Influence of artificial pinning centers on structural and superconducting properties of thick YBCO films on ABAD-YSZ templates

Two series of YBCO films with increasing thickness were grown on ABAD-YSZ templates using mixed targets with either 5 mol% (4.7 vol%) BYNTO or 6 mol% (2.6 vol%) BHO addition, respectively. Figure 1(a) shows XRD scans of selected samples of the BYNTO series. The YBCO (00ℓ) peaks are clearly visible for all films indicating a preferred c-axis-oriented growth of the superconductor. Additionally, the (004) peak of BYNTO was found at a 2θ angle of about 50.3° pointing to a biaxially oriented incorporation of these nanoparticles as already shown in previous studies [16, 23]. The broad peak indicates a small grain size, and the asymmetric form might be the result of either stoichiometric variations or different strain states due to the incorporation as plates or nanorods (see TEM results). Several peaks originating from the layers of the ABAD-YSZ template are still visible for the thinner films. They are no longer apparent for the thickest YBCO layer due to the absorption of the x-ray radiation in the superconductor. Instead, additional peaks appear for films with a thickness of 4 μm and above, which can be assigned to an a-axis-oriented texture component of YBCO as well as to other misorientations (mainly represented by the (103) peak at about 2θ = 38°). A similar behavior was found for thick BHO-doped YBCO films (not shown here), where misorientations are also present above a thickness of 4 μm. Nevertheless, these results indicate that a higher thickness of YBCO with preferential c-axis alignment is achieved by the incorporation of nanoscaled secondary phases in comparison to undoped films.

Zoom In Zoom Out Reset image size

Pole figure measurements confirmed the epitaxial growth of the YBCO layer on the metal-based template as shown for two samples in figure 1(b). The in-plane FWHM values are about 8° for both BYNTO-doped and BHO-doped films. These values are slightly higher compared to undoped films (showing a typical FWHM <6° [5]) and remain almost unchanged up to a thickness of about 4 μm. However, the in-plane distribution gets significantly broader with further thickness increase and reaches a value of about 13° for the 7 μm BYNTO-doped film (see upper right pole figure in figure 1(b)). The larger in-plane spread might be correlated to the appearance of additional texture components observed by XRD. The in-plane FWHM values for the BYNTO (220) planes show the same trend (pole figures not shown). Furthermore, the YBCO (102) pole figure gives a clear indication for the a-axis-oriented texture component by the appearance of the additional spot at ψ ∼ 33°.

Figure 2 compares the surface morphology and cross-section of thick YBCO films for selected samples. The undoped film with a thickness of about 4.2 μm shows a high number of pinholes at the surface, which decouple the individual YBCO grains. Additionally, a high density of misoriented grains is visible at the film surface. A BHO-doped film of similar thickness shows a smaller number of misoriented grains and a significant reduction of pores. The denser microstructure is also verified in the cross-sectional view (inset of figure 2(b)). This is a striking contrast to the undoped film having the same thickness, where deep pores, large Y₂O₃ precipitates (white spots) and misoriented regions are visible in the upper part of the film (compare inset of figure 2(a)). One reason for the denser structure might be a refinement of the Y₂O₃ nanoparticles by the secondary phase additions, which results in a lower probability for the nucleation of misoriented grains. BYNTO-doped films with a thickness of more than 4 μm also show a dense YBCO matrix, but larger misoriented clusters on the surface are clearly visible in figure 2(c). A FIB cut of such a region, as shown in the inset of this figure, reveals that some grains are aligned perpendicular to the substrate surface as expected for a preferentially grown a-axis component. Other grains are tilted and might be connected to the (103)-textured YBCO component visible in the θ-2θ scans. So far it is not clear what causes the nucleation of these large clusters in thick BYNTO-doped films.

Zoom In Zoom Out Reset image size

Figures 3 and 4 give exemplary TEM images for thick BHO-doped YBCO layers and BYNTO-doped YBCO layers, respectively. The doping with BHO results in extended nanorods with an alignment almost parallel to c_YBCO having an orientation spread to the substrate normal. The main component of the ensemble in figure 3(b) has an average tilt of 11° with a splay of ±3°, where occasional nanocolumns are within 4° of the c-axis direction. As shown before, the BHO nanocolumns are usually not perfectly straight but show a certain direction distribution along the length depending on the deposition conditions [16]. Additionally, there is a significant number of in-plane defects aligned in the (a–b)_YBCO direction as small plates or stacking faults (figure 3(b)). The in-plane TEM view (figure 3(c)) indicates a diameter of about 4.6 ± 0.8 nm and a high areal density of 3040 μm⁻² of these nanocolums, which corresponds to a matching field of 6.0 T. In some areas, a slightly lower nanocolumn density is observed, however also with matching fields around 5 T. The nanocolumns show a large splay of up to 20° ('fan-like structure') with occassional slight tilt of the main direction with respect to the substrate normal, figure 3(b).

