Photoluminescence study of aqueous-surfactant-wrapped single-walled carbon nanotubes under hydrostatic pressure

We present an experimental study of sodium-dodecylbenzene-sulfate-wrapped single-walled carbon nanotube (SWCNT) suspensions in ${\mathrm{D}}_2\mathrm{O}$ under hydrostatic pressure. Photoluminescence excitation mapping allows identification of the first and second Van Hove optical transitions ($E_{11}$ and $E_{22}$) in a pressure range of $P=0–9.1\mathrm{kbar}$ using a liquid clamp pressure cell. Shifts were observed in both $E_{11}$ and $E_{22}$ that allow the identification of two independent pressure effects. The first results from a uniaxial strain component with positive and negative energy shifts for structural factors $[q=\mathrm{mod}(n-m,3)]$ $q=2$ and 1 in agreement with theoretical band structure predictions. The second results from environmental effects that produce for all SWCNT’s a downshift in energy caused by significant changes in the Coulomb interactions for a higher-dielectric medium.

Figure 1

(Color online) (a) Contour plot of PLE map at $P=9.1\mathrm{kbar}$. Dark and light shades represent low and high intensity, respectively. Symbols indicate positions of peaks at $P=0\mathrm{kbar}$. (b) Quiver plot of PLE map pressure shifts from $0\text{to}9.1\mathrm{kbar}$. Shaded lines mark SWCNT family $(2n+m)$. Labels identify chiral indices $(n,m)$.

Figure 2

Plots of $\Delta E_{11}$ and $\Delta E_{22}$ as a function of pressure. (a) $\Delta E_{11}$ for PLE in the excitation range $700–850\mathrm{nm}$. (b), (c) $\Delta E_{11}$ and $\Delta E_{22}$, respectively, for PLE in the excitation range $850–940\mathrm{nm}$ in which $E_{22}$ shifts are better resolved. Lines are a guide to the eye only. Errors estimated from system resolution are $\pm 2$ and $\pm 3\mathrm{meV}$ for $E_{11}$ and $E_{22}$, respectively.

Figure 3

Plot of linear pressure coefficients from experiment and theory. Solid and dashed lines indicate linear best fits. Error bars are estimated from linear fits to data in Fig. 2.

Figure 4

Plot of the intensity ratio $I_{9.1\mathrm{kbar}}∕I_{0\mathrm{kbar}}$ as a function of SWCNT diameter. Errors are estimated from signal noise and reproducibility with correction for the transmission of the $1000\mathrm{nm}$ spectrometer filter.