Prof. Selvamanickam's Research Group
High Performance second-generation High Temperature Superconducting Wires

When produced in form of a thin film, it was demonstrated that YBa2Cu3Ox (YBCO) superconductors can sustain current densities of about 5 MA/cm2 compared to about 50,000 A/cm2 in first-generation (1G) HTS wires. However, such high current densities were achievable only on single crystalline substrates such as SrTiO3 and LaAlO3 which did not present a way to produce long wires. Flexible superconducting YBCO tapes were demonstrated by depositing the superconducting film atop nickel alloy substrates using intermediate buffer layers as diffusion barriers. However, such tapes were able to sustain current densities about 100 times lower than that achieved on single crystalline substrates. The reason was traced to grain to grain misorientation in polycrystalline films and their drastic negative influence on current density. The first innovative technique developed to solve this problem was ion beam assisted deposition (IBAD) of biaxially-textured buffer layers atop nickel alloy substrates [1].

When deposited under appropriate conditions, a high degree of cube texture of about 5 degrees within the film plane could be achieved within a thickness of 10 nm [2]. This biaxial texture is then epitaxially transferred to a superconducting film, where a grain-to-grain misorientation of just 2 degrees could be achieved. Such superconducting films on IBAD layers on nickel alloy substrates could sustain current densities of 5 MA/cm2 which is equal to that achieved on single crystal substrates. Hence a breakthrough was achieved in fabricating second-generation (2G) superconducting wires that can carry up to 200 times more current than a conventional copper conductor of the same cross section.

2G HTS Wire Architecture and Thin Film Processes :
A schematic of a complete stack of a type of thin film superconducting wire is schematically shown in the following figure. As shown in the figure, 90% of the wire is comprised of inexpensive substrate and copper stabilizer.


Schematic of the multilayered architecture of a second-generation high temperature superconducting wire

Spool of 2G HTS wire

Essentially, any substrate that is flexible and can withstand the high-temperature (up to 850°C) process conditions used in subsequent thin film processing can be used. Hastelloy C-276 substrate is typically used for its high temperature properties as well as its excellent strength that is valuable in fabrication of cables and coils from the final wire product. The substrate is typically polished by an electropolishing process with a surface roughness of 0.1 nm as measured by atomic force microscopy. Such a high degree of surface finish is due to the stringent requirement of the subsequent IBAD nucleation and growth process. An 80 nm thick alumina layer is deposited on the substrate by high-rate magnetron sputtering. The alumina layer is amorphous and serves as a barrier to cation diffusion from the substrate to the superconducting layer and resulting poisoning of the layer. About 7 nm thick yttria layer is deposited by magnetron sputtering atop the alumina and serves as a nucleation layer for the subsequent IBAD process. Under the appropriate conditions of the ion assist, the grains of the IBAD film (in this case MgO) are biaxially aligned both within the plane of the film and normal to the film. The biaxial texture originates essentially from the initial few nanometers during nucleation and subsequent growth of the film. Only 10 nm of IBAD MgO is needed to achieve a high degree of biaxial texture. The three layers, alumina, yttria, and MgO are all deposited at room temperature. This is an important advantage of this approach since essentially any substrate could be used, including metals, glasses and polymers. In the case of HTS of course, heat resistant high temperature alloys are necessary for the substrate.

All layers subsequently deposited on the IBAD MgO film are epitaxially grown at temperatures ranging from 600 to 850C. As shown in the Figure, a 30 nm thick film of MgO is homo-epitaxially grown on the IBAD MgO layer. This layer is deposited by magnetron sputtering and serves to sharpen the biaxial texture originated in the IBAD MgO film. The final layer in the 5-layer buffer stack is LaMnO3 (LMO) that is about 30 nm thick and deposited by magnetron sputtering. LMO provided a good lattice match with the following superconducting film. Without a lattice-matched film, the superconducting layer may grow in mutliple orientations within the plane, specifically at 0 and 45 degrees, which is deleterious to current flow. The final epitaxial layer that is grown in the stack is of course the superconducting layer, typically YBCO. Second-generation HTS wires are now produced by industry in lengths of a kilometer with a critical current performance of 200 A/cm. For initial commercial use, the price of 2G HTS wires needs to be reduced to $ 50/kA-m i.e. by an order of magnitude lower than the price today. Technological advances need to be made in order to reach critical current target levels of 700 A/cm and reduce costs (in $/m) by 3-fold.

High critical current performance in thick films :
A challenge with achieving high critical current is the reduction in critical current density with thickness of the superconducting layer as shown in the following figure. Since the superconductor layer is the thickest of all oxide layers by far (about 10x thickness of all buffer layers together), it is also important to achieve high critical current in thinner films in order to minimize the reduce overall cost of the wire as well as to increase the process speed (which decreases with increasing film thickness for a given deposition rate).
At the University of Houston, we are working on improving critical current performance levels in thick films by understanding current limiting mechanisms in superconducting films deposited by metal organic chemical vapor deposition (MOCVD). These include understanding of the initial nucleation stages in the MOCVD process, formation of secondary phases such as CuO and the associated misaligned grain growth, role of buffer texture and surface properties, rare-earth combination, overall composition of the superconducting layer such as rare-earth type and ratio, nature of interfacial defects such as misfit dislocations that result in threading dislocations in the bulk. Our goal is to identify the microstructural current limiting factors and then modify the processes and the conductor architecture and composition accordingly to eliminate those factors.

