Nanorod Self-Assembly in High Jc YBa2Cu3O7−x Films with Ru-Based Double Perovskites

Many second phase additions to YBa2Cu3O7−x (YBCO) films, in particular those that self-assemble into aligned nanorod and nanoparticle structures, enhance performance in self and applied fields. Of particular interest for additions are Ba-containing perovskites that are compatible with YBCO. In this report, we discuss the addition of Ba2YRuO6 to bulk and thick-film YBCO. Sub-micron, randomly oriented particles of this phase were found to form around grain boundaries and within YBCO grains in bulk sintered pellets. Within the limits of EDS, no Ru substitution into the YBCO was observed. Thick YBCO films were grown by pulsed laser deposition from a target consisting of YBa2Cu3Oy with 5 and 2.5 mole percent additions of Ba2YRuO6 and Y2O3, respectively. Films with enhanced in-field performance contained aligned, self-assembled Ba2YRuO6 nanorods and strained Y2O3 nanoparticle layers. A 0.9 µm thick film was found to have a self-field critical current density (Jc) of 5.1 MA/cm2 with minimum Jc(Θ, H=1T) of 0.75 MA/cm2. Conversely, Jc characteristics were similar to YBCO films without additions when these secondary phases formed as large, disordered phases within the film. A 2.3 µm thick film with such a distribution of secondary phases was found to have reduced self-field Jc values of 3.4 MA/cm2 at 75.5 K and Jc(min, Θ, 1T) of 0.4 MA/cm2.

to Y-123 films [28][29][30]. Recently, extremely high pinning forces (F p ) were obtained in a 0.5 μm thick film with J c (75.6 K, 1T, H||c) > 2 MA/cm 2 and F p (H||c) values in excess of 30 and 120 GN/m 3 at 75.5 K and 65 K, respectively [28][29][30]. Such results have spurred further efforts in exploring other possible perovskite-based additions. In this work, we report our results on the incorporation of a Ru-based double-perovskite phase into bulk YBCO and the deposition of YBCO films with Ba 2 YRuO y nanorod structures and enhanced J c performance in self and applied magnetic fields.

Results and Discussion
It is well known that a large number of perovskites can be formed with Ba [31][32][33] However, their incorporation into YBCO bulk and film forms must occur with minimal detrimental effects to the superconducting phase. Ruthenium was investigated as a potential material for addition to YBCO based on its known formation into the ordered, tilted perovskite A 2 BB'O 6 of composition Ba 2 YRuO 6 [32].
Its compatibility with YBCO was tested with mixtures of bulk YBCO and RuO 2 additions and its potential for performance improvements though its incorporation as nanorod secondary phases in YBCO films. It should be noted that Ru is an expensive material. Based on 5 mol % additions of Ba 2 YRuO 6 to YBCO in the fabrication of the targets used for PLD, costs for the HTS materials would increase by 13 to 22%, depending on the starting purity of the RuO 2 .

Ba 2 YRuO 6 Double-Perovskite
Ba 2 YRuO 6 was synthesized as a single phase material using conventional milling and sintering processes. Shown in Figure 1 is the XRD scan of the single-phase material and its ordered crystal structure. Table 1 contains the XRD data pertaining to observed and calculated peak positions and intensities. A single-phase diffraction pattern characteristic of a perovskite was obtained. The refined lattice parameter based on the integrated Bragg reflections obtained from diffraction experiments using Si (NIST SRM 640a) mixed in with the sample was a 0 = 8.3381(4). Observed intensities and peak positions are consistent with previous reported work on this phase which includes ordering of the Y and Ru atoms on the B sites of the perovskite structure [32].

Bulk YBCO with RuO 2 Additions
Shown in Figure 2 are SEM and STEM images of material taken from a pressed pellet containing 9 wt % RuO 2 additions to stoichiometric YBCO powder. The resulting material was multiphase containing primarily YBCO and a sub-micron sized secondary phase containing Ru along with small amounts of BaCu 2 O y and CuO, STEM/EDS analysis suggested an approximate ratio of 2:1:1 of Ba:Y:Ru with a small amount of Cu present in the material. However, the copper content in this phase was not consistent with the reported formation of the 2411 phase in other works [34]. Within the limits of EDS, no Ru was detected within the YBCO phase. Figure 3 contains the XRD analysis of the pellet which showed a multiphase material with the major peaks identified as either YBCO or the double-perovskite Ba 2 YRuO 6 . The intensities of the peaks associated with Ba 2 YRuO 6 are consistent with the observed amounts of the double-perovskite phase present in the sample.

