Raman Laser Polymerization of C60 Nanowhiskers

Photopolymerization of C60 nanowhiskers (C60NWs) was investigated by using a Raman spectrometer in air at room temperature, since the polymerized C60NWs are expected to exhibit a high mechanical strength and a thermal stability. Short C60NWs with a mean length of 4.4 μm were synthesized by LLIP method (liquid-liquid interfacial precipitation method). The Ag(2) peak of C60NWs shifted to the lower wavenumbers with increasing the laser beam energy dose, and an energy dose more than about 1520 J/mm2 was found necessary to obtain the photopolymerized C60NWs. However, excessive energy doses at high-power densities increased the sample temperature and lead to the thermal decomposition of polymerized C60 molecules.

C 60 molecules can be polymerized by electron beam irradiation [6].Although as-grown C 60 NWs are composed of the C 60 molecules that are weakly bonded via van der Waals forces [7], the C 60 NWs irradiated by electron beams showed the stronger thermal stability [8], the higher Young's modulus [9] than pristine van der Waals C 60 crystals.Hence, it is of great importance to study the polymerization of C 60 NWs in order to improve their mechanical and thermal properties.
Laser irradiation is a promising method to obtain the polymerized C 60 molecules [7,10].We first showed the photopolymerization of C 60 NWs by using the Raman laser beam irradiation [7].Rao et al. showed that the peak of A g (2) pentagonal pinch mode of C 60 shifts downward from 1469 cm −1 to 1459 cm −1 upon the photopolymerization [11], showing that the shift of A g (2) peak is a good indicator for the polymerization of C 60 .
Alvarez-Zauco et al. studied the polymerization of C 60 thin films in air by the ultraviolet (UV) laser irradiation as a function of laser energy dose (= fluence) from 10 to 50 mJ/cm 2 in order to optimize the photopolymerization of C 60 films [12].Likewise, the laser energy dose for the photopolymerization of C 60 NWs should be optimized.Hence, the present study aims to reveal how the polymerization of C 60 NWs proceeds as a function of the laser beam energy dose.

Experimental
C 60 NWs were synthesized by a modified liquid-liquid interfacial precipitation method.Isopropyl alcohol (IPA) was gently poured into a toluene solution saturated with C 60 (MTR Ltd. 99.5%) in a glass bottle to make a liquidliquid interface, and then the solution was subjected to ultrasonication and stored in an incubator at 10 • C to grow short C 60 NWs.The synthesized C 60 NWs were filtered and dried in vacuum at 100 • C for 120 min.to remove the solvents.In the Raman spectrometry analyses, the C 60 NWs dispersed in ethyl alcohol were mounted on a slide glass and dried in air.
A Raman spectrometer (JASCO, NRS-3100) with a green laser of 532 nm excitation wavelength was used for the polymerization and structural analysis of C 60 NWs in air.The power of laser light illuminated onto the specimens was measured by using a silicon photodiode (S2281, Hamamatsu Photonics K.K.).The laser beam power density (D) and

=
The power of laser beam (mW) the area of laser beam exposed on the sample (mm 2 ) . (1)

