Synthesis and Photophysics of Quaterrylene Molecules in Single

Sep 5, 2014 - Toyota Physical and Chemical Research Institute, Nagakute, Aichi 480-1192, Japan. •S Supporting Information. ABSTRACT: We report the ...
0 downloads 0 Views 4MB Size
Article pubs.acs.org/JPCC

Synthesis and Photophysics of Quaterrylene Molecules in SingleWalled Carbon Nanotubes: Excitation Energy Transfer between a Nanoscale Cylinder and Encapsulated Molecules Takeshi Koyama,*,† Takuya Tsunekawa,† Takeshi Saito,‡ Koji Asaka,§ Yahachi Saito,§ Hideo Kishida,† and Arao Nakamura†,∥ †

Department of Applied Physics, Graduate School of Engineering, Nagoya University, Chikusa, Nagoya, Aichi 464-8603, Japan Nanotube Research Center, National Institute of Advanced Industrial Science and Technology (AIST), Tsukuba, Ibaraki 305-8565, Japan § Department of Quantum Engineering, Graduate School of Engineering, Nagoya University, Chikusa, Nagoya, Aichi 464-8603, Japan ∥ Toyota Physical and Chemical Research Institute, Nagakute, Aichi 480-1192, Japan ‡

S Supporting Information *

ABSTRACT: We report the synthesis of quaterrylene monomers and dimers from perylene inside single-walled carbon nanotubes (SWNTs). Excitation energy transfer (EET) from the encapsulated molecule to the SWNT is investigated by using femtosecond pump−probe spectroscopy and time-resolved luminescence spectroscopy. Optical spectroscopy and transmission electron microscopy reveal the molecular arrangement of quaterrylene monomers and dimers within the SWNTs. The observed EET time from the dipole-allowed excited state of the quaterrylene monomer to the SWNT depends on the tube diameter and is found to be 1.1 ± 0.2 ps and 0.4 ± 0.1 ps for average diameters of 1.4 and 1.8 nm, respectively. These transfer times are ascribed to EET via a Coulomb interaction between transition dipoles. The transfer time from the dipole-forbidden state of the dimer to the SWNT with an average diameter of 1.8 nm is 3.8 ± 0.2 ps. The EET mechanism presumably corresponds to electron exchange or Coulomb interactions involving higher multipoles.



molecules,13,14 and the inner nanotubes in double-walled carbon nanotubes have been formed from fullerenes,11,15 perylene derivatives,16 ferrocenes,17 and twisted graphene nanoribbons.18 Polymerization of molecules inside SWNTs can lead to enhancement of the optical response of the lowest excitonic transition as well as control of the optical gap. The key to future applications of encapsulated molecule/ SWNT composites in optical devices is a fundamental understanding of the interactions that occur between the encapsulated molecules and SWNTs, as well as their dynamical properties. In general, SWNTs with diameters large enough to accommodate organic molecules have optical gaps in the nearinfrared region, while the offsets between the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO), the HOMO−LUMO gap, of organic molecules are in the visible region. Consequently, excitation energy transfer (EET) from encapsulated organic molecules to the encapsulating SWNTs can occur. In fact, EET

INTRODUCTION

Single-walled carbon nanotubes (SWNTs) are hollow cylinders of carbon atoms with a diameter of ∼1 nm and a length of ∼1 μm. When the chemical and thermal stability of SWNTs are utilized, many kinds of organic molecules can be encapsulated with them using simple procedures based on the reaction of SWNTs with the molecules in vapors1−8 and solutions.9,10 Molecules encapsulated within SWNTs are protected from environmental gases, e.g., oxygen,4 which can cause photodegradation of the molecules; hence, the photostability of the encapsulated molecules can be improved.9 As a consequence of this high photostability, the photofunctionalities of the encapsulated molecules can be retained for long durations. This is an important advantage for optical devices such as photosensors and photoswitches that operate in air. SWNTs with various diameters in the range of ∼1−3 nm can function as nanoscale one-dimensional reaction chambers for different kinds of molecules11 because of their chemical and thermal stability. Very simple but important and useful chemical reactions can be realized by heat treatment. For example, graphene nanoribbons have been synthesized from perylene and coronene molecules12 and tetrathiafulvalene © 2014 American Chemical Society

Received: June 26, 2014 Revised: August 25, 2014 Published: September 5, 2014 21671

dx.doi.org/10.1021/jp506361b | J. Phys. Chem. C 2014, 118, 21671−21681

The Journal of Physical Chemistry C

Article

heated in air, the HiPco-SWNTs were purified by heating in vacuum, refluxing in H2O2, and finally washing in HCl solution.2) The heated SWNTs (1 mg quantity) were dispersed in water (6 mL) with 1% weight sodium cholate by sonication for 20 min (Branson, Sonifier 450). The sonication simultaneously generates a large number of entrance holes at both ends of nanotubes, which permit the encapsulation of molecules. The dispersion was immediately centrifuged at 30 000g for 30 min (Hitachi Koki, 70P-72), and the upper 80% of the supernatant was then collected. In addition to the dispersants of the three kinds of SWNT materials, for the SO-SWNTs, we prepared semiconducting-SWNT-enhanced and metallic-SWNT-enhanced dispersants using gel column chromatography.24,25 The abundance ratios of semiconducting and metallic SWNTs in the two dispersants were estimated from the spectral weights of the exciton absorptions.26 Taking the ratio of the integrated molar absorption coefficients of the M11- and E22-exciton bands to be 1.2,26 the abundance ratios of semiconducting SWNTs were increased from 72% to 98% in the semiconductingSWNT-enhanced dispersant, and those of metallic SWNTs were increased from 28% to 71% in the metallic-SWNTenhanced dispersant after the gel column chromatography procedure. Five kinds of dispersed SWNTs were collected by vacuum filtration on membrane filters (Millipore, MF-Millipore, 0.22 μm pore) with the surfactants removed. These SWNT films on the membrane filters were transferred onto CaF2 substrates, and the membrane filters were removed by acetone, leaving behind the SWNT films on the CaF2 substrates. The diameter and typical thickness of SWNT films were ∼4 mm and ∼300 nm, respectively. The synthesis of quaterrylene molecules inside SWNTs was carried out by heat treatment of the mixture of perylene powder and SWNT films in vacuum, using a procedure similar to that utilized for fabrication of metallofullerene1 and fullerene2 peapods. The SWNT film on the CaF2 substrate was sealed with perylene powder (∼1 mg) at ∼3 × 10−5 Torr in a glass ampule. The sealed ampule was heated in an electric furnace (Isuzu, ETR-11K) at 400 °C for 15 h or 200 °C for 72 h. After the heat treatment, the SWNT film was washed thoroughly in toluene. The obtained film samples are referred to as rylene/ SWNT samples in this paper. Reference samples were also prepared using the same heat treatment procedure, but without the perylene powder. Experimental Equipment and Setup. Absorption spectra were obtained using a HITACHI U-3500 spectrophotometer. Raman spectra were measured by a Renishaw spectrometer with a He-Ne laser at 633 nm (corresponding photon energy 1.96 eV) and a diode laser at 488 nm (photon energy 2.54 eV). TEM observations were carried out employing a JEM-2010 instrument operated at 120 kV. The time evolutions of differential absorption signals (ΔA) were measured by pump−probe experiments.27 The light source was a regenerative amplifier (1 kHz, 800 nm, 120 fs), which was seeded by a mode-locked Ti:sapphire laser (82 MHz, 800 nm, 80 fs). The output of the amplifier was divided into two beams for preparing the pump and probe pulses. The pump pulse was generated by an optical parametric amplifier using the regenerative pulse, and the photon energy of the pump pulse was set at 1.86 eV. The probe pulse was a white continuum generated by self-phase modulation of the regenerative pulse in a water cell. The polarizations of the

from encapsulated molecules to SWNTs with time scales in the femtosecond and picosecond range has been observed for SWNTs encapsulating squarylium dye,10 β-carotene,19,20 and sexithiophene molecules.5 However, details of EET, including the dependence on the molecular structure and molecule−wall distance, and the transfer mechanism involved (Förster type21 or Dexter type22), have not been thoroughly investigated. In this study, starting with perylene, we synthesized quaterrylene molecules inside SWNTs, making use of the hollow space within SWNTs, as schematically shown in Figure 1. Quaterrylene has a high aspect ratio and is a “one-