YBCO with Self-Assembled Nanorods of Ba 2 YRuO y
Films containing the Ba 2 YRuO y (BRYO) secondary phase were grown from PLD targets consisting of YBa 2 Cu 3 O y with 2.5 mol % additions of Y 2 O 3 and 5 mol % additions of Ba 2 YRuO 6 . The addition of the extra Y 2 O 3 was included for the purposes of producing the two-phase pinning mesostructure of strained Y 2 O 3 nanoparticle layers and perovskite-based nanorods [19]. High self-field J c values of 5.0, 5.1, 4.5, and 3.4 MA/cm 2 were obtained in measurements at 75.5 K and self-field for films of thickness 0.85, 0.9, 0.95 and 2.3 μm, respectively.  [35]. While there was clearly an J c (H) improvement for the 0.95 micron thick film, there was essentially none above what would be expected for plain YBCO for the 2.3 micron thick film. Hence, a close comparison of the microstructural properties with J c performance is necessary for understanding these differences.   Figure 6 show the Ru to be associated with the BYRO nanorod structures that are continuous though the full thickness of the film. EDS analysis of the BYRO nanorods showed a small copper content. However, the amount detected was not consistent with the reported 2411 phase [36]. The continuous nature of these nanorod structures is confirmed with the Moire fringe analysis shown in Figure 7. Growth of the nanorods starts at the CeO 2 buffer layer. Indexing of the diffraction pattern the double perovskite in Figure 5(a) revealed the nanorod phases to have an alignment defined by [1][2][3][4][5][6][7][8][9][10] BRYO || [1][2][3][4][5][6][7][8][9][10] YBCO with (220) BRYO || (110) YBCO, identical to the one found for Ba 2 YNbO 6 nanorods [28]. The lobe-shaped contrast around the Y 2 O 3 particles in the TEM image (arrows Figure 5(b)) indicates that the alignment of the Y 2 O 3 particles with the YBCO is highly strained. Hence, these additions were nearly successful in the reproduction of the idealized, two phase, strained nanoparticle layer and nanorod pinning mesostructure as first reported by the Los Alamos group [14,19]. The only difference between the ideal structure and the mesostructure observed here being the continuous nature of the BYRO nanorods versus the short, segmented structure of the Zr-based nanorods.   The improvements in observed performance, relative to plain YBCO are then understood quite well. In this particular TEM FIB specimen, the distribution of the secondary phases was very uniform as indicated by the low magnification images of Figure 5(c,d). Figure 5(d), the dark-field, weak-beam TEM image shows this distribution quite well. Weak-beam diffraction contrast is well-suited for imaging the nano-sized phases and structures at relatively low magnifications as the strain associated with the phases allows the phases to be highlighted in the images. The arrows in Figure 5(c) indicate localized reactions between the substrate and CeO 2 buffer layer to form BaCeO 3 [37]. This latter phase is not aligned and appears dark against the contrast from the YBCO film in the weak-beam image of Figure 5(d). The low-magnification TEM images also show the surface porosity "P" that is commonly found in these PLD films when the surface morphology is examined in the SEM. While the porosity depth is small for these nominally 1 micron thick samples, it becomes more pronounced for the thicker films as shown below.
While an enhancement in the in-field performance of the films of thickness 0.85 to 0.95 µm in thickness was observed, the absolute level of performance was still not to the level of the best, optimized YBCO films with BaZrO 3 and Y 2 O 3 additions. Hence, further microstructural investigations were undertaken using plan view (S)TEM. Figure 8(a) shows a STEM plan view image where the observed material is located near the top surface of the 0.9 µm thick film described above. As indicated by the circles, areas could be identified that were devoid of any secondary phases. Further investigations of this film with additional cross-sectional specimens showed that indeed, these regions existed within the films. These clear areas were found to occur above regions where interruptions occurred in the development of the nanorod/strained nanoparticle mesostructure. As shown in the example of Figure 8(b), growth of the nanorod/nanoparticle mesostructure did not occur across the full thickness of the film. Instead, there was a change to a structure that ended with large Y 2 O 3 / Ba 2 YRuO 6 composite particles capping the growth of the nanorods. Above this structure, a region free of secondary phases could be found. This degradation of the mesostructure was present in moderate amounts in the submicron films and would explain why the in-field performance of these first films were not to the levels found in related films that have been optimized for additions of BaZrO 3 , for example [19]. The mechanism for this mesostructure degradation will be the focus of future work. The identification of the mesostructure degradation in regions of the nominally 1 µm thick films was important as it leads to the basis of understanding performance degradation in the 2.3 µm thick film. While the self-field J c value of 3.4 MA/cm 2 obtained in the latter film was very respectable, its J c performance in field was no different from that measured in plain YBCO films without c-axis aligned defects. Now it has been commonly observed in many works that J c decreases as a function of film thickness (See reference [9] and works cited therein). It is tempting to assume that this is an intrinsic property of these thick films. However, in recent works where detailed microstructural analysis was coupled to film performance in high quality PLD or ex situ metal orgranic deposition (MOD) films, it was shown that J c becomes independent of thickness when the microstural features are uniform though thickness [10,19,38]. The 2 micron, 5 MA/cm 2 (75K, SF) film reported by Feldmann et al. is a clear illustration of the effects of uniformity though thickness and its pronounced influence on film properties. It also strongly supports the notion that the J c dependence for film thickness is an extrinsic property.
The STEM images of Figure 9 show significant degradation of the film structure, especially for those sections of the film more than 0.5 microns from the CeO 2 buffer. The z-contrast image, Figure 9a, shows a distribution of large secondary phases that appear to be composites of multiple phases based on their z-contrast. The spectral imaging data of Figure 10 revealed these composite phases to be composed of Y 2 O 3 and Ba 2 YRuO 6 . Hence, we attribute the degradation of the in-field properties to the consolidation of the nanorod/strained nanoparticle mesostructure into large composite secondary phases. The remaining BYRO nanorods that are still present in this sample are much larger and irregularly shaped compared to those found in the nominally 1 µm thick films. Understanding the underlying causes to the changes in this mesostructure is important for incorporating these structures into continuous processing of YBCO films for long-length wires [39,40]. Related work by one of the authors suggests that these same degradation mechanisms can also occur in films with BaZrO 3 or Ba 2 YNbO 6 additions [41]. The difference in the mesostructure would also explain the early results with Ba 2 YNbO 6 additions in which the phases were present, but no appreciable increase in performance was observed [24]. Figure 9. (a) STEM Z-contrast image showing a distribution of large Y 2 O 3 particles, many of which appear to have a composite structure containing a second, higher Z material (arrows); (b) Bright field STEM image showing the splayed, ill-formed, and non-continuous network of BRYO nanorods (white dots) within this thick film. Surface pores, which extend into the sample for appreciable distances in this thick film, are marked with a "P".

