Surface resistance of epitaxial YBa2Cu3O7-x films on various substrates: Effects of pair condensation and quasiparticle scattering

The temperature dependent surface resistance R-s(T) of eight high-quality epitaxial YBa2CU3O7-x films on LaAlO3(Y/L), MgO (Y/M), and CeO2-buffered sapphire (Y/S) substrates was investigated at 19 GHz with a resolution of 20 mu Omega. The residual level R-s(T-->0) was (90-180) mu Omega for all films. The slope delta R-s(T)/delta T in the range (4-20) K decreased from (6-8) mu Omega/K for Y/L to (0-2) mu Omega/K for Y/M, and to slightly negative values for Y/S. This slope correlated with the transition temperature T-c, including published data of about (10-20) mu Omega/K for high-purity YBa2Cu3O7-x, single crystals. The increase of R-s from 4.2 showed power-law behavior Delta R-s(T)proportional to T-a up to 40 K with a approximate to 1.3 for Y/L and Y/M. However, it was exponential for Y/S, Delta R-s(T)proportional to exp (-delta XTc/T) with delta approximate to 0.8. The results can be described with the two-fluid model in terms of the quasiparticle density n(N)(T) and the scattering time tau(T). The high and reproducible residual resistance implies (magnetic) impurity scattering at a rate tau(-1)(0)proportional to n(N)(0), i.e., proportional to the density of quasiparticles. In crystals and unstrained films, the pair condensation is gapless, and the power-law temperature dependences of tau and n(N) are reflected in that of R-s. In contrast, in strained films, n(N)(T) displays activated behavior, and the low-temperature behavior of R-s changes to exponential. The formation of an energy gap is attributed to the interaction between the Cu-O planes and the chains, which is affected by strain. The existence of an energy range with zero density of states limits possible interpretations of the order parameter of YBaCuO. A two-band scenario with magnetic pairbreaking and two different order parameters (2 Delta/kT(c)=6-8 for the planes and greater than or equal to 0 for the chains) with s-wave symmetry, at least in the chains, appears to be an adequate explanation.