Figure 1. Structure of perylene (top left) and quaterrylene (top right) molecules. Schematic of synthesis of a quaterrylene molecule from perylene inside a SWNT (bottom). The schematic was drawn using VESTA software (URL: http://jp-minerals.org/vesta/en/).54

dimensional” planar molecule, leading to stable geometrical arrangement within SWNTs. In addition, perylene and quaterrylene are known as rylene nanoemitters that have been explored for applications in optoelectronics and photonic devices. Using optical absorption and Raman scattering measurements and transmission electron microscopy (TEM) observations, we revealed the formation and molecular arrangement of quaterrylene monomers and dimers inside the SWNTs. The dynamics of photoexcited states of the quaterrylene (perylene)/SWNT composites were investigated using femtosecond pump−probe spectroscopy and timeresolved luminescence spectroscopy. The observed decay of the photoexcited state of the encapsulated quaterrylene monomers is much faster than radiative recombination decay, and the time constant is dependent on the tube diameter. Therefore, it is found that EET from the dipole-allowed excited state to the SWNT occurs and the time for EET depends on the tube diameter (molecule−wall distance). The decay time constant for transient absorption due to quaterrylene dimers inside SWNTs is longer than that due to monomers, and is ascribed to EET from the dipole-forbidden excited state to the SWNT. We find that the faster EET from the dipole-allowed state to the SWNT is associated with a Coulomb interaction between transition dipoles and the slower EET from the dipoleforbidden state to the SWNT is associated with electron exchange or Coulomb interactions involving higher multipoles, which are effective at nanoscale distances.



EXPERIMENTAL SECTION Sample Preparation. We used three kinds of SWNT materials in this study, which were produced using (1) a highpressure carbon monoxide process (NanoIntegris; henceforth called HiPco), (2) an arc-discharge process (Meijo Nano Carbon; SO), and (3) an enhanced direct injection pyrolytic synthesis process (DIPS),23 with tube diameters of 1.0 ± 0.2, 1.4 ± 0.1, and 1.8 ± 0.1 nm, respectively. The SWNT materials were heated at 550 °C for 1 h in air to form a large number of entrance holes for the encapsulated molecules.1 (Before being 21672

dx.doi.org/10.1021/jp506361b | J. Phys. Chem. C 2014, 118, 21671−21681

The Journal of Physical Chemistry C

Article

Figure 2. (a) Absorption spectra for rylene/semiconducting-SO-SWNT (red curve) and reference (black curve) samples. The intensities of both spectra are normalized at 0.7 eV. (b) Differential absorption spectrum (red curve) obtained by subtracting the absorption spectrum for the reference sample from that for the rylene/semiconducting-SO-SWNT sample in panel a. (c) Absorption spectra for the rylene/DIPS-SWNT (red curve) and reference (black curve) samples. (d) Differential absorption spectrum (red curve) obtained by subtracting the absorption spectrum for the reference sample from that for the rylene/DIPS-SWNT sample shown in panel c. Absorption spectra for quaterrylene (orange curve) and perylene (green curve) molecules in 1,2,4-trichlorobenzene and tetrahydrofuran solvents, respectively, are also shown in panels b and d. The spectrum for the quaterrylene molecule is taken from ref 32.

2.5 eV are observed, and they were assigned to the E11-, E22-, and E33- (and E44-) exciton bands of ensembles of SWNTs with various chiralities.27,30,31 Because the SO-SWNTs have diameters of ∼1.4 nm, the exciton transition energies for different chiralities are close to each other. Thus, the exciton absorption bands due to SWNTs with different chiralities overlap and cannot be resolved, especially in the E11- and E22exciton energy regions. These broad absorption bands are also observed in the spectrum for the rylene/semiconducting-SOSWNT sample (red curve). However, the two spectra are very different at energies above ∼1.5 eV, with additional absorption bands observed in the rylene/semiconducting-SO-SWNT spectrum. To investigate these additional bands in more detail, the differential absorption spectrum was calculated by subtracting the spectrum for the reference sample from that for the rylene/ semiconducting-SO-SWNT sample. This is shown as the red curve in Figure 2b. Two principal bands can be seen, with peaks at 1.68 and 2.72 eV and fine structure on the high-energy side at 1.84 and 2.02 eV, and 2.90 and 3.07 eV, respectively. The shapes of the two absorption bands closely resemble those of the absorption spectra for quaterrylene in 1,2,4-trichlorobenzene (orange curve, data taken from ref 32) and perylene in tetrahydrofuran (green curve) with vibrational structures. Consequently, the two additional bands suggest the existence of quaterrylene and perylene molecules in the rylene/ semiconducting-SO-SWNT sample. The observed redshift (∼0.2 eV) of the absorption bands is probably caused by molecule−SWNT interactions and a change in the dielectric constant of the materials surrounding the molecules. To confirm the formation of quaterrylene molecules, we measured the Raman spectra for the two samples with an excitation photon energy of 1.96 eV, which is resonant with the lower absorption band with a peak at 1.68 eV and the fine structure at 1.84 and 2.02 eV in Figure 2b. Figure 3a shows the Raman spectra for the reference (black curve) and rylene/ semiconducting-SO-SWNT (red curve) samples in the range of

pump and probe pulses were perpendicular to each other. The probe pulse transmitted through the sample was focused into a fiber-coupled spectrograph and detected by a thermoelectrically cooled charge-coupled device (CCD) camera. The pump power was measured using a thermal power sensor. The spot size of the pump pulse was measured by capturing the spot image using a CCD camera and fitting it to a Gaussian function. The pump fluence was 1.6 × 1016 photons cm−2 per pulse (4.8 mJ cm−2 per pulse). Luminescence kinetics were measured using femtosecond time-resolved luminescence spectroscopy based on the frequency up-conversion technique.28 The light source was a mode-locked Ti:sapphire laser (82 MHz, 800 nm, and 80 fs), and the excitation density was 6.9 × 10−7 J cm−2 per pulse. The instrument response function of the measurement system was determined by measuring the cross-correlation trace between the gate pulse and the excitation pulse, which had a Gaussian shape with a full width at half-maximum of 120 fs. The spectral resolution was about 0.03 eV. As described in the Supporting Information for ref 29, we detected background signals in the frequency up-conversion measurements, which were observed even in the range of negative time delay because of the unphase-matched second harmonics of the strong gate pulse and other stray light. To obtain the luminescence signal component, we subtracted the average signal level in the negative time range from the measured signal of the decay curve. All optical measurements in this study were conducted at room temperature in air.



