Study of Grain Boundaries in Ca-Doped YbBCO Thin Films

While rare-earth-barium-copper-oxide (REBCO) coated conductor is a promising candidate as a practical superconductor for high magnetic field applications, its performance is largely limited by the depressed current-carrying capability of its grain boundaries (GBs). GBs with misorientation angles beyond relatively low values (3-5o) can substantially reduce the critical current density (Jc) of GBs, which makes the GB problem a critical one, since real, long-length coated conductors are made on polycrystalline substrates, where GBs are inevitable. The reduced critical current density of GBs has multiple origins, including the existence of non-superconducting dislocation cores, dislocation-strain driven high oxygen vacancies and lowered charge-carrier (hole) density, and excessive local charge due to deviations from stoichiometry that can only be poorly screened due to the low carrier density Since Ca-doping introduces extra holes into the REBCO compounds, it has long been considered a possible method to improve GB transparency. The positive effect of Ca-doping has been verified by a series of studies. Later detailed microstructural study showed that the underlying reason for improved GB properties is not simply due to a uniform Ca-doping because Ca was found to concentrate preferentially at GBs. Furthermore, the Ca distribution along and across the GB is highly non-uniform, especially for low angle grain boundaries that are critical in providing the high-Jc current path. A segregation model was proposed to explain the observed complex GB Ca distribution to dislocation cores and channels characteristic of such low angle grain boundaries (LAGB). This model shows that Ca segregation is driven by the combined effects of local strain and local charge at the grain boundary. This thesis work presents a systematic study of GBs in Ca-doped YbBCO thin films. While a number of works have been done on GBs in Ca-doped YBCO films, this work significantly expands understanding of doped GBs in following aspects. First, the strain and charge model developed in our group predicts a stronger segregation in Ca-doped REBCO for RE elements smaller than Y. We chose YbBCO because Yb is the smallest RE ion possible in REBCO. Second, previous work mainly focused on evaluating the GB properties at high temperatures in the vicinity of 60-80K. This work expands the study of Ca-doped GBs into the low temperature (4-30 K) regime relevant to the growing interest in using REBCO coated conductors for magnetic field applications. Other variables that also affect GB transparency but have been less studied before are also included in this study. These variables include the applied field orientation and the intra-grain oxygen content level. Third, this study uses Low Temperature Scanning Laser Microscopy (LTSLM) to study GB properties. LTSLM is a tool that detects the local electric field, a rather powerful tool for observing where the dominant source of E occurs in a thin film track. Fourth, thin film samples used in this study are grown using carefully optimized conditions. Effects of major parameters are studied systematically. Collectively this new study allows us to more completely evaluate the effect of Ca-doping in samples with both optimized and un-optimized intra-grain properties. The comparison of intra-grain and inter-grain Jc in Ca-doped YbBCO films shows that GBs become more transparent with Ca-doping at high temperatures (T=0.85Tc). For YbBCO, the optimum Ca-doping level is found to be lower than that for YBCO, a result consistent with the charge-segregation model. On the other hand, the beneficial effect of Ca-doping decreases rapidly as the temperature is lowered, becoming then of only marginal value. This temperature dependence is also analyzed in the framework of segregation model, which predicts a depressed segregation in the channels between dislocation where supercurrent is expected to flow. Thus the channels benefit from having a higher local Tc, the effect of such a Tc enhancement being naturally most pronounced at high temperatures. Using the LTSLM to explicitly observe the local electric field, we can for the first time unambiguously distinguish intra-grain and inter-grain Jc These results show that though GB dislocations reduce the current-carrying cross section, they can also act as pinning centers. When the direction of the applied field coincides with the orientation of the GB dislocations, which occurs for H parallel to the c-axis and the perpendicular to the film plane, the GB pinning strength is stronger than the intra-grain pinning due to naturally-occurring defects. More interestingly, depending on the strength of the intra-grain pinning, pinning by GB dislocations can result in the interesting condition of intra-grain Jc being greater than inter-grain Jc. Implied crossovers of this sort have been inferred before but the LTSLM has allowed us to provide the first experimental confirmation of this interesting state. However, pinning by GB dislocations is strongly anisotropic and becomes very weak in parallel field. The optimized film growth conditions for YbBCO are different from those needed for YBCO. YbBCO growth generally requires higher oxygen partial pressure and lower temperature due to its lower melting point. Dependencies of film morphology on different Ca-doping levels and substrate types are also found. An important conclusion of this work is that Ca-doping does not effectively improve GB properties at low temperature, even though it can do so at high temperatures near Tc. Nevertheless, this does not prove that the problem of GB limitation of the critical current density cannot be resolved. In the final part of the thesis, through comparison with literature, we discuss how some naturally-occurring GBs (usually meandered GBs of mixed type, far from the highly idealized [001] tilt geometry studied here) can be better than these artificial straight GBs typical of those grown by PLD, which dominate most present studies of GBs. A concluding recommendation is that future GB studies should focus more on real GBs.