 15
N Chemical Shiftsa
7Z calc 7Z obs 7E calc 7E obs 12 calc
Nb 1.2 -20.2 30.8 1.4 -115.4
Nc -39.7 -52.1 -23.9 -42 -164.3
a ppm from nitromethane standard.
b Nitrogen proximal to the tert-butyl group.
c N-O.
Although the calculated shifts are 10-20 ppm downfield from the observed value for
7Z, the predicted direction and magnitude of the changes upon Z E isomerization lie
within 8 ppm of the observed value. The chemical shifts of oxadiaziridine 12 are
calculated to fall drastically upfield from those of 7Z,E, ruling out 12 as the photoisomer.
Additional support for the structure of 7E was provided by its independent synthesis
from 10Z. A solution of 10E in toluene-d8 was irradiated in an NMR tube with a 150 W
213
xenon lamp until the 10Z:10E ratio was 0.35. The mixture was oxidized in situ with
MCPBA, resulting in the clean conversion of all 10Z to 7E within 10 min. The tert-butyl
group of 7E ( = 0.943 ppm) exhibited a sizable upfield shift in toluene-d8 relative to that
of 7Z (1.472 ppm), just as in the azo precursors (10E 1.311; 10Z 1.090).28
After 10Z had
disappeared, the oxidation of 10E continued at a much slower rate and oxygen was
introduced proximal15
to the tert-butyl group to yield major product 11 accompanied by
~ 6% 7Z. The reversal of oxidation regiochemistry between 10E and 10Z is noteworthy
and apparently unprecedented since the only acyclic cis azoalkane oxidized previously
was symmetrical.27
Secondary Photolysis of Azoesters. As seen in Table 1, the azoester products 8a,b
absorb enough light to undergo secondary photoreactions and in fact these can complicate
the mechanistic interpretation of azoxy ketone photolysis. Irradiation of authentic 8a,b
yielded nitrogen plus a mixture of products, as shown below (Scheme 6). In 8b only,
decomposition was accompanied by isomerization to the cis azo isomer 13. Because it
was initially surprising to find benzoic acid as the major product of 8a,b, the irradiation
of 8a was repeated in MeOH-d4. The benzoyloxy group was found partially replaced by
CD3O to yield 14, consistent with ionic dissociation of the substituent to the azo
group.42, 43
214
Scheme 6
1.4. Discussion
Mechanism of 4 to 8. The simplest mechanism for the photorearrangement begins
with -cleavage to benzoyl radical plus -azoxy radical 5a. These radicals recombine at
azoxy oxygen to afford intermediate 15, which then rearranges to the final product 8
(Scheme 7). However, careful scrutiny of the reaction by low-temperature NMR failed to
reveal the presence of 15.
215
Scheme 7
A literature search revealed no published structures quite like 15 44-46
and an attempt
to generate 15 from acetone tert-butyl hydrazone and benzoyl peroxide at 0 C led
cleanly to 8a and unknown NMR peaks not consistent with 15a. Having failed to observe
this postulated intermediate by NMR, the activation energy for Claisen rearrangement
was calculated theoretically at the B3LYP/6-31G* level by a collaborator.41
Table 5
shows that the activation enthalpy lies in the same range as that for the Cope
rearrangement.47
Radical delocalizing substituents R1 and methyl groups R
2 and R
3
decrease H but not enough to explain an inability to detect 15. Homolytic dissociation
of 15 was calculated to require more energy than the figures in Table 5. Perhaps 15
absorbs UV light as does a peroxide or N-chloroamine and 15 8 occurs
photochemically.
216
Table 5. Calculated Activation Energy for Claisen Rearrangement of 15 to 8*
R1
R2
R3
HF, hartrees
TS,
hartrees H ,
kcal/mol
H H H -338.465 327 5 -338.419 204 8 28.9
H2C=CH- H H -415.868 578 8 -415.828 300 7 25.3
Ph H H -569.530 707 4 -569.490 234 4 25.4
H H Me -417.103 647 5 -417.069 106 5 21.7
H Me Me -456.418 324 7 -456.387 535 6 19.3
H2C=CH- Me Me -533.822 631 1 -533.789 195 6 21.0
* - calculation results courtesy of William B. Smith, Texas Christian University, Fort
Worth, Texas.
Lacking experimental evidence for 15, a verification was needed that -cleavage of
4a really is the first step of the reaction. Irradiation of this azoxy ketone with tert-butyl
thiol led to benzaldehyde20
but it was not clear initially whether it came from 4a or from
secondary photolysis of 8a. However, a plot of benzaldehyde versus time showed a
sizable initial slope while a similar plot of acetone, a decomposition product of 8, showed
an initial slope of zero. If benzaldehyde arose solely from secondary photolysis of 8a,
both plots would have exhibited an initial slope of zero. The ESR experiment with 4a
also supports initial -cleavage since essentially the same spectrum was observed from
4a and from irradiation of 9 with di-tert-butyl peroxide. The nitrogen splitting constants
extracted from the latter spectrum (13.75, 1.91 G) are in accord with those of known
heavily substituted hydrazonyl oxide radicals (e.g. for t-Bu2C=N2-N
1(O•)-Ph, a(N
1) =
12.1 G, a(N2) = 2.7 G
48 and for t-BuPhC=N
2-N
1(O•)-t-Bu, a(N
1) = 12.5 G, a(N
2) = 1.7
G.49
The proton splittings of 5a are exceptionally small, reflecting the high degree of
radical stabilization that one might expect for this nitroxyl structure.50
217
There remains the possibility that -cleavage is a side reaction not on the path from
4a to 8a. However, in a t-BuSH trapping experiment where 60% of 4a reacted, the yield
of PhCHO was 25% while that of 8a and its photolysis products was 21% and 15%,
respectively. Thiol trap t-BuSH was found diminishing the amount of 8a by an amount
close to the yield of PhCHO, thus placing PhCO• on the reaction pathway. Since t-BuSH
failed to trap 5a, the only evidence for this radical comes from the ESR experiment.
The low quantum yield for photorearrangement of 4a,b may be attributed to efficient
recombination of benzoyl radicals at the original site of attachment. However, the spin
density of 5a is higher at oxygen than at carbon, as shown by the calculated spin densities
below (calculated by a collaborator). Moreover, cage recombination of other "allylic"
radicals occurs at both ends.51
Another explanation for the low quantum yield is radiationless decay of excited 4a,b,
which could be accompanied by Z E isomerization of the azoxy group. To explore the
importance of these factors, a brief investigation of the azoxy group as a triplet energy
acceptor was conducted.
Triplet Energy Transfer to Model Azoxy Compounds. Azoxyalkanes were found
to be surprisingly rapid triplet quenchers. The quenching rate constant of acetophenone
triplets was in the range of 109 M
-1s
-1, which is similar to that for azo-tert-butane.
52 The
"triplet sensitized" deoxygenation of azoxybenzene53
was shown to be a case of chemical
218
sensitization;54
that is, ketyl radicals derived from the sensitizer reduced the azoxy
group.55
For this reason, acetophenone-sensitized deoxygenation of 7Z in toluene (cf.
Table 3) was initially suspected to follow the same pathway. However, the product
distribution was unchanged in C6D6 solvent, suggesting that triplet sensitized
deoxygenation56
is a true photoreaction of 7Z. In contrast, direct irradiation gave cis-trans
isomerization, already a well-established process.27, 28, 57
Investigating the triplet energy
of azoxy compounds and the fate of oxygen are potential topics for further research.
Although the azoxy group quenches ketone triplets intermolecularly, it is
inappropriate to discuss 4a,b in terms of individual chromophores because the enhanced
UV absorption (Table 1) indicates mixing of electronic states. The experimental facts are
that upon direct irradiation in benzene-d6 neither compound undergoes deoxygenation,
only 4bZ exhibits cis-trans isomerization, and both compounds undergo inefficient
-cleavage of the benzoyl group.
Photolysis of -Benzoyloxyazoalkanes. The initial products 8a,b are themselves
photolabile, giving rise to a product mixture dominated by benzoic acid and acetone. Part
of the likely mechanism, shown in Scheme 8, involves the usual nitrogen loss from
azoalkanes, presumably via thermolysis of the labile cis isomer7 followed by plausible
reactions of the formed tert-butyl and acyloxyalkyl radicals 16. Published information on
1-acyloxyalkyl radicals is sparse and mainly concerns their inter- and intramolecular
addition to alkenes.58, 59
However, Wille recently reported that photolysis of 17 gave
cyclohexanone in 33% yield by fragmentation of radical 18,60
exactly analogous to the
behavior of 16.
219
Scheme 8.
The major photolysis product of 8a,b is benzoic acid but Scheme 8 cannot explain its
presence. Levi and Malament reported that acyclic azoalkanes containing -chloro or
-acyloxy groups underwent heterolytic cleavage even in nonpolar solvents.42, 43
The
same unusual mechanism can be applied to formation of benzoate and the known highly
stabilized -azo cation 19.61
Most likely, benzoic acid arises by proton transfer from
adventitious water to benzoate, as the ortho protons of the acid were clearly visible by
NMR of the unopened, degassed, sealed tubes. The ionic mechanism cannot operate
exclusively, even though acetone could arise when 19 traps water, because it cannot
rationalize the obviously free-radical-derived products isobutane, benzaldehyde,
benzophenone, and isopropyl benzoate. Thus 8a undergoes competitive C-N homolysis
and C-O heterolysis. To support the ionic mechanism, 8a was irradiated in CD3OD and
monitored by 1H NMR (Scheme 9). The starting material disappeared over 19 h and gave
220
mainly 14, as proven by comparison with an authentic sample of the nondeuterated
analogue. On further irradiation, 14 also decomposed.
Scheme 9
Irradiation of 8b in C6D6 caused azo trans-cis isomerization, in accord with the
known photochemistry of phenylazoalkanes.36
Several of the products were similar to
those of 8a, again suggesting a competition between ionic and homolytic decomposition.
However, the yield of benzophenone and benzaldehyde from 8a was much higher than
that from 8b, indicating that that the C-N homolysis pathway is more important in 8a (cf.
Scheme 8). The -azo cation from 8b is surely more stabilized than 19, favoring ionic
decomposition of 8b.
1.5. Conclusions
Despite our failure to identify any products from 5a,b, their involvement in the
photorearrangement of -azoxy ketones 4a, 4bZ to 8a, 8b and their observation by ESR
shows that sterically unhindered -azoxy radicals are viable intermediates. The small
hydrogen hyperfine splittings in the ESR spectrum of 5a indicate a very high degree of
resonance stabilization. Direct or acetone sensitized irradiation of 4bZ also induces azoxy
Z E isomerization, but in model compound 7Z acetophenone triplets cause
deoxygenation without Z E isomerization. Azoxy compounds are surprisingly rapid
quenchers of acetophenone triplets. In both 4bZ and 7, steric repulsion causes thermal
221
reversion of the E azoxy isomer to be much faster than in previously reported
homologues. Azoesters 8a,b undergo photochemical C-O heterolysis in competition with
C-N homolysis to 1-acyloxy radicals, which fragment to ketones plus acyl radicals.
1.6. Experimental Part
2-tert-Butyl(ONN)azoxy-2-benzoylpropane, PhCOCMe2-N=N(O)-Bu-t (4a). A
solution of nitroso-tert-butane dimer (250 mg, 1.43 mmol, 0.8 equiv) in CH3CN (10 mL)
was stirred for 3.5 h at 25 C in the dark to allow dissociation to monomer. Meanwhile, a
solution of 2-benzoyl-2-aminopropane hydrochloride62
(795 mg, 3.98 mmol, 1 equiv) in
1.5 N aq HCl (12 mL, 4.5 equiv) was added dropwise to a suspension of Ca(OCl)2 (1.424
g, 5.97 mmol, 60 wt %, 1.5 equiv) in CH2Cl2 (24 mL) and water (24 mL) at 5 C. After
the mixture was stirred for 1 h at 5 C, the organic layer was separated and the water
layer was extracted with CH2Cl2 (2 × 8 mL). The combined organic phase was dried over
MgSO4, filtered, and concentrated to yellow oil (0.806 g, 87% yield). The crude N,N-
dichloroamine was light sensitive and was used immediately in the next step. 1H NMR
(CDCl3) 8.23 (m, 2H), 7.56 (m, 1H), 7.45 (m, 2H), 1.74 (s, 6H). Following Nelson et
al.,24
KI (298 mg, 1.79 mmol) was added to the above nitroso-tert-butane solution at 25
C, the temperature was lowered to 0 C, and the N,N-dichloroamine (416 mg, 1.79
mmol) in MeCN (5 mL) was added. The mixture was stirred for 2-3 h at 5 C, then the
temperature was gradually raised to 25 C and the mixture was stirred overnight. Water
(50 mL) and ether (25 mL) were stirred in and the water layer was separated and
extracted with ether (2 × 8 mL). The combined organic layer was extracted with
222
sufficient aq Na2S2O3 (~ 312 mg, 1.97 mmol) to change the color from dark brown to
light green. After drying over MgSO4 and removal of the solvent, the crystalline product
was purified by silica gel chromatography, eluting with 4:1 hexane:ethyl acetate, Rf 0.52.
Solvent evaporation yielded 273 mg (61%) of -azoxy ketone 4a, mp 70.5-71 C. Further
purification was effected by recrystallization from a small amount of MeOH. 1H NMR
(CDCl3) 7.84 (m, 2H), 7.46 (m, 1H), 7.35 (m, 2H), 1.59 (s, 6H), 1.31 (s, 9H). NMR
(C6D6) 7.99 (m, 2H), 7.08 (m, 1H), 6.99 (m, 2H), 1.64 (s, 6H), 1.07 (s, 9H). NMR
(toluene-d8) 7.89 (m, 2H), 7.10 (m, 1H), 7.01 (m, 2H), 1.59 (s, 6H), 1.09 (s, 9H). 13
C
NMR (CDCl3) 199.7, 135.0, 132.2, 127.78, 127.72, 76.1, 68.6, 27.6, 22.5. NMR
(toluene-d8) 198.2, 136.0, 131.8, 128.2, 127.7, 75.8, 68.7, 27.4, 22.6.
2-Phenyl(ONN)azoxy-2-benzoylpropane (4bZ), PhCOCMe2-N=N(O)-Ph.
2-Benzoyl-2-(N,N-dichloroamino)propane was reacted with 1.1 equiv of nitrosobenzene
in MeCN as described above, except that the nitroso dimer dissociation step was omitted
because nitrosobenzene exists largely as the monomer. The product was purified on silica
gel, eluting with 5:1 hexane:ethyl acetate, Rf 0.43. Vacuum drying afforded 325 mg
(68%) of product that was further purified by recrystallization from hot hexane (2 mL),
mp 56-58 C. 1H NMR (toluene-d8) 8.02 (m, 2H), 7.86 (m, 2H), 6.80-6.96 (m, 6H),
1.72 (s, 6H). 13
C NMR (toluene-d8) 197.5, 147.3, 135.5, 132.2, 131.6, 128.7, 128.2,
127.9, 122.2, 69.6, 22.9.
2-Amino-2-phenylacetylpropane Hydrochloride, PhCH2-COCMe2-NH3Cl.
To a mixture of 3-methyl-1-phenyl-2-butene (0.7 g, 4.79 mmol) and isopentyl nitrite
(0.77 mL, 5.75 mmol) cooled to 5 C was added dropwise concentrated HCl (0.96 mL,
223
11.5 mmol, 12 N). After the solution was stirred for 15 min, the nitroso chloride dimer
precipitated as a greenish solid. This product was washed with a small portion of warm
acetone, dried in vacuo, and used in the next step without further purification. Yield 0.50
g (49%) of colorless crystals, mp 133-134 C (lit. mp 136-137 C).63
The nitroso dimer
(0.25 g, 1.18 mmol) was mixed with MeOH:EtOH (3 mL:3 mL) and the solution was
stirred for 24 h under under 10 psi of NH3 at 45 C,64
causing eventual dissolution of the
dimer. The solvent was evaporated and the solid was dissolved in 50 mL of 6 M HCl by
heating to 50 C. This solution was extracted with ether, and then the aqueous phase was
made alkaline with sodium carbonate (note: CO2 evolution!). Upon raising the pH to 10,
the color changed from yellow to blue/green. The precipitate was dissolved in ether and
dried over Na2SO4, and the solvent was evaporated. The residue (190 mg) was found by
NMR to be a mixture of oxime and ketone. A solution of 6 N aq HCl (10 mL) was added
and the mixture was stirred for 1.5 h at 50 C. The solvent was evaporated and the
residue dissolved in a small amount of isopropyl alcohol by heating to 50 C. The salt
was precipitated by addition of ether, filtered, and dried in vacuo. Yield 91 mg (36%) of
PhCH2-COCMe2-NH3Cl, mp 134-141 C. 1H NMR (MeOH-d4) 7.23-7.34 (m, 5H),
1.95 (s, 2H), 1.64 (s, 6H). 13
C NMR (MeOH-d4) 207.0, 134.8, 130.9, 129.7, 128.3,
63.4, 43.1, 23.1.
