Phase-coherent transport in ropes of single-wall carbon nanotubes

Abstract

To study the phase breaking scattering events in single-wall carbon nanotubes ͑SWNTs͒, ropes of SWNTs are intentionally damaged by Ar ϩ ion milling. Due to this treatment, the average distance an electron can travel before being elastically scattered is reduced to about 10 nm. This significantly increases the probability of one-dimensional localization and allows us to obtain the phase coherence length (L ⌽ ) in ropes of SWNTs as a function of temperature. We find that Nyquist scattering ( ⌽ ϳT Ϫ2/3 ) as well as another dephasing mechanism with a ⌽ ϳT Ϫ1 dependence are involved in limiting the phase-coherent transport. We also investigate the scattering of hot electrons in the system. The results support the statement that two different scattering mechanisms dominate the phase coherence length for different rope samples.Extensive studies of the electrical transport in singleand multiwall carbon nanotubes ͑SWNTs and MWNTs͒ have deepened our insight into the electrical properties of onedimensional metallic and semiconducting systems. The main impact of the reduced dimensionality on electrical transport is the decrease in the phase space available for different kinds of scattering events. In the case of metallic nanotubes it has been shown for example that the scattering rate in the case of the scattering of electrons by acoustic phonons follows a ϳT dependence 1 instead of a ϳT 5 power law as expected for three-dimensional metallic structures in the Grüneisen regime.While it is possible to investigate momentum changing scattering processes in a standard transport experiment, not every scattering event can be detected in this way. In particular, there are phase breaking mechanisms limiting the phase coherence length L ⌽ , such as Thouless scattering, which do not cause backscattering of the current carrying electrons and therefore cannot be measured in such an experiment. A common way to obtain information about L ⌽ in two-and threedimensional samples is to evaluate the magnetoresistance in a transport experiment. This approach was successfully used for MWNTs by different groups 2-5 and the analysis of the data clearly showed the two-dimensional character of electron motion in these systems. However, in one-dimensional systems such as SWNTs, no magnetic field dependence of resistance is expected, unless the fields are extremely high such that the dispersion relation of the sample is significantly changed, 6 or the tubes are used as building blocks for more complicated geometries, e.g., a ring geometry. 7 Because of this lack of a direct process for measuring L ⌽ , little is known about phase-coherent transport in SWNTs.This paper examines the nature of the major phase breaking scattering mechanisms in single-wall nanotubes and the dependence of L ⌽ on temperature. In order to perform a phase sensitive measurement, a rope of SWNTs is intentionally damaged by Ar ϩ ion bombarding at an energy of 500 eV. Simulations 8 suggest that this treatment would reduce the elastic mean free path L 0 from several micrometers 9 to L...

Temperature Dependence of Electron-to-Lattice Energy Transfer in Single-Wall Carbon Nanotube Bundles

“…The results suggest that room temperature scattering times τ e-ph and mean free paths l e-ph can exceed 1.5 ps and 1.5 µm, respectively. The latter observation is in qualitative agreement with the results by Frank et al [8] on MWNTs, Bachtold et al [9] and recent transport studies by Appenzeller et al [22,23] on SWNTs, which all estimate e-ph mean free path to significantly exceed 1 µm at room temperature. The experiments furthermore demonstrate that energy transfer between the electronic system and its host lattice depends strongly, i.e.…”

Phase coherent transport in graphene nanoribbons and graphene nanoribbon arrays

“…For Nyquist scattering τ ϕ is proportional to T −2/3 and therefore L ϕ is expected to be proportional to T −1/3 . [23][24][25] In our experiment the temperature dependence of L ϕ is about ∼T −0.3 at Kelvin temperatures and hence agrees with this model. But for milli-Kelvin temperature it gets weaker, suggesting a saturating behavior at a few hundred nanometers.…”

Dephasing in strongly anisotropic black phosphorus

“…A weaker loss of the dephasing length with temperature, as compared to the 2D case, has previously been observed in the presence of a strong one-dimensional (1D) confinement potential, e.g., in carbon nanotubes [5], and L ϕ was shown to be well described by the theoretical prediction of a T −1/3 power law. Remarkably, the transition from the 2D limit to the quasi-1D limit was observed in the systematic study undertaken by Natelson et al [6] whereby the dephasing time in metallic nanowires was measured as a function of the wire width down to 5 nm.…”