This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Nano Res
1
Radial deformation of single-walled carbon nanotubes
on quartz substrates and the resultant anomalous
diameter-dependent reaction selectivity
Juan Yang, Yu Liu, Daqi Zhang, Xiao Wang, Ruoming Li, and Yan Li ()
Nano Res., Just Accepted Manuscript • DOI 10.1007/s12274-015-0811-1
nanoelectronic devices [2], chemical and biosensors
[3, 4]. Previously, it was widely reported in many
literatures that small-diameter tubes show higher
reaction activity over the large ones. The reactants
included nitronium ions [5], methane plasma [6],
NO2 [7], fluorine gas [8], benzenediazonium salts [9],
lithium vapor [10], O2 [11], dichlorocarbene [12], etc.
As a representative example, it was found that three
diameter-dependent regimes exist for SWNTs in
treatment with methane plasma [6]. In the
small-diameter regime, both metallic (M) and
semiconducting (S) tubes were etched
nondiscriminately. In the medium-diameter regime
only M-SWNTs were selectively removed, whereas in
the large-diameter regime both M- and S-SWNTs
were not etched [6]. Therefore, diameter is a general
key factor for the chemical reactivity of SWNTs. This
higher reactivity of small-diameter tubes can be
mainly attributed to the higher curvature of
small-diameter tubes and the consequent
curvature-induced strain in sp2-hybridized graphene
sheet, leading to decreased stability of the
carbon-carbon bonds [6, 10, 11].
In all previous cases, small-diameter tubes showed
higher chemical reactivity over the large ones, and
the small-diameter tubes were always preferentially
etched from the sample via various chemical
reactions. However, photoelectronic applications of
SWNTs require tuning of bandgaps in various
spectral region. In some special cases, selective
removal of large-diameter small-bandgap tubes are
strongly needed. In order to preferentially etch those
normally less reactive tubes, it is necessary to
increase the reactivity of the large-diameter tubes
over the small ones. Our design is to change the
electron density distribution by varying the local
curvature of tubes with different diameter. This can
be realized by the strong interaction between SWNTs
and the substrate lattice, e.g. quartz lattice.
Previously, strong interaction between SWNTs and
the quartz lattice was reported by many researchers
[13-17]. A nonuniform axial compressive strain
arising from the difference in coefficients of thermal
expansion (CTE) between quartz and SWNTs was
believed to exist in horizontally aligned SWNTs
grown on quartz substrate [13]. However, axial strain
cannot lead to large changes in local curvature and
the electron density distribution. Only radial
deformation of SWNTs will result in variations in
local curvature and distorted partial electron density.
Therefore, radial deformation of SWNTs is necessary
to achieve different chemical reaction selectivities of
SWNTs.
Here in this article, we report that radial
deformation is indeed present for quartz
lattice-oriented SWNTs based on detailed Raman
mappings. We also reveal that the radial deformation
of SWNTs is non-destructive and recoverable. More
importantly, we demonstrate that due to this radial
deformation of SWNTs, more delocalized partial
electrons are distributed at high curvature sidewall
sites on large-diameter tubes. Consequently, those
large-diameter tubes distribute higher reaction
priority over the small ones in treatment with iodine
vapor (Scheme 1), which is distinctly different from
the widely reported and well accepted higher
reaction activity in small-diameter tubes over the
large ones. This anomalous reaction activity thus
offers a novel approach to selectively remove those
small-bandgap large-diameter tubes.
Scheme 1 Schematic diagram of iodine reacting with the radial-deformed SWNTs on quartz substrate. Red dimers denote iodine molecules and red dots denote iodine radicals.
Inc.) using chemical vapor deposition (CVD) method
described elsewhere [18, 19].
2.2 Iodine treatment procedure
The lattice-oriented SWNT arrays on quartz substrate
prepared by chemical vapor deposition (CVD) were
placed in the central heating zone of a quartz tube in
a 1 inch tube furnace. Iodine powders were put in a
small porcelain boat at the entrance of the tube
furnace inside the quartz tube. The entire system was
first purged with an argon flow of 500 sccm for 30
min to maximally remove the residual oxygen, then
heated up to 900 °C in an argon flow of 500 sccm. At
this high temperature iodine powders were
sublimated into vapor, and the iodine vapor was
mixed with the argon flow into the quartz tube. The
heating stopped after 15-30 min and the system was
then cooled down to room temperature in argon
atmosphere.
2.3 Raman and SEM measurements
The Raman spectra of SWNTs on quartz substrate
were collected using a Jobin Yvon LabRam ARAMIS
spectrometer with 532 nm laser excitation. The
Raman spectra were taken in a backscattering
configuration by a microscope using a 100× objective
with laser focal spot of ~1 μm in diameter and a
charge coupled device (CCD) detector. All laser
power was attenuated to be less than 1 mW to avoid
heating effects.
The SEM images were taken on a Hitachi S4800 at
acceleration voltage of 1 kV.