RESULTS AND DISCUSSION Characterization by Optical Spectroscopy and Transmission Electron Microscopy. Figure 2a shows absorption spectra for the rylene/semiconducting-SO-SWNT and reference samples. In the spectrum for the reference sample (black curve), broad absorption structures with peaks at ∼0.7, 1.2, and 21673

dx.doi.org/10.1021/jp506361b | J. Phys. Chem. C 2014, 118, 21671−21681

The Journal of Physical Chemistry C

Article

molecule-related Raman peaks. Therefore, the absorption band peak at 2.72 eV is attributable to perylene molecules. Here, we estimate the number ratio of perylene to quaterrylene molecules using the ratio of spectral weights of the two molecules in the differential absorption spectra in Figure 2b and the ratio of oscillator strengths of the two molecules. The ratio of the spectral weight obtained from Figure 2b is 0.86. Using the oscillator strength ratio of 0.34 (ref 36) or 0.42 (ref 37), the perylene-to-quaterrylene number ratio is estimated to be 2.5 or 2.0, respectively. TEM observations clearly revealed the molecular encapsulation. Figure 4a shows typical TEM images of the rylene/SOSWNT samples (tube diameter ∼1.3 nm), including a widearea view in the left panel and an expanded view in the right panel. In the vicinity of the point labeled with a triangle, a dark line is observed at an off-center position between two longer Figure 3. Raman scattering spectra for the reference (black curve) and rylene/semiconducting-SO-SWNT (blue and red curves) samples in the range 1000−1700 cm−1. The excitation photon energy is (a) 1.96 eV or (b) 2.55 eV. The intensities are normalized at 1590 cm−1.

1000−1700 cm−1. In the spectrum for the reference sample, three peaks are observed at 1325, 1575, and 1593 cm−1, which are assigned to the D, G−, and G+ bands, respectively.33 These bands are also observed in the Raman spectrum for the rylene/ semiconducting-SO-SWNT sample. In addition, there are five peaks at 1059, 1256, 1287, 1361, and 1546 cm−1 in this spectrum. The ratio of the intensity of the 1256 cm−1 peak to that of the G+ band peak is ∼2.7. In the theoretically calculated Raman spectrum for quaterrylene molecules,34 five peaks occur at 1055, 1256, 1291, 1363, and 1549 cm−1. All of these peaks are observed for the rylene/semiconducting-SO-SWNT sample, and their positions and relative intensities closely correspond to those in the literature. Therefore, these results indicate the formation of quaterrylene molecules inside the SWNTs in the present study. We also confirmed that the additional absorption band in Figure 2b with a peak at 1.68 eV can be attributed to quaterrylene molecules by comparing the Raman spectra measured under resonant (Figure 3a) and off-resonant excitation conditions. Figure 3b shows Raman spectra obtained at an excitation energy of 2.54 eV for the reference (black curve) and rylene/semiconducting-SO-SWNT (blue curve) samples. No peaks at 1055 and 1549 cm−1 due to quaterrylene are observed, and the intensity of the 1256 cm−1 peak is very low. The relative intensity of this band with respect to the G+ band is ∼0.1. Consequently, the Raman spectrum observed with an excitation energy of 1.96 eV exhibits resonant behavior with the optical transition of quaterrylene molecules corresponding to the absorption band observed in the spectral range of 1.6−2.1 eV. The spectrum for the rylene/semiconductingSO-SWNT sample in Figure 3b shows peaks at 1166, 1259, 1290, and 1369 cm−1 in addition to the D-, G−-, and G+-band peaks of the SWNTs observed in the spectrum for the reference sample. These additional peaks can be attributed to perylene molecules because the observed Raman shifts show a good correspondence with the experimental and calculated values for perylene molecules.35 The excitation energy of 2.54 eV is close to the peak energy of 2.72 eV for the upper absorption band in Figure 2b, which causes resonant enhancement of the perylene

Figure 4. Typical TEM images and corresponding schematics for (a) thin SWNT and (b) thick SWNT. TEM images are shown in the top panels: wide-area view (left panel) and expanded view in the vicinity of the point indicated by the triangle (right panel). Schematics of quaterrylene molecules encapsulated in an SWNT are shown in the bottom panels: lateral view (left panel) and cross-sectional view (right panel). Note that the molecules are located at off-center positions. 21674

dx.doi.org/10.1021/jp506361b | J. Phys. Chem. C 2014, 118, 21671−21681

The Journal of Physical Chemistry C

Article

spectra, broad absorption bands with peaks at ∼0.5, 0.9, and 1.4 eV are observed, which are attributed to E11-, E22-, and M11exciton bands, respectively, because of an ensemble of SWNTs with different chiralities.30,31 In the rylene/DIPS-SWNT sample, additional absorption bands are observed above ∼1.5 eV, similar to the spectrum for the rylene/semiconducting-SOSWNT sample in Figure 2a. The differential absorption spectrum for the rylene/DIPS-SWNT sample, following subtraction of the spectrum for the reference sample, is shown in red in Figure 2d, together with the absorption spectra for quaterrylene (orange, ref 32) and perylene (green) molecules in solution. In the differential absorption spectrum, the largest peak is at 1.90 eV, whereas in Figure 2b, the largest peak is at 1.68 eV. This difference suggests enrichment of Htype quaterrylene dimers in the rylene/DIPS-SWNT sample. Similar results were observed for H-type dimers of perylene derivatives in the form of cyclophanes (i.e., covalently bound perylene-derivative dimers).41 It is also confirmed that, for the rylene/HiPco-SWNT sample, in which the tube diameter is 1.0 ± 0.2 nm and the formation of the H-type dimer is impossible, the highest peak in the differential absorption spectrum is located at 1.68 eV, and a weak shoulder is observed at 1.90 eV (Figure S2 in the Supporting Information). For further confirmation of the formation of H-type quaterrylene dimers, we estimated the energy separation between the H-type dimer transition and the monomer transition and compared it with the experimental results (details are given in the Supporting Information). Although the face-to-face distance between the two molecules is measured to be 0.5 nm in Figure 4b, the angle θ of the molecular plane from a straight line drawn between two center points of the molecules cannot be resolved from the TEM observations. We first consider an angle θ of 80° as an example. Using the dielectric constant value of 1.5 (ref 42) and the theoretical values of the transition energy (1.88 eV) and oscillator strength per molecule (1.17) for the quaterrylene molecule,36 the energy of the dipole−dipole interaction is calculated to be 0.16 eV. When we use the experimental values of the transition energy (1.88 eV) and oscillator strength per molecule (0.87) for the quaterrylene derivative estimated from the molar absorption spectrum,37 the energy of the dipole−dipole interaction is estimated to be 0.12 eV. For angles in the range of 65−90°, the estimated energy of the dipole−dipole interaction is in the range of ∼0.1−0.2 eV. As there are two electrons in the HOMO state, the shift of the transition energy of the H-type dimer from the monomer is twice the dipole−dipole interaction energy and is in the range of ∼0.2−0.4 eV. Because H-type dimers generally show a broad and structureless absorption band compared to monomer, because of splitting of vibrational modes in the dimer, as observed in the perylene solution at low temperatures,43 it is difficult to determine the exact position of the absorption peak of the Htype quaterrylene dimers. We consider the energy position of the highest peak at 1.90 eV as the peak energy of the absorption band of H-type quaterrylene dimers in Figure 2d. The separation of the energy positions between the highest absorption peaks in panels b and d of Figure 2 (1.90 − 1.68 = 0.22 eV) is consistent with the estimated difference between the two transition energies (∼0.2−0.4 eV). Thus, this result supports the assignment of the additional absorption peak in Figure 2d to H-type quaterrylene dimers, i.e., the formation of H-type quaterrylene dimers.

