MS#: LY8629
Title: Absolute Potential of the Fermi Level of Isolated Single-Walled Carbon Nanotubes
Authors: Ken-ichi Okazaki, Kei Murakoshi, and Yoshihiro Nakato
APPENDIX A1: Preparation of highly dispersed SWNT on polycrystalline Au electrode
Highly–dispersed SWNT solution was prepared according to the method reported by O’Connell et al [5]. Commercially available SWNT (Carbolex Co., AP-Grade, purity
50-70%) was dispersed in aqueous sodium dodecyl sulfate (SDS) surfactant (1wt %). The resulting dispersion was sonicated for 10 min. Immediately after the sonication, this dispersion was centrifuged, and was then decanted to obtain the suspension in which
SWNT were enough isolated by SDS micelles. Adsorption of isolated SWNT onto a cleaned and mirror-finished polycrystalline Au surface was carried out by dropped this
SWNT/SDS solution. After drying deposited solution, the electrode was sonicated in ultra-pure water to remove excess SDS and the tubes.
APPENDIX A2 : Atomic force Microscope (AFM) image of isolated SWNT on polycrystalline Au electrode
Figure 1S shows the height image of tapping mode AFM measured by using
Nanoscope IIIa, Digital Instruments in Air. The arrow indicates the isolated SWNTs. The m 2 with tube lengths from 200 nm to 2000 nm on the Au substrate. In spite of careful sonication in ultra-pure water after the deposition, precipitated SDS crystals were still
A-1

observed on the surface. Isolated SWNT, however, were observed in Figure 1S. Apparent diameter of the present sample was in the range between 1 to 3 nm, reflecting the accuracy of the AFM diameter measurement. The grooves on the Au surface are scratches formed by the polishing at the preparation of the electrode before the adsorption of SWNT.
APPENDIX A3: Experimental conditions and procedure of electrochemical
experiment and in situ confocal micro Raman spectroscopy
SWNT adsorbed Au electrode was fixed at the bottom of a Teflon cell, and then deaerated aqueous 0.2 M Na
2
SO
4
electrolyte was poured into the cell. A platinum ring was set as a counter electrode such that it surrounded the electrode. The potential of the
Au working electrode was regulated with respect to the Ag/AgCl reference electrode. A potentiostat (Hokuto-Denko Co., HAB-151) were used to control the potential of the Au electrode. The potential was swept with the rate of 10 mV s
-1
in the range between -800 mV and +400 mV. Reactions of hydrogen evolution and oxidation of Au surface can be prevented if the electrode potential was kept in the potential range.
Confocal resonance near-infrared (NIR) Raman measurements carried out using commercial Raman microprobe spectrometer (Ramanscope System-2000, Renishaw) with a solid-state continuous-wave diode laser for 785 nm ( E ex
= 1.58 eV) excitation. The expanded NIR beam is focused onto the sample using water-immersion objective lens with 100x magnifications and a numerical aperture of 1.0 (Olympus). Estimated spot size
A-2

All of Raman spectra are measured by the exposure time for 5 sec under the potential sweep with the rate of 10 mV s -1 . The observed spectrum reflects the average of
Raman intensities in the potential rage of 50 mV. Potentials shown in the manuscript correspond to the average potentials at each spectral measurement.
APPENDIX A4: Characteristics of individual SWNT on Au electrode
Figure 2S shows that typical Raman spectrum of isolated single SWNT in the electrolyte solution. Observation of the sample at one spot gave typical Raman spectrum of isolated single SWNT, showing radial breathing mode (RBM) in the low frequency region around 100 - 300 cm
-1
, and D-, G-, and G’- bands in 1200 - 2600 cm
-1
[3,8].
Spectra at different spots gave different peaks of RBM. The full width at half maximum
(FWHM) at the RBM peaks of metallic tube around 140-180 cm
-1
and semiconducting tube around 200-240 cm -1 were approximately 9 cm -1 and 3 cm -1 excited at 785 nm, respectively[3]. These characteristics suggest that the observed spectrum at each spot reflect resonance near-infrared Raman scattering intensity of individual SWNT on the Au electrode. Apparent feature of asymmetric profile towards lower frequency at G-band, a
Breit-Wigner-Fano (BWF) profile, around 1580 cm
-1
was not observed at most of the spots, indicating that contribution of bundled tubes[3,8] and electron-phonon interaction between SWNT and Au electrode should be minor in the present system.
APPENDIX A5: Comparison between potential dependence of RBM Raman intensity and electronic structure
A-3