3 Results and Discussion
Fig. 1a shows the Raman G band mapping for several
lattice-oriented SWNTs on quartz substrate in the
spectral region of 1580-1620 cm-1, from which irregular variations in both G band frequency (ωG)
and intensity (IG) are clearly observed. In the
direction of nanotube growth, which is along the
quartz [100] direction, some parts of the SWNT arrays distribute low ωG at ~1590 cm-1 (position A),
close to that of a suspended SWNT free of interaction.
Figure 1 Raman G band mapping for several lattice-oriented SWNTs on quartz substrate (a) and a typical individual SWNT grown on
silicon substrate (b). The mapping regions are 1580-1620 cm-1 for (a) and 1570-1600 cm-1 for (b), respectively. (c) Normalized Raman G band spectra for the marked positions A, B, and C, respectively. (d) Raman G band mapping and the corresponding G band frequency for a partially bent individual SWNT on quartz substrate. The red arrow indicates the quartz [100] direction.
However, some other parts of the SWNT arrays show distinctly upshifted ωG at above 1600 cm-1 (position B).
In very rare cases ωG as high as ~1620 cm-1 is also
observed. As a comparison, SWNTs grown on silicon
substrate distribute nearly uniform IG and the ωG is
typically observed in the 1580-1590 cm-1 range (Fig.
1b, position C). Therefore, lattice-oriented SWNTs
directly grown on quartz substrate differ from
SWNTs grown on silicon substrate in significant ωG
upshift and irregular IG variation. This distinct ωG
upshift and irregular IG variation for quartz
lattice-oriented SWNTs are also reported previously
by many others [13-17].
According to literatures, at constant experimental
parameters such as temperature and laser power, the ωG upshift and IG variation of SWNTs on substrate
may arise either from defects [20, 21], doping or
charge carrier implantation [22-24], or mechanical
deformation [14, 15, 25, 26]. Since the quartz
lattice-oriented SWNTs do not show clear
disorder-related D band, and doping or charge
carrier implantation could hardly cause a G band
upshift of more than 10 cm-1 [22-24], the only possible reason accounting for the significant ωG upshift and
IG variation is mechanical deformation of SWNTs.
For a one-dimensional SWNT, the mechanical
deformation can be either in the axial direction or in
the radial direction. Previously, it is believed that an
axial compressive strain is resulted from the large
difference in CTE of SWNTs and quartz. A 27 cm-1 G
band upshift per 1% compressive strain with a
maximum compressive strain of up to ~1.1% at room
temperature is reported [15]. However, no radial
deformation of quartz lattice-oriented SWNTs is
reported previously. With this strong interaction
between SWNTs and the quartz lattice, is any
deformation in the radial direction of SWNTs also
present?
To answer this question, we first perform the G
band mapping for a partially bent individual SWNT
on quartz substrate. The G band mapping and the
corresponding G band frequency with respect to
positions of this SWNT are illustrated in Fig. 1d. The
red arrow indicates the quartz [100] direction. It is
evident that all the SWNT parts along the [100]
direction distribute large ωG upshifts with frequency
higher than 1600 cm-1, whereas all the rest parts along
other directions correspond to ωG lower than 1600
cm-1. As the interaction between SWNTs and the
quartz [100] direction is the strongest of all, the
upshifts in ωG thus can be directly related to the
SWNT-quartz interaction. More importantly, because
the CTE of SWNTs is expected to be at least one order
of magnitude less than that of single crystal quartz
[27-29], and the CTE of quartz for all directions only
differs from each other by a ratio of about 2 [27, 28],
if the ωG upshifts are resulted all from axial
compressive strain due to the large difference in CTE
of SWNTs and quartz, it would be expected that the ωG variation of this SWNT is more or less uniform in
all directions. The nonuniformity of ωG variations in
different directions then indicates that deformation
of SWNT other than axial compressive strain is
present, which is most likely to be deformation in the
radial direction.
As G band of SWNTs corresponds to tangential
vibrational modes, it is quite insensitive to the
changes in the radial direction. The radial breathing
mode (RBM), on the other hand, is a
diameter-dependent Raman active mode
corresponding to all carbon atoms moving
simultaneously in the radial direction. Thus it is a
very sensitive mode to study changes of SWNTs
happened in the radial direction.
Previously, it is reported that the RBM intensity (IRBM) but not the frequency (ωRBM) changes
significantly when axial strain is applied [25, 30, 31]. In situ Raman measurements of suspended
individual SWNTs under tensile strain do not show evident change in ωRBM, however, IRBM reduces
rapidly with increasing strain [32]. Force constant
calculations based on molecular dynamics (MD)
simulations suggest that ωRBM is insensitive to axial
strain and that the shift in ωRBM is very small (less
than ±1 cm-1) with ±1% axial strain [33]. A critical
compressive strain, at which the tube buckles and a
sudden lateral deflection appears, is found to be
inversely proportional to tube diameter (dt) and also
dependent on tube chirality. For all tubes with dt>0.7
nm, the calculated critical compressive strain is less
than -10%. Above the critical compressive strain the
tube undergoes unstable changing and consequently ωRBM drops rapidly [33]. Therefore, a maximum
compressive strain of ~1.1% at room temperature
arising from the CTE difference in SWNTs and quartz
would lead to little ωRBM variation but significant IRBM
www.theNanoResearch.com∣www.Springer.com/journal/12274 | Nano Research
5 Nano Res.
variation. Meanwhile, it is also reported that both IRBM and
ωRBM vary significantly when radial deformation is
applied, for example, by hydrostatic pressure [13, 34].