dark lines. When an electron beam travels through a region with a high atomic density, the scattering probability is high, and the region appears dark in the TEM image. This would be the case when the beam travels through the SWNT sidewalls or through a rylene molecule whose molecular plane is edge-on with respect to the beam direction. Thus, the TEM images in Figure 4a indicate the presence of a rylene molecule within a SWNT. The length of the inner line is 1.8−2.0 nm, which is consistent with the long axis of the quaterrylene molecule (1.90 nm).38 The encapsulated molecule is located at an off-center position within the SWNT as schematically shown in the bottom panels of Figure 4a; the lateral view is shown in the left panel, and the cross-sectional view is shown in the right panel. Earlier studies by Yanagi et al.10,19 and McIntosh et al.39 showed that encapsulated molecules with a π-conjugated structure are located at off-center positions, probably because of the π-stacking interaction between the molecule and the SWNT. To confirm the synthesis of quaterrylene molecules within SWNTs, we carried out sample preparation using a different procedure. We simply heated perylene powder in a sealed glass tube at 400 °C for 15 h. No quaterrylene molecules were synthesized using this procedure, indicating that they are difficult to form in an unconfined space. In contrast, when SWNTs encapsulating only perylene molecules were heated at 400 °C for 15 h, quaterrylene molecules were produced inside the SWNTs. (The accommodation of only perylene molecules inside SWNTs was accomplished by heat treatment of perylene powder and SWNTs at 200 °C for 72 h.) These results indicate that quaterrylene molecules can be synthesized when a onedimensional template is utilized. Details of this additional check and confirmation of the synthesis of quaterrylene molecules within SWNTs are described in the Supporting Information. Interestingly, a different molecular arrangement of encapsulated molecules was observed for SWNTs with large diameters. Figure 4b shows typical TEM images of rylene/SO-SWNT samples with a tube diameter of ∼1.5 nm; the left panel shows a wide-area view, and the right panel shows an expanded view. In the vicinity of the point labeled with a triangle, two quaterrylene molecules are arranged in two lines with their molecular planes parallel to each other along the sidewalls (face-to-face configuration), as schematically shown in the bottom panel of Figure 4b. Other TEM images (not shown) of thick SWNTs with diameters ≥1.5 nm showed a similar molecular arrangement. A parallel arrangement of π-conjugated molecules, such as perylene-3,4,9,10-tetracarboxylic dianhydride16 and α-sexithiophene,5 in the SWNTs has also been observed in earlier studies. The face-to-face configuration of the encapsulated molecules suggests the formation of H-type molecular dimers. For this type of molecular dimer, the excited state is split into two because of the Coulomb interaction between the molecules.40 In such a case, the dipole transition from the ground state to the lower state is forbidden and that to the upper state is allowed. Thus, the formation of H-type quaterrylene dimers can be tested using optical absorption measurements. For this, we prepared quaterrylene molecules encapsulated in DIPS-SWNTs with larger diameters. The tube diameter of the DIPS-SWNTs was 1.8 ± 0.1 nm, which is well above the value of 1.5 nm for the SO-SWNTs; hence, a large amount of H-type quaterrylene dimers can be produced in the rylene/DIPS-SWNT samples. Figure 2c shows the absorption spectra for the rylene/DIPSSWNT (red curve) and reference (black) samples. In both 21675

dx.doi.org/10.1021/jp506361b | J. Phys. Chem. C 2014, 118, 21671−21681

The Journal of Physical Chemistry C

Article

Photoexcited-State Dynamics of Quaterrylene Molecules in SWNTs. To investigate the dynamics of photoexcited states of quaterrylene molecules in the SWNTs, we conducted femtosecond pump−probe and time-resolved luminescence measurements. In the pump−probe measurements, we obtained differential absorption spectra, i.e., ΔA spectra, by subtracting the unpumped absorption spectrum from the pumped spectrum measured with different delay times between the pump and probe pulses. The ΔA spectra for the rylene/ semiconducting-SO-SWNT and reference samples measured with an excitation energy of 1.86 eV, resonant with the absorption band of the quaterrylene molecules, are shown in Figure S4 in the Supporting Information. To extract the absorption change in the quaterrylene molecules, we calculated the differential spectra of the ΔA spectra, i.e., Δ(ΔA) spectra, by subtracting the ΔA spectrum for the reference sample from that for the rylene/semiconducting-SO-SWNT sample: Δ(ΔA) = ΔArylene/SWNT − kΔAreference, where k is a normalization factor that equalizes the number of photons absorbed by the SWNTs in the rylene/semiconducting-SO-SWNT and reference samples. The obtained Δ(ΔA) spectra are shown in Figure 5b, and for comparison, the differential linear absorption spectrum in Figure 2b is replotted in Figure 5a. The Δ(ΔA) spectra just after the time origin show a decrease in absorption in the range of 1.6−2.1 eV while maintaining a shape similar (mirror image) to the differential linear absorption spectrum. This indicates that the absorption decrease is caused by spectral bleaching of the absorption band of the quaterrylene molecules. To more clearly display the decay behavior of the bleaching signal, we plotted the time evolution of Δ(ΔA) measured at 1.65 eV corresponding to the lowest absorption peak of quaterrylene molecules in Figure 5c (red circles). The decay curve is fitted using a single-exponential function (black curve) with a time constant of 1.1 ± 0.2 ps, and it is reproduced well by the function. Therefore, recovery of the bleaching signal occurs on a picosecond time scale. The recovery time constant of 1.1 ± 0.2 ps is 3 orders of magnitude shorter than those for the isolated quaterrylene derivatives (∼1 ns)44 and perylene monomers (∼5 ns).45 This fast decay behavior suggests EET from the quaterrylene molecule to the SWNT. When EET occurs between donor and acceptor materials, the decrease in the excited-state population in the donor material corresponds to the increase in the excited-state population in the acceptor material. This increase can be directly observed as a delayed rise in the luminescence kinetics of the acceptor material because the intensity of instantaneous luminescence is proportional to the population at a certain time. Panels a and b of Figures 6 show the luminescence kinetics at 0.70 eV in the E11-exciton band of the SWNTs in the reference sample and the rylene/semiconducting-SO-SWNT sample, respectively. Because the excitation energy of 1.55 eV is situated in the higher energy tail of the E22-exciton band of the SWNTs and in the lower energy tail of the quaterrylene absorption band, both the SWNTs and quaterrylene molecules are directly excited by the excitation pulse. First, we examine the luminescence kinetics in the reference sample (Figure 6a), wherein only SWNTs are present. The luminescence kinetics shows fast decay behavior and is fitted to a single-exponential function with a time constant of 0.23 ± 0.08 ps (black curve). The ultrafast decay is caused by EET between neighboring SWNTs in bundles as observed in earlier studies.28,29,46−49 Next, we analyze the luminescence kinetics in the rylene/ semiconducting-SO-SWNT sample (Figure 6b). The observed

Figure 5. (a) Differential absorption spectrum for the rylene/ semiconducting-SO-SWNT sample (replot of the spectrum in Figure 2b). (b) Δ(ΔA) spectra measured at different delay times as indicated on the right of the figure. The Δ(ΔA) spectra were obtained by subtracting the ΔA spectrum for the reference sample from those for the rylene/semiconducting-SO-SWNT sample in Figure S4 in the Supporting Information. The excitation photon energy of 1.86 eV is resonant with the quaterrylene absorption band. Because of the strong scattering of pump pulses, the change in absorption in the region of ∼1.8−1.9 eV cannot be observed. (c) Time evolution of Δ(ΔA) at 1.65 eV. The fitted result using a single-exponential function is plotted with the black curve.

