Do the stripes explain the c-axis transport in cuprates?

Abstract

We propose that the stripes running differently in different CuO 2-planes cause large impurity and diffusion terms to the c-axis scattering rate τ c −1 in addition to the quadratic Fermi-liquid term. The main assumption made here is that the stripes of localized charge are formed by the bosons localized into rows of tilted CuO 6 octahedra in CuO 2-planes (D-stripes). They can scatter the 3D holes traversing into c-direction which experience the stripes as impurities giving rise to a constant scattering rate and the disorder gives linear temperature dependence. With the Fermi-liquid term added the scattering rate becomes τ c −1( t)= a c + b c t+ d c t 2, with t= T/ T *, where T * is the scaling temperature. In the ab-plane direction the situation is different since the holes and the bosons coexist within the realm of the chemical equilibrium reaction B ++ ⇌ 2h + and the localized bosons do not scatter the holes. Some disorder due to the vacancies in the 2D boson lattice causes still a small linear diffusion term and the rate becomes τ ab −1( t)= bt+ dt 2. Assuming the same carrier system (the holes), the ratio of resistivities for T> T c becomes ρ c / ρ ab ∼( m c / m ab )( a c / b) t −1, where m c and m ab are the effective masses. Since a c can be large and b small, the coefficient of the leading temperature dependence t −1 can be much larger than the mass ratio. This result is in agreement with experiments. Since the carrier density of holes obtained from the chemical equilibrium has linear temperature dependence we obtain linear planar resistivity ρ ab ( t)= A+ Bt and in the c-direction the same singular behavior ρ c ∼ t −1.