Conductors of HTS: Key Problems David Larbalestier Applied Superconductivity Center Department of Materials Science and Engineering Department of Physics University of Wisconsin General reference: David Larbalestier, Alex Gurevich, Matthew Feldmann and Anatoly Polyanskii, “High Transition Temperature Superconducting Materials For Electric Power Applications”, Nature 414, 368-377, (2001).
Generations of Conductors Generation I: Bi-2223 – available now Generation II: Coated YBCO conductor. (IBAD, RABiTS or ISD) – available 2005?
Underlying Issues Physics Materials Science Fabrication Anisotropy and flux pinning Materials Science Grain boundaries Epitaxial growth Phase control and conversion Fabrication Metal working for BSCCO Epitaxial multilayers for YBCO BSCCO-2223 available now, YBCO in development
Practical HTS Materials Bi-Sr-Ca-Cu-O (Bi-2212, Bi-2223) Tc ~92K and ~108 K Basis for ALL present applications micaceous double Bi-O charge reservoir layer that is almost insulating, making the compounds an SIS stack Y-Ba-Cu-O (Y-123) Tc ~92K basis for new generation II technology less anisotropic because charge reservoir layer is metallic grain boundaries less transparent than BSCCO
YBCO Cu-O chain layer CuO2 layer Y Ba YBCO (YBa2Cu3O7-x) possesses the first crystal structure with Tc > 77K It has defined cation stoichiometry 1:2:3, but O content is variable from 6-7 Tc > 90K demands x<0.05 YBCO is often thought of as being archetypal but in fact the Cu-O chain layers are very unusual Make charge reservoir layer metallic Most HTS are 2D, but YBCO is anisotropic 3D with electron mass anisotropy g = (mc/ma)0.5 of ~7
The BSCCO Family g ~ 200-300 g may be ~ 50 Tc ~30K ~70-95K ~105-110K Y. Zhu (BNL) CuO2 Ca Bi-O double layer Sr g ~ 200-300 g may be ~ 50 Tc ~30K ~70-95K ~105-110K
H-T Plane of Superconductors 10 20 30 40 60 90 120 Temperature (K) Field (T) Helium Nb-Ti Nb3Sn MgB2 film MgB2 bulk Hydrogen Neon Nitrogen BSCCO YBCO Cooling liquids
The Irreversibility Line (H*(T)): Material Anisotropy Problem After Suenaga et al. 1991 After Tallon et al. 1996 Thick charge reservoir layers between the CuO2 planes decouple
Basic Materials Issues High anisotropy ratios ~7 YBCO to ~200 Bi-2212 Irreversibility fields much lower than Hc2 Easy thermal flux depinning Tendency for finite dissipation at all J, even below Jc! Very short coherence lengths xab 1-2 nm, xc < 1 nm grain boundaries and other interfaces strongly perturbed grain boundaries become large barrier to current flow except at very small misorientations Grain boundaries are strong obstacles to current Critical current densities in low temperature limit are very high, 106-107 A/cm2 within the grains, but the overall Jc averaged over large length scales is much less!
Jc Issues Ic is measured quantity Jc is desired quantity Ic/A = Jc ?? Jc is desired quantity Fp = S np fp = Jc x B np density of flux pinning defects fp calculable to factors of a few Strong and dense pinning in HTS diminished by thermal activation as T increases Single crystals: Aeffective = A All polycrystals, Aeffective < A, sometimes <<
Electronic State of Low-Angle GB SC Carrier density Temperature Optimum doping Sub-optimum, carrier-depleted GB state Insulating dislocation cores J A-J vortices at GB Abrikosov vortices GB Lattice dislocations enable the misorientation BUT destroy local superconductivity Non-stoichiometry at GB produces charge imbalance required screening depresses superconducting gap further HRTEM image of 8°[001] tilt GB in Bi2Sr2CaCu2Ox …increase carrier density of GB…. Gurevich PRB 98, 99, Babcock, Larbalestier JMR 96
Dislocation core overlap models explain a decrease in Jb/Jc with increasing q as a decrease in superconducting area at the GB. Low angle tilt grain boundary showing GB dislocations. Dislocation cores are highly distorted regions and are likely to be insulating. Jb/Jc = (D - 2rm)/D = 1 - 2rmq/|b| D.Dimos, P.Chaudhari, J.Mannhart, F.K.LeGoues, PRL 61, 219, (1988). M.F.Chisholm, S.J. Pennycook, Nature 351, 47 (1991).