luminescence kinetics yields a longer decay than that in Figure 6a, suggesting an additional component in the kinetics. Because the luminescence band for quaterrylene molecules32 lies in the energy region much higher than 0.70 eV, the observed luminescence originates from the SWNTs. Because the pump−probe data in Figure 5b suggest EET from the quaterrylene molecule to the SWNT, we consider that the additional component is caused by the excited-state population transferred from the quaterrylene molecules. We performed exponential function fitting considering a rise component using the following function obtained by analysis of the rate equations:50 21676

dx.doi.org/10.1021/jp506361b | J. Phys. Chem. C 2014, 118, 21671−21681

The Journal of Physical Chemistry C

Article

Here, α and β are the constants; ΓQ is the EET rate from quaterrylene molecule to SWNT, γQ the decay rate of the photoexcited state of quaterrylene molecule except for the EET rate, and γS the decay rate of the photoexcited state of SWNT. The fitting function consists of two components: a singleexponential function without a rise term and a singleexponential function with a rise term, which represent luminescence from the SWNTs excited directly by the excitation pulse and by EET from the quaterrylene molecules, respectively. Here, the decay time constant for the exponential function γS−1 is set at 0.23 ps, which was obtained by fitting the SWNT luminescence in Figure 6a, and γQ−1 is set at 1 ns.44 The fitting results are shown by the black solid curve in Figure 6b; the components with and without the rise term are plotted by the black dash-dotted and dashed curves, respectively. The fitted curve reproduces the observed luminescence kinetics fairly well. The obtained rise time constant ΓQ−1 is 1.3 ± 0.2 ps, in agreement with the recovery time of the bleaching of quaterrylene absorption (1.1 ± 0.2 ps). Therefore, this result supports EET from the quaterrylene molecule to the SWNT. Considering the energy-level scheme for quaterrylene and semiconducting SWNTs with diameters of 1.0−1.8 nm, the HOMO−LUMO gap of the quaterrylene lies between the E22and E33-exciton transition energies of the semiconducting SWNTs. Therefore, EET can take place from the excited state of the quaterrylene to the E22-exciton state of the semiconducting SWNTs. In general, the EET time depends on the distance of the molecule face from the tube wall (molecule− tube distance). The dependence of the EET time on the molecule−tube distance will be discussed later.

Figure 6. Luminescence kinetics at 0.70 eV in the E11-exciton band for SWNTs in the (a) reference and (b) rylene/semiconducting-SOSWNT samples. The circles and solid curves in panels a and b represent the results of the experiment, single-exponential fitting, and exponential fitting considering a rise component, respectively. The dash-dotted and dashed curves in panel b represent the components with and without the rise term, respectively. α exp(− γSt ) +

β ΓQ (γQ + ΓQ ) − γS

{exp( − γSt ) − exp[ −(γQ + ΓQ )t ]}

(1)

Figure 7. (a) Differential absorption spectrum for the rylene/DIPS-SWNT sample (replot of the spectrum in Figure 2d). (b) Δ(ΔA) spectra measured at different delay times as indicated on the right of the figure. The Δ(ΔA) spectra were obtained by subtracting the ΔA spectrum for the reference sample from those for the rylene/DIPS-SWNT sample in Figure S5 in the Supporting Information. The excitation photon energy is 1.86 eV and is resonant with the quaterrylene absorption band. Because of the strong scattering of pump pulses, the absorption change in the photon energy region of ∼1.8−1.9 eV cannot be observed. (c) Time evolution of Δ(ΔA) at 1.65 eV. The fitted result using a single-exponential function is plotted by the black curve. (d) Time evolution of Δ(ΔA) at 2.15 eV. The fitted result using a double-exponential function is plotted by the black solid curve; the black dashed and dash-dotted curves represent the fast and slow components, respectively. 21677

dx.doi.org/10.1021/jp506361b | J. Phys. Chem. C 2014, 118, 21671−21681

The Journal of Physical Chemistry C

Article

the internal conversion in H-type quaterrylene dimers is expected to occur on a similar time scale and is a process which competes with EET to the SWNTs. In addition, the contribution of quaterrylene monomers is also involved in the fast decay component (see Figure S6 and Table S2 in the Supporting Information). We next discuss the slow decay component observed at 2.15 eV. The fast decay with a time constant of 0.4 ps observed for quaterrylene monomers is caused by EET to the SWNTs. Because the reduction in the excited-state population in the monomers occurs with this time constant (Figure 7c), the slow decay component reflects the population reduction in the Htype dimers. It is to be noted that spectral bleaching of the absorption band of the H-type dimers arises from both filling of the upper state and depletion of the ground state. In the present case, the spectral bleaching is partially recovered after relaxation from the dipole-allowed upper state to the dipoleforbidden lower state; however, the bleaching signal can still be observed until the relaxation to the ground state is completed. As discussed above, the time constant for relaxation from the dipole-allowed upper state to the dipole-forbidden lower state via internal conversion should be less than several hundred femtoseconds. Thus, the slow decay component with a time constant of 3.8 ps is governed by the population relaxation from the dipole-forbidden lower state to the ground state. The slow decay component with a similar time constant is also observed for the rylene/semiconducting-SO-SWNT sample. As SWNTs encapsulating quaterrylene dimers are less abundant in the sample, the Δ(ΔA) signal is very weak but the time evolution exhibits two-component decay (see Figure S7 in the Supporting Information). The possible relaxation processes involved in this decay are internal conversion, radiative decay via perturbation-induced allowed transitions, and the EET to SWNT process. Relaxation via triplet states is negligible because the rate of intersystem crossing in the H-type quaterrylene dimers is low owing to the low mass of the constituent atoms. If internal conversion or radiative decay via perturbation-induced allowed transition is involved in the population decay of the dipole-forbidden state, decay behavior with a time constant of several picoseconds should also be observed for H-type quaterrylene dimers in solution. Although time-resolved measurements on the H-type quaterrylene dimers in solution have not been carried out, πstacked aromatic molecular dimers52,53 showed no such fast decay to the ground state in the time range of several picoseconds. Therefore, internal conversion and radiative decay can be ruled out as a mechanism of population decay from the lower state to the ground state. The most plausible process is EET to SWNTs. Because the lower state is dipole-forbidden, the population in the lower state cannot decay by EET via transition dipole coupling, and the possible EET mechanisms are electron exchange and Coulomb interactions involving higher multipoles.22 When the donor−acceptor distance is short, typically on the nanoscale, overlap of the electron wave function between the donor and acceptor and multipolar charge distribution need to be considered. In fact, the estimated molecule−wall distance for DIPS-SWNT is as short as 0.39 nm, and a quadrupolar charge distribution is realized in the dipoleforbidden lower state of the H-type quaterrylene dimers. Consequently, we attribute the observed slow time constant of 3.8 ps to the time constant for EET via electron exchange or Coulomb interactions involving higher multipoles. In contrast,

