DYNAMIC CONTRIBUTION TO THE FISHTAIL EFFECT IN A TWIN-FREE DYBA2CU3O7-DELTA SINGLE-CRYSTAL

Superconducting current densities j(s) and dynamic relaxation rates Q = d ln j(s)/d In(dB(e)/dt), where dB(e)/dt is the sweep rate of the ex ternal magnetic field B-e, were measured as a function of temperature (5 K < T < 65 K) in magnetic fields up to 7 T on a twin-free DyBa2Cu3O 7-delta single crystal by means of a high-sensitivity capacitance torq ue magnetometer. Above 15 K, we observe a ''fishtail'' effect, i.e. a pronounced minimum in the j(s)(B-e) curve at fields around B-e = 1 T. The relaxation rate Q shows an anomalous increase at low fields which is correlated to the minimum in the j(s)(B-e) curve. Both the j(s) ver sus B-e and Q versus B-e data are used as input parameters into the ge neralized inversion scheme developed by Schnack et al. [Phys. Rev. B 4 8 (1993) 13178] to calculate the true critical current density j(c) wh ich is by definition independent of relaxation effects. Interestingly, the j(c)(B-e, T) curves derived in this way do not show a minimum. Th is points clearly to a dynamic contribution to the fishtail effect. Th e true critical current density j(c)(B-e, T) decreases weakly with inc reasing B-e over the entire measured temperature and field range, as e xpected for single-vortex pinning. This indicates that the observed fi shtail effect is not caused by a crossover from single-vortex pinning to pinning of flux bundles. The temperature dependence of j(c) is in g ood agreement with the predictions of a model based on single-vortex p inning caused by spatial fluctuations in the charge-carrier mean free path.