Experimental Section
Ba 2 YRuO 6 was prepared by mixing individual elemental oxide precursors and milling for 8 h in a Retsch mortar grinder with distilled H 2 O, Isopropanol, and KD-1 dispersant. After milling, the slurry was dried on a Schlink line. The dried powder was hand-ground with a mortar and pestle and pressed at 5 K psi into 1.25 cm diameter pellets. The pellets were heat treated in flowing O 2 in a hightemperature box furnace. The pellets were ramped to 1000 °C at 10 °C/h and held for 15 h. They were then ramped at 1 °C/h to 1550 °C and held for 15 h. The pellets were then cooled to 1000 °C at 2 °C/h for a final dwell of 2 h, and then ramped to room temperature at 15 °C/h. Bulk YBa 2 Cu 3 O 7 (YBCO) pellets with additions of 9 wt % RuO 2 were prepared by milling stoichiometric mixtures of prepared YBCO with RuO 2 . The powders were pressed into 1.25 cm diameter pellets which were then sintered at 1000 °C for 50 h, followed by an annealing step at 400 °C for 50 h in flowing oxygen. Powder Xray diffraction (XRD) patterns were obtained on a Scintag X-1 Advanced Diffraction System using filtered Cu Ka radiation (λ = 1.5406 Å). XRD patterns were taken with and without a Si reference material (NIST SRM 640a); scans with Si were used for quantitative structure refinements. Scans were collected at room temperature in Θ-Θ mode (2Θ = 90°) with a step size of 0.02 and collection time of 10 sec/step. The XRD data collected was analyzed though the Jade 6.5 software package which as used for phase identification and determination of refined unit cell parameters [42]. Crystal structures and calculated diffraction patterns were obtained using CrystalMaker ® [43]. Targets for pulsed laser deposition (PLD) were made by mixing precursor powders in the desired ratios, followed by milling in isopropanol in a Retsch motar grinder for 8 h. Powders were dried on a shlenk line and then calcined as a loose powder for 15 h at 900 °C. This powder was then lightly ground in a glove box, pressed into a 5.5 cm diameter target 0.5 cm thick, and then sintered at 925 °C and 950 °C for 15 h each with a furnace cool to 400 °C for 5 h in between steps. The sintering process ended with a low temperature anneal at 400 °C for 50 h.
Films of thickness 0.8 to 2.3 µm were grown by PLD using a KrF laser (248 nm). The target material was YBa 2 Cu 3 O 7 (YBCO) with 5 mol % additions of Ba 2 YRuO 6 and 2.5 mol % additions of Y 2 O 3 . Substrates were single crystals of either SrTiO 3 or Y-stabilized Zr 2 O 3 with a 40 nm thick CeO 2 buffer layer. The deposition temperature during growth was either 775 °C or 795 °C. Films were patterned for transport measurement using standard photolithography and wet etching techniques. Before photolithography, Ag pads were deposited by thermal evaporation and the films subsequently annealed at 500 °C for 30 minutes in flowing O 2 . Values of critical temperature (T c ) were determined by an inductive method and values of J c were determined using the standard 1 µV/cm electric field criterion and a four-point transport measurement. All bridges were 1.5 mm long and approximately 200 µm wide. Measurements of J c as a function of field orientation were performed in the maximum Lorentz force configuration, with Θ being the angle between the film normal and field vector. J c measurements in applied magnetic fields were performed in a 1 T split coil electromagnet or in a 7 T split coil superconducting magnet. (Scanning) transmission electron microscopy ((S)TEM) characterizations were performed on foils thinned by either conventional dimple-polishing and ion-milling or foils prepared with a focused ion beam (FIB) technique.