In the low pressure regime, upshifts in ωRBM are
linear and reversible, and the pressure derivative of
ωRBM increases with increasing dt [13]. Classical
constant-pressure MD simulations and the force
constants model indicate the existence of a critical
pressure, below which the cross-section of SWNTs
remains circular but above which the tube shape
changes to ellipse [35]. Calculations also show that
ωRBM increases linearly with increasing pressure blow
the critical pressure [35]. Although the strong
interaction between SWNT and the quartz lattice will
probably cause asymmetric radial deformation of the
SWNT (presumably forming an asymmetric elliptical
cross-section, flatter on the side facing quartz), the
experiments and calculations based on uniform
deformation of SWNTs under pressure may still
serve as important guides for our quartz
lattice-oriented SWNTs. And because the asymmetric
deformation caused by the strong interaction
between SWNTs and quartz lattice is in the radial
direction, both IRBM and ωRBM are expected to vary
significantly.
According to the above experimental and
theoretical analysis in literatures, the variation in ωRBM is a key factor in order to determine whether or
not radial deformation is present for quartz lattice-oriented SWNTs. Therefore, detailed in situ
Raman mappings for the RBM are performed.
However, due to the strong interaction between
SWNTs and the quartz lattice, only very weak or
nearly no RBM signal can be observed, allowing very
limited in situ RBM data to be collected. Fig. 2a plots
the in situ RBM mapping of a typical quartz
lattice-oriented SWNT as an example. Significant variations in IRBM are evidently observed for this
SWNT. More importantly, ωRBM also distributes
significant variations, the data of which are shown in
green dots with ωRBM at 154±4 cm-1 in Fig. 2b. A 5.2%
large relative RBM shift is observed for this particular
SWNT. In Fig. 2b, the variations in ωRBM with respect
to positions for 5 individual quartz lattice-oriented
SWNTs (colored dots) and 3 individual SWNTs
transferred from quartz to silicon substrate (black
squares) with different diameters are illustrated.
Clearly, the 5 quartz lattice-oriented SWNTs all show
significant ωRBM variations whereas the 3 transferred
Figure 2 (a) In situ RBM mapping for a typical quartz lattice-oriented SWNT. The mapping region is 140-168 cm-1. (b) In situ RBM frequency variations for 5 individual quartz lattice-oriented SWNTs (colored dots) and 3 individual SWNTs transferred from quartz to silicon substrate (black squares) with different diameters. The RBM mapping in (a) corresponds to data in green dots with RBM frequencies at 154±4 cm-1. (c) Relative RBM shifts for the 5 individual quartz lattice-oriented SWNTs with respect to their average
RBM frequencies.
SWNTs on silicon substrate distribute no shift at all.
This means the deformation of SWNTs disappears
when transferred from quartz to silicon substrate. In
other words, it demonstrates the deformation of
SWNTs on quartz are non-destructive and recoverable. Assuming that this large ωRBM variations
for quartz lattice-oriented SWNTs all arise from axial strain, in order to cause an ωRBM variation as large as
±4 cm-1 the axial strain needs to reach at least an
order of ±10% according to MD calculations [33]. On
one hand, a compressive stain as large as -10% on
SWNTs is far beyond the reported maximum
compressive strain of up to ~1.1% at room
temperature [15]. On the other hand, at this large
compressive strain almost all SWNTs would have
already buckled and undergone unrecoverable
structural transitions. Therefore, the large ωRBM
variations can only be resulted from radial
deformation of SWNTs.
Based on the above experimental results, we
conclude that the mechanical deformation of SWNTs
on quartz is mainly radial deformation, arising from
the strong interaction between SWNTs and the
quartz lattice. This radial deformation is found to be
nonuniform along the nanotube growth direction.
Previously, the nonuniformity of G band was
believed to arise from the polishing induced surface
roughness of the quartz substrate [15]. We also
believe nanometer scale surface fluctuation of quartz
substrate can be caused by parallel scratches during
the standard surface polishing procedure. Stronger
interactions between SWNTs and the quartz lattice
are expected for the surface convexes and weaker
interactions for the surface concaves.