Results and Discussion
Figure 1 shows examples of scanning electron microscopy (SEM) images and the size distributions of the synthesized C 60 NWs with a mean length of 4.4 ± 2.7 µm and a mean diameter of 540 ± 161 nm.The distribution of aspect ratios (length/diameter) is also shown.Most of the C 60 NWs were found to possess the aspect ratios less than 15.The power of excitation laser beam can be changed by selecting ND filters.Figure 2 shows the relationship between the ND filter number and the power of laser beam irradiated on samples.The laser beam power could be widely changed The exposed area (y, µm 2 ) and the defocus value (x, µm) were plotted as shown in Figure 3(g).The plotted points can be approximated by the fitted quadratic curve, y = 0.88x 2 + 6.8x + 36.Figure 4 summarizes the relationship among the laser beam power density, ND filter number, and the defocus value.
Figure 5 shows examples of the Raman spectra of C 60 NWs taken by using the ND filters of OD1, OD2, and OD3 for an exposure time of about 220 s, where the spot size of laser beam on samples was 9 µm in diameter.Each power density of the excitation laser beam was (a) 11800, (b) 1660, and (c) 71.5 mW/mm 2 , respectively.The A g (2) peak around 1468 cm −1 sifted to the lower wavenumbers with increasing the laser beam power density.
Figure 6 shows the A g (2) peak positions of the Raman spectra of C 60 NWs as a function of energy dose of the laser beam for each defocus value from 100 µm to 0 µm (just focus).The power density of laser beam on samples was changed by changing the defocus value and the ND filter number as shown in Figure 4.The energy dose was changed by setting the beam exposure time at 215 ± 6 s, 441 ± 10 s, 665 ± 9 s, and 899 ± 29 s for each power density.Hence, as a whole, 72 data points are plotted in Figure 6.As shown in Figure 5, the Raman shifts are found to generally decrease to the lower values with increasing the energy dose.However, the Raman shifts were observed to increase along 710 H g (3) 765 H g ( 4) 1417 H g (7) 1460 A g (2) 1568 H g (8) Raman shift (cm −1 ) the red arrows for the high energy doses in Figures 6(c), 6(d), 6(e), and 6(f).These phenomena are supposed to be explained by the temperature rise of the C 60 NWs exposed to the laser beams, since it is known that the photopolymerized C 60 molecules decompose into their primary monomers and dimers by heating at temperatures higher than about 100 • C [13].
The data points obtained using the highest power densities are indicated in each graph of Figure 6 by the black arrows for the exposure time of about 220 s. Figure 7 shows the relationship between the laser beam energy dose and the A g (2) peak position for the arrowed data points of Figure 6.The fitted curve of semilog plot is expressed as y = −2.2x+ 1467, where x represents log 10 (laser beam energy dose) and y represents the Raman shift of A g (2) peak.Using this experimental formula, the energy dose more than about 1520 J/mm 2 is found to be necessary for the photopolymerization of C 60 NWs in air, when the laser light with a wavelength of 532 nm is used.
Since it is known that the photopolymerization of C 60 progresses through the formation of four-membered rings between adjacent C 60 molecules [11], it is considered that  C 60 molecules are linearly polymerized by forming the fourmembered rings along the growth axis of C 60 NWs, as was shown in Figure 6 of [2].In the gas chromatography-mass spectrometry (GC-MS) measurement of solvents contained in the C 60 NWs that were prepared by use of toluene and IPA, the major residual solvent was toluene, and the content of IPA was very small compared with toluene [14].Since the residual toluene of C 60 NWs was measured to be about 0.2% after drying in an Ar atmosphere at 100 • C for 30 min.[14], it is considered that the residual toluene of the vacuum-dried samples of C 60 NWs in the present experiment is negligible and does not influence the Raman profiles.

Conclusions
The photopolymerization of C 60 NWs was investigated by using the Raman laser beam of 532 nm wavelength at various exposure conditions for the power density and the exposure time in air.
The A g (2) peak of C 60 NWs shifted to the lower wavenumbers from that of the as-grown dried C 60 NWs.However, the A g (2) peaks were found to move to the higher wavenumbers from the polymerized positions by the irradiation of laser beams for high energy doses at high-power densities, indicating the thermal dissociation of polymerized C 60 molecules owing to the temperature rise.
An energy dose larger than about 1520 J/mm 2 was found to be necessary for the laser beam of 532 nm wavelength to obtain the photopolymerized C 60 NWs.

Figure 2 :
Figure 2: Relationship between the neutral density (ND) filter number and the laser beam power.

Figure 4 :
Figure 4: Power density of the Raman excitation laser beam measured as a function of ND filter number and the defocus value.