Introduction

YBa2Cu3O7−δ (YBCO) based coated conductors have large potential in such diverse applications as wires/cables, motors/generators, high-field coils and superconducting permanent magnets, each of them with a certain range of temperature and magnetic field and a certain need in magnitude and isotropicity in critical current density Jc . In order to use the full potential of YBCO, it is mandatory to tailor its transport properties for the envisaged application. It is therefore necessary to understand how the critical current density Jc behaves over wide ranges of magnetic field strength H, temperature T and angle θ between H and the crystallographic c-axis. The Jc(H,T,θ) dependence, being determined by the underlying pinning landscape in a complex way1, can be raised and adjusted by the creation of artificial defects within the superconductor which act as pinning centers. This can be achieved by e.g. irradiation2, substrate decoration3,4, rare-earth substitution5,6 and incorporation of secondary phases. The latter one is relatively inexpensive and easy and therefore commonly used. After initial studies on naturally growing nanoparticles in YBCO thin films, such as Y2O37 (recently investigated again more closely as artificial pinning centers8,9), first investigations on artificial nanoparticles were made on BaZrO310. Jc could be strongly improved through incorporating this or related barium perovskites BaMO3 (M transition metal) because they grow as globular or columnar structures which act as pinning centers. This has been reported for various techniques such as pulsed laser deposition (PLD)11,12,13,14,15 or metal-organic chemical vapor deposition (MOCVD)16. In chemical deposition methods (CSD, MOD), usually randomly oriented, more or less isotropic nanoparticles are formed17,18,19. Secondary phases can furthermore introduce concurrent defects in the YBCO matrix. Strong positive correlations between nanostrain and Jc for example suggest that nanostrain and strain-induced defects such as stacking faults20 or dislocations21 act as pinning centers. Nanoinclusions of the double perovskites Ba2YNbO6 (BYNO) and Ba2YTaO6 (BYTO) have been suggested22 as promising pinning centers due to several advantages compared to BZO. Nb or Ta ions are less likely to substitute for Y in YBCO, which should lead to a smaller Tc reduction22, (a minute Nb substitution on Cu was even reported to slightly increase Tc23). Furthermore, according to their melting points (ZrO2 2715 °C, Nb2O5 1512 °C and Ta2O5 1872 °C), Ta-O and Nb-O species presumably have a larger mobility on the substrate surface and within the film during growth than Zr-O, leading to well aligned BYNO nanocolumns24,25 even at high deposition rates26. BYTO27 and Ba2Y(Nb/Ru)O628 have been shown to form nanocolumns in PLD-grown films, whereas BYTO forms extended nanoparticles in CSD-grown films29, very similar to the single perosvkites. Mixing BYNO and BYTO effectively adjusts interfacial energies and strain as well as diffusion lengths.

As shown by Ercolano et al.30, double-doping of YBCO with BYNO and BYTO leads to a further improvement of the transport properties. Furthermore, these mixed-double perovskite Ba2Y(Nb/Ta)O6-doped YBCO films (BYNTO:YBCO) seem to show a more complex Jc anisotropy than single-doped samples. Related structures are the so-called 2411-phases, Cu-containing mixed double perovskites, which also were shown to improve Jc, however with less chemical stability and less tendency to form nanocolumns31,32,33. A deeper investigation of the material system BYNTO:YBCO is needed to provide a better understanding of its complex pinning landscape and the resulting effects on Jc (H,T,θ). This detailed study shows the influence of the microstructure of such BYNTO:YBCO thin films grown under various deposition conditions on Jc and N-value anisotropy as well as the field dependence of the pinning force density, Fp(H), at liquid nitrogen temperatures. We explain the appearance of symmetric shoulders in Jc (θ) and N(θ) by a matching effect which does not depend on the absolute magnitude of the applied magnetic field but rather on its c-component.

