Temperature-dependent photoluminescence from single-walled carbon nanotubes

Photoluminescence (PL) and photoluminescence excitation (PLE) spectroscopy of pillar-suspended single-walled carbon nanotubes has been measured for temperatures between 300 and $5\mathrm{K}$. The atmospheric environment strongly affects the low-temperature luminescence. The PL intensity is quenched at temperatures below $\sim 40\mathrm{K}$ for nanotubes in high vacuum, while nanotubes in helium ambient remain luminescent. The PL peak emission energy is only very weakly dependent on temperature, with a species-dependent blueshift upon cooling corresponding to a relative shift in bandgap of $-3\times {10}^{-5}{\mathrm{K}}^{-1}$ or less. The integrated peak intensities change by only a factor of 2, with linewidths showing a moderate temperature dependence. In PLE, the second absorption peak energy $\left(E_{22}\right)$ is also only weakly temperature dependent, with no significant shift and a limited reduction in linewidth upon cooling to $20\mathrm{K}$. In addition to the previously assigned nanotube PL peaks seen at room temperature, at least two distinct new classes of PL peaks are observed at cryogenic temperatures.

Figure 1

(Color online) Temperature-dependent photoluminescence spectra from an ensemble of single-walled carbon nanotubes. (a) Excitation at $E_{\text{laser}}=1.959\mathrm{eV}$ $\left(632.8\mathrm{nm}\right)$. (b) Excitation at $E_{\text{laser}}=1.621\mathrm{eV}$ $\left(765\mathrm{nm}\right)$. The topmost curve (red) is at the lowest temperature, with temperature increasing as labeled for each successive curve. The peak around $1.1\mathrm{eV}$ is emission from the Si substrate. Explicitly assigned SWNT peaks are labeled with their $\left(n,m\right)$ indices. White stars at the bottom of (a) indicate tentatively assigned $\left(n,m\right)$ peaks (see text). Emergent peaks labeled “MT” are discussed in the text. The red hatched squares highlight the average expected position of $E_{11}$ transitions for SWNT species having $E_{22}\approx E_{\text{laser}}$, using an average $E_{22}∕E_{11}$ ratio of 1.7.

Figure 2

(Color online) Blueshift of photoluminescence peak position for an ensemble of single-walled carbon nanotubes. The peak positions were extracted from Fig. 1. The blueshift is defined to be zero at room temperature.

Figure 3

(Color online) Temperature dependence of photoluminescence properties for ensembles of five nanotube species. The nanotube species are those assigned in Fig. 1. (a) The integrated PL peak intensity. (b) The PL full width at half maximum (FWHM). The dashed line corresponds to the thermal energy.

Figure 4

(Color online) Temperature dependence of MT peak intensity. The integrated PL peak intensity for two MT peaks is plotted as a function of temperature between 10 to $300\mathrm{K}$. The data were extracted from the MT peaks shown in Fig. 1b.

Figure 5

(Color online) Low-temperature PLE maps. (a) A PLE map for an ensemble of SWNTs at $20\mathrm{K}$. (b) The same PLE map including peak assignments. Chirality assigned peak positions are labeled with an open circle. The (9,8) and (9,7) species show the brightest PL. The MT peaks are labeled in white. The dotted line overlaps the position of the $2\mathrm{D}+2\mathrm{G}$ Raman mode. The horizontal dashed line indicates the excitation energy used for Fig. 1b.

Figure 6

(Color online) Temperature-dependent photoluminescence spectra from an individual single-walled carbon nanotube. The laser excitation was at $1.959\mathrm{eV}$ $\left(632.8\mathrm{nm}\right)$ and the emission is tentatively assigned to the (12,2) SWNT species. (a) PL spectra at temperatures from 300 to $5\mathrm{K}$. (b),(c) Parameters extracted from the spectra displayed in (a). (b) The measured blueshift in emission energy is plotted as a function of temperature, including a fit using the Varshni empirical functional form (dotted line). (c) The measured full width half maximum (FWHM) as a function of temperature with a linear fit (continuous line) and the thermal energy $k_{B}T$ (dotted line).

Figure 7

Comparison of ensemble and individual SWNT photoluminescence spectra. The PL spectra are shown for the (9,8) SWNT species at $20\mathrm{K}$, for resonant excitation at $1.55\mathrm{eV}$. The dark gray curve is an ensemble of many (9,8) nanotubes. The light gray curves show separate individual (9,8) nanotubes.

Figure 8

Photoluminescence intensity as a function of excitation energy for an individual (12,1) nanotube at 295 and $22\mathrm{K}$. The continuous lines are Lorentzian fits to the data, with a $25\mathrm{meV}$ width at $295\mathrm{K}$ and a $15\mathrm{meV}$ width at $22\mathrm{K}$. The curve at $295\mathrm{K}$ is offset and the dotted line indicates the baseline. The inset shows the emission spectrum at $22\mathrm{K}$. For reference, the shift in emission energy between 300 and $20\mathrm{K}$ is about $5\mathrm{meV}$.

Figure 9

(Color online) Low-temperature luminescence from an individual (9,8) SWNT. (a) PL spectra are shown at five different temperatures. The laser excitation is at $1.558\mathrm{eV}$ $\left(796\mathrm{nm}\right)$. (b) The corresponding PLE map taken at $5\mathrm{K}$.