2-tert-Butyl(ONN)azoxy-2-phenylacetylpropane, PhCH2-COCMe2-N=N(O)-Bu-t
(4c), was prepared in the same manner as PhCO-CMe2-N=N(O)-Bu-t. Purification on
silica gel eluting with 4:1 hexane:ethyl acetate (Rf 0.51) yielded 32 mg of clear, oily
product (42% based on PhCH2-CO-CMe2-NH3Cl). 1H NMR (toluene-d8) 7.00-7.21 (m,
224
5H), 3.45 (s, 2H), 1.34 (s, 6H), 1.29 (s, 9H). 13
C NMR (toluene-d8) 203.8, 137.5, 130.3,
128.4, 126.7, 76.1, 69.3, 43.5, 27.8, 21.5.
Azoxy-tert-butane was made according to the method of Freeman15
and fractionally
distilled twice, bp 66 C/40 mm. 1H NMR (C6D6) 1.40 (s, 9H), 1.35 (s, 9H).
13C NMR
(C6D6) 76.39, 57.83, 28.25, 25.69.
tert-Butyl(O,N,N)azoxy-2-propane, i-Pr-N=N(O)-Bu-t,8 was synthesized by a
modified literature procedure.65
Nitroso-tert-butane dimer (217.5 mg, 1.25 mmol) in
absolute EtOH (2.5 mL) was stirred for 3 h at 25 C in the dark. In a separate vessel,
i-PrNHOH·HCl (290 mg, 2.6 mmol) was added in one portion to a solution of KOH (154
mg, 2.75 mmol) in absolute EtOH (2.5 mL). After brief stirring, the ethanolic suspension
of i-PrNHOH was added to a dark blue solution of t-BuNO. The mixture was stirred for
2 h at 25 C and for 16 h at 38 C, after which it was diluted with 1 N aq HCl (5 mL),
extracted with pentane (10 × 3 mL), and dried over Na2SO4. The solvent was removed by
careful bulb-to-bulb distillation at 50 mmHg, cooling the receiver in a dry ice-2-propanol
bath. Since the azoxyalkane product was highly volatile, some loss was unavoidable. The
residue was distilled by reducing the pressure and the product was purified by preparative
HPLC on a silica gel column, eluting with 10:1 pentane:ethyl ether. Bulb-to-bulb
distillation at 50 mmHg removed the solvent, after which the residue was distilled at
0.001 mm to a trap at -196 C. Yield 55 uL. GC analysis of the collected solvent
indicated that it contained approximately 67 uL of i-Pr-N=N(O)Bu-t. 1H NMR (C6D6)
4.32 (septet, 1H, J = 6.4 Hz), 1.36 (s, 9H), 1.15 (d, 6H, J = 6.4 Hz). 13
C NMR (C6D6)
75.5, 51.0, 28.1, 19.5.
225
t-Bu-N=N(O)-Ph (7Z) was made according to Sullivan et al.26
and distilled, bp 44-
62 C / 0.001 mm. Further purification was effected by silica gel chromatography, eluting
with 9:1 hexane:EtOAc, Rf 0.54. The pure compound retains a yellow coloration. 1H
NMR (toluene-d8) 8.13 (m, 2H), 6.90-7.02 (m, 3H), 1.47 (s, 9H). 13
C NMR (toluene-d8)
149.6, 131.4, 128.9, 122.8, 59.2, 26.3. 15
N NMR ( vs MeNO2) -52.1 (N-O), -20.2. The
15N HMBC experiment was optimized for a long-range N-H coupling of 3 Hz.
t-Bu-N=N(O)-Ph (7E): 1H NMR (toluene-d8) 6.92-7.03 (m, 3H), 6.78 (m, 2H),
0.94 (s, 9H). 15
N NMR ( vs MeNO2) - 42.0 (N-O), 1.4.
2-Phenylazo-2-benzoyloxypropane (8b), PhCOO-C(Me)2-N=N-Ph. To freshly
distilled PhNHNH2 (3.85 g, 35.6 mmol) and water (29 mL) at 5 C was added acetic acid
(0.95 mL, 16.6 mmol), then dropwise 5.8 mL of acetone. After the solution was stirred
for 1.5 h, the air-sensitive hydrazone was quickly filtered off, washed twice with ice
water, and dried in vacuo. Yield 90%. NMR (C6D6) 1.04 (s, 3H), 1.75 (s, 3H), 6.30-
6.50 (br, 1H), 6.72-7.30 (m. 5H). To the freshly prepared hydrazone (1.446 g, 9.77 mmol)
in CH2Cl2 (8 mL) at -78 C was added dropwise t-BuOCl (1.219 g, 11.24 mmol). The
mixture was stirred for 2.5 h as the temperature rose to 25 C. The solvent and tert-butyl
alcohol byproduct were evaporated and the residual Ph-N=N-CMe2-Cl was dissolved in
benzene (5 mL). Because neat Ph-N=N-CMe2-Cl decomposes within 30 min at 25 C,
this freshly prepared solution was added dropwise to silver benzoate (1.79 g, 7.80 mmol)
in benzene (10 mL) at 5 C. The heterogeneous mixture was stirred for 1 h at 5 C, then
at 25 C for 3 days. The azoester suspension was filtered through a cotton plug and then
through filter paper and the filtrate was concentrated to deep yellow oil. The crude
226
product was purified on silica gel, eluting with 9:1 hexane/ethyl acetate, Rf 0.47. Yield
1.71 g (81%) of deep yellow solid with mp 48-49 C. The compound crystallized from
hexane after maintaining the solution at -20 C for several weeks. 1H NMR (C6D6) 8.26
(m, 2H), 7.79 (m, 2H), 7.03-7.16 (m, 6H), 1.78 (s, 6H). 13
C NMR (C6D6) 165.3, 152.3,
133.2, 132.2, 131.4, 130.5, 129.5, 128.9, 123.4, 102.6, 25.2.
2-tert-Butylazo-2-benzoyloxypropane, PhCOO-CMe2-N=N-Bu-t (8a). Crude 2-
tert-butylazo-2-chloropropane, t-Bu-N=N-CMe2-Cl,66
was treated with silver benzoate as
in the synthesis of 8b above. The product was purified on silica gel, eluting with 9:1
hexane:ethyl acetate, Rf 0.48. Yield 0.956 g (53% from acetone-tert-butyl hydrazone).
The azo compound crystallized at 25 C over time, yielding ice-like, transparent crystals
that melted at 29-30 C to light beige oil. 1H NMR (C6D6) 8.20-8.22 (m, 2H), 7.02-7.12
(m, 3H), 1.65 (s, 6H), 1.20 (s, 9H). 13
C NMR (C6D6) 164.75, 132.67, 132.02, 130.04,
128.42, 101.32, 67.09, 26.69, 24.59.
2-tert-Butylazo-2-methoxypropane, t-Bu-N=N-C(Me)2-OMe (protiated 14).67
To t-Bu-N=N-CMe2-Cl (0.934 g, 7.29 mmol) prepared as above and cooled to -78 C was
added 2 mL of MeOH.68
In a separate vessel, sodium methoxide (0.45 g, 8.39 mmol) was
dissolved in MeOH (4 mL) and the solution was added to the reaction mixture dropwise
at -78 C. The mixture was stirred at -78 C for 10 min and the temperature was allowed
to increase gradually to 25 C over 3 h. Water and CH2Cl2 were added, the mixture was
shaken, and the aqueous phase was extracted with CH2Cl2 several times. The combined
organic phase was dried over Na2SO4 and the solvent was evaporated. Removal of
227
residual solvent in vacuo at ~ 40 mmHg yielded 0.509 g (44% from t-Bu-NH-N=CMe2)
of volatile liquid. 1H NMR (MeOH-d4) 3.39 (s, 3H), 1.21 (s, 6H), 1.19 (s, 9H).
1.7. Supporting information
Calculated isotropic Fermi contact couplings for 5a, ESR spectra of 5a, UV spectra
of 4a, 6, 8a, 4bZ, 7Z, and 8b, computed structures of 15, 8, 5a, and NMR spectra are
provided in Appendix D.
1.8. References and Notes
1. Engel, P. S.; Duan, S. M.; He, S. L.; Tsvaygboym, K.; Wang, C. R.; Wu, A. Y.;
Ying, Y. M.; Smith, W. B., The free radical chemistry of the azoxy group.
Arkivoc 2003, (12), 89-108.
2. Brough, J. N.; Lythgoe, B.; Waterhouse, P., Macrozamin. IV. Positional and
Geometrical Isomerism in Mixed Aliphatic-Aromatic Azoxy-Compounds.
Journal of the Chemical Society 1954, (12), 4069-4079.
3. Matsumoto, H.; Nagahama, T.; Larson, H. O., Studies on Methylazoxymethanol
Aglycone of Cycasin - a Synthesis of Methylazoxymethyl Acetate. Biochemical
Journal 1965, 95, (2), 13C-14C.
4. Fujiu, M.; Sawairi, S.; Shimada, H.; Takaya, H.; Aoki, Y.; Okuda, T.; Yokose, K.,
Azoxybacilin, a Novel Antifungal Agent Produced by Bacillus-Cereus NR2991 -
Production, Isolation and Structure Elucidation. Journal of Antibiotics 1994, 47,
(7), 833-835.
5. Ohwada, J.; Umeda, I.; Ontsuka, H.; Aoki, Y.; Shimma, N., Synthesis and
Structure-Activity Relationships of a Novel Antifungal Agent, Azoxybacilin.
Chemical & Pharmaceutical Bulletin 1994, 42, (8), 1703-1705.
6. Parry, R. J.; Li, Y.; Lii, F. L., Biosynthesis of Azoxy Compounds - Investigations
of Valanimycin Biosynthesis. Journal of the American Chemical Society 1992,
114, (25), 10062-10064.
7. Engel, P. S., Mechanism of the Thermal and Photochemical Decomposition of
Azoalkanes. Chemical Reviews 1980, 80, (2), 99-150.
228
8. Engel, P. S.; He, S. L.; Wang, C. R.; Duan, S. M.; Smith, W. B., Radical
cyclization and fragmentation of azoxy compounds. Journal of the American
Chemical Society 1999, 121, (27), 6367-6374.
9. Taylor, K. G.; Clark, M. S., Aliphatic Azoxy Compounds. V. Functionalization of
(Z)-Phenylmethyldiazene 1-Oxide. Journal of Organic Chemistry 1976, 41, (7),
1141-1145.
10. Gould, I. R.; Zimmt, M. B.; Turro, N. J.; Baretz, B. H.; Lehr, G. F., Dynamics of
Radical Pair Reactions in Micelles. Journal of the American Chemical Society
1985, 107, (16), 4607-4612.
11. Lewis, F. D.; Magyar, J. G., Photoreduction and Alpha Cleavage of Aryl Alkyl
Ketones. Journal of Organic Chemistry 1972, 37, (13), 2102-2107.
12. Encina, M. V.; Lissi, E. A.; Lemp, E.; Zanocco, A.; Scaiano, J. C., Temperature-
Dependence of the Photochemistry of Aryl-Alkyl Ketones. Journal of the
American Chemical Society 1983, 105, (7), 1856-1860.
13. Leopold, D.; Fischer, H., EPR Studies on the Photofragmentation of 2,2-Dialkyl-
2-Alkylaminoacetophenones. Journal of the Chemical Society - Perkin
Transactions 2 1992, 513-517.
14. Turro, N. J.; Mattay, J., Photochemistry of 1,2-diphenyl-2,2-dimethyl-propanone-
1 in micellar solutions. Cage effects, isotope effects and magnetic field effects.
Tetrahedron Letters 1980, 21, (19), 1799-1802.
15. Freeman, J. P., Nuclear Magnetic Resonance Spectra and Structure of Aliphatic
Azoxy Compounds. Journal of Organic Chemistry 1963, 28, (10), 2508-2511.
16. Wagner, P. J., Conformational Changes Involved in Singlet-Triplet Transitions of
Biphenyl. Journal of the American Chemical Society 1967, 89, (12), 2820-2825.
17. Engel, P. S.; Monroe, B. M., Complications in Photosensitized Reactions. In
Advances in Photochemistry; Pitts, J. N. H.; Hammond, G. S.; Noyes, W. A.,
Eds.; John Wiley and Sons: New York, 1971; Vol. 8, pp 245-313.
18. Lamola, A. A.; Turro, N. J., Energy Transfer and Organic Photochemistry.
Wiley: New York, 1969; Vol. 14, p. 374.
19. Caldwell, R. A.; Singh, M., Lifetimes of Conjugated Diene Triplets. Journal of
the American Chemical Society 1982, 104, (22), 6121-6122.
20. Colley, C. S.; Grills, D. C.; Besley, N. A.; Jockusch, S.; Matousek, P.; Parker, A.
W.; Towrie, M.; Turro, N. J.; Gill, P. M. W.; George, M. W., Probing the
reactivity of photoinitiators for free radical polymerization: Time-resolved
229
infrared spectroscopic study of benzoyl radicals. Journal of the American
Chemical Society 2002, 124, (50), 14952-14958.
21. Kochi, J. K.; Krusic, P. J., Isomerization and Electron Spin Resonance of Allylic
Radicals. Journal of the American Chemical Society 1968, 90, (25), 7157-7159.
22. Brocks, J. J.; Beckhaus, H.-D.; Beckwith, A. L. J.; Ruchardt, C., Estimation of
Bond Dissociation Energies and Radical Stabilization Energies by ESR
Spectroscopy. J. Org. Chem. 1998, 63, (6), 1935-1943.
23. http://epr.niehs.nih.gov/
24. Nelson, V.; Serianz, A.; Kovacic, P., Chemistry of N-Halo Compounds. 25.
Regiospecific Synthesis of Unsymmetrical Azoxy Compounds (Diazene N-
Oxides). Journal of Organic Chemistry 1976, 41, (10), 1751-1754.
25. Zawalski, R. C.; Kovacic, P., Chemistry of N-Halo Compounds. 30. Regiospecific
Formation of Azoxyaralkanes (Diazene N-Oxides) from N,N-Dibromo
Compounds and Nitrosobenzene. Journal of Organic Chemistry 1979, 44, (13),
2130-2133.
26. Sullivan, F. R.; Luck, E.; Kovacic, P., Novel, Directed Synthesis of
Unsymmetrical Azoxyalkanes and Azoxyaralkanes from N,N-Dihaloamine and
Nitroso Precursors. Journal of Organic Chemistry 1974, 39, (20), 2967-2970.
27. Taylor, K. G.; Isaac, S. R.; Swigert, J. L., Aliphatic Azoxy Compounds. VI.
Photolytic Isomerization of Azoxyalkanes and Thermal Ring-Opening of 2-
Cyclohexyl-3-Methyloxadiaziridine. Journal of Organic Chemistry 1976, 41, (7),
1146-1152.
28. Taylor, K. G.; Riehl, T., Aliphatic Azoxy Compounds. II. Synthesis of New
Azoxy Compounds by Photolytic Isomerizations. Journal of the American
Chemical Society 1972, 94, (1), 250-255.
29. Engel, P. S.; Schexnayder, M. A., Photochemistry of an Acyclic Bis
Beta,Gamma-Unsaturated Ketone. Journal of the American Chemical Society
1972, 94, (12), 4357-4359.