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

In contrast, a more step-like pattern of c_YBCO-parallel nanorods and (a–b)_YBCO-oriented platelets is clearly visible for the BYNTO-doped film (figure 4). Whereas the fine nanorods with a diameter of about 7 nm show a pure BYNTO composition, the larger platelets of length up to 200 nm contain a significant amount of yttrium. A detailed analysis showed mainly pure Y₂O₃ platelets decorated with BYNTO particles. These individual platelets are biaxially oriented and grow parallel to the (a–b)-planes. However, ensembles of these platelets tend to have a certain angle (between ±4° and ±10°, figure 4) to the (a–b)-planes of the YBCO matrix due to a terrace-like configuration, as seen in figure 4(b). Such a microstructure of terrace-like 'composite defects' has been observed also in MOCVD-grown coated conductors [28], where an IBAD-MgO template was used showing an additional tilt of the matrix crystal structure [29]. However, no clear indication for a similar tilt of the YBCO matrix structure was found for the ABAD-YSZ templates used in this study. Finally, the few c-axis-oriented BYNTO nanorods have an areal density of around 1500 μm⁻², corresponding to a matching field of around 3.1 T. An overview on the dependence of the nanoparticle distribution on the deposition condition can be found in our previous publication [16].

The resistively measured transition temperatures of T_c,0 = 89.3 ± 0.3 K (BYNTO-doped) and T_c,0 = 88.0 ± 0.4 K (BHO-doped YBCO), remain almost constant up to a thickness of 7 μm and are only slightly lower than the values of undoped films. In general, a constant J_c value (as obtained by scanning Hall probe microscopy) of about 0.6 MA cm⁻² was determined for BHO-doped films with a thickness of 1–5 μm resulting in an I_c/w value of more than 300 A cm⁻¹ (figure 5(a)). These J_c values are significantly lower compared to the undoped film, where a drop from 2.5 to 1.5 MA cm⁻² was observed for the same thickness region [5]. The lower J_c values might partially originate from the lower T_c values; however, other effects as an insufficient oxygeneation or a different strain state due to the denser microstructure can not be excluded. Nevertheless, the I_c/w values are increasing almost linearly until a thickness of 5 μm in contrast to the undoped films, where no significant further increase was found above 3 μm [5]. Furthermore, the BHO-doped YBCO layers revealed a homogeneous J_c distribution in scanning Hall probe microscopy in liquid nitrogen temperature as shown for the 5 μm thick film (inset in figure 5(a)).

Zoom In Zoom Out Reset image size

The BYNTO-doped YBCO films show a slightly different behavior. In this case, J_c values of about 1 MA cm⁻² and 0.8 MA cm⁻² were determined for the 1.2 μm and 3.9 μm thick film, respectively (figure 5(b)). The J_c distribution remains homogeneous; however, some regions with lower values are visible. Nevertheless, the resulting I_c/w value of 310 A cm⁻¹ for the 3.9 μm thick film is even higher than for the BHO-doped sample with the same thickness. However, J_c of thicker films drops significantly to a value of about 0.3 MA cm⁻² for 7 μm. This reduction is most probably correlated with the microstructural changes described before.

In summary, the incorporation of BHO or BYNTO secondary phase particles in the YBCO matrix leads to a denser microstructure and enables a preferential c-axis-oriented growth of the superconductor up to a thickness of more than 4 μm. This value is significantly higher than the typically reported values, showing the potential for a further improvement in I_c. Above 4 μm, more and more a-axis-aligned and misoriented grains result in lower J_c values. Possibly, this critical thickness for the nucleation of misoriented YBCO grains can be shifted to even higher values by careful adjustment of the deposition parameters.

Three different films, i.e. pure YBCO, BHO-doped YBCO, and BYNTO-doped YBCO, with a thickness of about 1.3 μm were deposited on 20 mm long ABAD-YSZ templates to study the influence of the dopant on the microstructure and the anisotropy in more detail. The deposition conditions were further optimized to achieve an homogeneous film thickness on the longer substrates. All samples showed a homogeneous microstructure with a purely c-axis-oriented growth. The nanoparticles incorporated are biaxially oriented as discussed in section 3.1 and shown in [16].

All three samples reveal a homogeneous J_c distribution with some smaller defect areas, as checked by scanning Hall probe microscopy (figure 6). The average J_c value at 77 K in self-field reaches 2.2 MA cm⁻² for the undoped film, whereas 0.9 MA cm⁻² and 1.8 MA cm⁻² were determined for the BHO- and the BYNTO-doped YBCO layer, respectively. The reduced J_c value for the BHO-doped film might be mainly due to the smaller T_c of about 87 K measured inductively for this sample in comparison to 89 K for the undoped and 90 K for the BYNTO-doped film, respectively. A similar trend for BHO-doped films was already found in previous studies, e.g. [30].