In the rylene/DIPS-SWNT sample, wherein SWNTs encapsulating H-type dimers are more abundant, quite different relaxation dynamics are observed. Panels a and b of Figure 7 show the differential linear absorption spectrum (replot of the spectrum in Figure 2d) and the Δ(ΔA) spectra for the rylene/ DIPS-SWNT sample, respectively. The Δ(ΔA) spectra were measured at an excitation energy of 1.86 eV. (The ΔA spectra for rylene/DIPS-SWNT and the reference samples are shown in Figure S5 in the Supporting Information.) Figure 7b shows a striking difference between the rylene/DIPS-SWNT and rylene/semiconducting-SO-SWNT samples. The Δ(ΔA) spectra for the rylene/DIPS-SWNT sample show decay above ∼2.0 eV that is slower than that at 1.65 eV, while those for the rylene/semiconducting-SO-SWNT sample in Figure 5b show that the same decay behavior is observed at both 1.65 eV and ∼2.0−2.2 eV. We note that above ∼2.2 eV, the Δ(ΔA) spectra are strongly affected by the absorption changes due to the SWNTs (Figure S4 in the Supporting Information). To understand the reason for this behavior, we analyzed the Δ(ΔA) decay curve measured at 1.65 and 2.15 eV for the rylene/DIPS-SWNT sample. Figure 7c shows the decay curve for Δ(ΔA) at 1.65 eV, which exhibits fast decay and is fitted to a single-exponential function with a time constant of 0.4 ± 0.1 ps. As the photon energy of 1.65 eV is situated in the absorption band of the quaterrylene monomers, Δ(ΔA) at 1.65 eV is attributed to bleaching of this absorption band. The time constant of 0.4 ps is shorter than that of 1.1 ps observed for the rylene/semiconducting-SO-SWNT sample. This difference can be explained by the decrease in the molecule−wall distance21,22 because of the smaller curvature of the DIPS-SWNTs, which have larger diameters than the SO-SWNTs. TEM observations indicate π stacking between the quaterrylene molecules and the tube wall, which suggests that the molecule−wall distance depends on the curvature of the SWNT, i.e., the diameter of the SWNTs. The average diameters of the SO-SWNTs and DIPS-SWNTs are 1.4 and 1.8 nm, respectively. Assuming that hydrogen atoms at the two long sides of a quaterrylene molecule can approach the tube wall at a distance of the sum of the van der Waals radii of hydrogen (0.12 nm) and carbon (0.17 nm) atoms, the distance between the molecule face and the tube wall is estimated to be 0.47 and 0.39 nm for the SO-SWNTs and DIPS-SWNTs, respectively. Figure 7d shows the time evolution of Δ(ΔA) at 2.15 eV for the rylene/DIPS-SWNT sample, which exhibits doubleexponential decay with fast and slow time constants of 0.3 ± 0.1 and 3.8 ± 0.2 ps, respectively. First, we discuss the origin of the fast decay component. The excitation photon energy of 1.86 eV is situated in the absorption band of the H-type quaterrylene dimers, and the probed photon energy of 2.15 eV is also situated in this absorption band. Thus, the fast decay component involves the contribution of H-type quaterrylene dimers. When the dipole-allowed upper state in the H-type dimers is excited by the pump pulse, decay of the population in this state occurs because of both EET to the SWNT and internal conversion to the dipole-forbidden lower state in the H-type dimers. Because the time constant for EET from the quaterrylene monomer to the SWNT is 0.4 ps for the rylene/ DIPS-SWNT sample (Figure 7c), EET from the upper state in the H-type dimer to the SWNT is expected to occur on a similar time scale. The internal conversion between the two excited states in π-stacked aromatic molecular dimers such as perylene derivatives,51 porphyrin,52 and doxorubicin53 dimers takes place in less than several hundred femtoseconds. Thus, 21678

dx.doi.org/10.1021/jp506361b | J. Phys. Chem. C 2014, 118, 21671−21681

The Journal of Physical Chemistry C

Article

reference samples, estimation of the separation of transition energy between the H-type dimer and the monomer, ΔA spectra for rylene/semiconducting-SO-SWNT and reference samples, ΔA spectra for rylene/DIPS-SWNT and reference samples, time evolution of Δ(ΔA) at various photon energies for rylene/DIPS-SWNT samples, and time evolution of Δ(ΔA) at 2.15 eV for rylene/semiconducting-SO-SWNT sample. This material is available free of charge via the Internet at http:// pubs.acs.org.

the time constant of 0.4 ps reflecting EET from the monomer excited state, wherein Coulomb interaction between transition dipoles is effective, is an order of magnitude shorter. Finally, we discuss the time constants for EET observed for quaterrylene monomers and dimers, comparing with reported values for squarylium dye and β-carotene encapsulated in SWNTs. The time constant for EET from the dipole-allowed excited state to the SWNT for the squarylium dye is less than 190 fs (instrument response time) for a molecule−wall distance of 0.42 ± 0.01 nm.10 The time constants for EET from the dipole-allowed excited state and the dipole-forbidden excited state to the SWNT for β-carotene are 50 fs and 6.6 ps, respectively, and the molecule−wall distance is 0.4 nm.20 The molecule−wall distance for quaterrylene molecules in the rylene/DIPS-SWNT sample is 0.39 nm, which is almost the same as that for squarylium dye and β-carotene molecules, but the time constant for EET from the dipole-allowed state for the quaterrylene molecule in our study is 5−10 times longer than those for the squarylium dye and β-carotene molecules. This difference may be related to the molecular structure of the molecules; the quaterrylene consists of fused benzene rings and is a planar molecule. In contrast, the time constant for EET from the dipole-forbidden state is almost the same for both quaterrylene and β-carotene. These results suggest that EET via electron exchange or Coulomb interactions involving higher multipoles is less affected by the molecular structure of the encapsulated molecules than the Coulomb interaction between transition dipoles.



Corresponding Author

*E-mail: [email protected]. Phone: +81-52-7894450. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work was supported by MEXT KAKENHI Grant 22016004; JSPS KAKENHI Grants 21340081, 25400332; The Tatematsu Foundation; and The Ogasawara Foundation for the Promotion of Science and Engineering.



REFERENCES

(1) Hirahara, K.; Suenaga, K.; Bandow, S.; Kato, H.; Okazaki, T.; Shinohara, H.; Iijima, S. One-Dimensional Metallofullerene Crystal Generated Inside Single-Wall Carbon Nanotubes. Phys. Rev. Lett. 2000, 85, 5384−5387. (2) Kataura, H.; Maniwa, Y.; Kodama, T.; Kikuchi, K.; Hirahara, K.; Suenaga, K.; Iijima, S.; Suzuki, S.; Achiba, Y.; Krätschmer, W. HighYield Fullerene Encapsulation in Single-Wall Carbon Nanotubes. Synth. Met. 2001, 121, 1195−1196. (3) Kataura, H.; Maniwa, Y.; Abe, M.; Fujiwara, A.; Kodama, T.; Kikuchi, K.; Imahori, H.; Misaki, Y.; Suzuki, S.; Achiba, Y. Optical Properties of Fullerene and Non-Fullerene Peapods. Appl. Phys. A: Mater. Sci. Process. 2002, 74, 349−354. (4) Takenobu, T.; Takano, T.; Shiraishi, M.; Murakami, Y.; Ata, M.; Kataura, H.; Achiba, Y.; Iwasa, Y. Stable and Controlled Amphoteric Doping by Encapsulation of Organic Molecules Inside Carbon Nanotubes. Nat. Mater. 2003, 2, 683−688. (5) Loi, M. A.; Gao, J.; Cordella, F.; Blondeau, P.; Menna, E.; Bártová, B.; Hébert, C.; Lazar, S.; Botton, G. A.; Milko, M.; et al. Encapsulation of Conjugated Oligomers in Single-Walled Carbon Nanotubes: Towards Nanohybrids for Photonic Devices. Adv. Mater. (Weinheim, Ger.) 2009, 22, 1635−1639. (6) Alvarez, L.; Almadori, Y.; Arenal, R.; Babaa, R.; Michel, T.; Le Parc, R.; Bantignies, J.-L.; Jousselme, B.; Palacin, S.; Hermet, P.; et al. Charge Transfer Evidence between Carbon Nanotubes and Encapsulated Conjugated Oligomers. J. Phys. Chem. C 2011, 115, 11898− 11905. (7) Okazaki, T.; Iizumi, Y.; Okubo, S.; Kataura, H.; Liu, Z.; Suenaga, K.; Tahara, Y.; Yudasaka, M.; Okada, S.; Iijima, S. Coaxially Stacked Coronene Columns Inside Single-Walled Carbon Nanotubes. Angew. Chem., Int. Ed. 2011, 50, 4853−4857. (8) Fujihara, M.; Miyata, Y.; Kitaura, R.; Nishimura, Y.; Camacho, C.; Irle, S.; Iizumi, Y.; Okazaki, T.; Shinohara, H. Dimerization-Initiated Preferential Formation of Coronene-Based Graphene Nanoribbons in Carbon Nanotubes. J. Phys. Chem. C 2013, 116, 15141−15145. (9) Yanagi, K.; Miyata, Y.; Kataura, H. Highly Stabilized β-Carotene in Carbon Nanotubes. Adv. Mater. (Weinheim, Ger.) 2006, 18, 437− 441. (10) Yanagi, K.; Iakoubovskii, K.; Matsui, H.; Matsuzaki, H.; Okamoto, H.; Miyata, Y.; Maniwa, Y.; Kazaoui, S.; Minami, N.; Kataura, H. Photosensitive Function of Encapsulated Dye in Carbon Nanotubes. J. Am. Chem. Soc. 2007, 129, 4992−4997.