30. Engel, P. S.; Schexnayder, M. A.; Ziffer, H.; Seeman, J. I., Effect of Alpha-
Methyl Groups on Photochemistry of 3,4,5,6,7,8-Hexahydronaphthalen-2(1H)-
One. Journal of the American Chemical Society 1974, 96, (3), 924-925.
31. Warzecha, K. D.; Leitich, J.; Demuth, M., Intramolecular 2+2 cycloaddition of
some 1,n-diene-1,1-dicarbonitriles - difference between singlet and triplet
reactivity. Journal of Photochemistry and Photobiology A: Chemistry 2002, 152,
(1-3), 103-108.
230
32. Cristol, S. J.; Kaufman, R. L.; Opitz, S. M.; Szalecki, W.; Bindel, T. H.,
Photochemical-Transformations. XXXIII. Some Studies of the Photo-
Rearrangements of Dibenzobarrelenes - a Novel Excitation-Transfer Relay
Mechanism. Journal of the American Chemical Society 1983, 105, (10), 3226-
3234.
33. Suehiro, T.; Tashiro, T.; Nakausa, R., Electron-Spin Resonance Study of
Aryldiazenyl Radicals in Solution. Chemistry Letters 1980, (11), 1339-1342.
34. Takemura, T.; Aikawa, M.; Baba, H., Phosphorimetric Investigation of Triplet-
Triplet Energy Transfer in Fluid Solution. Bulletin of the Chemical Society of
Japan 1974, 47, (10), 2476-2481.
35. Porter, N. A.; Funk, M. O., cis-Intermediates in Azo-Compound Photolysis.
Journal of the Chemical Society, Chemical Communications 1973, (8), 263-264.
36. Porter, N. A.; Marnett, L. J., Photolysis of Unsymmetric Azo-Compounds - Cis
Azo Compound Intermediates. Journal of the American Chemical Society 1973,
95, (13), 4361-4367.
37. Curtin, D. Y.; Ursprung, J. A., Reaction of Organometallic Compounds with
Diazonium Salts - Synthesis of Arylazoalkanes. Journal of Organic Chemistry
1956, 21, (11), 1221-1225.
38. Greene, F. D.; Hecht, S. S., Oxadiaziridines, Cyclic Form of an Azoxy Group -
Synthesis, Valence Isomersim, and Reactivity. Journal of Organic Chemistry
1970, 35, (8), 2482-2486.
39. Swigert, J.; Taylor, K. G., Aliphatic Azoxy Compounds. I. Photolytic
Isomerization Azoxyalkanes. Journal of the American Chemical Society 1971, 93,
(26), 7337-7338.
40. Olsen, H.; Pedersen, C. L., Rearrangement of 1,2-Diazene N-Oxides. 3.
Photochemical Isomerization of Bicyclic cis-1,2-Diazene N-Oxides to
Oxadiaziridines. Acta Chemica Scandinavica, Series B, Organic Chemistry and
Biochemistry 1982, 36, (10), 701-706.
41. Frisch, M. J. T., G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.;
Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A., Jr.; Stratmann, R. E.;
Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, A. D.; Kudin, K. N.; Strain,
M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi, M.; Cammi, R.; Mennucci, B.;
Pomelli, C.; Adamo, C.; Clifford, S.; Ochterski, J.; Petersson, G. A.; Ayala, P. Y.;
Cui, Q.; Morokuma, K.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;
Foresman, J. B.; Cioslowski, J.; Ortiz, J. V.; Stefanov, B. B.; Liu, G.; Liashenko,
A.; Piskorz, P.; Komaromi, I.; Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.;
Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.;
231
Gill, P. M. W.; Johnson, B. G.; Chen, W.; Wong, M. W.; Andres, J. L.; Head-
Gordon, M.; Replogle, E. S.; Pople, J. A. Gaussian 98, revision A.4, Gaussian,
Inc.: Pittsburgh, PA, 1998.
42. Levi, N.; Malament, D. S., Photochemistry of Alpha,Alpha'-Disubstituted
Azoalkanes. I. Photosolvolytic Inversion and Substitution of Alpha,Alpha'-
Dichlorooxyazoalkanes and Alpha,Alpha'-Dialkoyloxyazoalkanes. Israel Journal
of Chemistry 1974, 12, (5), 925-936.
43. Levi, N.; Malament, D. S., Photochemistry of alpha,alpha'-Disubstituted
Azoalkanes. II. Photolytic Dechlorination, Decomposition, and Oxidation of
alpha,alpha'-Dichloro-Azoalkanes and alpha,alpha'-Diacyloxy-Azoalkanes.
Journal of the Chemical Society, Perkin Transactions 2 1976, (11), 1249-1256.
44. D'Anello, M.; Erba, E.; Gelmi, M. L.; Pocar, D., 5-Oxazolones. II. 2,4-Diaryl-4-
(2,4-Dinitroaryl)-5(4H)-Oxazolones - Synthesis and Acid-Catalyzed
Transformation into 1-Hydroxy-1H-Indazole-Derivatives. Chemische Berichte-
Recueil 1988, 121, (1), 67-73.
45. Steglich, W.; Kubel, B.; Gruber, P., Thermolysis of Oxazolin-5-Ones. IV.
Conversion of Ortho Nitrophenyl Substituted 3-Oxazolin-5-Ones into 1-
Acyloxyindazoles and (Ortho Nitrophenyl) Ketones. Chemische Berichte-Recueil
1973, 106, (9), 2870-2876.
46. Yanai, M.; Kinoshita, T., Studies on Synthesis of Pyridazine Derivatives. X.
Reaction of 3,6-Dimethoxy-4-Nitropyridazine 1-Oxide with Alkyl Halides.
Chemical & Pharmaceutical Bulletin 1968, 16, (7), 1221-1227.
47. Hrovat, D. A.; Beno, B. R.; Lange, H.; Yoo, H. Y.; Houk, K. N.; Borden, W. T.,
A Becke3LYP/6-31G*study of the cope rearrangements of substituted 1,5-
hexadienes provides computational evidence for a chameleonic transition state.
Journal of the American Chemical Society 1999, 121, (45), 10529-10537.
48. Ahrens, W.; Berndt, A., 1,2-Diaza-Allyl-Radikale and 1,3-Diaza-Allyl Radicals.
Tetrahedron Letters 1974, (43), 3741-3742.
49. Aurich, H. G.; Lotz, I.; Weiss, W., Nitroxides. XXVI. ESR Study of Hydrazonyl
Oxides - Determination of Spin-Density. Tetrahedron 1978, 34, (7), 879-885.
50. Creary, X.; Engel, P. S.; Kavaluskas, N.; Pan, L.; Wolf, A.,
Methylenecyclopropane Rearrangement as a Probe for Free Radical Substituent
Effects. Values for Potent Radical-Stabilizing Nitrogen-Containing
Substituents. Journal of Organic Chemistry 1999, 64, (15), 5634-5643.
232
51. Engel, P. S.; Gerth, D. B., Thermolysis of Acyclic Azoalkanes - Simultaneous or
Stepwise C-N Homolysis. Journal of the American Chemical Society 1983, 105,
(23), 6849-6851.
52. Wamser, C. C.; Medary, R. T.; Kochevar, I. E.; Turro, N. J.; Chang, P. L., Steric
effects in singlet and triplet electronic energy transfer to azo compounds. J. Am.
Chem. Soc. 1975, 97, (17), 4864-4869.
53. Tanikaga, R., Photoreduction of Azoxybenzene to Azobenzene. Bulletin of the
Chemical Society of Japan 1968, 41, (7), 1664-1668.
54. Monroe, B. M.; Wamser, C. C., Mechanism of photochemical reactions in
solution. LXIV. Photoreduction of azoxybenzene by chemical sensitization.
Molecular Photochemistry 1970, 2, (3), 213-223.
55. Engel, P. S.; Wu, W. X., Reduction of Azoalkanes by Benzhydryl Radicals.
Journal of the American Chemical Society 1989, 111, (5), 1830-1835.
56. Hata, N.; Ono, I.; Kawasaki, M., Photoinduced Deoxygenation Reaction of
Heterocyclic N-Oxides. Chemistry Letters 1975, (1), 25-28.
57. Basch, H.; Hoz, T., Azoxy rearrangement reactions. Molecular Physics 1997, 91,
(5), 789-803.
58. Beckwith, A. L. J.; Glover, S. A., Determination of the Rates of Ring-Closure of
Oxygen-Containing Analogs of Hex-5-Enyl Radical by Kinetic Electron Spin
Resonance Spectroscopy. Australian Journal of Chemistry 1987, 40, (1), 157-173.
59. Giese, B.; Harnisch, H.; Luning, U., The Influence of Steric Effects on the
Selectivity of Radical CC Bond Formation Reactions. Chemische Berichte-
Recueil 1985, 118, (4), 1345-1351.
60. Wille, U., Self-terminating, oxidative radical cyclizations: A novel reaction of
acyloxyl radicals. Journal of the American Chemical Society 2002, 124, (1), 14-
15.
61. Malament, D. S.; McBride, J. M., Alpha,Alpha'-Dichloroazoalkanes. I. Synthesis
- Stereospecificity and Side Reactions - Crystal Structure of 1,1'-Dichloro-
1,1'diphenyl-1,1'-Azopropane. Journal of the American Chemical Society 1970,
92, (15), 4586-4593.
62. Zawalski, R. C.; Lisiak, M.; Kovacic, P.; Luedtke, A.; Timberlake, J. W., Radical
Stabilization - Synthesis and Decomposition of a Beta-Ketodiazene. Tetrahedron
Letters 1980, 21, (5), 425-428.
233
63. Gnichtel, H.; Beier, M., N-Oxide-Enhydroxylamine Tautomerism of 4-Benzyl-2-
Oxo-3-Imidazoline 3-Oxides. Liebigs Annalen Der Chemie 1981, (2), 312-316.
64. Al-Hassan, S. S.; Cameron, R. J.; Curran, A. W. C.; Lyall, W. J. S.; Nicholson, S.
H.; Robinson, D. R.; Stuart, A.; Suckling, C. J.; Stirling, I.; Wood, H. C. S.,
Specific Inhibitors in Vitamin Biosynthesis. VII. Syntheses of Blocked 7,8-
Dihydropteridines Via Alpha-Amino Ketones. Journal of the Chemical Society-
Perkin Transactions 1 1985, (8), 1645-1659.
65. Taylor, K. G.; Isaac, S. R.; Clark, M. S., Aliphatic Azoxy Compounds. IV.
Reaction of Nitrosoalkanes with Hydroxylamines - Synthesis of Unsymmetrical
Primary and Secondary Azoxyalkanes by N-N Bond Formation. Journal of
Organic Chemistry 1976, 41, (7), 1135-1140.
66. Engel, P. S.; Wu, W. X., Thermolysis, Photolysis, and Acid Catalysis of an
Alpha-(Dimethylamino)Azoalkane - Amino Stabilization of a Carbon Radical
Center. Journal of Organic Chemistry 1990, 55, (9), 2720-2725.
67. MacLeay, R. E., Pennwalt Corporation, Azoxy compound for preparing foamed
structures. US Patent no. 4228069, 1980.
68. MacLeay, R. E.; Sheppard, C. S., Pennwalt Corporation, Tertiary-aliphatic
amidazo compounds. US Patent no. 4001207, 1977.
RICE UNIVERSITY
Photochemical Studies of
Single-Walled Carbon Nanotube Ozonides and -Azoxy Ketones
by
Konstantin Tsvaygboym
Volume II of II
HOUSTON, TEXAS
APRIL 2007
235
Appendix A
Mathematics for regression analysis of fluorescence and NIR
absorbance data
236
A1. General Information
Ozonation of single-walled carbon nanotubes has a dramatic influence on their
spectral properties. Fluorescence and NIR absorbance changes of ozonated SWNT were
examined in this work. Five formulas were tested for elucidation of ozonide decay rates,
b. All regression runs discussed in this chapter were performed on either normalized or
inverted normalized data sets. Normalization procedure facilitated comparison of tubes
emitting light at different wavelengths and with different intensities. Thus, tube (8,3),
having its maximum emission near 954 nm, had about 1.5 times stronger intensity than
tube (7,5) with emmax
at 1027 nm. To adequately compare these two tubes, data sets
would have to be normalized.
Normalized set was calculated by division of data points by a corresponding
maximum value. Inverted normalized set was computed by division of a maximum value
by data points from the same set. Points ranged from 0 to 1 in normalized sets and from 1
and higher in inverted normalized sets. For example, if ozonation caused fluorescence
intensity to drop to 0.05 level of its initial value, then the lowest point in normalized set
would be 0.05 and the highest point in inverted set would be 20.
Points acquired before and during ozone injections, being irrelevant to ozonide
decay, were excluded from regression runs. Further in the text inverted normalized data
sets will be referred to as inverted data sets. Data discussed in this appendix were
acquired on NS1 Nanospectralyzer with 660 nm excitation source. SWNT samples were
in the form of aqueous SDS suspensions.
237
A2. Regression models for normalized data
A fluorescence recovery rate of ozonated SWNT varied with temperature and
wavelength. Tube (8,3) with its emmax
at 954 nm was calculated to recover to
1/yfinal ~ 75 % level of its original intensity with rate b ~ 0.0123 s-1
, or lifetime ~ 81 s at
r. t., while other tubes recovered to different levels and with different rates. Fluorescence
recovery at four distinct wavelengths is shown in Figure 1.
Time (s)
0 1000 2000 3000
No
rma
lize
d F
luo
resce
nce
In
ten
sity
0.0
0.5
1.0
954 nm
1027 nm
1125 nm
1251 nm
Figure 1. Fluorescence recovery of ozonated SWNT (shown with symbols) at r. t. and
corresponding regression curves (solid lines) calculated with formula F2. Kinetics shown
for four characteristic wavelengths. Points before and during ozone addition are depicted
with dotted lines.
Three formulas, with 3, 5 and 6 variables, were employed for regression on
normalized fluorescence data set and results are summarized in Table 1.
238
minmin1
y
ceaey
yy
n
bt
bt
final
(F1)
n
bt
bt
final ceaey
y
1
(F2)
bt
final aeyy
1 (F3)
Formulas F1 and F2 include two exponential terms, a fast term btae
and a slow
term n
bt
ce
, where parameter n reflects how many times the fast term is faster than the
slow one.
Table 1. Regression results calculated with formulas F1, F2 and F3 for normalized
fluorescence data recorded at 1027 nm emission wavelength*
Data Set Full Truncated
Parameter F1, n > 10 F2, n > 10 F2, n > 1 F3 F2, n > 10 F3
a 16.91 16.91 17.41 8.91 17.73 14.63
b 0.008172 0.008172 0.008473 0.004985 0.008580 0.007113
yfinal 1.331 1.331 1.366 1.467 1.255 1.632
c 0.6509 0.6509 0.7444 0.8534
n 10.00 10.00 7.942 10.06
ymin 0.0000
r2 0.9996 0.9996 0.9997 0.9763 0.9998 0.9976
* Formula number and lower boundary for variable n are written at the head of each
column. Constraints used for regression: 0 < ymin < 1; 0 < a < 1000; 0 < b < 100;
0 < c < a; n < 100; yfinal > 1.05.
Formula F1 is a derivative of:
239
n
bt
bt
final ceaeyyy
yty
1)(
minmax
min
where right denominator is a two exponential decay expression. A substitution of ymax by
1 made formula applicable for all normalized data sets.
Parameter b is a decay rate and it is always a positive number. The upper constraint
was set to b = 100, or lifetime = 1/b = 0.01 s. Typical ozonide decay rates observed in
this work were between 0.005 and 0.05 s-1
, which corresponds to in a range 200 to
20 s.
Calculated rates were independent of time values, as expected. Formulas were tested
with first data point starting at 0 and 100 sec and the same rates and r2 values obtained for
regression curves. Shifting time values from 0 to 100 sec resulted in a change of
exponential prefactors a and c, but not rates b or any other parameters.
Fluorescence curve can be divided into four sections: original intensity (i), ozone
addition and equilibration (ii), fast recovery (iii) and slow recovery (iv) as shown in
Figure 2.