Zoom In Zoom Out Reset image size

Self-field J_c values at 77 K of 3.6 MA cm⁻², 1.6 MA cm⁻² and 2.1 MA cm⁻² for the undoped, BHO-doped and BYNTO-doped YBCO layer, respectively, were determined with transport measurements on microbridges. These values are slightly higher than the integral values calculated from the scanning Hall probe microscopy measurements, which originate both from the different electric field in the two types of measurements as well as from local variations of J_c.

The J_c(B, θ) anisotropy, θ being the angle between B and (a–b)_YBCO-planes, for different temperatures and magnetic fields shows distinct differences between the different dopings, as summarized in figure 7 for selected datasets. The undoped YBCO film (red) shows high J_c values for fields parallel to the (a–b)-planes (θ = 90° and 270°) and a small peak for B∣∣c. The latter is related to aligned growth defects as for example grain boundaries originating from the granularity of the ABAD-YSZ template. In addition to the intrinsic electronic anisotropy, which accounts for a J_c anisotropy J_c^{B∣∣(a−b)}/J_c^B∣∣c of around 2 at 2 T, 77 K if only small, uncorrelated pinning centers were present (and hence a J_c of ∼0.3 MA cm⁻² for B∣∣(a–b), stacking faults as well as (a–b)-aligned Y₂O₃ particles lead to a significant pinning contribution for B∣∣(a–b).

Zoom In Zoom Out Reset image size

The BHO-doped sample (green) reveals smaller J_c values at 77 K due to the smaller T_c value. However, a strong peak for B∣∣c is visible at 2 and 5 T for both temperatures. This peak is shifted to higher angles due to the tilted orientation of the BHO nanocolumns. The shift of the (a–b)-peak to lower angles at low applied fields (where the direction of the applied magnetic field and the average vortex direction are not necessarily parallel anymore because of energy minimization [31]) has the same origin. The reason is mainly an increased pinning probability in the directions of smaller angles between the two sets of defect structures, c-axis columns and (a–b)-planar defects, called mixed pinning [32]. Apparently, the J_c values for B∣∣c at 64 K are significantly higher than for B∣∣(a–b). In general, BHO tends to form extended nanocolumns having some spread as shown in more detail in our previous publications (see figure 7 in [5]). These nanocolums have typical diameters of about 4 nm and are densely arranged in the YBCO matrix. Therefore, these nanocolums are strong pinning centers for B∣∣c in particular at lower temperatures (i.e. 40 K [5]).

The J_c(θ) curves of the BYNTO-doped films show a significantly reduced anisotropy. At 2 T, the BYNTO-doped film outperformes the undoped YBCO film for most field directions except near B∣∣(a–b). Additionally, a broad peak for B∣∣c is visible. This behavior results from the typical distribution of artificial pinning centers with short c-axis-aligned BYNTO nanocolumns and additional (a–b)-oriented BYNTO or Y₂O₃ platelets. At a closer look, both main directions, B∣∣(a–b) and B∣∣c, show a double peak structure at low magnetic fields (2 T) and more so at lower temperatures. This is explained by the effectively tilted defect microstructure of more or less well oriented short defects and the different elasticities of the vortices and hence trapping lengths in the (a–b)-planar defects. At high magnetic fields close to the (a–b)-direction, the flux lines follow the macroscopic direction of the applied field and align with the (a–b)-defects for B∣∣(a–b), which leads to a single peak at B∣∣(a–b). At low fields, they can be trapped by the individual defects and gain the largest energy now for fields applied parallel to the macroscopic direction of the defect structure, which leads to a double peak around 90°. The peaks are positioned at around ± 8°, which is comparable to the average angle of the platelets of 4°–10° from TEM, figure 4. Similarly, a broad double peak around 0° is observed at 2 T, 65 K, whereas at high fields, which exceed the matching field of the nanorod segments, no c-axis peak is visible. Similar double peak structures have been observed recently for (a–b)-planar SrTiO₃ structures by Crisan et al [33] and for c-axis-oriented BaZrO₃ nanorod segments by Malmivirta et al [34].

The technologically interesting minimum value of J_c for a certain field and temperature, ${{J}_{{\rm{c}}}}^{{\rm{\min }}},$ is slightly decreased by the BHO addition, but significantly increased for the BYNTO addition with respect to the pure sample at 2 T. This is a direct consequence of the complex pinning landscape with c-axis nanorod segements and (a–b)-planar platelet structures in the latter samples. At 5 T, ${{J}_{{\rm{c}}}}^{{\rm{\min }}}$ is nearly the same in all three samples.