CONCLUSIONS We successfully synthesized quaterrylene molecules from perylene inside SWNTs. Encapsulation of quaterrylene molecules inside the SWNTs was confirmed by absorption and Raman scattering measurements as well as TEM observations. The relaxation dynamics of photoexcited states in the quaterrylene molecules were observed by femtosecond pump−probe and time-resolved luminescence spectroscopy, and ultrafast EET from the quaterrylene molecule to the SWNT was demonstrated. The time constants for EET from the quaterrylene monomers to the SWNTs with average diameters of 1.4 and 1.8 nm were 1.1 ± 0.2 and 0.4 ± 0.1 ps, respectively, and the estimated values for the corresponding molecule−wall distances were 0.47 and 0.39 nm, respectively. Because the lowest excited state in the monomer is dipoleallowed, the EET mechanism is a Coulomb interaction between transition dipoles. The time constant for EET from the quaterrylene H-type dimer to the SWNT with an average diameter of 1.8 nm was 3.8 ± 0.2 ps. For the H-type dimers, the excited state is split into two and the lowest excited state is dipole-forbidden. It was thus found that EET from the lowest state occurs via electron exchange or Coulomb interactions involving higher multipoles with a time constant of 3.8 ± 0.2 ps, which is an order of magnitude longer than EET via a Coulomb interaction between transition dipoles. Our findings demonstrate the great advantage of SWNTs as a nanoscale reaction chamber to accommodate organic molecules, yielding a new composite system that could lead to new types of optoelectronic devices.



AUTHOR INFORMATION

ASSOCIATED CONTENT

S Supporting Information *

Supporting results for encapsulation of quaterrylene molecules in SWNTs, absorption spectra for rylene/HiPco-SWNT and 21679

dx.doi.org/10.1021/jp506361b | J. Phys. Chem. C 2014, 118, 21671−21681

The Journal of Physical Chemistry C

Article

(11) Bandow, S.; Takizawa, M.; Hirahara, K.; Yudasaka, M.; Iijima, S. Raman Scattering Study of Double-Wall Carbon Nanotubes Derived from the Chains of Fullerenes in Single-Wall Carbon Nanotubes. Chem. Phys. Lett. 2001, 337, 48−54. (12) Talyzin, A. V.; Anoshkin, I. V.; Krasheninnikov, A. V.; Nieminen, R. M.; Nasibulin, A. G.; Jiang, H.; Kauppinen, E. I. Synthesis of Graphene Nanoribbons Encapsulated in Single-Walled Carbon Nanotubes. Nano Lett. 2011, 11, 4352−4356. (13) Chuvilin, A.; Bichoutskaia, E.; Gimenez-Lopez, M. C.; Chamberlain, T. W.; Rance, G. A.; Kuganathan, N.; Biskupek, J.; Kaiser, U.; Khlobystov, A. N. Self-Assembly of a Sulphur-Terminated Graphene Nanoribbon within a Single-Walled Carbon Nanotube. Nat. Mater. 2011, 10, 687−692. (14) Chamberlain, T. W.; Biskupek, J.; Rance, G. A.; Chuvilin, A.; Alexander, T. J.; Bichoutskaia, E.; Kaiser, U.; Khlobystov, A. N. Size, Structure, and Helical Twist of Graphene Nanoribbons Controlled by Confinement in Carbon Nanotubes. ACS Nano 2012, 6, 3943−3953. (15) Zhang, J.; Miyata, Y.; Kitaura, R.; Shinohara, H. Preferential Synthesis and Isolation of (6,5) Single-Wall Nanotubes from OneDimensional C60 Coalescence. Nanoscale 2011, 3, 4190−4194. (16) Fujita, Y.; Bandow, S.; Iijima, S. Formation of Small-Diameter Carbon Nanotubes from PTCDA Arranged Inside the Single-Wall Carbon Nanotubes. Chem. Phys. Lett. 2005, 413, 410−414. (17) Shiozawa, H.; Pichler, T.; Grüneis, A.; Pfeiffer, R.; Kuzmany, H.; Liu, Z.; Suenaga, K.; Kataura, H. A Catalytic Reaction Inside a SingleWalled Carbon Nanotube. Adv. Mater. (Weinheim, Ger.) 2008, 20, 1443−1449. (18) Lim, H. E.; Miyata, Y.; Kitaura, R.; Nishimura, Y.; Nishimoto, Y.; Irle, S.; Warner, J. H.; Kataura, H.; Shinohara, H. Growth of Carbon Nanotubes via Twisted Graphene Nanoribbons. Nat. Commun. 2013, 4, 2548. (19) Yanagi, K.; Iakoubovskii, K.; Kazaoui, S.; Minami, N.; Maniwa, Y.; Miyata, Y.; Kataura, H. Light-Harvesting Function of β-Carotene Inside Carbon Nanotubes. Phys. Rev. B: Condens. Matter Mater. Phys. 2006, 74, 155420. (20) Abe, K.; Kosumi, D.; Yanagi, K.; Miyata, Y.; Kataura, H.; Yoshizawa, M. Light-Harvesting Function of β-Carotene Inside Carbon Nanotubes Explored by Femtosecond Absorption Spectroscopy. Phys. Rev. B: Condens. Matter Mater. Phys. 2008, 77, 165436. (21) Fö rster, T. Zwischenmolekulare Energiewanderung und Fluoreszenz. Ann. Phys. (Berlin, Ger.) 1948, 2, 55−75. (22) Dexter, D. L. A Theory of Sensitized Luminescence in Solids. J. Chem. Phys. 1953, 21, 836−850. (23) Saito, T.; Ohshima, S.; Okazaki, T.; Ohmori, S.; Yumura, M.; Iijima, S. Selective Diameter Control of Single-Walled Carbon Nanotubes in the Gas-Phase Synthesis. J. Nanosci. Nanotechnol. 2008, 8, 6153−6157. (24) Liu, H.; Nishide, D.; Tanaka, T.; Kataura, H. Large-Scale SingleChirality Separation of Single-Wall Carbon Nanotubes by Simple Gel Chromatography. Nat. Commun. 2011, 2, 309. (25) Miyata, Y.; Shiozawa, K.; Asada, Y.; Ohno, Y.; Kitaura, R.; Mizutani, T.; Shinohara, H. Length-Sorted Semiconducting Carbon Nanotubes for High-Mobility Thin Film Transistors. Nano Res. 2011, 4, 963−970. (26) Miyata, Y.; Yanagi, K.; Maniwa, Y.; Kataura, H. Optical Evaluation of the Metal-to-Semiconductor Ratio of Single-Wall Carbon Nanotubes. J. Phys. Chem. C 2008, 112, 13187−13191. (27) Koyama, T.; Yoshimitsu, S.; Miyata, Y.; Shinohara, H.; Kishida, H.; Nakamura, A. Transient Absorption Kinetics Associated with Higher Exciton States in Semiconducting Single-Walled Carbon Nanotubes: Relaxation of Excitons and Phonons. J. Phys. Chem. C 2013, 117, 20289−20299. (28) Koyama, T.; Miyata, Y.; Asaka, K.; Shinohara, H.; Saito, Y.; Nakamura, A. Ultrafast Energy Transfer of One-Dimensional Excitons between Carbon Nanotubes: A Femtosecond Time-Resolved Luminescence Study. Phys. Chem. Chem. Phys. 2012, 14, 1070−1084. (29) Koyama, T.; Asaka, K.; Hikosaka, N.; Kishida, H.; Saito, Y.; Nakamura, A. Ultrafast Exciton Energy Transfer in Bundles of SingleWalled Carbon Nanotubes. J. Phys. Chem. Lett. 2011, 2, 127−132.