240
Time (s)
0 1000 2000 3000
No
rmaliz
ed
Flu
ore
scen
ce
Inte
nsity
0.0
0.3
0.6
0.9
1/ yfinal
i ii iii iv
fast
slow
ymax
Figure 2. Normalized fluorescence recovery of ozonated SWNT at 1027 nm (shown with
symbols), corresponding regression curve (solid line) and a schematic diagram. Points
before and during ozone addition are depicted with a dotted line. Fluorescence curve is
divided into four sections: i) original intensity, ii) ozone addition and equilibration, iii)
fast recovery and iv) slow recovery.
Values ymin for fluorescence data were calculated with F1 to be approximating zero
(with ymin > 0 constraint) for sets recorded at different wavelengths. This result may be
interpreted such that the degree of SWNT functionalization with ozonides was high
enough to affect all “sections” of SWNT capable to fluoresce. In a view of this result, ymin
parameter was found to be unnecessary in fluorescence data regression. The same
formula used for NIR Abs recovery yielded ymin greater than zero.
Formulas F1 and F2 have fast and slow exponential components. Assumption was
made that the slow component should be n times slower than the fast one. With all
constraints being inactive, the resulted n value for long tail normalized data set recorded
at 1027 nm emission wavelength was n ~ 8. For 954, 1125 and 1251 nm wavelengths
241
values were approximately 10, 8, and 10 correspondingly. In case of 1251 nm emission,
regression model was trying to make a straight line for a slow component, and yfinal > 1.3
was enforced. The value yfinal ~ 1.3 was obtained for other three wavelengths. The
problem with yfinal in regression model indicated that there were not enough "tail" data
points at 1250 nm wavelength (see curve in Figure 1).
Parameter yfinal describes intensity recovery level at time t ; yfinal = 1 would
mean a full recovery. Constraint yfinal > 1.05 for regression performed on fluorescence
data seemed reasonable, because recovery to a level higher than 1/yfinal = 95% after
epoxide formation is unlikely and was not observed in this work.
NIR absorbance of SWNT was affected by ozonation in the same manner as
fluorescence. NIR absorbance recovery is shown in Figure 3.
Time (s)
0 500 1000 1500 2000 2500
No
rma
lize
d N
IR A
bso
rba
nce
0.50
1.00
953.9 nm
1027.2 nm
1125.4 nm
1251.3 nm
0.760.72
0.39
0.29
Figure 3. NIR absorbance recovery of ozonated SWNT (shown with symbols) and
corresponding regression curves (solid lines) calculated with formula F1. Kinetics shown
for four characteristic wavelengths. Points before, during ozone addition and during
equilibration are depicted with dotted lines.
242
NIR Absorbance signal drop below regression curves right after bubbling of O2/O3
gaseous mixture is not seen on fluorescence curves (Figure 1). There are several factors
that could contribute to such behavior: vigorous stirring, heat produced from reaction
with ozone, dissolved gases, influence of one ozonide on another, presence of unreacted
ozone in solution and NIR absorbance bleaching of metallic SWNT.
As explained in Chapter 2, a number of 1,2,3-trioxolanes that can be on a tube at any
time is limited. Parameter yinert was introduced to describe the lowest possible NIR
absorption at any given wavelength. Absorption equaled to yinert would mean that SWNT
is completely saturated with 1,2,3-trioxolanes. yinert values were measured experimentally
and are summarized in Table 3.
Table 3. Comparison of ymin calculated for normalized NIR Absorbance of SWNT and
experimentally determined yinert at four different emission wavelengths
Wavelength 954 nm 1027 nm 1125 nm 1251 nm
yinert 0.4486 0.3335 0.2263 0.1815
ymin 0.7633 0.7226 0.3932 0.2866
Measured yinert values are useful for establishing constraints in regression analysis,
though their physical meaning is a subject for further studies. Analogously to
fluorescence, NIR absorbance curve can be broken down into four sections as shown in
Figure 4.
243
Time (s)0 500 1000 1500 2000 2500
Norm
aliz
ed
NIR
Abso
rbance
0.30
0.60
0.90
ymax
1 / yfinal
yinert
ymin
i ii iii iv
fast
slow
Figure 4. NIR absorbance recovery of ozonated SWNT at 1125 nm (shown with
symbols), corresponding regression curve (solid line) and a schematic diagram. Points
before, during ozone addition and during equilibration are depicted with a dotted line.
NIR absorbance divided into four sections: i) original intensity, ii) ozone addition and
equilibration, iii) fast recovery and iv) slow recovery. Horizontal dash lines represent
yinert, ymin, 1/yfinal, and ymax values.
Formula F1 was employed for NIR absorbance regression with constraints
mentioned in Table 1, also 0 < ymin < 1 and yfinal > 1.015. By definition, ymin is a positive
number greater than or equal to yinert. Regression yielded ymin greater than yinert values
determined experimentally (Table 3). Parameter ymin was introduced to describe NIR
absorbance of SWNT “sections” that did not get “bleached” by ozone. For example,
normalized value of yinert at 1027 nm was determined experimentally as 0.33 (Table 3). If
normalized NIR absorption at 1027 nm dropped from 1.00 to 0.79 after ozonation, it
would imply that ymin was somewhere between 0.33 and 0.79 (see Figure 4). Introduction
of ymin was necessary to separate decay rate b from NIR absorbance caused by unreacted
segments of SWNT. While the lower boundary for normalized ymin was estimated as 0.33
244
at 1027 nm, the upper boundary dependes solely on the amount of ozone reacted with
SWNTs. The more ozone reacted, the lower ymin would be. Variable ymin indirectly
accounts for a quenching degree, or an amount of injected ozone. Alternative physical
interpretations of parameter ymin are also possible.
yfinal was held at 1.015 due to nearly complete NIR absorbance recovery (see Figure
3 and 4). It is likely that highly conjugated -bond structure of SWNT was only
marginally affected by random epoxides on its surface. Figure 5 shows many possible -
conjugation routes:
O O
Figure 5. Multiple ways for NIR light to interact with SWNT conjugated bond
system. Formation of an epoxide should have marginal effect on overall NIR absorbance.
Assumption was made that epoxides per se do not absorb NIR light and therefore do
not affect absorbance values. No additional parameter was introduced to account for NIR
absorbance of epoxides. With respect to NIR absorbance, decomposition of electron-
withdrawing ozonides increased electron density on SWNT and was expressed with a
term btae
in all formulae. Such decomposition is considered a fast component. Term
n
bt
ce
represents secondary, or a slow component. It should be noted that physical
nature of the slow term is not totally understood. Vacuum line studies indicated that
245
oxygen evolves from ozonated SWNT for at least 20 min at r. t. Thus slow component is
likely to be a combination of „slow‟ ozonide decomposition and possible structural
rearrangements of SWNT.
For NIR absorbance n was calculated to be greater than 25, along with very small
exponential prefactors for all four wavelengths in Figure 3.
A3. Regression models for inverted normalized data
Inverted normalized fluorescence data sets recorded at four distinct wavelengths are
shown in Figure 6.
Time (s)
0 1000 2000 3000
Invert
ed N
orm
aliz
ed F
luore
scen
ce
10
25
40 954 nm
1027 nm
1125 nm
1251 nm
1
Figure 6. Inverted normalized fluorescence recovery of ozonated SWNT (shown with
symbols) and corresponding regression curves (solid lines) calculated with formula F4.
Kinetics are shown for four characteristic wavelengths. Points before, during ozone
addition and equilibration are depicted with dotted lines. Horizontal dash line represents
original fluorescence intensity.
Two formulas with 3 and 5 variables were used for regression on inverted
normalized sets:
246
n
bt
bt
final ceaeyy
(F4)
bt
final aeyy (F5)
Fluorescence curve can be broken down into four distinct sections (Figure 7):
Time (s)0 1000 2000 3000
Inve
rte
d N
orm
aliz
ed
Flu
ore
scen
ce
5
15
25
yfinal
i ii iii iv
fast
slow
1
Figure 7. Inverter normalized fluorescence recovery of ozonated SWNT at 1027 nm
(shown with symbols), corresponding regression curve (solid line) and a schematic
diagram are shown. Points before and during ozone addition and equilibration are
depicted with a dotted line. Fluorescence curve is divided into four sections: i) original
intensity, ii) ozone addition and equilibration, iii) fast recovery, iv) slow recovery.
Regression results are summarized in Table 4.
247
Table 4. Regression results calculated with formulas F4 and F5 for inverted normalized
fluorescence data recorded at 1027 nm emission wavelength*
Data set Full Truncated
Parameter F4, n > 10 F4, n > 1 F5 F4, n > 10 F5
a 18.29 17.61 18.61 18.08 18.59
b 0.008972 0.009499 0.008219 0.009233 0.008570
yfinal 1.313 1.424 1.528 1.050 1.722
c 0.838 1.524 1.382
n 10.000 4.513 10.000
r2 0.9991 0.9992 0.9969 0.9991 0.9984
* Formula number and lower boundary for variable n are written at the head of each
column. Constraints used for regression: 0 < ymin < 1; 0 < a < 1000; 0 < b < 100;
0 < c < a; n < 100; yfinal > 1.05.
A4. Comparison of normalized and inverted normalized regression models
Regression results for rates b and coefficients of determination r2 are summarized in
Figure 8.
248
Invert
ed N
orm
aliz
ed F
luore
scence
1
3
5
7
9
Time (s)0 500 1000 1500 2000
Norm
aliz
ed F
luore
scence
0.00
0.25
0.50
0.75ozonide decay rearrangement
btaefinal
yy
1
btaefinal
yy
r 2
= 0.9977b = 0.00835 a = 18.60
r 2
= 0.9882b = 0.00585 a = 11.08
Figure 8. A comparison of rates b and coefficients of determination r2 obtained with
3-parameter regression models on truncated normalized and inverted normalized data sets
from the same experiment. Ozonide decay, a fast component, and some secondary
processes, represented by a slow component, are separated by a dotted vertical line.
Rate b obtained for normalized data set was 1.4 times slower than that for inverted
set. Regression model adjusted for slowly changing tail, brining the overall rate down.
249
The coefficient of determination r 2
for the normalized model was lower than that for
inverted model.
Having a goal to calculate ozonide decay rate, short-tail, or truncated inverted
normalized data set was considered as the most useful for such regression. Both 5 and 3-
parameter formulas could be utilized. Value n would be needed for 5 - parameter
formula. An approximate value of n for slow component was obtained from regressions
performed on long-tail normalized fluorescence data sets. Value n > 10 deemed a
reasonably accurate constraint for 5-parameter regression on short-tail inverted data set.
Resulted ozonide decay rate b was found to be 8% greater for a 5-parameter model. From
this it can be concluded that error for calculated ozonide decomposition rate is at least
8%.
A5. Conclusions
As deduced from regression modeling, an ozonide decay curve had fast and slow
components and could not be fitted with a single exponent. Introduction of a second
exponent or cutting off “tail” deemed necessary to get a better fit.
Truncating data sets, or shortening “tails”, gave faster rates for both normalized and
inverted normalized sets. Rates obtained for one exponent formulae were slower than
those for two exponent formulas because model had to adjust for a slow component.
With the same number of variables, regression produced slightly different decay
rates for normalized and inverted normalized data sets. This happened because all points
were considered equally weighted. For that reason, the most accurate rate for slow
component is considered the one in long tail normalized data set. Rate calculations of
slow component on truncated datasets gave no meaningful results due to insignificant
250
number of "tail" points. The most accurate rate for fast component is considered the one
in short tail inverted normalized data set.
Values ymin for fluorescence data were calculated to be approximating zero (with ymin
> 0 constraint) for sets recorded at different wavelengths. This result is interpreted such
that degree of SWNT functionalization was high enough to affect all “sections” of SWNT
capable to fluoresce. In a view of this finding, ymin parameter was found to be
unnecessary in fluorescence data regression. The same formula used for NIR Abs
recovery yielded ymin greater than zero.
Slow term n
bt
ce
for NIR absorbance was found negligible; n was calculated to be
greater than 25, along with very small exponential prefactors for all four wavelengths, i.e.
954, 1026, 1123 and 1250 nm. Slow term for fluorescence had n values in a range
between 8 and 10 for the same wavelengths.
To avoid meaningless numbers, ozonide decay rate calculation with one exponent
formula should be performed on inverted normalized data set with a truncated tail.
Ideally, data set should have no “tail” attributable to the slow component. Utilization of a
single exponential formula and a truncated inverted data set is beneficial for approximate
estimation of ozonide decay rates, but longer acquisition times are needed to determine
dependence of the slow component from the fast one.
