When double-wall carbon nanotubes are combined with iodine and chlorosulfonic acid, the charge transfer increases the conductivity by a factor of 7. The doping species move the Fermi level up to 1.1 eV while not changing the average electronic mean free path at 30 nm. Works combining both optical and transport experiments, and simulations reveal the underlying mechanism.
Two strategies are possible to obtain high electrical conductivity values in carbon nanotube (CNT) fibers,: (1) increasing the electronic mean free path and (2) increasing the number of electronic conduction channels in each NTC by moving the Fermi level through charge transfer.
Chlorosulfonic acid (CSA) is a true solvent of CNT. If, during CNT fiber synthesis, a part of CSA remains in the material, a charge transfer thus occurs and leads to an up-shift of the Fermi level of 0.7 eV. The consequence is an increase of the conductivity by a factor of 5 compared to a system composed of CNT only. Similar porperties can be obtained in pure CNT fiber impregnate with iodine. In that case, the conductivity gain is a slightly larger, approaching factor of 7, and the Fermi level up-shift is 1.1 eV.
At high currents, the CNT fibers are heated by Joule effect and the chemical species (iodine or CSA) evaporate. The pure (dedoped) NTC fibers are then obtained allowing the evaluation of the conductivity gain.
Raman spectrometry permits to measure the Fermi level displacement in double-wall CNT. The high expertise of the CEMES group on optical spectroscopy and on C-based materials has allowed designing accurate fitting strategies to extract reliable information. By analyzing the Raman signals from both the inner and outer tubes of the double-wall CNT, they succedded in measuring the charge transfer and, in the case of high doping, to convert it into the Fermi level up-shift. For the first time, it was possible to go further in the analysis in terms of average free path for non-defective CNTs. The value of about 30 nm was obtained independent of the Fermi level position.
The charge transfer process in CSA doped CTNs is relatively obvious while no convincing explanation has been published yet for iodine doped CNTs. Quantum calculations were then carried out to systems typically composed of 500 atoms to state how iodine atoms mutually arranged. Constrained by the material geometry to remain aligned, they form chains which are polarized (6 iodine atoms form 2 tri-iodides in contact for example) and thus, allow the formation of a very large number of ions, the role of counter ion being played by the CNT and explaining the charge transfert experimental result.
These works involving synthesis, optical and electrical characterizations, and numerical simulation were carried out within the framework of three PhD projects and have combined the specific fields of expertise of three French laboratories in Toulouse: CEMES, LPCNO and CIRIMAT and two groupes in United States : Rensselaer and Rice Universities.
Pascal Puech, Pascal.Puech at cemes.fr