Most fullerenes are wide band gap semiconductors. The C60 conduction band accommodates six electrons per molecule. On doping with alkali metals (A) one electron should be transferred per alkali metal ion. We expect C60 and A6C60 (filled conduction band) to be insulators, while intermediate compounds AxC60 (x=0-6) should be metals. A3C60 compounds are indeed the "best" metals and even superconductors.
This model fails for A4C60 compounds. Filling the fullerene interstitials with four heavy alkali metal ions per C60 gives a body-centered tetragonal lattice with no metallic properties, while for the light compounds Li4C60 and Na4C60 covalent bonds form between fullerene ions, creating semiconducting or semimetallic polymers.
NMR measurements on Rb4C60 to 1.2 GPa have indicated the existence of a high pressure insulator-to-metal transition. Because the semiconducting A4C60 compounds are not very well understood, this transition into the "normal" metallic state is very interesting. In spite of this, few experimental studies have been made on Rb4C60 under pressure, and the results are contradictory.
We have now measured the electrical resistance of Rb4C60 as a function of pressure and temperature. To avoid complications due to metallic Rb3C60 the samples were intentionally produced to contain a mixture of Rb4C60 and Rb6C60, both of which are insulators at zero pressure. The air sensitive samples were loaded into the pressure cell in an Ar atmosphere, using hexagonal boron nitride as pressure medium. The measured resistance decreases with increasing pressure up to 2 GPa, but there is no sharp transition to a "metallic" state (with positive temperature coefficient). At all temperatures and pressures investigated (150-400 K, 0-2 GPa) the resistivity is best described by a hopping model, with fitted parameters that change continuously with pressure. We cannot rule out an initial, sluggish drop in the magnitude of the resistivity with increasing pressure below 300 MPa, possibly connected with the structural changes observed in other experiments, but the effect is weak and could equally well be an effect of the initial compaction of the powder sample.
|