In Fig. 2b, one may also notice that the ωRBM
variations for large-diameter tubes are slightly larger than that for the small ones, given that ωRBM is
linearly related to the reciprocal of dt [36]. As the
curvature change of a nanotube can be reflected by
Δdt/dt, which is closely related to the relative RBM shift defined as ΔωRBM/ωRBM, we then plot the relative
RBM shifts for the 5 individual quartz
lattice-oriented SWNTs with respect to their average
ωRBM in Fig. 2c. It is found that the relative RBM shifts
for large-diameter SWNTs (roughly with ωRBM<180
cm-1) are higher than that for the small ones (roughly with ωRBM>180 cm-1). Calculations indicate that large
SWNTs are found to distribute larger degree of radial
deformation than the small ones.
As mentioned earlier, small-diameter tubes
showed higher chemical reactivity over the large
ones in all previous cases. In order to selectively
remove those normally less reactive large-diameter
tubes, our design is to utilize the strong interaction
between SWNTs and the quartz lattice. Now that we
already demonstrate radial deformation do exist for
quartz lattice-oriented SWNTs, and that larger degree
of radial deformation is present for large-diameter
tubes, we would like then to further explore the
possible changes in chemical reaction selectivities of
the radial-deformed SWNTs.
The quartz lattice-oriented horizontally aligned
SWNT arrays are heated at 900 °C in iodine vapor for
a time period of 15-30 min, the detailed procedure of
which is described in the experimental section. We
do not observe any encapsulation of iodine into the
tubes [37]. It is found by scanning electron
microscopy (SEM) that after a heating time of 30 min
at 900 °C in iodine vapor, all SWNT arrays on quartz
substrate disappear, however, there are still some
SWNTs remaining if the heating time is reduced to 15
min (Fig. 3a and 3b). As a control, SWNT arrays on
quartz substrate treated under exactly the same
experimental conditions, only without introducing
iodine vapor, do not show any distinguishable
changes (Fig. S1a and S1b), suggesting that iodine
vapor somehow reacts with the SWNT arrays on
quartz substrate at 900 °C.
Figure 3 SEM images of SWNTs on quartz substrate. Lattice-oriented SWNTs before (a) and after (b) a heating time of 15 min at 900 °C in iodine vapor. A specifically prepared sample
with both lattice-oriented SWNT arrays and floating-growth SWNTs on the same quartz substrate before (c) and after (d) a heating time of 30 min at 900 °C in iodine vapor.
www.theNanoResearch.com∣www.Springer.com/journal/12274 | Nano Research
7 Nano Res.
Moreover, it is observed in a specifically prepared
sample with both lattice-oriented SWNT arrays and
floating-growth SWNTs on the same quartz substrate,
that with a heating time of 30 min at 900 °C in iodine
vapor, only the lattice-oriented SWNT arrays
disappear whereas the floating-growth SWNTs are
kept intact (Fig. 3c and 3d). Further experiments
demonstrate that all SWNTs on silicon substrate, no
matter the floating-growth ultralong tubes, or
random tubes, or SWNT arrays transferred from
quartz substrate with the nanotransfer printing
technique [38], as well as bulk SWNT samples, do not
react with iodine under the previously described
experimental conditions. Therefore, it can be
summarized that in all above SWNT samples only
quartz lattice-oriented SWNTs react with iodine
under the previously described experimental
conditions.
In Fig. 3b, after a reaction time of 15 min at 900 °C
in iodine vapor, the sample shows much lower
SWNT density in the middle area (position 2) than in
the area (position 1) close to the pattered catalyst
regions. On the other hand, the control sample
treated under exactly the same experimental
conditions only without introducing iodine vapor
distributes similar SWNT density at the
corresponding positions 1’ and 2’ (Fig. S1b). This
demonstrates that the reaction of SWNTs with iodine
starts from the tip of the SWNTs, and that some
SWNTs react faster, or are of higher reactivity, than
others. Therefore, special reaction selectivity might
exist for those SWNTs in reaction with iodine.
Hereafter, we refer position 1 (or 1’) to the area close
to the pattered catalyst regions and refer position 2
(or 2’) to the middle area away from the pattered
catalyst regions.
In order to obtain information about the selectivity
of this reaction, Raman spectra in the RBM region are
collected. Again the strong interaction between
SWNTs and the quartz lattice quenches most RBM
signals. We then have to transfer the SWNT arrays
from quartz to silicon substrate with the nanotransfer
printing technique [38]. After transferring, the
sample gives more and stronger RBM signal, and a
statistical analysis of the RBM shifts (ωRBM) can thus
be obtained. Fig. 4a and 4b plot the statistical
distributions of the observed ωRBM with 532 nm laser
excitation at positions 1 and 2, respectively, for the
sample after a reaction time of 15 min at 900 °C in
iodine vapor. These two distributions clearly differ in the percentage of SWNTs with ωRBM <175 cm-1, i.e.,
50% at position 1 and only 15% at position 2. As a
comparison, the corresponding values are 50% at
position 1’ (Fig. 4c) and 41% at position 2’ (Fig. 4d) for
the control sample treated under exactly the same
experimental conditions only without introducing
iodine vapor. Therefore, it is found that after the
reaction with iodine the percentage of SWNTs with ωRBM<175 cm-1 is largely decreased in the middle area,
where more SWNT arrays react with iodine.
