Chirality-Selective Optical Scattering Force on Single-Walled Carbon Nanotubes

We show that optical forces acting on carbon nanotubes are substantially enhanced when the optical wavelength is tuned to resonances in the electronic band structure. Using a tunable laser source, we experimentally demonstrate a resonant optical scattering force on single-walled carbon nanotubes and, by tuning the wavelength, exploit this force to achieve chirality enrichment of four chiralities of nanotubes. Our results represent a significant step towards optical manipulation and sorting of carbon nanotubes based on their chiral vector.

Figure 1

The experimental setup used. (a) The SWCNT dispersion is confined within a microcapillary chamber. A tunable Ti:sapphire laser (100 mW) is coupled in to one end of the microcapillary. An objective lens collects the Raman scattered light which is analyzed by using a spectrometer. The inset images show optical images of the boundary between the oil and SWCNT suspension at each end of the microcapillary chamber. (b),(c) Scanning-electron-microscope images showing the dimensions of the microcapillary.

Figure 2

Experimental spectral characterization of the SWCNT suspension. (a)–(c) Radial-breathing-mode region of the Raman spectrum for an excitation wavelength range 700–805 nm; (d) the SWCNT chiralities detected in our experiment: (12,1), (10,5), (9,7), (10,2), (9,4), and (8,6); (e) measured Raman excitation profiles for the same SWCNT chiralities.

Figure 3

Raman data obtained by using a Raman excitation wavelength ${\lambda }_{\text{exc}}=775\mathrm{nm}$ and an optical pushing wavelength ${\lambda }_p=795\mathrm{nm}$. (a)–(e) Time-lapse Raman spectra obtained at the far end of the microcapillary. The thick black line shows the raw Raman spectra; the thinner colored curves show Lorentzian fits to the Raman peaks. (f) Ratios of the height of the Lorentzian fits of the (10,2) (black), (10,5) (blue), and (12,1) (red) nanotubes, normalized by the height of the (9,7) chirality peak, versus the pushing time. The ratios are normalized to the initial ratio at $t=0\mathrm{h}$. The error bars are derived from the errors in the fits.

Figure 4

Raman data obtained by using a Raman excitation wavelength ${\lambda }_{\text{exc}}=775\mathrm{nm}$ and an optical pushing wavelength ${\lambda }_p=805\mathrm{nm}$. (a)–(c) Time-lapse Raman spectra obtained at the far end of the microcapillary. (d) Ratios of the height of the Lorentzian fits of the (12,1) (black) and (10,5) (red) nanotubes, normalized by the height of the (10,2) chirality peak, versus the pushing time. The ratios are normalized to the initial ratio at $t=0\mathrm{h}$. The error bars are derived from the errors in the fits.

Figure 5

Raman data obtained by using a Raman excitation wavelength ${\lambda }_{\text{exc}}=730\mathrm{nm}$ and an optical pushing wavelength ${\lambda }_p=740\mathrm{nm}$. (a)–(d) Time-lapse Raman spectra obtained at the far end of the microcapillary. (e) Ratios of the height of the Lorentzian fits of the (9,4) to the (10,2) chirality peak versus the pushing time. (f) Ratios of the height of the Lorentzian fits of the (8,6) to the (10,2) chirality peak versus the pushing time. The ratios are normalized to the initial ratio at $t=0\mathrm{h}$. The error bars are derived from the errors in the fits.