Results and Discussion

The growth of self-assembled nanocolumns is not only driven by interfacial energies and strain34,35,36 but also by diffusion of the respective atomic species, which can be controlled by temperature37,38,39,40, time (i.e. laser repetition rate)41,13 and diffusion lengths (i.e. distances between deposited material). Furthermore, secondary phases39 as well as growth direction (vicinality of the substrate)42 influence the growth kinetics of nanocolumns.

In this study, films with various laser repetition rates fDep between 1/2 Hz and 50 Hz were grown at fixed (optimized) temperature in order to investigate the influence of different microstructures inside the BYNTO:YBCO matrix on the transport properties.

Microstructure

The YBCO matrix shows epitaxial growth on the STO substrates for all prepared films. As an example, Fig. 1 shows the Θ-2Θ scan for the film grown with a laser repetition rate of 1 Hz. For YBCO and BYNTO, just (00) and (00 2) peaks are visible, respectively. Y2O3 could be found by the appearance of its (004) peak and Y2BaCuO5 by its (220) peak. BYNTO nanoparticles are aligned cube-on-cube with the YBCO for all samples. This was confirmed by selected area electron diffraction (SAED) as well as pole figure measurements of YBCO (102) and BYNTO (220), Supp. Fig. S1. The intensity of the BYNTO (00) peaks and therefore the amount of biaxially incorporated BYNTO is increasing with decreasing (Supp. Fig. S2). TEM cross section images of the films grown at 1 Hz (Fig. 2a,b and Supp. Fig. S4) and 5 Hz (Supp. Fig. S3) show three types of defects: BYNTO nanocolumns (horizontal), Y2O3 plates (vertical) and defects in the YBCO lattice like stacking faults (extra Y or CuO2 planes) or anti-phase boundaries (APB). The density and morphology of the nanoparticles depend strongly on the laser repetition rate. With increasing the density of the nanorods increases (1 Hz: 1 column per 29.9 nm, 5 Hz: 1 column per 14.5 nm), their diameter decreases (1 Hz: (10 ± 2) nm, 5 Hz: (8 ± 4) nm) and the density of Y2O3 plates seems to increase. Fig. 2c shows a large-area TEM cross section of 3200 nm width of the 1 Hz sample. Some of the nanorods seem to start or end within the film (red dots). However, if all columns are counted (black arrows plus red dots) an average column distance of d = 29.9 nm is calculated. This corresponds nicely to the value d = 30.5 nm obtained by plain-view TEM (Supp. Fig. S6). Therefore, we conclude that for up to 5 Hz, most of the nanorods are penetrating the whole film and some are slightly inclined and cut by the lamella preparation. The components, BYNTO nanorods and Y2O3 particle, were further confirmed by EDX mapping of the contained elements (Supp. Fig. S5).

Figure 1
figure 1

Θ-2Θ scan of the BYNTO:YBCO sample grown with fDep = 1 Hz.

Co Kα was used for the measurement.

Full size image
Figure 2
figure 2

(a,b) HRTEM image of the 1 Hz sample at substrate-film interface (a) and near the film surface (b). Pinning centers such as BYNTO columns, Y2O3 particles and planar intergrowths are marked. (c) BFTEM cross section of the 1 Hz sample over an area of 3200 nm width. BYNTO columns are marked with arrows. Columns which are tilted (approx. 12% of them) are marked by red dots.

Full size image

Critical current density and Pinning force

Films grown at fDep of 1 Hz or 1/2 Hz show a transition temperature Tc of ∼90 K, which is in the experimental range of undoped samples (90…92 K). In general, Tc decreases for higher repetition rates (Fig. 3), while the transition width ΔTc (error bars in inset of Fig. 3) is increasing by a factor of 7 (ΔTc (1 Hz) = 0.5 K, ΔTc (50 Hz) = 3.3 K). The reduction in Tc might be avoidable by adjusting other PLD parameters such as deposition temperature, energy density, oxygen partial pressure and target-substrate distance. Whereas the samples with 1 Hz and 1/2 Hz are similar in their current carrying capability, the samples grown at 5 Hz or higher show much smaller Jc values at 77 K (Fig. 4a and Table 1). This reduction is mainly due to the reduced Tc value because the decrease in Jc is smaller at lower temperatures such as 30 K, Fig. 4b.