251
Appendix B
Supporting Information for Part I, Chapter 5. 1H NMR spectrum
252
1H NMR for reaction of C60 with Et3N
253
Appendix C
XPS spectra for reactions of ozonated SWNT with different classes of
compounds
254
swnt_059.spe: Rice University
2007 Jan 28 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 5.2243e+004 max 2.64 min
Su1/Area29: Guanidine ref/1
0100200300400500600700800900100011000
1
2
3
4
5
6x 10
4 swnt_059.spe
Binding Energy (eV)
c/s
Atomic %
C1s 92.3
O1s 7.1
Fe2p3 0.6Guanidine + SWNT
-F
e L
MM
-F
e L
MM
-F
e2p
3 -O
1s
-C
1s
-F
e3p
swnt_060.spe: Rice University
2007 Jan 28 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 3.7060e+004 max 2.64 min
Su1/Area30: Guanidine ox/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4
4.5x 10
4 swnt_060.spe
Binding Energy (eV)
c/s
Atomic %
C1s 75.3
O1s 18.7
N1s 6.0
Guanidine + ozonated SWNT
-O
KL
L
-O
1s
-N
1s
-C
1s
-F
KL
L -F
1s
255
swnt_065.spe: Rice University
2007 Jan 28 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 4.7052e+004 max 2.64 min
Su1/Area35: Folic/a ref/1
0100200300400500600700800900100011000
1
2
3
4
5
6x 10
4 swnt_065.spe
Binding Energy (eV)
c/s
Atomic %
C1s 94.0
O1s 6.0Folic acid + SWNT
-O
1s
-C
1s
swnt_066.spe: Rice University
2007 Jan 28 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 3.4625e+004 max 2.64 min
Su1/Area36: Folic/a ox/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4x 10
4 swnt_066.spe
Binding Energy (eV)
c/s
Atomic %
C1s 73.9
O1s 19.6
N1s 3.5
Fe2p3 1.6
Na1s 1.3
Folic acid + ozonated SWNT
-N
a1s
-O
KL
L
-F
e2p
3
-O
1s
-N
a K
LL
-N
1s
-C
1s
-F
e3p
-F
KL
L
-F
1s
256
wurstr_01.spe: none Rice University
2007 Feb 1 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 5.1603e+004 max 2.64 min
Su1/Area1: TMPD ref/1
0100200300400500600700800900100011000
1
2
3
4
5
6x 10
4 wurstr_01.spe
Binding Energy (eV)
c/s
Atomic %
C1s 95.2
O1s 3.1
Fe2p3 1.7
Wurster reagent (TMPD) + SWNT
-F
e L
MM
-F
e2p
3
-O
1s
-C
1s
-F
e3p
wurstr_02.spe: none Rice University
2007 Feb 1 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 3.9482e+004 max 2.64 min
Su1/Area2: TMPD ox/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4
4.5x 10
4 wurstr_02.spe
Binding Energy (eV)
c/s
Atomic %
C1s 80.9
O1s 18.6
N1s 0.6
Wurster reagent + ozonated SWNT
-O
KL
L
-O
1s
-N
1s
-C
1s
257
swnt_001.spe: Rice University
2007 Jan 27 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 5.3593e+004 max 2.64 min
Su1/Area1: Alanine ref/1
0100200300400500600700800900100011000
1
2
3
4
5
6x 10
4 swnt_001.spe
Binding Energy (eV)
c/s
Alanine + SWNT
Atomic %
C1s 96.1
O1s 3.9
-O
KL
L
-O
1s
-C
1s
swnt_002.spe: Rice University
2007 Jan 27 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 4.5174e+004 max 2.64 min
Su1/Area2: Alanine ox/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5x 10
4 swnt_002.spe
Binding Energy (eV)
c/s
Alanine + ozonated SWNT
Atomic %
C1s 74.0
O1s 21.0
Na1s 2.8
N1s 2.3
-N
a1s
-O
KL
L
-O
KL
L
-O
1s
-N
a K
LL
-N
1s
-C
1s
258
swnt_003.spe: Rice University
2007 Jan 27 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 5.0253e+004 max 2.64 min
Su1/Area3: Cysteine ref/1
0100200300400500600700800900100011000
1
2
3
4
5
6x 10
4 swnt_003.spe
Binding Energy (eV)
c/s
Cysteine + SWNT
Atomic %
C1s 95.4
O1s 3.5
Fe2p3 1.1
-F
e L
MM
-F
e L
MM
-F
e2p
3
-O
1s
-C
1s
-F
e3p
swnt_004.spe: Rice University
2007 Jan 27 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 4.2877e+004 max 2.64 min
Su1/Area4: Cysteine ox/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5x 10
4 swnt_004.spe
Binding Energy (eV)
c/s
Cysteine + ozonated SWNT
Atomic %
C1s 78.4
O1s 17.9
Na1s 2.4
Fe2p3 1.3
-N
a1s
-O
KL
L
-F
e2p
3
-O
1s
-N
a K
LL
-C
1s
-F
e3p
-N
KL
L
-N
1s
-S
2s
-S
2p
259
swnt_005.spe: Rice University
2007 Jan 27 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 3.2565e+004 max 2.64 min
Su1/Area5: Cysteine no NaOH ref/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4x 10
4 swnt_005.spe
Binding Energy (eV)
c/s
Cysteine no NaOH + SWNTAtomic %
C1s 58.0
O1s 23.3
N1s 10.6
S2p 8.1
-O
KL
L
-O
1s
-N
1s
-C
1s
-S
2s
-S
2p
swnt_006.spe: Rice University
2007 Jan 27 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 3.8772e+004 max 2.64 min
Su1/Area6: Cysteine no NaOH ox/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4
4.5x 10
4 swnt_006.spe
Binding Energy (eV)
c/s
Cysteine no NaOH+ ozonated SWNT
Atomic %
C1s 52.9
O1s 27.2
N1s 12.0
S2p 8.0
-O
KL
L
-O
1s
-N
1s
-C
1s
-S
2s -
S2
p
260
swnt_007.spe: Rice University
2007 Jan 27 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 5.2939e+004 max 2.64 min
Su1/Area7: Aspartic acid ref/1
0100200300400500600700800900100011000
1
2
3
4
5
6x 10
4 swnt_007.spe
Binding Energy (eV)
c/s
Aspartic acid + SWNT
Atomic %
C1s 97.4
O1s 2.6
-O
KL
L
-O
1s
-C
1s
swnt_008.spe: Rice University
2007 Jan 27 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 3.4124e+004 max 2.64 min
Su1/Area8: Aspartic acid ox/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4x 10
4 swnt_008.spe
Binding Energy (eV)
c/s
Aspartic acid + ozonated SWNT
Atomic %
C1s 74.1
O1s 20.7
Na1s 3.8
N1s 1.3
-N
a1s
-O
KL
L
-O
1s
-N
a K
LL
-N
1s
-C
1s
261
swnt_009.spe: Rice University
2007 Jan 27 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 5.4776e+004 max 2.64 min
Su1/Area9: Glutamic acid ref/1
0100200300400500600700800900100011000
1
2
3
4
5
6x 10
4 swnt_009.spe
Binding Energy (eV)
c/s
Glutamic acid + SWNT
Atomic %
C1s 92.2
O1s 6.3
Fe2p3 1.5
-F
e L
MM
-F
e2p
3
-O
1s
-C
1s
-F
e3p
swnt_010.spe: Rice University
2007 Jan 27 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 3.9732e+004 max 2.64 min
Su1/Area10: Glutamic acid ox/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4
4.5x 10
4 swnt_010.spe
Binding Energy (eV)
c/s
Glutamic acid + ozonated SWNT
Atomic %
C1s 73.8
O1s 21.7
Na1s 3.8
N1s 0.7
-N
a1s
-O
KL
L
-O
1s
-N
a K
LL
-N
1s
-C
1s
262
swnt_011.spe: Rice University
2007 Jan 27 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 5.2438e+004 max 2.64 min
Su1/Area11: Phenylalanine ref/1
0100200300400500600700800900100011000
1
2
3
4
5
6x 10
4 swnt_011.spe
Binding Energy (eV)
c/s
Phenylalanine + SWNT
Atomic %
C1s 89.2
O1s 7.0
Fe2p3 2.4
N1s 1.3
-F
e2p
3
-O
1s
-N
1s
-C
1s
-F
e3p
swnt_012.spe: Rice University
2007 Jan 27 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 4.0178e+004 max 2.64 min
Su1/Area12: Phenylalanine ox/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4
4.5x 10
4 swnt_012.spe
Binding Energy (eV)
c/s
Phenylalanine + ozonated SWNT
Atomic %
C1s 75.7
O1s 19.7
Na1s 2.5
N1s 2.0
-N
a1s
-O
KL
L
-O
1s
-N
a K
LL
-N
1s
-C
1s
263
swnt_013.spe: Rice University
2007 Jan 27 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 5.5709e+004 max 2.64 min
Su1/Area13: Glycine ref/1
0100200300400500600700800900100011000
1
2
3
4
5
6
7x 10
4 swnt_013.spe
Binding Energy (eV)
c/s
Glycine + SWNT
Atomic %
C1s 93.1
O1s 5.0
Fe2p3 1.9
-F
e2p
3
-O
1s
-C
1s
-F
e3p
swnt_029.spe: Rice University
2007 Jan 27 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 3.4291e+004 max 2.64 min
Su1/Area29: Glycine ox 2/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4x 10
4 swnt_029.spe
Binding Energy (eV)
c/s
Glycine + ozonated SWNT
Atomic %
C1s 73.7
O1s 20.5
N1s 2.9
Na1s 2.8
-N
a1s
-O
KL
L
-N
a K
LL
-O
1s
-N
a K
LL
-N
1s
-C
1s
264
swnt_015.spe: Rice University
2007 Jan 27 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 4.9780e+004 max 2.64 min
Su1/Area15: Histidine ref/1
0100200300400500600700800900100011000
1
2
3
4
5
6x 10
4 swnt_015.spe
Binding Energy (eV)
c/s
Histidine + SWNT
Atomic %
C1s 92.6
O1s 5.7
Fe2p3 1.7
-F
e2p
3
-O
1s
-C
1s
-F
e3p
swnt_030.spe: Rice University
2007 Jan 27 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 3.8410e+004 max 2.64 min
Su1/Area30: Histidine ox 2/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4
4.5x 10
4 swnt_030.spe
Binding Energy (eV)
c/s
Histidine + ozonated SWNT
Atomic %
C1s 73.0
O1s 20.1
N1s 4.5
Na1s 1.9
Fe2p3 0.6
-N
a1s -
O K
LL
-F
e L
MM
-F
e2p
3
-O
1s
-N
a K
LL
-N
1s
-C
1s
-F
e3p
265
swnt_017.spe: Rice University
2007 Jan 27 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 5.2703e+004 max 2.64 min
Su1/Area17: Isoleucine ref/1
0100200300400500600700800900100011000
1
2
3
4
5
6x 10
4 swnt_017.spe
Binding Energy (eV)
c/s
Isoleucine + SWNT
Atomic %
C1s 92.7
O1s 5.8
Fe2p3 1.5
-F
e L
MM
-F
e2p
3
-O
1s
-C
1s
-F
e3p
swnt_018.spe: Rice University
2007 Jan 27 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 3.9022e+004 max 2.64 min
Su1/Area18: Isoleucine ox/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4
4.5x 10
4 swnt_018.spe
Binding Energy (eV)
c/s
Isoleucine + ozonated SWNT
Atomic %
C1s 76.3
O1s 18.8
Na1s 2.4
N1s 1.5
Fe2p3 1.1
-N
a1s
-O
KL
L
-F
e2p
3
-O
1s
-N
a K
LL
-C
1s
-F
e3p
-N
KL
L
-N
1s
266
swnt_019.spe: Rice University
2007 Jan 27 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 5.2939e+004 max 2.64 min
Su1/Area19: Lysine ref/1
0100200300400500600700800900100011000
1
2
3
4
5
6x 10
4 swnt_019.spe
Binding Energy (eV)
c/s
Lysine + SWNT
Atomic %
C1s 93.4
O1s 4.4
Fe2p3 2.2
-F
e L
MM
-F
e L
MM
-F
e L
MM
-F
e2p
3
-O
1s
-C
1s
-F
e3p
swnt_020.spe: Rice University
2007 Jan 27 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 3.7631e+004 max 2.64 min
Su1/Area20: Lysine ox/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4
4.5x 10
4 swnt_020.spe
Binding Energy (eV)
c/s
Lysine + ozonated SWNT
Atomic %
C1s 77.3
O1s 20.0
N1s 2.0
Na1s 0.7
-N
a1s -O
KL
L
-O
1s
-N
a K
LL
-N
1s
-C
1s
267
swnt_021.spe: Rice University
2007 Jan 27 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 5.5291e+004 max 2.64 min
Su1/Area21: Lysine no NaOH ref/1
0100200300400500600700800900100011000
1
2
3
4
5
6
7x 10
4 swnt_021.spe
Binding Energy (eV)
c/s
Lysine no NaOH + SWNT
Atomic %
C1s 91.5
O1s 6.0
Fe2p3 2.5
-F
e L
MM
-F
e2p
3
-O
1s
-C
1s
-F
e3p
swnt_022.spe: Rice University
2007 Jan 27 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 3.7867e+004 max 2.64 min
Su1/Area22: Lysine no NaOH ox/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4
4.5x 10
4 swnt_022.spe
Binding Energy (eV)
c/s
Lysine no NaOH + ozonated SWNT
Atomic %
C1s 75.4
O1s 20.4
N1s 4.2
-O
KL
L
-O
1s
-N
1s
-C
1s
268
swnt_023.spe: Rice University
2007 Jan 27 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 5.4734e+004 max 2.64 min
Su1/Area23: Methionine ref/1
0100200300400500600700800900100011000
1
2
3
4
5
6x 10
4 swnt_023.spe
Binding Energy (eV)
c/s
Methionine + SWNT
Atomic %
C1s 93.5
O1s 4.4
Fe2p3 2.1
-F
e L
MM
-F
e L
MM
-F
e2p
3
-O
1s
-C
1s
-F
e3p
swnt_024.spe: Rice University
2007 Jan 27 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 3.5181e+004 max 2.64 min
Su1/Area24: Methionine ox/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4x 10
4 swnt_024.spe
Binding Energy (eV)
c/s
Atomic %
C1s 72.5
O1s 21.5
N1s 3.0
Na1s 2.6
S2p 0.4
-N
a1s
-O
KL
L
-O
1s
-N
a K
LL
-N
1s
-C
1s
-S
2s
-S
2p
-F
KL
L
-F
1s
269
swnt_025.spe: Rice University
2007 Jan 27 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 5.4136e+004 max 2.64 min
Su1/Area25: Asparagine ref/1
0100200300400500600700800900100011000
1
2
3
4
5
6x 10
4 swnt_025.spe
Binding Energy (eV)
c/s
Asparagine + SWNT
Atomic %
C1s 92.9
O1s 5.4
Fe2p3 1.7
-F
e L
MM
-F
e L
MM
-F
e2p
3
-O
1s
-C
1s
-F
e3p
swnt_026.spe: Rice University
2007 Jan 27 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 3.7826e+004 max 2.64 min
Su1/Area26: Asparagine ox/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4
4.5x 10
4 swnt_026.spe
Binding Energy (eV)
c/s
Asparagine + ozonated SWNT
Atomic %
C1s 76.9
O1s 21.4
Na1s 1.8
N1s <.1
-N
a1s
-O
KL
L
-O
1s
-N
a K
LL
-N
1s
-C
1s
270
swnt_027.spe: Rice University