Conclusions
We have shown that the double-perovskite Ba 2 YRuO 6 can be an effective and compatible addition for both bulk and thick film YBCO materials. In the former case, this secondary phase distributed itself within the YBCO material as small, sub-micron randomly-oriented secondary phases. Additions of 5 mol % Ba 2 YRuO 6 and 2.5 mol % Y 2 O 3 to YBCO thick films resulted in mesostructures within the film of aligned, continuous BYRO nanorod and strained Y 2 O 3 nanoparticle structures. The departure from the idealized mesostructure being the presence of the continuous, as opposed to the optimal short segmented, nanorods. Self-field J c values up to 5.2 MA/cm 2 were obtained in 0.95 µm thick films with improved in-field performance as indicated by the measured minimum J c (Θ, H=1T) of 0.75 MA/cm 2 We also showed how the degradation to the mesostructure can lead to performance that is not very different from YBCO films without additions. High self-field J c values are still possible as indicated by the measurement of 3.4 MA/cm 2 (self-field, 75.5 K) in a 2.3 µm thick film with the BYRO and Y 2 O 3 additions. However, the degradation of the mesostructure within this thick film results in a measured minimum J c (Θ, H=1T) of only 0.4 MA/cm 2 . The observed mesostructure degradation is important in understanding the variable range of performance one can obtain when investigating pinning additions to YBCO films. It is believed that continued optimization of the deposition process and chemistry for Ba 2 YRuO 6 and Y 2 O 3 additions would lead to performance levels comparable to the best achieved with BaZrO 3 and Y 2 O 3 additions.