Table 1 Overview of the superconductivity properties of the discussed films.
Full size table
Figure 3
figure 3

Relative resistance over applied temperature for BYNTO:YBCO films grown at laser repetition rates between 0.5 Hz and 50 Hz.

Inset: Transition temperature versus laser repetition rate. Error bars indicate the transition width which was determined by taking the difference between Tc,90 and Tc,10.

Full size image
Figure 4
figure 4

Jc versus applied magnetic field, , at 77 K (a) and 30 K (b) for the BYNTO:YBCO samples grown at 1/2 Hz, 1 Hz and 5 Hz as well as the corresponding FP (μ0H) at 77 K (c) and 30 K (d). The thickness of these films is around 250 nm. Green datapoints belong to a pure YBCO film without any additions. In this case FP (μ0H) at 77 K can be described with equation 1 (bold green line in (c), p = 0.5, q = 2).

Full size image

The critical current density in self field Jc,sf at 77 K reaches values of 3.2 MA/cm2 (1 Hz) and 4.0 MA/cm2 (1/2 Hz) and decreases in an external field as it is expected for a high-temperature superconductors in the strong pinning limit43. Jc,sf values at 30 K as high as 28 MA/cm2 can be achieved in BYNTO:YBCO films grown with laser repetition rates of 1 Hz or below. The critical current density is above 1 MA/cm2 even for an external magnetic field of 9 T for all samples at this temperature (Fig. 4).The field dependence of Jc at 77 K shows an unusual behavior of samples with  = 1 Hz and 1/2 Hz (Fig. 4a and Supp. Fig. S7). Above approx. 500 mT, Jc stays almost constant for increasing fields.

The field dependence of the pinning force density is equally unusual. For these BYNTO:YBCO thin films, it is not possible to describe by one Dew-Hughes function44

alone as it is possible for a YBCO film without any additions (bold green line in Fig. 4c). The field  = μ0H(FP,max) at which the maximum pinning force density is reached increases with laser repetition rate (1.6 T for 1/2 Hz, 2.3 T for 1 Hz and 3.2 T for 5 Hz). Similar Jc(H) curves have recently been reported for BaHfO3(BHO)-doped SmBa2Cu3O7−δ45. The pinning force density maximum for the 1 Hz sample (25 GN/m3 for the 250 nm thick film at 77 K), Fig. 4c, is among the highest values in YBCO at 77 K measured so far (e.g. 240 nm BaSnO3-doped: 28.3 GN/m3 12, 500 nm BYNO-doped: 32.3 GN/m3 at 75.5 K25, 200 nm BaZrO3-doped: 21 GN/m3 17). Other REBCO films with slightly higher Tc values doped with BHO (370 nm GdBCO: 23.5 GN/m3 46, 300 nm SmBCO: 28 GN/m3 45) show similar FP,max values at 77 K. The previously mentioned plateaus in Jc(H) at 77 K also end at these fields . Furthermore, is temperature independent as can be seen by the dashed lines in Fig. 4c,d for the pinning force density at 77 K and 30 K (see also Supp. Fig. S8). Therefore, this plateau can be explained by a matching effect. Up to this field, every flux line can be pinned by an individual BYNTO column. The mean column distance of d = 29.9 nm, as observed in TEM for the 1 Hz sample, corresponds to a matching field of approx. 2.3 T which corresponds quite well to the end position of the constant Jc region, respective peak position in (H), dashed black line Fig. 4c,d (2.3 T). This value was also confirmed by a TEM plain view image where 312 columns are visible in a 0.289 μm2 wide area which leads to a matching field of (2.23 ± 0.14) T (Supp. Fig. S6). The same correspondence between end of Jc plateau, and has been observed for BHO-doped SmBa2Cu3O7−δ/LAO45 and GdBa2Cu3O7−δ/IBAD-MgO46.