2007 Jan 27 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 5.3385e+004 max 2.64 min
Su1/Area27: Proline ref/1
0100200300400500600700800900100011000
1
2
3
4
5
6x 10
4 swnt_027.spe
Binding Energy (eV)
c/s
Proline + SWNT
Atomic %
C1s 93.7
O1s 6.3
-O
KL
L
-O
1s
-C
1s
swnt_028.spe: Rice University
2007 Jan 27 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 3.9802e+004 max 2.64 min
Su1/Area28: Proline ox/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4
4.5x 10
4 swnt_028.spe
Binding Energy (eV)
c/s
Proline + ozonated SWNT
Atomic %
C1s 75.7
O1s 21.5
Na1s 2.8
-N
a1s
-O
KL
L
-O
1s
-N
a K
LL
-C
1s
271
swnt_031.spe: Rice University
2007 Jan 28 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 5.6293e+004 max 2.64 min
Su1/Area1: Glutamine ref/1
0100200300400500600700800900100011000
1
2
3
4
5
6
7x 10
4 swnt_031.spe
Binding Energy (eV)
c/s
Atomic %
C1s 91.6
O1s 6.6
Fe2p3 1.8
-F
e L
MM
Atomic %
C1s 91.6
O1s 6.6
Fe2p3 1.8Glutamine + SWNT
-F
e L
MM
-F
e2p
3
-O
1s
-C
1s
-F
e3p
swnt_032.spe: Rice University
2007 Jan 28 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 3.7686e+004 max 2.64 min
Su1/Area2: Glutamine ox/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4
4.5x 10
4 swnt_032.spe
Binding Energy (eV)
c/s
Atomic %
C1s 77.6
O1s 19.2
N1s 2.1
Na1s 1.0
Glutamine + ozonated SWNT
-N
a1s
-O
KL
L
-O
1s
-N
a K
LL
-N
1s
-C
1s
272
swnt_033.spe: Rice University
2007 Jan 28 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 4.8973e+004 max 2.64 min
Su1/Area3: Arginine ref/1
0100200300400500600700800900100011000
1
2
3
4
5
6x 10
4 swnt_033.spe
Binding Energy (eV)
c/s
Atomic %
C1s 92.1
O1s 7.9Arginine + SWNT
-O
1s
-C
1s
swnt_034.spe: Rice University
2007 Jan 28 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 3.6712e+004 max 2.64 min
Su1/Area4: Arginine ox/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4x 10
4 swnt_034.spe
Binding Energy (eV)
c/s
Atomic %
C1s 72.9
O1s 20.0
N1s 3.4
Na1s 2.6
Fe2p3 1.0
Arginine+ ozonated SWNT
-N
a1s
-O
KL
L
-O
KL
L
-F
e2p
3
-O
1s
-N
a K
LL
-N
1s
-C
1s
-F
e3p
-F
KL
L1
-F
KL
L
-F
1s
273
swnt_035.spe: Rice University
2007 Jan 28 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 5.0713e+004 max 2.64 min
Su1/Area5: Serine ref/1
0100200300400500600700800900100011000
1
2
3
4
5
6x 10
4 swnt_035.spe
Binding Energy (eV)
c/s
Atomic %
C1s 91.8
O1s 6.9
Fe2p3 1.2Serine + SWNT
-O
KL
L
-F
e2p
3 -O
1s
-C
1s
-F
e3p
swnt_036.spe: Rice University
2007 Jan 28 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 3.7185e+004 max 2.64 min
Su1/Area6: Serine ox/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4
4.5x 10
4 swnt_036.spe
Binding Energy (eV)
c/s
Atomic %
C1s 75.1
O1s 21.8
Na1s 3.2
N1s <.1
Serine + ozonated SWNT
-N
a1s
-O
KL
L
-O
KL
L
-O
1s
-N
a K
LL
-C
1s
-N
KL
L
-N
1s
274
swnt_037.spe: Rice University
2007 Jan 28 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 5.3663e+004 max 2.64 min
Su1/Area7: Threonine ref/1
0100200300400500600700800900100011000
1
2
3
4
5
6x 10
4 swnt_037.spe
Binding Energy (eV)
c/s
Atomic %
C1s 90.6
O1s 7.2
Fe2p3 2.1
Threonine + SWNT
-O
KL
L
-F
e2p
3 -O
1s
-C
1s
-F
e3p
swnt_038.spe: Rice University
2007 Jan 28 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 3.6504e+004 max 2.64 min
Su1/Area8: Threonine ox/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4x 10
4 swnt_038.spe
Binding Energy (eV)
c/s
Atomic %
C1s 79.0
O1s 18.5
Na1s 1.6
N1s 0.8
Threonine + ozonated SWNT
-N
a1s
-O
KL
L
-O
1s
-N
a K
LL
-N
1s
-C
1s
-F
1s
275
swnt_039.spe: Rice University
2007 Jan 28 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 5.6571e+004 max 2.64 min
Su1/Area9: Valine ref/1
0100200300400500600700800900100011000
1
2
3
4
5
6
7x 10
4 swnt_039.spe
Binding Energy (eV)
c/s
Atomic %
C1s 91.8
O1s 5.4
Fe2p3 2.8Valine + SWNT
-O
KL
L
-F
e L
MM
-F
e2p
3
-O
1s
-C
1s
-F
e3p
swnt_040.spe: Rice University
2007 Jan 28 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 3.9092e+004 max 2.64 min
Su1/Area10: Valine ox/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4
4.5x 10
4 swnt_040.spe
Binding Energy (eV)
c/s
Valine + ozonated SWNT
Atomic %
C1s 77.8
O1s 19.5
Na1s 1.5
N1s 1.2
Valine + ozonated SWNT
-N
a1s
-O
KL
L
-O
1s
-N
a K
LL
-C
1s
-N
KL
L
-N
1s
276
swnt_041.spe: Rice University
2007 Jan 28 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 5.3552e+004 max 2.64 min
Su1/Area11: Tyrosine ref/1
0100200300400500600700800900100011000
1
2
3
4
5
6x 10
4 swnt_041.spe
Binding Energy (eV)
c/s
Atomic %
C1s 90.8
O1s 6.7
Fe2p3 2.6Tyrosine + SWNT
-F
e2p
3 -O
1s
-C
1s
-F
e3p
swnt_042.spe: Rice University
2007 Jan 28 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 3.8730e+004 max 2.64 min
Su1/Area12: Tyrosine ox/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4
4.5x 10
4 swnt_042.spe
Binding Energy (eV)
c/s
Atomic %
C1s 73.8
O1s 19.2
N1s 4.9
Na1s 2.0
Tyrosine + ozonated SWNT
-N
a1s
-O
KL
L
-O
1s
-N
a K
LL
-N
1s
-C
1s
-F
KL
L
-F
1s
277
SWNT_007.spe: none Rice University
2007 Jan 12 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 5.0754e+004 max 2.64 min
Su1/Area7: GABA Na(+) H2O ref/1
0100200300400500600700800900100011000
1
2
3
4
5
6x 10
4 SWNT_007.spe
Binding Energy (eV)
c/s
Atomic %
C1s 89.5
O1s 8.1
Fe2p3 2.4GABA Na(+) / H2O + SWNT
-O
KL
L
-F
e L
MM
-F
e2p
3 -O
1s
-C
1s
-F
e3p
SWNT_008.spe: none Rice University
2007 Jan 12 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 3.9927e+004 max 2.64 min
Su1/Area8: GABA Na(+) H2O ox/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4
4.5x 10
4 SWNT_008.spe
Binding Energy (eV)
c/s
Atomic %
C1s 79.3
O1s 18.2
Na1s 2.3
N1s 0.2
GABA Na(+) / H2O + ozonated SWNT
-N
a1s
-O
KL
L
-O
1s
-N
a K
LL
-N
1s
-C
1s
278
SWNT_07.spe: none Rice University
2006 Dec 22 Al mono 39.3 W 200.0 µ 45.0° 140.00 eV 1.3325e+004 max 3.20 min
Su1/Point7: 4-aminobutyric acid/1
0200400600800100012000
5000
10000
15000SWNT_07.spe
Binding Energy (eV)
c/s
Atomic %
C1s 94.0
O1s 4.1
Fe2p3 1.8
GABA + SWNT
-F
e2p
3
-O
1s
-C
1s
-F
e3p
SWNT_08.spe: none Rice University
2006 Dec 22 Al mono 39.3 W 200.0 µ 45.0° 140.00 eV 9.3375e+003 max 3.20 min
Su1/Point8: 4-aminobutyric acid ox/1
0200400600800100012000
2000
4000
6000
8000
10000
12000SWNT_08.spe
Binding Energy (eV)
c/s
GABA + ozonated SWNTAtomic %
C1s 73.0
O1s 20.6
Na1s 3.4
N1s 3.0
-N
a1s
-O
KL
L -O
1s
-N
a K
LL
-N
1s
-C
1s
279
SWNT_52.spe: none Rice University
2006 Dec 22 Al mono 39.3 W 200.0 µ 45.0° 26.00 eV 1.5688e+002 max 11.28 min
O1s/Point7: 4-aminobutyric acid/1
2503003504004505005500
100
200
300
400
500
600
700SWNT_52.spe
Binding Energy (eV)
c/s
Atomic %
C1s 95.4
N1s 2.6
O1s 2.0
GABA + SWNT
SWNT_52.spe: none Rice University
2006 Dec 22 Al mono 39.3 W 200.0 µ 45.0° 26.00 eV 1.5688e+002 max 11.28 min
O1s/Point7: 4-aminobutyric acid/1
2802852902950
100
200
300
400
500
600
Binding Energy (eV)
c/s
C1s
3943963984004024040
5
10
15
20
25
30
Binding Energy (eV)
c/s
N1s
5305355400
10
20
30
40
50
60
Binding Energy (eV)
c/s
O1s
280
SWNT_53.spe: none Rice University
2006 Dec 22 Al mono 39.3 W 200.0 µ 45.0° 26.00 eV 2.8417e+002 max 11.28 min
O1s/Point8: 4-aminobutyric acid ox/1
2503003504004505005500
50
100
150
200
250
300
350
400
450
500SWNT_53.spe
Binding Energy (eV)
c/s
Atomic %
C1s 76.7
O1s 22.2
N1s 1.1
GABA + ozonated SWNT
SWNT_53.spe: none Rice University
2006 Dec 22 Al mono 39.3 W 200.0 µ 45.0° 26.00 eV 2.8417e+002 max 11.28 min
O1s/Point8: 4-aminobutyric acid ox/1
2802852902950
100
200
300
400
500
Binding Energy (eV)
c/s
C1s
3943963984004024040
10
20
30
40
Binding Energy (eV)
c/s
N1s
5305355400
50
100
150
200
Binding Energy (eV)
c/s
O1s
281
swnt_045.spe: Rice University
2007 Jan 28 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 5.2814e+004 max 2.64 min
Su1/Area15: Urea ref/1
0100200300400500600700800900100011000
1
2
3
4
5
6x 10
4 swnt_045.spe
Binding Energy (eV)
c/s
Atomic %
C1s 90.0
O1s 8.4
Fe2p3 1.6Urea + SWNT
-O
KL
L
-F
e L
MM
-F
e2p
3
-O
1s
-C
1s
-F
e3p
swnt_046.spe: Rice University
2007 Jan 28 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 3.7853e+004 max 2.64 min
Su1/Area16: Urea ox/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4
4.5x 10
4 swnt_046.spe
Binding Energy (eV)
c/s
Atomic %
C1s 78.3
O1s 21.7Urea + ozonated SWNT
-O
KL
L
-O
KL
L
-O
1s
-C
1s
282
swnt_047.spe: Rice University
2007 Jan 28 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 5.4261e+004 max 2.64 min
Su1/Area17: uracil ref/1
0100200300400500600700800900100011000
1
2
3
4
5
6x 10
4 swnt_047.spe
Binding Energy (eV)
c/s
Atomic %
C1s 93.3
O1s 6.7Uracil + SWNT
-O
KL
L
-O
1s
-C
1s
swnt_048.spe: Rice University
2007 Jan 28 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 4.5396e+004 max 2.64 min
Su1/Area18: Uracil ox/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5x 10
4 swnt_048.spe
Binding Energy (eV)
c/s
Atomic %
C1s 74.4
O1s 23.0
N1s 1.3
Na1s 1.3
Uracil + ozonated SWNT
-N
a1s
-O
KL
L
-O
KL
L
-O
1s
-N
a K
LL
-N
1s
-C
1s
-F
KL
L1
-F
KL
L
-F
2s
-F
1s
283
swnt_049.spe: Rice University
2007 Jan 28 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 5.4359e+004 max 2.64 min
Su1/Area19: adenine ref/1
0100200300400500600700800900100011000
1
2
3
4
5
6x 10
4 swnt_049.spe
Binding Energy (eV)
c/s
Atomic %
C1s 91.8
O1s 8.2Adenine + SWNT
-O
KL
L
-O
1s
-C
1s
swnt_050.spe: Rice University
2007 Jan 28 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 3.4750e+004 max 2.64 min
Su1/Area20: adenine ox/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4x 10
4 swnt_050.spe
Binding Energy (eV)
c/s
Atomic %
C1s 76.1
O1s 21.2
Na1s 2.7Adenine + ozonated SWNT
-N
a1s
-O
KL
L
-O
1s
-N
a K
LL
-C
1s
-F
KL
L
-F
1s
284
swnt_061.spe: Rice University
2007 Jan 28 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 5.2480e+004 max 2.64 min
Su1/Area31: 2-mercaptoPy ref/1
0100200300400500600700800900100011000
1
2
3
4
5
6x 10
4 swnt_061.spe
Binding Energy (eV)
c/s
Atomic %
C1s 91.7
O1s 6.8
Fe2p3 1.42-mercaptopyridine + SWNT
-F
e L
MM
-F
e L
MM
-F
e2p
3
-O
1s
-C
1s
-F
e3p
swnt_062.spe: Rice University
2007 Jan 28 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 4.2919e+004 max 2.64 min
Su1/Area32: 2-mercaptoPy ox/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5x 10
4 swnt_062.spe
Binding Energy (eV)
c/s
Atomic %
C1s 84.3
O1s 15.7
N1s <.12-mercaptopyridine + ozonated SWNT
-O
KL
L
-O
1s
-N
1s
-C
1s
-F
KL
L -F
1s
285
swnt_051.spe: Rice University
2007 Jan 28 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 5.5208e+004 max 2.64 min
Su1/Area21: Nh4OH 5x dil ref/1
0100200300400500600700800900100011000
1
2
3
4
5
6
7x 10
4 swnt_051.spe
Binding Energy (eV)
c/s
Atomic %
C1s 89.9
O1s 6.7
Fe2p3 3.4NH4OH 5x diluted + SWNT
-F
e L
MM
-F
e2p
3
-O
1s
-C
1s
-F
e3p
swnt_052.spe: Rice University
2007 Jan 28 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 3.6865e+004 max 2.64 min
Su1/Area22: NH4OH 5x dl ox/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4
4.5x 10
4 swnt_052.spe
Binding Energy (eV)
c/s
Atomic %
C1s 77.9
O1s 17.6
N1s 4.6
NH4OH 5x diluted + ozonated SWNT
-O
KL
L
-O
1s
-N
1s
-C
1s
-F
KL
L
-F
1s
286
swnt_01.spe: none Rice University
2006 Nov 11 Al mono 42.0 W 200.0 µ 45.0° 140.00 eV 2.9513e+004 max 2.93 min
Su1/Point1: No2/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5x 10
4 swnt_01.spe
Binding Energy (eV)
c/s
Atomic %
C1s 98.1
O1s 1.9
MeOCH2CH2NH2 + SWNT
-O
1s
-C
1s
swnt_02.spe: none Rice University
2006 Nov 11 Al mono 42.0 W 200.0 µ 45.0° 140.00 eV 2.1363e+004 max 2.93 min
Su1/Point2: No2ox/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5x 10
4 swnt_02.spe
Binding Energy (eV)
c/s
Atomic %
C1s 83.7
O1s 14.3
N1s 2.0
MeOCH2CH2NH2 + ozonated SWNT
-O
KL
L
-O
1s
-N
1s
-C
1s
287
swnt_09.spe: none Rice University