According to our specifically derived relation [39] of
ωRBM=222.0/dt+8.0 for SWNTs transferred from quartz
to silicon substrate by nanotransfer printing
technique, the statistical data of ωRBM indicate that the
reaction rates of large-diameter tubes with dt>1.33 nm
is predominant. In other words, large-diameter tubes
indeed distribute higher reactivity over the small
ones in reaction with iodine as we expected. This
reaction priority in large-diameter tubes is distinctly
differs from the previously reported reaction
priorities in small-diameter tubes. For 633 nm
excitation, with only very limited SWNTs excited by
the 633 nm laser, we still observe a similar trend that
the percentage of SWNTs with ωRBM <175 cm-1 is
much less at the middle position 2 (0%) than at
position 1 (62%) close to the pattered catalyst regions,
shown as in Fig. S2. Because semiconducting tubes
are expected to appear in the ωRBM regions of roughly
about 125-195 cm-1 for 532 nm excitation, and of
about 170-215 cm-1 for 633 nm excitation, this reaction
selectivity is clearly diameter-dependent, not
conductivity-dependent. Unambiguously, the ωRBM
boundary of 175 cm-1 nearly perfectly matches with
the previously mentioned boundary value of ~180
cm-1 between large and small relative RBM shifts in
Fig. 2c, indicating the intrinsic connections between
the anomalous reactivity and the radial deformation
of SWNTs, as we expected in our initial design.
| www.editorialmanager.com/nare/default.asp
8 Nano Res.
S MM
50%
S MM
15%
S MM
50%
S MM
41%
Figure 4 Statistical distributions of RBM frequencies with 532 nm exciation at positions 1 (a) and 2 (b) in Fig. 3b for quartz lattice-oriented SWNTs after a reaction time of 15 min at 900 °C
in iodine vapor, and at the corresponding positions 1’ (c) and 2’
(d) in Fig. S1b for the control sample treated under exactly the same experimental conditions only without introducing iodine vapor. Metallic and semiconducting tubes are denoted as M and S, respectively.
To further explore this anomalous reactivity,
detailed Raman mappings for the tangential G band
are performed. The G band mapping for a sample of
quartz lattice-oriented SWNT arrays clearly
distributes irregular G band variations in both
frequency and intensity (Fig. 5a). For the same
sample after a heating time of 15 min at 900 °C in
iodine vapor, the G band mapping shows evident
disappearance of G band at many locations (Fig. 5d).
Comparing to the G band mapping of the same
sample before reaction, the percentage of SWNTs
with large G band upshifts (in the spectral region of
1600-1620 cm-1) reduces greatly, indicating the
reaction happens selectively from the SWNT parts
with large G band upshifts, i.e., the SWNT parts with
large degree of radial deformation. Randomly
selected G band spectra of some individual tubes
from the stacked spectra in Fig. 5e and 5f are listed in
Fig. S3.
Figure 5 G band mapping and spectra for quartz lattice-oriented SWNTs before (a-c) and after (d-f) a reaction
time of 15 min with iodine. The G band mapping regions are
1580-1620 cm-1 for (a) and (d), and 1600-1620 cm-1 for (b) and (d), respectively. As can be seen, disappearance of G band at many locations is clearly observed, and the percentage of SWNTs with G band intensity in 1600-1620 cm-1 reduces greatly after
reaction. The black and red guiding lines in (c) and (f) correspond to Raman frequencies at 1590 and 1610 cm-1,
respectively.
To verify how radial deformation of a SWNT
affects its reaction activity, Fig. 6 plots the electron
localization functions (ELF) of a circular and a radial-deformed (11,11) tubes with dt of 1.49 nm.
Higher values of ELF implies more localized electron
density distribution. As can be seen, the circular tube
distribute uniform curvatures whereas for the
radial-deformed tube, sidewall sites with high and
low local curvatures are introduced. At the sites with
low local curvature, most electron density focuses
between the two neighboring carbon atoms. At the
high local curvature sites, however, more partial
electron density distributes around the carbon atoms
and less in the carbon-carbon neighboring section
than at the low local curvature sites. Those
delocalized partial electrons are more active and
easier to be attacked upon chemical reactions [40]. Meanwhile, as radial deformation affects the s and p
orbital hybridization on carbon atoms, there is less sp2 component on the carbon-carbon bonds at the
high local curvature sites than at the low local
curvature sites, resulting in weaker carbon-carbon
bonds and higher reaction priority at those high local
www.theNanoResearch.com∣www.Springer.com/journal/12274 | Nano Research
9 Nano Res.
curvature sites. Once the reaction is initiated from
those high local curvature sites, the SWNTs become
much more reactive. On the other hand, uniform
curvatures on a circular tube without radial
deformation lead to uniform electron density on the
sidewalls, and consequently the reactivity is low.
Figure 6 ELF plots of (a) a circular and (b) a radial-deformed (11,11) tubes.