The N value

The electric field-current density characteristics E(J) are well described by a power-law dependence, E ~ JN, over a wide electric-field range in the vicinity of Jc. The N value has, in general, statistical and microscopic explanations. In inhomogeneous superconductors, 1/N is proportional to the variance in Jc as described by e.g. Warnes and Larbalestier47 in statistical models. The good reproducibility of N and Jc in our measurements however points to a microscopic explanation for N. It has been shown by Griessen et al. that in high-Tc superconductors N is strongly influenced if not determined by flux creep processes. Here, the flux creep rate (M … magnetization)48. Strictly, this relation only holds if S and N are measured at the same electrical field E. In general, N values determined from creep rates (equation above) are larger than from magnetization measurements49 and those are larger than determined from transport E(J) curves50 due to the negative curvature of E(J) below the glass-liquid transition. For the considerations below, this effect can be ignored, however.

For flux creep processes determined by the pinning potential and under the assumption of a logarithmic J-dependence of U051, and, hence, for N ≫ 1 (i.e. neglecting a log. time dependence term). This in turn means that Jc, being determined itself by U0, scales with N. This is indeed found on a variety of samples in a wide range of N values with an empirical relation (N–1) ~ Jc1/p, p ~ 3, Fig. 5b. The origin of this power law is beyond the scope of this paper. As seen in Fig. 5, N is anticorrelated to Jc in a wide magnetic field region below the matching field. The explanation for this peculiarity is a second mechanism for flux creep effects. In the presence of strongly correlated 1D or 2D pinning centers and for μ0H < Bm, the flux creep rate is determined by the ability of flux half loops and double kink structures to form and move between the pinning centers, which consequently leads to an effective “hopping” of the flux lines from nanocolumn to nanocolumn. The net velocity of the flux lines and so the creep rate depends on the probability of finding a free nanocolumn in the neighborhood. Since this probability decreases with increasing magnetic field μ0H < Bm, S decreases and N increases for μ0H . This effect is the equivalent of a Mott insulator. As observed in magnetization measurements on irradiated YBCO single crystals with columnar defects52, matching effects are only visible for a very narrow orientation distribution (splay) of the correlated defects, in that case of less than 4°. This, in turn, leads to an overall decreased Jc due to the increased creep rates. Larger splay values lead to lower creep rates and therefore higher Jc values but no observable matching effects. Similar reductions in creep rate due to additional defects in YBCO films have been reported recently for increased film thickness and substrate decoration53 and nanoparticle addition (Y211, BZO, BYTO) in films grown by metall-organic deposition (MOD)54,55. In our samples, we observe high Jc values and matching effects. This is a consequence of the intermediate Y2O3 particles and stacking faults, which hinder the half loops and double kink structures from running parallel the nanocolumns and effectively jump to the next column13.

Figure 5
figure 5

(a) N-value over field at 77 K for different samples. (b) Jc over N illustrating the reduced N values below . (c) Derivation of the N value for the 5 Hz sample. (5 Hz) is visible as step in the derivative.

Full size image

Critical current density anisotropy

The anisotropy of the critical current density Jc(θ) at 77 K shows a maximum at 90° due to the electronic anisotropy of YBCO56 and one at 180° due to c-axis correlated pinning57 for all BYNTO:YBCO samples. Besides these well understood features of Jc(θ) additional maxima are visible. They appear as pairs which are oriented symmetrically around the c-axis peak similar to the shoulder formation seen by Ercolano et al.30 but more pronounced. The shoulder position with respect to the c-axis depends on the used laser repetition rate as well as the applied field, Fig. 6. Only the samples grown at a laser repetition rate of 1 Hz and 1/2 Hz will be discussed in the following. Because of the lower Jc values, the shoulders of the 5 Hz sample are barely visible. The appearance and the position of these shoulders becomes comprehensible if N value and Jc are plotted versus the c-axis component of the applied magnetic field, , Fig. 7a,b. For both samples, the N value maxima (and hence minima in S) appear once reaches the matching field just the same as for Jc at . A similar effect has recently been observed by Trastoy et al.58 on YBCO films with a periodic square pattern of artificial pin holes produced by ion irradiation. They have seen strong matching effects with μ0H  in field and angular dependence of resistivity and glass-liquid transition, concurrent with an increased mass anisotropy due to partial deoxygenation and a 2D glass-liquid transition. Our results show that this type of matching is independent of dimensionality (our samples are 3D) and periodicity (the BYNTO nanocolumns show a certain degree of density variance).