Thickness dependence of critical current density of second-generation high temperature superconducting wire fabricated by metal organic chemical vapor deposition

Growth of misaligned a-axis grain in a superconducting film. It appears that the grain nucleated from a CuO grain on the buffer

High magnetic field performance in 2G HTS wires :
 Since superconductors are typically used in presence of a magnetic field, their critical current performance in a magnetic field is an important metric. In a typical second-generation HTS wire, the critical current density decreases by a factor of 7 to 10 when a magnetic field of 1 T is applied perpendicular to the c-axis (see figure). Recently, substantial progress with doping of foreign elements such as Zr, Sn, Ti have been made to reduce the drop in the critical current density with field. Most of the advances have been made using pulsed laser deposition (PLD) which is generally considered to be a laboratory technique. At the University of Houston, we are working on understanding the flux pinning centers responsible for in-field performance and anisotropy of MOCVD-derived films, which are used in industrial 2G HTS wires. We are working on the influence of rare-earth type and composition, foreign element doping such as Zr, modification of buffer layer surface on the microstructure to achieve superior in-field performance of critical current and reduced anisotropy. The following figure shows the performance in magnetic field of the state-of-the-art MOCVD films and the target performance. The state-of-the-art MOCVD films made with Zr doping shows columnar defects of BaZrO3 (BZO) primarily in the vertical direction, leading to improved pinning in this orientation. As indicated in the Figure, our target is to achieve 50% retention in critical current at 77 K and 1 T, from the zero field values.

Performance of a typical second-generation superconducting wire in a magnetic field at 77 K (red curve) and the target performance (blue curve).

Cross-sectional microstructure of a Zr-doped superconducting film synthesized by MOCVD showing abundant columnar structures of self-assembled BaZrO3.

The process and microstructural modifications made in the state-of-the-art MOCVD films have been found to reduce the overall anisotropy of the 2G HTS wire which is also an important metric. The lowest critical current which occurs at an intermediate angle of magnetic field between parallel and perpendicular to the a-b plane is the limiting factor in performance of a coil and hence it is important to reduce the anisotropy. The following figure shows the critical current and anisotropy performance of state-of-the-art MOCVD-derived films as compared to a baseline film. The figure also displays the target performance and anisotropy levels. As shown, the target is to achieve a completely isotropically-performing material with 50% retention in critical current at 77 K and 1 T, uniformly at all field angles.

Anisotropy in critical current density of second-generation high temperature superconducting wire

Simplified 2G HTS multilayer architecture :
A key factor that determines the cost of 2G HTS wires is the yield of the manufacturing process which in turn is related to the number and complexity of manufacturing steps. The architecture of 2G HTS wire is complicated containing several layers which have to be carefully grown over kilometer lengths. Simplification of the architecture is bound to result in fewer problems and hence higher yield. At University of Houston we are evaluating approaches such as planarization of substrate to eliminate electropolishing of substrate as well as vacuum deposition of initial barrier and nucleation layer, identify and synthesize multifunctional layers that would lead to a reduction in number of buffer layers from 5 to 3 or even 2. In order for the modified architecture to perform as well as the standard architecture, it is important to study barrier properties to cation diffusion, the influence of nucleation of key template layer deposited by IBAD, in-plane and out-of-plane texture and surface roughness of modified buffers, and lattice match with superconducting layer.

High-rate MOCVD processes :
Another factor that affects the cost of 2G HTS wires as well as their production capacity is the deposition rate of the superconducting layer, which is the thickest and most complex ceramic layer in the 2G HTS wire architecture. MOCVD has been used to demonstrate the most superior wire performance as essentially all length scales. Further, deposition rates of about 0.5 to 1 micrometers/minute have been demonstrated with MOCVD. This rate can be further increased if the reaction kinetics of REBa2Cu3Ox formation from the complex MOCVD precursors can be increased. We are investigating the use of plasma assist to increase REBCO formation kinetics which will also increase the efficiency of conversion of precursor to film.

Advanced practical 2G HTS wire :
Beyond critical current, in-field performance, simpler/fewer buffers and deposition rate, it is important to develop superconductors which can meet other application requirements such as low ac losses, quench stability, robust mechanical properties, persistent joints, fault current limitation and recovery to name a few. Reduction in ac losses requires low hysteretic, eddy current, and coupling losses. Hysteretic losses can be minimized by subdividing the superconductor into fine filaments. At UH, we are working on dry etch and wet etch techniques to create multifilamentary 2G HTS wire architectures that lead to reduction in ac losses.

Multifilamentary second-generation high temperature superconducting tape and lower ac losses

Quench stability of superconductors is critical for all coil-based applications. The stabilizer plays a vital role in quench stability and we are investigating the influence of material, thickness, design (such as filamentization) as well as its effect on ac losses. 2G HTS wires have been shown to exhibit excellent mechanical properties such as high critical tensile stress as well as tensile, compressive and bend strains that meet requirements of most applications. There are other properties such as stiffness to withstand buckling, torsion, vibration to name a few that need to be well quantified and improved if needed. Further, mechanical properties after modification to the wire architecture such as filamentization as well as those of joints and splices are being studied.

Persistent joints are required in applications such as NMR and MRI where an excellent uniformity in magnetic field has to be achieved. Persistent joints are nearly superconducting with a very low resistance and hence require a superconductor-to-superconductor joint. Creation and testing of such joints is an area of research that is being conducted.

Fault Current Limiters (FCL) is an application area that is being pursued by several institutions for limiting surges in electric power transmission to protect downstream electrical equipment. Resistive FCLs are based on superconductor property of a sharp increase in resistance as the superconductor transitions to a normal state from a superconducting state. There are several opportunities exist for materials research in this area such as modifying material architecture to enhance heat dissipation in transient regimes of a millisecond scale, increase voltage drop across the FCL element, and recover under load after the fault passes.

References :