For the reaction of SWNTs with iodine vapor, we
have previously concluded that quartz
lattice-oriented SWNTs can react with iodine and that
large-diameter tubes distribute higher reaction
priority over the small ones. Based on the above
experimental data and analysis, we now believe that
our design of varying the reaction selectivity of
SWNTs by the radial deformation of SWNTs and the
consequent changes in the local curvature of tubes
with different diameter is successful. As radial
deformation results in distorted partial electron
distribution and weaker carbon-carbon bonds at
certain high local curvature sites, it is thus easier for
those weaker carbon-carbon bonds to break down
upon reaction. Since large-diameter tubes present
larger degree of radial deformation, the local
curvature of their high curvature sites could become
even higher than that for the small ones. Therefore, it
is easier for the carbon-carbon bonds at the sidewall
sites with high local curvature on large-diameter
tubes to break down upon reaction. And the
anomalous reaction priority for large-diameter
quartz lattice-oriented SWNTs over the small ones
can be readily expected.
The reaction of iodine with the radial-deformed
SWNTs on quartz is likely a radical addition reaction.
For the lattice-oriented SWNTs on quartz substrates,
different degree of radial deformation in SWNTs, i.e.,
larger for large-diameter tubes and smaller for the
small ones, is present due to the strong SWNT-quartz
interaction. Consequently, the SWNTs with different
degree of radial deformation distribute different
chemical reaction activities. When iodine vapor
which serves as a mild reactant, is introduced to the
SWNTs, first, iodine molecules break down into
iodine radicals by homolysis at 900 oC. Second, the
iodine radicals selectively attack the radial-deformed
SWNT sidewall sites with high local curvature,
presumably on the large-diameter tubes. Then,
iodine leaves with carbon in a form of CIx from the
SWNT, and the thermally unstable CIx quickly
decomposes upon heating. As a consequence,
large-diameter tubes are selectively etched by iodine
vapor, and the percentage of small-diameter tubes
increases significantly after this reaction.
4 Conclusion
Based on the large variations in both intensity and frequency from the in situ Raman RBM mappings of
SWNTs, we distinctly observe radial deformation for
quartz lattice-oriented SWNTs, and attribute this
radial deformation to the strong interaction between
[3] Oh, J.; Chang, Y. W.; Kim, H. J.; Yoo, S.; Kim, D. J.; Im, S.; Park, Y. J.; Kim, D.; Yoo, K.-H. Carbon nanotube-based
dual-mode biosensor for electrical and surface plasmon resonance measurements. Nano Lett. 2010, 10, 2755-2760.
[4] Wohlstadter, J. N.; Wilbur, J. L.; Sigal, G. B.; Biebuyck, H. A.; Billadeau, M. A.; Dong, L. W.; Fischer, A. B.; Gudibande, S. R.; Jamieson, S. H.; Kenten, J. H. et al.
[5] An, K. H.; Park, J. S.; Yang, C. M.; Jeong, S. Y.; Lim, S. C.; Kang, C.; Son, J. H.; Jeong, M. S.; Lee, Y. H. A
diameter-selective attack of metallic carbon nanotubes by nitronium ions. J. Am. Chem. Soc. 2005, 127, 5196-5203.
[6] Zhang, G. Y.; Qi, P. F.; Wang, X. R.; Lu, Y. R.; Li, X. L.; Tu, R.; Bangsaruntip, S.; Mann, D.; Zhang, L.; Dai, H. J. Selective etching of metallic carbon nanotubes by gas-phase reaction. Science 2006, 314, 974-977.
[7] Seo, K.; Park, K. A.; Kim, C.; Han, S.; Kim, B.; Lee, Y. H. Chirality- and diameter-dependent reactivity of no2 on
carbon nanotube walls. J. Am. Chem. Soc. 2005, 127, 15724-15729.
[8] Yang, C.-M.; An, K. H.; Park, J. S.; Park, K. A.; Lim, S. C.; Cho, S.-H.; Lee, Y. S.; Park, W.; Park, C. Y.; Lee, Y. H.
Preferential etching of metallic single-walled carbon
nanotubes with small diameter by fluorine gas. Phys. Rev. B 2006, 73, 075419.
[9] Doyle, C. D.; Rocha, J.-D. R.; Weisman, R. B.; Tour, J. M. Structure-dependent reactivity of semiconducting single-walled carbon nanotubes with benzenediazonium salts. J. Am. Chem. Soc. 2008, 130, 6795-6800.
[10] Kalbac, M.; Kavan, L.; Dunsch, L. Selective etching of thin single-walled carbon nanotubes. J. Am. Chem. Soc. 2009, 131, 4529-4534.
[11] Yu, B.; Liu, C.; Hou, P.-X.; Tian, Y.; Li, S.; Liu, B.; Li, F.; Kauppinen, E. I.; Cheng, H.-M. Bulk synthesis of large
diameter semiconducting single-walled carbon nanotubes by oxygen-assisted floating catalyst chemical vapor deposition. J. Am. Chem. Soc. 2011, 133, 5232-5235.
