We present a resonance Raman study of electrochemical charge transfer doping on polyelectrolyte single-walled carbon nanotubes (SWNTs) in solution. Changes in the intensity of the radial breathing modes of well-identified SWNTs are measured as a function of the electrochemical potential. The intensity is maximum when the nanotubes are neutral. Unexpectedly, the Raman signal decreases as soon as charges are transferred to the nanotubes, leading to intensity profiles that are triangular for metallic and trapezoidal for semiconducting nanotubes. A key result is that the width in energy of the plateaus for the semiconducting nanotubes is roughly equal to the optical gap (rather than the free carrier gap). While these experiments can be used to estimate the energy levels of individual nanotubes, strong dynamical screening appears to dominate in individual SWNT polyelectrolytes so that only screened energy levels are being probed.

An assembly of packed and oriented rodlike particles exhibit anisotropic physical properties. We investigate in the present work the anisotropic conductivity of films made of intrinsically conducting rods. These films are obtained from more or less ordered carbon nanotube liquid crystals. Their orientational order parameter is measured by polarized Raman spectroscopy. A relationship between the anisotropy of surface conductivity and orientational order parameter is determined. The experimental results are accounted for by a model that takes into account the number of intertube contacts and density of conductive pathways in different directions, as introduced by J. Fischer et al. for magnetically aligned nanotubes. We find that a good agreement, without any fitting parameter, of the proposed model and experiments is obtained when we consider a two-dimensional (2D) Gaussian distribution of the nanotube orientation. The conductivities parallel and perpendicular to the nematic director differ by almost an order of magnitude. This anisotropy is much greater than that of conventional dielectric liquid crystals, where the behavior is governed by the mobility anisotropy of ionic current carriers. The present results do not depend on the intrinsic properties of the nanotubes and are expected to be relevant for other assemblies of conducting rodlike particles, such as metallic or semi-conducting nanowires and ribbons.

Liquid crystal ordering is an opportunity to develop novel materials and applications with carbon nanotubes spontaneously aligned on macroscopic scales. Nevertheless, achievement of large orientational order parameter and extended monodomains remains challenging. In this work, we show that shortening nanotubes allows the formation of liquid crystals that can easily be oriented under the form of large macroscopic monodomains. The orientational order parameter of single-wall nanotube liquid crystals measured by polarized Raman spectroscopy at the isotropic-nematic transition exceeds by far the value reported in previous experiments. The presently measured order parameter approaches the value theoretically expected for liquid crystals made of rigid rods in solution. This finding suggests that the production of highly ordered nanotubebased liquid crystals was presumably limited in earlier contributions by the length and waviness of long nanotubes. Both factors increase the material viscosity, can yield some elasticity, and stabilize topological defects.

Présentation des objectifs et des premier résultats de la thèse

Catalytic chemical vapor deposition (CCVD) is a major method for the synthesis of carbon nanotubes (CNTs) and nanowires. Despite considerable progress in the control of the carbon nanotube length, orientation and structure, the fundamental processes involved in the nanotube nucleation, growth and termination still remain poorly understood. Rationalization is made difficult because CCVD combines the difficulties of different fields (heterogeneous catalysis, crystal growth, thin film deposition). The complexity of the problem especially arises when one considers the different chemical and physical processes involved, operating both on surface and in gas-phase, intricate at different time and spatial scales (molecular, nano, macro). Developing realistic growth models clearly requires the combination of theoretical and simulation works with reliable in situ data. We chose to investigate the growth of SWNT using in situ Raman measurements because SWNT provide an intense, specific and structure-dependent Raman signature. We initiated a systematic study of the growth conditions for different catalyst-precursor couples. From our observations, at least two conditions are required for nucleating SWNT: the catalyst metal particle must be in a reduced state and they must reach a critical carbon concentration. Once reduced, catalyst particles quickly restructure, most noticeably at high temperatures. Being in kinetic competition with SWNT growth, catalyst restructuring reduces the lifetime and the yield of SWNT growth and shifts the SWNT distribution toward larger diameters. In addition to SWNT, we evidenced the by-formation of disordered carbon nanostructures (DCN) by Raman and HRTEM. Their relative proportion is dependent on the kinetic equilibrium between the supply of carbon atoms from the gas precursor and their rearrangement at the surface of the catalyst particle. These processes also influence the kinetics of SWNT growth and its deactivation due to catalyst encapsulation by DCN. By studying the defect-induced D band of each structure (SWNT and DCN) as a function of the growth temperature, activation energies for the healing of defects can be extracted (figure 1). Radial Breathing Modes (RBM) can also be exploited to monitor the distribution of SWNT diameter. In addition to the catalyst preparation, we observed that the growth conditions (temperature and precursor pressure) significantly influence the diameter distribution (figure 2). Most noticeably, the growth of small diameter nanotubes is promoted in a narrow domain of the parameters of growth. In situ Raman experiments were performed to identify the processes influencing the diameter distribution. References [1] M. Picher, E. Anglaret, V. Jourdain, Diamond, 19 (2010), pp. 581-585. [2] M. Picher, R. Arenal, V. Jourdain, E. Anglaret, submitted. [3] M. Picher, E. Anglaret, R. Arenal, V. Jourdain, Nano Letters, 9 (2009), pp. 542–547. [4] M. Picher, E. Anglaret, R. Arenal, V. Jourdain, submitted.