Figure 6
figure 6

Anisotropy of the critical current density at 77 K at several fields for samples grown at low fDep, red 0.5 Hz and black 1 Hz.

Large arrows indicate the position of the main off-axis maximum, where equals (2.0 T for the red arrows and 2.7 T for the black arrows), small arrows indicate a second off-axis maximum.

Full size image
Figure 7
figure 7

N value respectively Jc versus the c-axis component of the applied magnetic field Bc at 77 K.

Open circles show the according data for . Shaded region indicates anti-correlation between N and Jc.

Full size image

As calculated by Paulius et al.59 for correlated irradiation defects and described by e.g. Ercolano et al.30 and Jha et al.60 for nanocolumns, the vortex is not pinned by one column alone but rather by several columns through a staircase-like path for angles θ smaller than a certain trapping or accommodation angle56. The flux line segments in between the vortices are pinned by additional defects, such as Y2O3, stacking faults and antiphase boundaries, as observed in TEM. The specific position , where the curves have their maximum in Jc() is slightly raised compared to the matching field and the maxima in the N value (2.0 T for  = 1/2 Hz and 2.7 T for  = 1 Hz, Fig. 7c,d). Apparently, the extrema in creep rate S and hence the N value are determined solely by the occupation of the c-axis aligned correlated defects. Therefore, they appear at the matching field . Additional intermediate defects contribute to Jc for field directions closer to c and lead to a further increase in Jc in a small region of . This region (shaded area in Fig. 7) shows again anti-correlation between N and Jc (i.e. creep rate S not determined by pinning potential U0). However, this configuration is different from the case , since here the anti-correlation appears above , whereas it appears below for .

Besides the strongly pronounced shoulders, a second pair of shoulders can be seen for fields up to 8 T (Fig. 6, small arrows). They have a much smaller amplitude and their position is much closer to 90°. The position of those peaks does not scale with the c-axis component of the applied magnetic field, . A possible scenario for those peaks might be a combined effect of fractional occupation of the correlated defects and anisotropic pinning of the flux line segments at intermediate extended defects. This has to be investigated in more detail in future studies.

Conclusion

BYNTO-doped YBCO films were successfully grown using different laser repetition rates. Only for sufficiently long diffusion times ( = 1 Hz or lower), very high critical current densities were achieved. All films have a rich microstructure consisting of pinning centers with different size, shape and orientation distribution. The interplay between strongly c-axis correlated BYNTO nanocolumns, ab-stacking faults, c-axis oriented APBs, atomic disorder and biaxially oriented but randomly distributed Y2O3 nanoparticles leads to additional new features in FP(H) and Jc(θ). The critical current density shows almost constant values for magnetic fields up to several Tesla. The field where this plateau ends and where FP(H) has its highest value are equal to the matching field determined by TEM. A pinning force density of 25 GN/m3 at 77 K at the matching field of 2.3 T was achieved for the film grown at 1 Hz, which is among the highest values reported for YBCO. At 30 K, deposition rates up to  = 5 Hz lead to critical current densities of more than 1 MA/cm2 even in external fields of up to 9 T for .

The field dependence of the N value (the exponent of the E(J) curves) and its scaling behavior with the corresponding Jc value clearly show different regions of flux depinning mechanisms, being dominated by the creation of half loop and double kink structures below the matching field and by usual flux creep processes due to weaker, uncorrelated, pinning centers above the matching field.