Diameter effect on the sidewall functionalization of single
‐walled carbon nanotubes by addition of dichlorocarbene. Adv. Funct. Mater. 2012, 22, 5216-5223.
[13] Lebedkin, S.; Arnold, K.; Kiowski, O.; Hennrich, F.; Kappes, M. M. Raman study of individually dispersed
single-walled carbon nanotubes under pressure. Phys Rev B 2006, 73, 094109.
[14] Ding, L.; Zhou, W. W.; McNicholas, T. P.; Wang, J. Y.; Chu, H. B.; Li, Y.; Liu, J. Direct observation of the strong
interaction between carbon nanotubes and quartz substrate. Nano Res. 2009, 2, 903-910.
[15] Ozel, T.; Abdula, D.; Hwang, E.; Shim, M. Nonuniform
compressive strain in horizontally aligned single-walled carbon nanotubes grown on single crystal quartz. ACS Nano 2009, 3, 2217-2224.
[16] Soares, J. S.; Barros, E. B.; Shadmi, N.; Joselevich, E.; Jorio, A. Raman study of nanotube–substrate interaction
using single‐wall carbon nanotubes grown on crystalline quartz. phys. stat. sol. (b) 2011, 248, 2536-2539.
[17] Soares, J.; Barboza, A.; Araujo, P.; Barbosa, N. N. M.; Nakabayashi, D.; Shadmi, N.; Yarden, T.; Ismach, A.;
Geblinger, N.; Joselevich, E. Modulating the electronic properties along carbon nanotubes via tube-substrate interaction. Nano Lett. 2010, 10, 5043-5048.
[18] Kocabas, C.; Hur, S.; Gaur, A.; Meitl, M. A.; Shim, M.;
Rogers, J. A. Guided growth of large-scale, horizontally aligned arrays of single-walled carbon nanotubes and their
www.theNanoResearch.com∣www.Springer.com/journal/12274 | Nano Research
11 Nano Res.
use in thin-film transistors. Small 2005, 1, 1110-1116.
[19] Li, Y.; Cui, R.; Ding, L.; Liu, Y.; Zhou, W.; Zhang, Y.; Jin, Z.; Peng, F.; Liu, J. How catalysts affect the growth of single-walled carbon nanotubes on substrates. Adv. Mater. 2010, 22, 1508-1515.
[20] Wang, C. J.; Cao, Q.; Ozel, T.; Gaur, A.; Rogers, J. A.;
Shim, M. Electronically selective chemical functionalization of carbon nanotubes: Correlation between raman spectral and electrical responses. J. Am. Chem. Soc. 2005, 127, 11460-11468.
[21] Nguyen, K. T.; Shim, M. Role of covalent defects on phonon softening in metallic carbon nanotubes. J. Am. Chem. Soc. 2009, 131, 7103-7106.
[22] Rao, A. M.; Eklund, P. C.; Bandow, S.; Thess, A.; Smalley, R. E. Evidence for charge transfer in doped carbon nanotube bundles from raman scattering. Nature 1997, 388, 257-259.
[23] Shim, M.; Ozel, T.; Gaur, A.; Wang, C. J. Insights on charge transfer doping and intrinsic phonon line shape of
carbon nanotubes by simple polymer adsorption. J. Am. Chem. Soc. 2006, 128, 7522-7530.
[24] Tsang, J. C.; Freitag, M.; Perebeinos, V.; Liu, J.; Avouris, P.
Doping and phonon renormalization in carbon nanotubes. Nat. Nanotechnol. 2007, 2, 725-730.
[25] Cronin, S. B.; Swan, A. K.; Unlu, M. S.; Goldberg, B. B.; Dresselhaus, M. S.; Tinkham, M. Measuring the uniaxial
strain of individual single-wall carbon nanotubes: Resonance raman spectra of atomic-force-microscope
modified single-wall nanotubes. Phys. Rev. Lett. 2004, 93, 167401.
[26] Cronin, S. B.; Swan, A. K.; Unlu, M. S.; Goldberg, B. B.;
Dresselhaus, M. S.; Tinkham, M. Resonant raman
spectroscopy of individual metallic and semiconducting single-wall carbon nanotubes under uniaxial strain. Phys. Rev. B 2005, 72, 035425.
[27] Jay, A. The thermal expansion of quartz by x-ray measurements. Proc. R. Soc. London, A 1933, 142, 237-247.
[28] Rosenholtz, J. L.; Smith, D. T. Linear thermal expansion
and inversions of quartz, var. Rock crystal. Am. Mineral. 1941, 26, 103.
[29] Jiang, H.; Hwang, K.; Liu, B.; Huang, Y. Thermal expansion of single wall carbon nanotubes. J. Eng. Mater. Technol. 2004, 126, 265-270.
[30] Duan, X. J.; Son, H. B.; Gao, B.; Zhang, J.; Wu, T. J.; Samsonidze, G. G.; Dresselhaus, M. S.; Liu, Z. F.; Kong, J. Resonant raman spectroscopy of individual strained
single-wall carbon nanotubes. Nano Lett. 2007, 7, 2116-2121.
