Modeling the current‐voltage characteristics of silver‐sheathed Bi‐Sr‐Ca‐Cu‐O tapes

A. Gurevich, A. E. Pashitski, H. S. Edelman, D. C. Larbalestier
1993 Applied Physics Letters  
Measurements of voltage-current (V-I) characteristics and critical currents of 2212 and 2223 Ag-clad Bi-Sr-Ca-Cu-0 tapes at 0 < B < 7 T and 4.2 and 77 K are presented. We show that the V-1 curves are nonlinear at I < 1, and become linear above the critical current I > 1, due to the effect of the Ag cladding. It is shown that the V-1 curves can be well described by a simple universal formula, which enables one to extract I,( T, B) and the flux creep rate s( T, B) from the resistive measurements,
more » ... stive measurements, taking into account the effects of voltage criterion, strong nonlinearity of V(I) and I < I, and the dependence of the resistivity of Ag upon T and B. For a 2212 tape, I,(B) is shown to decrease as B-" with a=O.15 at 4.2 K, and exponentially at 77 K. For a 2223 tape, I,(B) displays a power law dependence with a=O.1-0.3 at 4.2 K, whereas at 77 K I,(B) can be well described by the formula I"b-"'( 1 -b), where b= B/B*, and B* is the irreversibility field. The results are interpreted in terms of strong bulk pinning modified by thermal fluctuations. Recent reports of high critical-current densities, J, in Bi-Sr-Ca-Cu-O/Ag (BSCCO) tapeslw3 have focused more attention on large-scale applications of high T, superconductors. At the same time, the mechanisms which determine J, in BSCCO tapes remain unclear due to their complicated multiphase microstructure"5 and to the specific features of flux dynamics and pinning which result from short coherence length, high anisotropy, and significant thermal fluctuations6 This leads to giant flux creep and a strong dependence of J, upon the voltage criterion, V, in resistive measurements or on the sweep rate in magnetization measurements,7'8 which manifests itself in a vanishing of J,( T, B) at the irreversibility field B* ( T) well below the upper critical field Bc2( T). At B < B*(T) the critical current can exhibit qualitatively different behavior in various regions of temperature T and magnetic induction B. For instance, J,(B) of BSCCO tapes at 4.2 K sharply drops at low B and then displays a plateau up to B -20 T, whereas at higher T the J,(B) dependence becomes exponential.'-3 These quite different characteristics appear to indicate that there is no universal mechanism of critical current control, although some factors which determine J,( T,B) may dominate in certain regions of T and B. For example, J,(B) reveals a Josephson-like behavior at low B and T, which may result from a weak-link structure along the c-axis,g whereas the exponential decrease of J,( T,B) at larger T and B may correspond to a smearing of the bulk pinning potential by thermal fluctuations." Under these circumstances, any particular operational definition of J, cannot be expected to be equally valid in all domains of T and B. These features of J,( T, B, V,) can be extracted from the voltage-current ( V-I) characteristics which, together with the Maxwell equations, determine the current-carrying capacity and relaxation of the critical state. In this letter we present measurements of the V-I characteristics of Bi,Sr,Ca,Cu30X( 2223) and Bi,Sr,CaCu,O, (22 12) Agclad tapes at 4.2 and 77 K for 0 < B < 7 T. A universal scheme is proposed to describe V(I) of the BSCCO tapes and to study the dependencies of J, and flux creep parameters upon T, B, V" and the properties of the normal cladding. Different regimes of critical-current control are shown to exist, depending on T and B. The tapes were prepared by the oxide-powder-in-tube method as described previously,4*5 then rolled, sintered, and cut into 2 cm long pieces. The samples had a BSCCO core ~25-50 pm thick and -2 mm wide, surrounded by Ag sheath of thickness e-254 pm. The field B was perpendicular to the plane of the tapes and was thus approximately parallel to the c-axis. The I-V curves of the 2212 and 2223 tapes measured by the standard four-probe method at various T and B are shown in Figs. 1 and 2. Our analysis of the experimental data exploits the large difference between the resistivity of the Ag cladding, pAs, and the flux flow resistivity, pf = pnB/B" where p" is the normal state resistivity of BSCCO (p,). In this case, all current in excess of I, is shunted into the Ag sheath, and the superconducting core remains in the flux creep regime, even for I considerably exceeding I,. The absence of flux flow in BSCCO allows one to write the electric field E(J) at J< J, in a fairly general form E(J) =E, exp[-U(J)/kT]. Here U(J) is the flux creep barrier which is a nonlinear function of J and vanishes at J= J, E, is a crossover electric field between the flux flow and flux creep regimes, and J is the current density in the superconductor. At small J the function U(J) can be qualitatively different in various models of flux dynamics,6 however, in resistive measurements limited by a particular voltage sensitivity (usually -0.1 pV), the details of U(J) at J<J, are irrelevant, and the observed V(I) at B < B* ( T) is determined by the dependence of U(J) near J-J,. Then E(J) can be obtained by expanding U(J) in a power series in J,-J, keeping only the first term, i.e., U(J)/kT= (J,--J)/J,." This yields E(J)=E,exp[(J-J,)/JI] (1) where J,= kTJdUo can be expressed via the observed flux 1688
doi:10.1063/1.109577 fatcat:i6e4vmv72ng33iskwcqfdtta4q