The anisotropy of the critical current density Jc(θ) as well as N(θ) are composed of two additional shoulder formations besides the maxima at 0° and 90° which appear at different angles if the applied magnetic field strength is changed. The existence of these shoulders can be explained by a staircase-like path of the vortices and additional pinning at intermediate defects. At a certain angle, the magnetic fields c-axis component μ0H ·cos θ reaches the matching field Bm which is accompanied by a maximum of the N value. Because of further contribution of intermediate defects to Jc but not to N, the maximum at the critical current density anisotropy is slightly higher than the matching field Bm.

Due to their ability to grow as very uniformly sized, distributed and oriented nanocolumns, Ba2Y(Nb/Ta)O6 nano-precipitates in YBa2Cu3O7−δ thin films are not only effective artificial pinning centers for increased Jc and FP values but also a model system to study matching effects and flux creep.

Methods

BYNTO:YBCO films were prepared by pulsed laser deposition (PLD) using a YBCO target with 2.5 mol% BYNO and 2.5 mol% BYTO. The target was prepared by mixing the precursor oxides (barium oxide, yttrium oxide, tantalum oxide and niobium oxide) in the appropriate amount with YBCO powder and grinding the mixture in an agate mortar. The powder was pressed in pellets and sintered at 950 °C in flowing O2 for 24 h. A Lambda Physics LPX305i KrF excimer laser (λ = 248 nm,  = 25 ns) was used with an energy density of 1.6 J/cm2 at the target surface to grow BYNTO:YBCO films of approx. 250 nm thickness on single crystalline (100)-oriented SrTiO3 (STO). An oxygen partial pressure of 0.4 mbar was maintained during the deposition process. The substrate temperature was set to 840 °C and checked with a HEITRONICS ceramic pyrometer. The laser repetition rate fDep was varied between 0.5 Hz and 50 Hz. Oxygen loading of the YBCO films took place in 400 mbar O2 for 1 hour at 770 °C. A silver cap layer of several nm thickness was deposited afterwards to improve the contact resistance.

Surface morphology and film thickness were analyzed by scanning electron microscopy using a JEOL JSM-6510 and an FEI Helios Nanolab 600i to cut cross sections by a focused gallium ion beam. Transmission electron microscopy (TEM) analysis was carried out on an FEI Osiris microscope, equipped with a ‘Super-X’ wide solid angle EDX detector, operated at 200 kV for bright field transmission electron microscopy (BFTEM), high angle annual dark field scanning transmission electron microscopy (HAADF STEM) and energy dispersive X-ray spectroscopy (EDX). FEI Titan ‘cubed’ electron microscope operated at 300 kV, equipped with an aberration corrector for the probe-forming lens as well as a high-brightness gun was used for high-resolution transmission electron microscopy (HRTEM). X-ray diffraction was carried out on a Bruker D8 Advance with a Co anode in a modified parallel-beam geometry for Θ-2Θ scans. Pole figures were measured at a Philips X’Pert PW3040 with a Cu anode in Bragg-Brentano geometry.

Bridges of 800 nm length and approx. 20 μm width for transport current measurements were prepared by laser cutting. Field and angular dependencies of critical current density Jc and exponent N of the V(I) curves were measured in a four-point assembly in magnetic fields up to 9 T with a Quantum Design physical properties measurement system (PPMS). Jc was defined by an electrical field criterion of 1 μV/cm on V(I) curves fitted as in the first decade of E above Ec. The angular dependence of the critical current density was measured under maximum Lorentz force configuration. To correct for small heating effects, the absolute value of some of those measurements were corrected to the value measured at where thermal contact was better.

Additional Information

How to cite this article: Opherden, L. et al. Large pinning forces and matching effects in YBa2Cu3O7-δ thin films with Ba2Y(Nb/Ta)O6 nano-precipitates. Sci. Rep. 6, 21188; doi: 10.1038/srep21188 (2016).