2006 Nov 11 Al mono 42.0 W 200.0 µ 45.0° 26.00 eV 6.8229e+002 max 9.68 min
N1s/Point1: No2/1
2503003504004505005500
500
1000
1500
2000
2500
3000
3500
4000
4500
5000swnt_09.spe
Binding Energy (eV)
c/s
Atomic %
C1s 96.9
O1s 3.1
N1s <.1
MeOCH2CH2NH2 + SWNT
swnt_09.spe: none Rice University
2006 Nov 11 Al mono 42.0 W 200.0 µ 45.0° 26.00 eV 6.8229e+002 max 9.68 min
N1s/Point1: No2/1
2802852902950
1000
2000
3000
4000
5000
Binding Energy (eV)
c/s
C1s
5255305355400
50
100
150
200
Binding Energy (eV)
c/s
O1s
3943963984004024040
20
40
60
80
Binding Energy (eV)
c/s
N1s
288
swnt_10.spe: none Rice University
2006 Nov 11 Al mono 42.0 W 200.0 µ 45.0° 26.00 eV 7.2771e+002 max 9.68 min
N1s/Point2: No2ox/1
2503003504004505005500
500
1000
1500
2000
2500
3000
3500swnt_10.spe
Binding Energy (eV)
c/s
Atomic %
C1s 83.1
O1s 15.3
N1s 1.6
MeOCH2CH2NH2 + ozonated SWNT
swnt_10.spe: none Rice University
2006 Nov 11 Al mono 42.0 W 200.0 µ 45.0° 26.00 eV 7.2771e+002 max 9.68 min
N1s/Point2: No2ox/1
2802852902950
500
1000
1500
2000
2500
3000
Binding Energy (eV)
c/s
C1s
5255305355400
200
400
600
800
1000
Binding Energy (eV)
c/s
O1s
3943963984004024040
50
100
150
200
Binding Energy (eV)
c/s
N1s
289
swnt_053.spe: Rice University
2007 Jan 28 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 4.9850e+004 max 2.64 min
Su1/Area23: MeOCH2CH2NH2 ref/1
0100200300400500600700800900100011000
1
2
3
4
5
6x 10
4 swnt_053.spe
Binding Energy (eV)
c/s
Atomic %
C1s 92.7
O1s 7.32-methoxyethylamine + SWNT
-O
1s
-C
1s
swnt_054.spe: Rice University
2007 Jan 28 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 3.5766e+004 max 2.64 min
Su1/Area24: MeOCH2CH2NH2 1ox/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4x 10
4 swnt_054.spe
Binding Energy (eV)
c/s
Atomic %
C1s 80.5
O1s 16.7
N1s 2.8
MeO-CH2CH2-NH2 + ozonated SWNT (1st cycle)
-O
KL
L
-O
1s
-N
1s
-C
1s
290
swnt_055.spe: Rice University
2007 Jan 28 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 2.9796e+004 max 2.64 min
Su1/Area25: MeOCH2CH2NH2 2ox/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5x 10
4 swnt_055.spe
Binding Energy (eV)
c/s
Atomic %
C1s 72.5
O1s 23.0
N1s 4.5
MeOCH2CH2NH2+ozonated SWNT (2nd cycle)
-O
KL
L
-O
1s
-N
1s
-C
1s
-F
KL
L
-F
1s
swnt_056.spe: Rice University
2007 Jan 28 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 2.9907e+004 max 2.64 min
Su1/Area26: MeOCH2CH2NH2 3ox/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5x 10
4 swnt_056.spe
Binding Energy (eV)
c/s
Atomic %
C1s 69.0
O1s 25.0
N1s 6.0
MeOCH2CH2NH2+ ozonated SWNT (3rd cycle)
-O
KL
L
-O
1s
-N
1s
-C
1s
-F
KL
L
-F
1s
291
swnt_057.spe: Rice University
2007 Jan 28 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 2.9462e+004 max 2.64 min
Su1/Area27: MeOCH2CH2NH2 4ox/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5x 10
4 swnt_057.spe
Binding Energy (eV)
c/s
Atomic %
C1s 66.0
O1s 26.2
N1s 7.8
MeOCH2CH2NH2+ ozonated SWNT (4th cycle)
-O
KL
L
-O
KL
L
-O
1s
-N
1s
-C
1s
-F
KL
L
-F
1s
292
SWNT_001.spe: none Rice University
2007 Jan 12 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 4.2502e+004 max 2.64 min
Su1/Area1: MeOCH2CH2NH2 ref/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5x 10
4 SWNT_001.spe
Binding Energy (eV)
c/s
Atomic %
C1s 94.2
O1s 5.8MeOCH2CH2NH2 + SWNT
-O
1s
-C
1s
SWNT_002.spe: none Rice University
2007 Jan 12 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 3.9287e+004 max 2.64 min
Su1/Area2: MeOCH2CH2NH2 ox 0min /1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4
4.5x 10
4 SWNT_002.spe
Binding Energy (eV)
c/s
Atomic %
C1s 82.4
O1s 14.1
N1s 3.5
MeOCH2CH2NH2 + ozonated SWNT, mixed at 0 min
-O
KL
L -O
1s
-C
1s
-N
KL
L
-N
1s
293
SWNT_003.spe: none Rice University
2007 Jan 12 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 4.0665e+004 max 2.64 min
Su1/Area3: MeOCH2CH2NH2 ox 10min /1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4
4.5x 10
4 SWNT_003.spe
Binding Energy (eV)
c/s
Atomic %
C1s 83.3
O1s 15.0
N1s 1.7
MeOCH2CH2NH2 + ozonated SWNT, mixed after 10 min
-O
KL
L -O
1s
-N
1s
-C
1s
SWNT_004.spe: none Rice University
2007 Jan 12 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 4.0136e+004 max 2.64 min
Su1/Area4: MeOCH2CH2NH2 ox 20min /1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4
4.5x 10
4 SWNT_004.spe
Binding Energy (eV)
c/s
Atomic %
C1s 81.5
O1s 15.2
N1s 3.2
MeOCH2CH2NH2+ ozonated SWNT, mixed after 20 min
-O
KL
L
-O
1s
-N
1s
-C
1s
294
SWNT_005.spe: none Rice University
2007 Jan 12 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 2.7959e+004 max 2.64 min
Su1/Area5: MeOCH2CH2NH2 ox 40min /1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5x 10
4 SWNT_005.spe
Binding Energy (eV)
c/s
Atomic %
C1s 80.8
O1s 15.5
N1s 3.7
MeOCH2CH2NH2+ ozonated SWNT, mixed after 40 min
-O
KL
L -O
1s
-N
1s
-C
1s
SWNT_006.spe: none Rice University
2007 Jan 12 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 3.7519e+004 max 2.64 min
Su1/Area6: MeOCH2CH2NH2 ox 1hr /1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4
4.5x 10
4 SWNT_006.spe
Binding Energy (eV)
c/s
Atomic %
C1s 84.0
O1s 15.1
N1s 0.8
Fe2p3 0.2
MeOCH2CH2NH2 + ozonated SWNT, mixed after 60 min
-O
KL
L
-F
e2p
3
-O
1s
-N
1s
-C
1s
-F
e3p
295
swnt_063.spe: Rice University
2007 Jan 28 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 5.8784e+004 max 2.64 min
Su1/Area33: HEPES ref/1
0100200300400500600700800900100011000
1
2
3
4
5
6
7x 10
4 swnt_063.spe
Binding Energy (eV)
c/s
Atomic %
C1s 90.7
O1s 7.3
Fe2p3 2.0
HEPES + SWNT
-F
e2p
3
-O
1s
-C
1s
-F
e3p
swnt_064.spe: Rice University
2007 Jan 28 Al mono 99.9 W 100.0 µ 45.0° 140.00 eV 4.0609e+004 max 2.64 min
Su1/Area34: HEPES ox/1
0100200300400500600700800900100011000
0.5
1
1.5
2
2.5
3
3.5
4
4.5x 10
4 swnt_064.spe
Binding Energy (eV)
c/s
Atomic %
C1s 74.9
O1s 20.2
N1s 3.7
Na1s 1.2
HEPES + ozonated SWNT
-N
a1s
-O
KL
L
-O
1s
-N
a K
LL
-N
1s
-C
1s
-F
1s
296
swnt_01.spe: none Rice University
2006 Nov 13 Al mono 42.0 W 200.0 µ 45.0° 140.00 eV 3.4575e+004 max 3.47 min
Su1/Point1: aniline/1
0200400600800100012000
0.5
1
1.5
2
2.5
3
3.5
4x 10
4 swnt_01.spe
Binding Energy (eV)
c/s
Atomic %
C1s 97.4
O1s 2.6
Atomic %
C1s 97.4
O1s 2.6aniline + SWNT
-O
1s
-C
1s
swnt_02.spe: none Rice University
2006 Nov 13 Al mono 42.0 W 200.0 µ 45.0° 140.00 eV 2.3500e+004 max 3.47 min
Su1/Point2: aniline ox/1
0200400600800100012000
0.5
1
1.5
2
2.5
3x 10
4 swnt_02.spe
Binding Energy (eV)
c/s
Atomic %
C1s 85.0
O1s 13.9
N1s 1.1aniline + ozonated SWNT
-O
KL
L -O
1s
-N
1s
-C
1s
297
swnt_07.spe: none Rice University
2006 Nov 13 Al mono 42.0 W 200.0 µ 45.0° 140.00 eV 3.0913e+004 max 3.47 min
Su1/Point7: Et3N/1
0200400600800100012000
0.5
1
1.5
2
2.5
3
3.5x 10
4 swnt_07.spe
Binding Energy (eV)
c/s
Atomic %
C1s 94.0
O1s 4.5
Fe2p3 1.5Et3N + SWNT
-C
KL
L
-F
e2p
3
-O
1s
-C
1s
-F
e3p
swnt_08.spe: none Rice University
2006 Nov 13 Al mono 42.0 W 200.0 µ 45.0° 140.00 eV 2.7075e+004 max 3.47 min
Su1/Point8: Et3N ox/1
0200400600800100012000
0.5
1
1.5
2
2.5
3x 10
4 swnt_08.spe
Binding Energy (eV)
c/s
Atomic %
C1s 85.4
O1s 11.7
N1s 2.6
Et3N + ozonated SWNT
-C
KL
L
-O
KL
L -O
1s
-C
1s
-N
KL
L
-N
1s
298
swnt_15.spe: none Rice University
2006 Nov 14 Al mono 42.0 W 200.0 µ 45.0° 140.00 eV 2.6600e+004 max 3.47 min
Su1/Point15: allyl amine/1
0200400600800100012000
0.5
1
1.5
2
2.5
3x 10
4 swnt_15.spe
Binding Energy (eV)
c/s
Atomic %
C1s 94.0
O1s 5.6
I3d5 0.4
N1s <.1
allyl amine + SWNT
-I
MN
N
-I3
d3
-I3
d5
-O
1s
-C
1s
-N
KL
L
-N
1s
swnt_16.spe: none Rice University
2006 Nov 14 Al mono 42.0 W 200.0 µ 45.0° 140.00 eV 2.5925e+004 max 3.47 min
Su1/Point16: allyl amine ox/1
0200400600800100012000
0.5
1
1.5
2
2.5
3x 10
4 swnt_16.spe
Binding Energy (eV)
c/s
Atomic %
C1s 83.8
O1s 13.2
N1s 3.1
allyl amine + ozonated SWNT
-C
KL
L
-O
KL
L
-O
1s
-N
1s
-C
1s
299
swnt_08.spe: none Rice University
2006 Nov 13 Al mono 42.0 W 200.0 µ 45.0° 140.00 eV 3.0925e+004 max 3.47 min
Su1/Point7: F-C6H4-SH/1
0200400600800100012000
0.5
1
1.5
2
2.5
3
3.5x 10
4 swnt_08.spe
Binding Energy (eV)
c/s
Atomic %
C1s 93.4
O1s 4.7
Fe2p3 2.0
F-C6H4-SH + SWNT
-F
e L
MM
-F
e2p
3
-O
1s
-C
1s
-F
e3p
swnt_09.spe: none Rice University
2006 Nov 13 Al mono 42.0 W 200.0 µ 45.0° 140.00 eV 3.1213e+004 max 3.47 min
Su1/Point8: F-C6H4-SH ox/1
0200400600800100012000
0.5
1
1.5
2
2.5
3
3.5x 10
4 swnt_09.spe
Binding Energy (eV)
c/s
Atomic %
C1s 91.9
O1s 7.9
S2p 0.2
-F
KL
L1
-F
KL
L
F-C6H4-SH + ozonated SWNT
-C
KL
L
-O
1s
-C
1s
-S
2s
-S
2p
-F
1s
Experiment with F-C6H4-SH was recorded on indium substrate.
300
swnt_19.spe: none Rice University
2006 Nov 14 Al mono 42.0 W 200.0 µ 45.0° 140.00 eV 3.4300e+004 max 3.47 min
Su1/Point19: F-C6H4-SH/1
0200400600800100012000
0.5
1
1.5
2
2.5
3
3.5
4x 10
4 swnt_19.spe
Binding Energy (eV)
c/s
Atomic %
C1s 94.9
O1s 5.1F-C6H4-SH + SWNT
-C
KL
L
-O
1s
-C
1s
swnt_20.spe: none Rice University
2006 Nov 14 Al mono 42.0 W 200.0 µ 45.0° 140.00 eV 3.3863e+004 max 3.47 min
Su1/Point20: F-C6H4-SH ox/1
0200400600800100012000
0.5
1
1.5
2
2.5
3
3.5
4x 10
4 swnt_20.spe
Binding Energy (eV)
c/s
-F
KL
L
Atomic %
C1s 89.4
O1s 7.1
F1s 2.0
S2p 1.5
F-C6H4-SH + ozonated SWNT
-C
KL
L
-O
1s
-C
1s
-S
2s
-S
2p
-F
1s
301
swnt_21.spe: none Rice University
2006 Nov 14 Al mono 42.0 W 200.0 µ 45.0° 26.00 eV 1.4563e+002 max 12.88 min
S2p/Point19: F-C6H4-SH/1
1002003004005006007000
500
1000
1500
2000
2500
3000
3500
4000
4500swnt_21.spe
Binding Energy (eV)
c/s
Atomic %
C1s 95.6
O1s 4.4
F1s 0.1
S2p <.1
F-C6H4-SH + SWNT
swnt_21.spe: none Rice University
2006 Nov 14 Al mono 42.0 W 200.0 µ 45.0° 26.00 eV 1.4563e+002 max 12.88 min
S2p/Point19: F-C6H4-SH/1
2802852902950
1000
2000
3000
4000
5000
Binding Energy (eV)
c/s
C1s
5305355400
50
100
150
200
250
300
Binding Energy (eV)
c/s
O1s
6856906950
20
40
60
80
Binding Energy (eV)
c/s
F1s
1601651700
10
20
30
40
50
Binding Energy (eV)
c/s
S2p
302
swnt_22.spe: none Rice University
2006 Nov 14 Al mono 42.0 W 200.0 µ 45.0° 26.00 eV 2.0896e+002 max 12.88 min
S2p/Point20: F-C6H4-SH ox/1
1002003004005006007000
500
1000
1500
2000
2500
3000
3500
4000
4500swnt_22.spe
Binding Energy (eV)
c/s
Atomic %
C1s 89.7
O1s 7.5
F1s 1.9
S2p 1.0
F-C6H4-SH + ozonated SWNT
swnt_22.spe: none Rice University
2006 Nov 14 Al mono 42.0 W 200.0 µ 45.0° 26.00 eV 2.0896e+002 max 12.88 min
S2p/Point20: F-C6H4-SH ox/1
2802852902950
1000
2000
3000
4000
5000
Binding Energy (eV)
c/s
C1s
5305355400
100
200
300
400
500
Binding Energy (eV)
c/s
O1s
6856906950
50
100
150
200
Binding Energy (eV)
c/s
F1s
1601651700
20
40
60
80
Binding Energy (eV)
c/s
S2p
303
swnt_17.spe: none Rice University
2006 Nov 14 Al mono 42.0 W 200.0 µ 45.0° 140.00 eV 3.2725e+004 max 3.47 min
Su1/Point17: ethylene diamine/1
0200400600800100012000
0.5
1
1.5
2
2.5
3
3.5
4x 10
4 swnt_17.spe
Binding Energy (eV)
c/s
Atomic %
C1s 97.9
O1s 2.1
ethylene diamine + SWNT
-O
1s
-C
1s
swnt_18.spe: none Rice University
2006 Nov 14 Al mono 42.0 W 200.0 µ 45.0° 140.00 eV 2.3975e+004 max 3.47 min
Su1/Point18: ethylene diamine ox/1
0200400600800100012000
0.5
1
1.5
2
2.5
3x 10
4 swnt_18.spe
Binding Energy (eV)
c/s
Atomic %
C1s 82.9
O1s 10.3
N1s 4.6
Fe2p3 2.2
ethylene diamine + ozonated SWNT
-C
KL
L
-O
KL
L
-F
e2p
3
-O
1s
-N
1s
-C
1s
-F
e3p
304
swnt_23.spe: none Rice University
2006 Nov 15 Al mono 42.0 W 200.0 µ 45.0° 26.00 eV 5.8563e+002 max 8.08 min
N1s/Point17: ethylene diamine/1
2503003504004505005500
500
1000
1500
2000
2500
3000
3500
4000
4500
5000swnt_23.spe
Binding Energy (eV)
c/s
Atomic %
C1s 96.6
O1s 2.3
N1s 1.1
ethylene diamine + SWNT
swnt_23.spe: none Rice University
2006 Nov 15 Al mono 42.0 W 200.0 µ 45.0° 26.00 eV 5.8563e+002 max 8.08 min
N1s/Point17: ethylene diamine/1
2802852902950
1000
