PHYSICAL REVIEW B VOLUME 47, NUMBER 13 1 APRIL 1993-I Identifying the loss of critical current density in YBa2Cu307 thin films J. Deak and M. McElfresh* Physics Department, Purdue University, West 'Lafayette, Indiana 47907 John R. Clem and Zhidong Hao Ames Laboratory and Department of Physics and Astronomy, Iowa State University, Ames, Iowa 500II M. Konczykowski Laboratoire des Solides Irradies, CEREM, Ecole Polytechnique, 91128 Palaiseau, France R. Muenchausen, S. Foltyn, and R. Dye ERDC, Los Alamos National Laboratory, Los Alamos, New Mexico 87545 (Received 6 October 1992) The magnetic and magnetotransport behavior of a YBazCu307 thin-film sample was studied by a variety of methods that have been used previously to determine the irreversibility line (IRL). From transport measurements it is possible to identify at least two regions separated by a boundary identified as a vortex-glass transition, with the region below the boundary having zero linear resistance and that above having a low-current linear resistivity which seems to be associated with thermally activated Aux motion. Evidence for Aux pinning above the glass transition suggests that there is a loss of critical current density (J, ) rather than an onset of reversible magnetic behavior at the glass transition. The complications of using first-harmonic ac susceptibility (g„) to define the IRL are discussed. In addition, the third-harmonic y„ is shown to measure the same property that the first-harmonic y„measures: ac magnetic-field penetration rather than the loss of J, . Results and a theoretical model showing that there are significant differences between field-cooled dc magnetization measurements when the data are col- lected on warming versus collecting data on cooling are presented. One of the more surprising features of high- temperature superconductors has been the loss of critical current density (J, ) in modest applied magnetic fields at temperatures well below the apparent superconducting transition temperature (T, ). This loss of J, has serious implications for those attempting to use these materials in many types of applications. Many models have been put forth to describe this so-called "irreversibility line" (IRL). Some of these models described the IRL as a dy- namic crossover, ' while others have considered it to be a thermodynamic transition. ' The models based on stan- dard flux creep theories suggested that a dissipative state always existed, but the results of Koch et al. provided strong support for the vortex-glass transition models in which, below a phase transition (Ts), a nondissipative state exists. Further evidence for this transition has made the prospects for applications much more prornis- ing. The measurement of the IRL has been a fairly contro- versial problem. The measurement of the IRL by Miiller, Tagashige, and Bednorz using dc magnetization measure- rnents seemed to provide a reliable method. However, we will show that a commonly used variation of this method does not provide reliable results. The use of fundamental-frequency ac susceptibility (y, ) became a popular method to measure the IRL, even after the warn- ings of Shaulov and Dorman. More recently, the realiza- tion that g& is probably a measure of ac flux penetra- tion ' rather than irreversibility has led many to choose the third-harmonic ac susceptibility response to deter- rnine the IRL. The basis for using this method is that the existence of irreversibility will produce harmonics. ' However, we will show that there is a coincidence of both the fundamental and third-harmonic response onsets that results from ac flux penetration occurring in a region in which the current-voltage behavior is nonlinear. Epitaxial Y-Ba-Cu-0 thin films, with the c axis perpen- dicular to the film plane, were prepared by pulsed laser deposition onto heated (001) LaA103 substrates using a method described previously. " The superconducting transition temperatures (T, ) ranged from 89.8 to 90.3 K, and the transition width 6 T, was about 2 K as measured by ac susceptibility at 2. 5 MHz. Voltage as a function of current (V I) measurem-ents were made on films that had been patterned using laser ablation, with typical bridge dimensions of 70 pm wide by 300 pm long. A standard four-point probe technique was used for the V-I measurements. For all measure- ments, the dc magnetic field for the rneasurernents ranged from 0 to 55 kOe and was applied perpendicular to the plane of the film. To avoid contact heating and eliminate contact potentials, the current was applied for 2 periods of a 1. 3-Hz square wave and followed by a 5-s pause. The temperature associated with each V-I curve is an average of the temperatures recorded at each data point of the curve with maximum temperature drift of 0.1 K ob- served. All measurements were zero-field cooled, and data were collected on warming. ac susceptibility (y„) measurements of the fundamental frequency and the third harmonic were made on the same films used in the V-I measurements. The y„measure- ments utilized a single-coil self-inductance technique for 47 1993 The American Physical Society