(30) 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 Lett. 2003, 3, 1235−1238. (31) Araujo, P. T.; Doorn, S. K.; Kilina, S.; Tretiak, S.; Einarsson, E.; Maruyama, S.; Chacham, H.; Pimenta, M. A.; Jorio, A. Third and Fourth Optical Transitions in Semiconducting Carbon Nanotubes. Phys. Rev. Lett. 2007, 98, 067401. (32) Lempka, H. J.; Obenland, S.; Schmidt, W. The Molecular Structure of Boente’s “Dicoronylene”, as Deduced from PE and UV Spectroscopy. Chem. Phys. 1985, 96, 349−360. (33) Jorio, A.; Pimenta, M. A.; Filho, A. G. S.; Saito, R.; Dresselhaus, G.; Dresselhaus, M. S. Characterizing Carbon Nanotube Samples with Resonance Raman Scattering. New J. Phys. 2003, 5, 139. (34) Zhampeisov, N. U.; Nishio, S.; Fukumura, H. Density Functional Theory Study of Vibrational Properties of the 3,4,9,10Perylene Tetracarboxylic Dianhydride (PTCDA) Molecule: IR, Raman, and UV-Vis Spectra. Int. J. Quantum Chem. 2005, 105, 368− 375. (35) Shinohara, H.; Yamakita, Y.; Ohno, K. Raman Spectra of Polycyclic Aromatic Hydrocarbons. Comparison of Calculated Raman Intensity Distributions with Observed Spectra for Naphthalene, Anthracence, Pyrene, and Perylene. J. Mol. Struct. 1998, 442, 221−234. (36) Minami, T.; Ito, S.; Nakano, M. Theoretical Study of Singlet Fission in Oligorylenes. J. Phys. Chem. Lett. 2012, 3, 2719−2723. (37) Koch, K.-H.; Müllen, K. Synthesis of Tetraalkyl-Substituted Oligo(1,4-naphthylene)s and Cyclization to Soluble Oligo(perinaphthylene)s. Chem. Ber. 1991, 124, 2091−2100. (38) Hayakawa, R.; Petit, M.; Wakayama, Y.; Chikyow, T. Growth of Quaterrylene Thin Films on a Silicon Dioxide Surface Using Vacuum Deposition. Org. Electron. 2007, 8, 631−634. (39) McIntosh, G. C.; Tománek, D.; Park, Y. W. Energetics and Electronic Structure of a Polyacetylene Chain Contained in a Carbon Nanotube. Phys. Rev. B: Condens. Matter Mater. Phys. 2003, 67, 125419. (40) Kasha, M.; Rawls, H. R.; El-Bayoumi, M. A. The Exciton Model in Molecular Spectroscopy. Pure Appl. Chem. 1965, 11, 371−392. (41) Langhals, H.; Ismael, R. Cyclophanes as Model Compounds for Permanent, Dynamic Aggregates − Induced Chirality with Strong CD Effects. Eur. J. Org. Chem. 1998, 1915−1917. (42) Pichler, T.; Knupfer, M.; Golden, M. S.; Fink, J.; Rinzler, A.; Smalley, R. E. Localized and Delocalized Electronic States in SingleWall Carbon Nanotubes. Phys. Rev. Lett. 1998, 80, 4729−4732. (43) Ferguson, J. Absorption and Emission Spectra of the Perylene Dimer. J. Chem. Phys. 1966, 44, 2677−2683. (44) Meyer, Y. H.; Plaza, P.; Müllen, K. Ultrafast Spectroscopy of Soluble Terrylene and Quaterrylene. Chem. Phys. Lett. 1997, 264, 643−648. (45) Piston, D. W.; Bilash, T.; Gratton, E. Compartmental Analysis Approach to Fluorescence Anisotropy: Perylene in Viscous Solvents. J. Phys. Chem. 1989, 93, 3963−3967. (46) Lauret, J.-S.; Voisin, C.; Cassabois, G.; Delalande, C.; Roussignol, Ph.; Jost, O.; Capes, L. Ultrafast Carrier Dynamics in Single-Walled Carbon Nanotubes. Phys. Rev. Lett. 2003, 90, 057404/ 1−4. (47) Rubtsov, I. V.; Russo, R. M.; Albers, T.; Deria, P.; Luzzi, D. E.; Therien, M. J. Visible and Near-Infrared Excited-State Dynamics of Single-Walled Carbon Nanotubes. Appl. Phys. A: Mater. Sci. Process. 2004, 79, 1747−1751. (48) Maeda, A.; Matsumoto, S.; Kishida, H.; Takenobu, T.; Iwasa, Y.; Shimoda, H.; Zhou, O.; Shiraishi, M.; Okamoto, H. Gigantic Optical Stark Effect and Ultrafast Relaxation of Excitons in Single-Walled Carbon Nanotubes. J. Phys. Soc. Jpn. 2006, 75, 043709/1−4. (49) Lüer, L.; Crochet, J.; Hertel, T.; Cerullo, G.; Lanzani, G. Ultrafast Excitation Energy Transfer in Small Semiconducting Carbon Nanotube Aggregates. ACS Nano 2010, 4, 4265−4273. (50) Koyama, T.; Asada, Y.; Hikosaka, N.; Miyata, Y.; Shinohara, H.; Nakamura, A. Ultrafast Exciton Energy Transfer between Nanoscale Coaxial Cylinders: Intertube Transfer and Luminescence Quenching in Double-Walled Carbon Nanotubes. ACS Nano 2011, 5, 5881−5887. 21680

dx.doi.org/10.1021/jp506361b | J. Phys. Chem. C 2014, 118, 21671−21681

The Journal of Physical Chemistry C

Article

(51) Schröter, M.; Kühn, O. Interplay Between Nonadiabatic Dynamics and Frenkel Exciton Transfer in Molecular Aggregates: Formulation and Application to a Perylene Bismide Model. J. Phys. Chem. A 2013, 117, 7580−7588. (52) Kullmann, M.; Hipke, A.; Nuernberger, P.; Bruhn, T.; Götz, D. C. G.; Sekita, M.; Guldi, D. M.; Bringmann, G.; Brixner, T. Ultrafast Exciton Dynamics after Soret- or Q-band Excitation of a Directly β,β′Linked Bisporphyrin. Phys. Chem. Chem. Phys. 2012, 14, 8038−8050. (53) Changenet-Barret, P.; Gustavsson, T.; Markovitsi, D.; Manet, I.; Monti, S. Unravelling Molecular Mechanisms in the Fluorescence Spectra of Doxorubicin in Aqueous Solution by Femtosecond Fluorescence Spectroscopy. Phys. Chem. Chem. Phys. 2013, 15, 2937−2944. (54) Momma, K.; Izumi, F. VESTA 3 for Three-Dimensional Visualization of Crystal, Volumetric and Morphology Data. J. Appl. Crystallogr. 2011, 44, 1272−1276.

21681

dx.doi.org/10.1021/jp506361b | J. Phys. Chem. C 2014, 118, 21671−21681