The Grain Boundary Problem Grain Boundary Weak Links Grain Boundaries obstruct current except when low angle Strong texture is needed! v1 SrTiO3 Common [hkl] v2 YBCO, BSCCO Dimos et al. 1990 Heinig et al. 1996, 1999
Current percolates in HTS Jc is really defined by Ic/A Local Jc is much higher than Ic/A, if A is taken as the whole cross-section How much cross-section does the transport current use? advances in Jc are coming more from elimination of barriers to current flow than by raising flux pinning YBCO conductors may have higher percolation efficiency but are also subject to the same limit just beginning to understand barriers in YBCO
Two Conductor Generations First is based on Bi-Sr-Ca-Cu-O (BSCCO) 2212 has Tc ~90 K 2223 has Tc ~107 K Second is based on biaxially aligned YBCO Tc ~ 90 K 3 types of coated conductors IBAD (Fujikura, LANL) RABiTS™ (ORNL) Inclined Substrate Sputtering (Sumitomo, ANL) Generation I: 85 filament BSCCO-2223 (American Superconductor Corp.) Generation II: YBCO tape (LANL)
Conventional Bi-2223 Process n(DS) where n 2 Rolling to tape results in densification (~90% density) Heat Treat 1 results in dedensification and growth of secondary phase particles (~70% density) Intermediate Rolling results in crack and Hillock formation (~88% density) Final Heat Treat results in partial crack healing, dedensification and 3221 formation (~73% density in product)
Magneto-Optical Imaging Technique Indicator film GGG YIG Aluminum Hard protective layer
BSCCO varies on Individual Filament Scale 300 mm wide filament fed by 12 mm diameter Au leads 55 kA/cm2 AMSC tape Van der Meer, Cai unpublished Tc Jc
Central, best filament Optical image 200 mm MO image ZFC T=12K H=40 mT SEM image Van der Meer, Cai, Jiang, Polyanskii unpublished
Central, highest Jc filament Top Surface MO image ZFC T=12K H=40 mT MO and SEM correlation Central, highest Jc filament Top Surface MO image ZFC T=12K H=40 mT Van der Meer, Cai, Jiang, Polyanskii unpublished Lighter regions admit flux and are weaker
MO reconstructions in Slab Geometry x y z MO reconstructions in Slab Geometry Requirements for CR: 2D current flow Sample infinite in -Z direction aspect ration of BSCCO tape is excellent approximation reconstructed here z = 0 z > 0 measured here If Bz in the plane z=0 is known, then AND Feldmann
MO reconstructions Highest values 184 kA/cm2 (36mT||ab), Jc(0T) 35 kA/cm2 Feldmann
Evaluation of Local Maxima Ha 36 mT Ha 45 mT Transport: Jc(0T) = 35 kA/cm2 Reconstruction: Jc (36mT) = 184 kA/cm2, |Jc|(36mT) ~ 40 kA/cm2 Feldmann, Cai
MO reconstructions HT1 and FHT HT 1 7 kA/cm2 After Final HT 30 kA/cm2 OB8050 77K, 36 mT H parallel to ab
MO and SEM Comparison FHT Uncertain due to imaging film domain structure OB8050 Final state 30 kA/cm2 150A David C Larbalestier: Identify the black
High Jc Tapes Still Contain “Invisible” 2212 Rolling reduction 5 gauss field parallel, so current flowing along ab planes Jc ~35 kA/cm2 Do the 2212 limit current? Quasi-Ideal moment as measured Large moment loss at 77 K (30% rolling reduction) Source is 2212 intergrowths Huang, Cai et al. ICMC 2001 to appear
Where is the 2212? 2212 seen as light-colored streaks in darker 2223 43,200 A/cm2 AMSC cable tape 47,800 A/cm2 2212 seen as light-colored streaks in darker 2223 Better seen in TEM as by Holesinger Huang, Cai et al. ICMC 2001 to appear
Correlation: 2212 Kink and Jc(77 K) Fragmented state 10 20 30 40 50 60 0.05 0.1 0.15 0.2 0.25 A-B (emu at 80 K) Jc (kA/cm2, 77 K, sf) “2212 kink” vs. Jc A B Simple correlation suggests 25% improvement from 2212 removal alone! Jc is a new benchmark for cable wire, Ic = 170A! Huang, Cai et al. ICMC 2001 to appear
BSCCO-2223 Summary Material is still improving strongly BSCCO now ~40% cross-section, balance Ag Very complex material system slowly being understood (Nb-Ti took 20 years – 1964-1984!) The conductor of choice for power applications.
Acknowledgements To my colleagues in UW-ASC (S. Babcock, X. Y. Cai, E. Hellstrom, A. Gurevich, A. Polyanskii) and present (J. Chandler, O van der Meer, Yuan) and former students J. Parrell (now OST), J. Anderson (now Motorola), M. Feldmann, and J. Jiang. To collaborators in the Wire Development Group at ANL (V. Maroni), LANL (T. Holesinger), ORNL (D. Kroeger) and ASC (Y. Huang, R. Parrella, and B. Riley). To AFOSR, DOE, and NSF-MRSEC for support.