[31] Gao, B.; Duan, X. J.; Zhang, J.; Wu, T. J.; Son, H. B.; Kong,
J.; Liu, Z. F. Raman spectral probing of electronic transition energy e-ii variation of individual swnts under torsional strain. Nano Lett. 2007, 7, 750-753.
[32] Lee, S. W.; Jeong, G. H.; Campbell, E. E. B. In situ raman measurements of suspended individual single-walled carbon nanotubes under strain. Nano Lett. 2007, 7,
2590-2595.
[33] Yang, W.; Wang, R. Z.; Yan, H. Strain-induced raman-mode shift in single-wall carbon nanotubes: Calculation of force constants from molecular-dynamics simulations. Phys. Rev. B 2008, 77, 195440.
[34] Amer, M. S.; El-Ashry, M. M.; Maguire, J. F. Study of the hydrostatic pressure dependence of the raman spectrum of
single-walled carbon nanotubes and nanospheres. J. of Chem. Phys. 2004, 121, 2752-2757.
[35] Yang, W.; Wang, R. Z.; Song, X. M.; Wang, B.; Yan, H. Pressure-induced raman-active radial breathing mode transition in single-wall carbon nanotubes. Phys. Rev. B 2007, 75, 045425.
[36] Dresselhaus, M. S.; Dresselhaus, G.; Saito, R.; Jorio, A.
Raman spectroscopy of carbon nanotubes. Phys. Rep. 2005, 409, 47-99.
[37] Guan, L. H.; Suenaga, K.; Shi, Z. J.; Gu, Z. N.; Iijima, S. Polymorphic structures of iodine and their phase transition in confined nanospace. Nano Lett. 2007, 7, 1532-1535.
[38] Ahn, J.-H.; Kim, H.-S.; Lee, K. J.; Jeon, S.; Kang, S. J.; Sun, Y.; Nuzzo, R. G.; Rogers, J. A. Heterogeneous three-dimensional electronics by use of printed
[39] Wang, J. Y.; Yang, J.; Zhang, D. Q.; Li, Y. Structure
dependence of the intermediate-frequency raman modes in isolated single-walled carbon nanotubes. J. Phys. Chem. C 2012, 116, 23826-23832.
[40] Lara, I. V.; Zanella, I.; Fagan, S. B. Functionalization of
carbon nanotube by carboxyl group under radial deformation. Chem. Phys. 2014, 428, 117-120.
[41] Perdew, J. P.; Burke, K.; Ernzerhof, M. Generalized gradient approximation made simple. Phys. Rev. Lett. 1996, 77, 3865-3868.
[42] Clark, S. J.; Segall, M. D.; Pickard, C. J.; Hasnip, P. J.; Probert, M. J.; Refson, K.; Payne, M. C. First principles methods using castep. Z Kristallogr 2005, 220, 567-570.
[43] Blochl, P. E. Projector augmented-wave method. Phys. Rev. B 1994, 50, 17953-17979.
[44] Perdew, J. P. Accurate density functional for the energy -
real-space cutoff of the gradient expansion for the exchange hole. Phys. Rev. Lett. 1985, 55, 1665-1668.
[45] Becke, A. D.; Edgecombe, K. E. A simple measure of electron localization in atomic and molecular-systems. J Chem. Phys. 1990, 92, 5397-5403.
[46] Silvi, B.; Savin, A. Classification of chemical-bonds based on topological analysis of electron localization functions. Nature 1994, 371, 683-686.
www.theNanoResearch.com∣www.Springer.com/journal/12274 | Nano Research
Nano Res.
Electronic Supplementary Material
Radial deformation of single-walled carbon nanotubes
on quartz substrates and the resultant anomalous
diameter-dependent reaction selectivity
Juan Yang, Yu Liu, Daqi Zhang, Xiao Wang, and Yan Li ( )
Supporting information to DOI 10.1007/s12274-****-****-* (automatically inserted by the publisher)
SUPPLEMENTARY FIGURES S1-S5
Figure S1 SEM images of quartz lattice-oriented SWNTs before (a) and after (b) a heating time of 15 min at
900 °C without introducing iodine, serving as a control to Figures 4a and 4b. No distinguishable changes before
and after heating are observed.
Figure S3 SEM images of quartz lattice-oriented SWNTs before (a) and after (b) a heating time
of 15 min at 900 °C without introducing iodine, serving as a control to Figures 4a and 4b. No
distinguishable changes before and after heating are observed.
| www.editorialmanager.com/nare/default.asp
Nano Res.
Control
MS
62%
MS
0%
Figure S2 Statistical distributions of RBM frequencies with 633 nm exciation at positions 1 (a) and 2 (b) in
Figure 3b for quartz lattice-oriented SWNTs after a reaction time of 15 min at 900 °C in iodine vapor. Metallic
and semiconducting tubes are denoted as M and S, respectively.
Figure S3 Randomly selected Raman G band spectra of some individual tubes from Figures 6e (before) and 6f