2000
3000
4000
5000
Binding Energy (eV)
c/s
C1s
5305355400
50
100
150
200
Binding Energy (eV)
c/s
O1s
3943963984004024040
20
40
60
80
100
Binding Energy (eV)
c/s
N1s
305
swnt_24.spe: none Rice University
2006 Nov 15 Al mono 42.0 W 200.0 µ 45.0° 26.00 eV 7.9146e+002 max 8.08 min
N1s/Point18: ethylene diamine ox/1
2503003504004505005500
500
1000
1500
2000
2500
3000
3500
4000swnt_24.spe
Binding Energy (eV)
c/s
Atomic %
C1s 83.0
O1s 11.2
N1s 5.9
ethylene diamine + ozonated SWNT
swnt_24.spe: none Rice University
2006 Nov 15 Al mono 42.0 W 200.0 µ 45.0° 26.00 eV 7.9146e+002 max 8.08 min
N1s/Point18: ethylene diamine ox/1
2802852902950
1000
2000
3000
4000
Binding Energy (eV)
c/s
C1s
5305355400
100
200
300
400
500
600
Binding Energy (eV)
c/s
O1s
3943963984004024040
100
200
300
400
Binding Energy (eV)
c/s
N1s
306
swnt_03.spe: none Rice University
2006 Nov 14 Al mono 42.0 W 200.0 µ 45.0° 140.00 eV 3.4188e+004 max 3.47 min
Su1/Point3: H2NCH2CH2OH/1
0200400600800100012000
0.5
1
1.5
2
2.5
3
3.5
4x 10
4 swnt_03.spe
Binding Energy (eV)
c/s
2-hydroxyethylamine + SWNT
Atomic %
C1s 95.8
O1s 2.5
Fe2p3 1.7
-C
KL
L
-F
e2p
3
-O
1s
-C
1s
-F
e3p
swnt_04.spe: none Rice University
2006 Nov 14 Al mono 42.0 W 200.0 µ 45.0° 140.00 eV 2.3613e+004 max 3.47 min
Su1/Point4: H2NCH2CH2OH ox/1
0200400600800100012000
0.5
1
1.5
2
2.5
3x 10
4 swnt_04.spe
Binding Energy (eV)
c/s
2-hydroxyethylamine + ozonated SWNTAtomic %
C1s 81.6
O1s 13.8
N1s 4.6
-C
KL
L
-O
KL
L
-O
1s
-N
1s
-C
1s
307
swnt_25.spe: none Rice University
2006 Nov 15 Al mono 42.0 W 200.0 µ 45.0° 26.00 eV 6.7604e+002 max 8.08 min
N1s/Point3: H2NCH2CH2OH/1
2503003504004505005500
1000
2000
3000
4000
5000
6000swnt_25.spe
Binding Energy (eV)
c/s
Atomic %
C1s 96.6
O1s 3.0
N1s 0.5
ethanolamine + SWNT
swnt_25.spe: none Rice University
2006 Nov 15 Al mono 42.0 W 200.0 µ 45.0° 26.00 eV 6.7604e+002 max 8.08 min
N1s/Point3: H2NCH2CH2OH/1
2802852902950
1000
2000
3000
4000
5000
Binding Energy (eV)
c/s
C1s
5305355400
50
100
150
200
250
Binding Energy (eV)
c/s
O1s
3943963984004024040
20
40
60
80
100
Binding Energy (eV)
c/s
N1s
308
swnt_26.spe: none Rice University
2006 Nov 15 Al mono 42.0 W 200.0 µ 45.0° 26.00 eV 6.9208e+002 max 8.08 min
N1s/Point4: H2NCH2CH2OH ox/1
2503003504004505005500
500
1000
1500
2000
2500
3000
3500swnt_26.spe
Binding Energy (eV)
c/s
Atomic %
C1s 81.7
O1s 15.1
N1s 3.2
ethanolamine + ozonated SWNT
swnt_26.spe: none Rice University
2006 Nov 15 Al mono 42.0 W 200.0 µ 45.0° 26.00 eV 6.9208e+002 max 8.08 min
N1s/Point4: H2NCH2CH2OH ox/1
2802852902950
1000
2000
3000
4000
Binding Energy (eV)
c/s
C1s
5305355400
200
400
600
800
1000
Binding Energy (eV)
c/s
O1s
3943963984004024040
50
100
150
200
Binding Energy (eV)
c/s
N1s
309
swnt_13.spe: none Rice University
2006 Nov 14 Al mono 42.0 W 200.0 µ 45.0° 140.00 eV 3.4375e+004 max 3.47 min
Su1/Point13: isoamyl amine/1
0200400600800100012000
0.5
1
1.5
2
2.5
3
3.5
4x 10
4 swnt_13.spe
Binding Energy (eV)
c/s
Atomic %
C1s 96.2
Fe2p3 2.5
O1s 1.3
isoamyl amine + SWNT
-C
KL
L
-F
e2p
3
-O
1s
-C
1s
-F
e3p
swnt_14.spe: none Rice University
2006 Nov 14 Al mono 42.0 W 200.0 µ 45.0° 140.00 eV 2.6350e+004 max 3.47 min
Su1/Point14: isoamyl amine ox/1
0200400600800100012000
0.5
1
1.5
2
2.5
3x 10
4 swnt_14.spe
Binding Energy (eV)
c/s
Atomic %
C1s 84.9
O1s 10.6
N1s 4.5
isoamyl amine + ozonated SWNT
-C
KL
L
-O
KL
L
-O
1s
-N
1s
-C
1s
310
swnt_27.spe: none Rice University
2006 Nov 15 Al mono 42.0 W 200.0 µ 45.0° 26.00 eV 6.1063e+002 max 8.08 min
N1s/Point13: isoamyl amine/1
2503003504004505005500
500
1000
1500
2000
2500
3000
3500
4000
4500
5000swnt_27.spe
Binding Energy (eV)
c/s
Atomic %
C1s 96.7
O1s 2.7
N1s 0.6
isoamyl amine + SWNT
swnt_27.spe: none Rice University
2006 Nov 15 Al mono 42.0 W 200.0 µ 45.0° 26.00 eV 6.1063e+002 max 8.08 min
N1s/Point13: isoamyl amine/1
2802852902950
1000
2000
3000
4000
5000
Binding Energy (eV)
c/s
C1s
5305355400
50
100
150
Binding Energy (eV)
c/s
O1s
3943963984004024040
20
40
60
80
Binding Energy (eV)
c/s
N1s
311
swnt_28.spe: none Rice University
2006 Nov 15 Al mono 42.0 W 200.0 µ 45.0° 26.00 eV 6.3938e+002 max 8.08 min
N1s/Point14: isoamyl amine ox/1
2503003504004505005500
500
1000
1500
2000
2500
3000
3500
4000swnt_28.spe
Binding Energy (eV)
c/s
Atomic %
C1s 87.3
O1s 11.1
N1s 1.7
Isoamyl amine + ozonated SWNT
swnt_28.spe: none Rice University
2006 Nov 15 Al mono 42.0 W 200.0 µ 45.0° 26.00 eV 6.3938e+002 max 8.08 min
N1s/Point14: isoamyl amine ox/1
2802852902950
1000
2000
3000
4000
Binding Energy (eV)
c/s
C1s
5305355400
100
200
300
400
500
600
Binding Energy (eV)
c/s
O1s
3943963984004024040
50
100
150
Binding Energy (eV)
c/s
N1s
312
swnt_01.spe: none Rice University
2006 Nov 14 Al mono 42.0 W 200.0 µ 45.0° 140.00 eV 3.7950e+004 max 3.47 min
Su1/Point1: PhSH/1
0200400600800100012000
0.5
1
1.5
2
2.5
3
3.5
4
4.5x 10
4 swnt_01.spe
Binding Energy (eV)
c/s
PhSH + SWNT Atomic %
C1s 96.6
O1s 1.9
S2p 1.5
-C
KL
L
-O
1s
-C
1s
-S
2s
-S
2p
swnt_02.spe: none Rice University
2006 Nov 14 Al mono 42.0 W 200.0 µ 45.0° 140.00 eV 3.2175e+004 max 3.47 min
Su1/Point2: PhSH ox/1
0200400600800100012000
0.5
1
1.5
2
2.5
3
3.5
4x 10
4 swnt_02.spe
Binding Energy (eV)
c/s
Atomic %
C1s 92.9
O1s 6.1
S2p 0.9
PhSH + ozonated SWNT
-C
KL
L
-O
1s
-C
1s
-S
2s
-S
2p
313
swnt_29.spe: none Rice University
2006 Nov 15 Al mono 42.0 W 200.0 µ 45.0° 26.00 eV 5.4188e+002 max 12.88 min
S2p/Point1: PhSH/1
1502002503003504004505005500
1000
2000
3000
4000
5000
6000swnt_29.spe
Binding Energy (eV)
c/s
Atomic %
C1s 97.1
O1s 2.0
S2p 0.9
PhSH + SWNT
swnt_29.spe: none Rice University
2006 Nov 15 Al mono 42.0 W 200.0 µ 45.0° 26.00 eV 5.4188e+002 max 12.88 min
S2p/Point1: PhSH/1
2802852902950
1000
2000
3000
4000
5000
6000
Binding Energy (eV)
c/s
C1s
5305355400
50
100
150
Binding Energy (eV)
c/s
O1s
1601651700
50
100
150
200
Binding Energy (eV)
c/s
S2p
314
swnt_30.spe: none Rice University
2006 Nov 15 Al mono 42.0 W 200.0 µ 45.0° 26.00 eV 3.7542e+002 max 12.88 min
S2p/Point2: PhSH ox/1
1502002503003504004505005500
500
1000
1500
2000
2500
3000
3500
4000
4500
5000swnt_30.spe
Binding Energy (eV)
c/s
Atomic %
C1s 92.1
O1s 6.2
S2p 1.7
PhSH + ozonated SWNT
swnt_30.spe: none Rice University
2006 Nov 15 Al mono 42.0 W 200.0 µ 45.0° 26.00 eV 3.7542e+002 max 12.88 min
S2p/Point2: PhSH ox/1
2802852902950
1000
2000
3000
4000
5000
Binding Energy (eV)
c/s
C1s
5305355400
100
200
300
400
Binding Energy (eV)
c/s
O1s
1601651700
50
100
150
Binding Energy (eV)
c/s
S2p
315
SWNT_09.spe: none Rice University
2006 Dec 22 Al mono 39.3 W 200.0 µ 45.0° 140.00 eV 1.3088e+004 max 3.20 min
Su1/Point9: 2-aminoethanol/1
0200400600800100012000
5000
10000
15000SWNT_09.spe
Binding Energy (eV)
c/s
ethanolamine + SWNT
Atomic %
C1s 92.8
O1s 7.2
-O
1s
-C
1s
SWNT_10.spe: none Rice University
2006 Dec 22 Al mono 39.3 W 200.0 µ 45.0° 140.00 eV 8.7750e+003 max 3.20 min
Su1/Point10: 2-aminoethanol ox/1
0200400600800100012000
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000SWNT_10.spe
Binding Energy (eV)
c/s
Atomic %
C1s 77.0
O1s 18.1
N1s 4.9
ethanolamine + ozonated SWNT
-O
KL
L
-O
1s
-N
1s
-C
1s
316
SWNT_54.spe: none Rice University
2006 Dec 22 Al mono 39.3 W 200.0 µ 45.0° 26.00 eV 2.2187e+002 max 11.28 min
O1s/Point9: 2-aminoethanol/1
2503003504004505005500
100
200
300
400
500
600
700SWNT_54.spe
Binding Energy (eV)
c/s
Atomic %
C1s 93.2
O1s 6.3
N1s 0.6
ethanolamine + SWNT
SWNT_54.spe: none Rice University
2006 Dec 22 Al mono 39.3 W 200.0 µ 45.0° 26.00 eV 2.2187e+002 max 11.28 min
O1s/Point9: 2-aminoethanol/1
2802852902950
200
400
600
800
Binding Energy (eV)
c/s
C1s
3943963984004024040
5
10
15
20
25
30
Binding Energy (eV)
c/s
N1s
5305355400
20
40
60
80
100
Binding Energy (eV)
c/s
O1s
317
SWNT_55.spe: none Rice University
2006 Dec 22 Al mono 39.3 W 200.0 µ 45.0° 26.00 eV 2.8958e+002 max 11.28 min
O1s/Point10: 2-aminoethanol ox/1
2503003504004505005500
50
100
150
200
250
300
350
400
450
500SWNT_55.spe
Binding Energy (eV)
c/s
Atomic %
C1s 78.9
O1s 18.9
N1s 2.2
ethanolamine + ozonated SWNT
SWNT_55.spe: none Rice University
2006 Dec 22 Al mono 39.3 W 200.0 µ 45.0° 26.00 eV 2.8958e+002 max 11.28 min
O1s/Point10: 2-aminoethanol ox/1
2802852902950
100
200
300
400
500
Binding Energy (eV)
c/s
C1s
3943963984004024040
10
20
30
40
50
60
Binding Energy (eV)
c/s
N1s
5305355400
50
100
150
200
Binding Energy (eV)
c/s
O1s
318
swnt_05.spe: none Rice University
2006 Nov 16 Al mono 42.0 W 200.0 µ 45.0° 140.00 eV 3.0800e+004 max 3.47 min
Su1/Point5: NaCN/1
0200400600800100012000
0.5
1
1.5
2
2.5
3
3.5x 10
4 swnt_05.spe
Binding Energy (eV)
c/s
Atomic %
C1s 90.6
O1s 7.9
Fe2p3 1.5NaCN + SWNT
-F
e2p
3
-O
1s
-C
1s
-F
e3p
swnt_06.spe: none Rice University
2006 Nov 16 Al mono 42.0 W 200.0 µ 45.0° 140.00 eV 2.5275e+004 max 3.47 min
Su1/Point6: NaCN ox/1
0200400600800100012000
0.5
1
1.5
2
2.5
3x 10
4 swnt_06.spe
Binding Energy (eV)
c/s
Atomic %
C1s 81.8
O1s 16.1
Cl2p 1.2
Na1s 0.9
NaCN + ozonated SWNT
-C
KL
L
-N
a1s
-O
KL
L
-O
1s
-N
a K
LL
-C
1s
-C
l2p
319
swnt_11.spe: none Rice University
2006 Nov 16 Al mono 42.0 W 200.0 µ 45.0° 140.00 eV 2.9413e+004 max 3.47 min
Su1/Point11: NaBr/1
0200400600800100012000
0.5
1
1.5
2
2.5
3
3.5x 10
4 swnt_11.spe
Binding Energy (eV)
c/s
Atomic %
C1s 90.4
O1s 9.6
aq. NaBr + SWNT -
C K
LL
-O
KL
L -O
1s
-C
1s
swnt_12.spe: none Rice University
2006 Nov 16 Al mono 42.0 W 200.0 µ 45.0° 140.00 eV 2.5638e+004 max 3.47 min
Su1/Point12: NaBr ox/1
0200400600800100012000
0.5
1
1.5
2
2.5
3x 10
4 swnt_12.spe
Binding Energy (eV)
c/s
NaBr + ozonated SWNTAtomic %
C1s 79.7
O1s 17.5
Na1s 1.6
Cl2p 1.2
-N
a1s
-O
KL
L
-O
1s
-C
1s
-C
l2s
-C
l2p
320
swnt_23.spe: none Rice University
2006 Nov 16 Al mono 42.0 W 200.0 µ 45.0° 140.00 eV 2.2700e+004 max 3.47 min
Su1/Point23: TsCl/1
0200400600800100012000
0.5
1
1.5
2
2.5x 10
4 swnt_23.spe
Binding Energy (eV)
c/s
Atomic %
C1s 88.5
O1s 10.7
Cl2p 0.8TsCl + SWNT
-C
KL
L
-O
1s
-C
1s
-C
l2p
swnt_24.spe: none Rice University
2006 Nov 16 Al mono 42.0 W 200.0 µ 45.0° 140.00 eV 2.8888e+004 max 3.47 min
Su1/Point24: TsCl ox/1
0200400600800100012000
0.5
1
1.5
2
2.5
3
3.5x 10
4 swnt_24.spe
Binding Energy (eV)
c/s
Atomic %
C1s 85.8
O1s 13.4
Cl2p 0.8TsCl + ozonated SWNT
-C
KL
L
-O
KL
L
-O
1s
-C
1s
-C
l2p
321
swnt_06.spe: none Rice University
2006 Nov 13 Al mono 42.0 W 200.0 µ 45.0° 140.00 eV 3.1425e+004 max 3.47 min
Su1/Point5: CF3CH2OH/1
0200400600800100012000
0.5
1
1.5
2
2.5
3
3.5x 10
4 swnt_06.spe
Binding Energy (eV)
c/s
CF3CH2OH + SWNT
Atomic %
C1s 93.1
O1s 6.9
-O
1s
-C
1s
swnt_07.spe: none Rice University
2006 Nov 13 Al mono 42.0 W 200.0 µ 45.0° 140.00 eV 2.9200e+004 max 3.47 min
Su1/Point6: CF3CH2OH ox/1
0200400600800100012000
0.5
1
1.5
2
2.5
3
3.5x 10
4 swnt_07.spe
Binding Energy (eV)
c/s
Atomic %
C1s 90.4
O1s 9.6CF3CH2OH + ozonated SWNT
-C
KL
L
-O
KL
L
-O
1s
-C
1s
322
swnt1628_11.spe: none Rice University
2006 Nov 11 Al mono 42.0 W 200.0 µ 45.0° 140.00 eV 3.2550e+004 max 3.20 min
Su1/Point1: not ox/1
0200400600800100012000
0.5
1
1.5
2
2.5
3
3.5
4x 10
4 swnt1628_11.spe
Binding Energy (eV)
c/s
Atomic %
C1s 97.2
O1s 1.7
Fe2p3 1.1
SWNT before O3
-F
e2p
3
-O
1s
-C
1s
-F
e3p
swnt1628_03.spe: none Rice University
2006 Nov 10 Al mono 42.0 W 200.0 µ 45.0° 140.00 eV 1.7988e+004 max 3.47 min
Su1/Point2: swnt after 30 ox/1
0200400600800100012000
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2x 10
4 swnt1628_03.spe
Binding Energy (eV)
c/s
Atomic %
C1s 70.7
O1s 29.1
In3d5 0.2SWNT after O3
-O
KL
L
-O
1s
-In
3d
3 -
In3
d5
-C
1s
323
Appendix D
Supporting Information for Part II, Chapter 1. Calculated isotropic
Fermi contact couplings, computed structures, ESR, UV and NMR
spectra
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367