Figure 3S depicts the potential dependence of Raman intensity at semiconducting

RBM
= 208 cm -1

RBM
= 156 cm -1 ) superimposed on respective electronic structures of each tube (semiconducting (11, 6) and metallic (13, 10) [18]).
Narrower distribution of the potential dependence on the Raman intensity at the semiconducting tube in comparison to the metallic tube appears to reflect smaller E ii
of the semiconducting tube. In this study, we propose that average values of the peak potentials at negative and positive scan can reflect the energy level of the center position between van Hove singularities, i.e., the Fermi level of isolated SWNT as shown in
Figure 3S. The value of -0.55 V for the semiconducting (11, 6) tube was more negative than that of -0.05 V observed for the metallic (13, 10) tube, for example. This result implies that the Fermi level of the semiconducting (11, 6) tube locates at higher energy than that of the metallic (13, 10) tube.
APPENDIX A6: Time-dependent change in the resonant Raman intensity at the RBM of isolated SWNT during the potential change

RBM
= 156 cm -1 of isolated SWNT on the Au electrode in electrolyte solution. The potential was stepped between 0 V and +0.6 V (vs. Ag/AgCl). The intensity measurements were performed at intervals of 1 sec. It was shown that the intensity was changed within 1 sec, and was recovered to the initial values after the potential step. The result proves that the change in the Raman intensity caused by the electrode potential modulation is fast enough, and shows good reversibility. The difference in the peak potentials at the potential scanning experiments (Figure 3 and Figure 3S) seems to be caused by the polarization of SWNT at
A-4

the positive or negative potential for relatively long time during the slow potential scanning. Generally, it is known that change in the charge of SWNT results in the macroscopic structural change. If the macroscopic structures are different at the positive and negative potentials, the difference may be reflected to the potential dependence of the
Raman intensity. To prove the hypothesis, further experiment is needed using in-situ structural observation of isolated single SWNT on Au electrode in electrolyte solution.
APPENDIX A7: Assignments of individual SWNT observed in the present expmeriment
Table 1S listed the possible chiralities of the isolated SWNTs observed in the present experiment. The diameter ( d t

RBM
) by using the relation of d t

RBM
. These values were used to assign the tubes. The

0
= 2.9 eV, bond length a
C-C
=
 ex
= 1.58 eV) [9,10]. In the present system, the energy levels of SWNT should be broadened by the solvation as discussed in the manuscript.
Thus, the assignment shown here contains the tubes with large E ii
.
A-5

Figure 1S. AFM image of SWNT adsorbed on Au electrode
A-6

Figure 2S. Typical Raman spectrum of isolated SWNT adsorbed on Au electrode in aqueous 0.2 M Na
2
SO
4
solution.; Electrode potential was 0 V.
A-7

Figure 3S Potential dependence of RBM Raman intensity observed at the semiconducting tube (

RBM
= 208 cm
-1
) and the metallic (

RBM
= 156 cm
-1
)
SWNT during anodic (open circle) and cathodic (closed circle) scanning.
Proposed potential of the Fermi level of each tube is presented as dotted lines.
Density of the states of assigned tubes, which are (11, 6) semiconducting tube
( d t
= 1.19 nm) and (13.10) metallic tube ( d t
= 1.59 nm), are also displayed. The center position between van Hove singularities was adjusted to the proposed
Fermi level of each tube.
A-8

Figure 4S. Time course of the Raman intensity observed at

RBM
= 156 cm
-1
of isolated SWNT on the Au electrode in electrolyte solution.
A-9

TABLE 1S Possible chiralities for metallic and semiconducting nanotubes and their

RBM

RBM
were also shown.
( n, m ) d t
[nm]

[deg]

RBM
(cal.)
[cm -1 ] d t
[nm]
(exp.) (exp.) a E ii
[eV]
Metallic
(15, 9) 1.67
(13,10) 1.59
(17, 5) 1.59
(18, 3)
(14, 8)
1.56
1.53
21.8
25.7
12.5
7.6
21.1
148.8
156.4
156.4
158.8
162.0
149
156
158
158
162
1.67
1.59
1.57
1.57
1.53
1.46
1.54
1.51
1.53
1.58
(11,11) 1.51
(15, 6) 1.49
(13, 7) 1.40
30.0
16.1
20.2
164.0
166.7
177.7
163
167
176
1.52
1.48
1.41
1.62
1.73
1.63
Semiconducting
(11, 6)
(12, 4)
( 9, 7)
(11, 4)
(12, 2)
( 8, 7)
1.19
1.14
1.10
1.07
1.04
1.03
20.4
13.9
25.9
14.9
7.6
27.8
209.2
216.6
224.9
232.2
238.2
240.3
208
215
225
229
236
239
1.19
1.15
1.10
1.08
1.05
1.04
1.37
1.40
1.48
1.60
1.66
1.60 a and were shown for metallic and semiconducting tubes, respectively.
1 1
E S
2 2
1.51
1.58
1.62
1.66
1.65
1.72
1.81
A-10