Nanostructured templates for critical current density enhancement in YBa₂Cu₃O_7–x films

The PAD technique involves the preparation of a water based solution containing polymer, a chelating agent and the metals precursors [33]. The procedure for the PAD solution with BaZrO₃ (BZO) and ZrO₂ (ZO) precursors (BZO and ZO PAD solution) starts with the preparation of the individual metal-polymer aqueous solutions: Ba(NO₃)₂, ZrO(NO₃)₂ were dissolved in water together with ethylenediaminetetraacetic acid (EDTA) in a 1:1 molar ratio. Branched polyethylenimine (PEI, average M_n = 70 000) was added to reach a pH value of 5–5.5. Any unbound ion and the counter ions in the solution are removed via centrifugation in the Amicon® ultra filtration units up to constant volume of the retained fraction.

Inductively Coupled Plasma spectroscopy was used to determine the single metal precursor solutions. Then these solutions are mixed and the final concentration was adjusted to the desired value by adding ultra-pure water to (Oxide) = 0.003 M. Due to the low substrate wettability, in order to ensure homogeneous deposition, the (001) oriented STO single crystal substrates were treated at 600 °C for 2 h in O₂ flow. The final polymeric solutions were spin coated on (100) SrTiO₃ (STO) substrates at 5000 rpm for 60 s. The samples were then annealed at 1000 °C for 4 min in air by directly introducing the samples into the furnace, followed by quenching to room temperature. More detail on STO surface decoration has been already presented in [31, 33].

The YBCO coating solution was prepared with a low-fluorine MOD approach. Details on the preparation and thermal processing conditions are described elsewhere [34, 35]. The resulting film thickness was about 90 nm. The YBCO films analyzed in this paper on bare and on BZO or ZO decorated STO substrates were grown during the same thermal treatment using the same precursor solution, so as to avoid problems related to solution or treatment reproducibility.

The x-ray diffraction (XRD) measurements, to analyze the structural properties of the films, were performed using a Rigaku Geigerflex diffractometer working with Cu-Kα radiation.

A LEO 1525 field emission high-resolution scanning electron microscopy (SEM) was used to investigate the surface morphology. Nanoparticles size and distribution were analyzed through atomic force microscopy (AFM). Analyses were performed using a Park Systems XE-150 AFM operating in non-contact mode with a pre-mounted non-contact high-resolution cantilever working at 309 MHz with a nominal tip radius below 10 nm. Images were flattened by subtracting a linear background in the fast scan direction and a quadratic background in the slow scan direction.

The TEM analysis was performed at the Electron Microscopy for Materials Science at the University of Antwerp. Plan view and cross-section transmission electron microscopy (TEM) lamellae were prepared via Focused Ion Beam (FIB). During the preparation, Pt protective layers were deposited on top of the film. The probe aberration–corrected high angular annular dark field scanning measurement transmission electron microscopy (HAADF-STEM) were carried out on an FEI Titan electron microscope operated at 300 kV.

In order to perform dc transport measurements, for the current-voltage characteristics (I–V), the films were patterned, using standard UV photolithography and wet etching to obtain 1 mm long strips with a width of 30 μm and 50 μm. The patterned samples were mounted on a sample holder and loaded in a He gas flow cryostat provided with a 7 T superconducting magnet. The sample can rotate in a wide angular range changing the magnetic field incident angle θ, with θ= 180° and θ= 90° correspond in H parallel to the YBCO c-axis and ab-plane, respectively. Critical current values as a function of the applied magnetic field or the angle value have been obtained from the I–V characteristics measurements using the 5 µV · cm⁻¹ criterion.

Figure 1(a) shows the AFM scans of a ZO-decorated STO substrate with well-defined nanometric NPs nanometric in size, uniformly distributed over the whole substrate surface. For BZO decorated STO substrate similar features are observed.

Zoom In Zoom Out Reset image size

Through the AFM image processing more quantitative analyses of nanoparticles are derived. The thus obtained NPs are characterized by typical values (averaged over several similarly processed samples) for the radius R = (17 ± 5) nm and a height h = (5 ± 1) nm for the BZO and R = (19 ± 9) nm and h = (5 ± 1) nm for the ZO, with an average particle density of approximately 200 ÷ 300 NPs μm⁻² for BZO and 300 ÷ 400 NPs μm⁻² for ZO [31]. More detailed characterizations of BZO and ZO NPs systems can be found in [21, 26]. The calculated density is in line with values previously reported in other works using different methods and NP systems [23, 36]. Higher values can be achieved, for example, by increasing the solution concentration or with different systems as shown in our previous work [29].

On the other hand, it has to be taken into account that an increased density, resulting in a larger coverage of the substrate surface, could ultimately be detrimental for the nucleation and epitaxial growth of the YBCO films, and then leading to poorer film quality, unless NPs radius was not progressively reduced. Both BZO and ZO NP systems studied in this work exhibit the optimal characteristics obtained so far.

In figure 1, SEM images show the surface morphology of YBCO grown on bare (b) and on ZO decorated STO substrate (c), respectively. For both films, the surface appears flat and compact with some outgrowths. It is worth considering that no changes on the surface morphology, due to the presence of the ZO nanoparticle, is observed. The same consideration can be done also on the film grown on BZO decorated substrate (SEM image not reported).

X-ray diffraction analysis was performed on all samples: YBCO, and YBCO on BZO and ZO decorated substrates diffraction patterns are shown in figure 2. The spectra show an epitaxial growth of the three samples and no difference with respect to the pristine YBCO sample is noticed. The calculated c-axis parameters of all the samples are in agreement with typical YBCO values reported in previous work [37], showing that the presence of BZO or ZO nanoparticle on STO surface does not affect the c-axis film growth. The full width at half maximum (FWHM) of the (005) peaks ω-scan are also reported in table 1. In addition, traces of Y₂O₃ are also observed. Moreover, no evidence of BZO or ZO peaks can be detected, probably because the intensity is below our experimental detection limit.

Zoom In Zoom Out Reset image size

In figure 3(a), the TEM cross section of the YBCO bare sample is reported. To make sure that the obtained microstructural features could be directly correlated with the transport properties, the TEM lamellae were taken from the strips used for the I–V measurement.

Zoom In Zoom Out Reset image size

It is clear that the YBCO grows without major structural defects apart from the region close to surface, where stacking faults can be recognized. Furthermore, the presence of twin boundaries, visible as the succession of darker and brighter zones in the TEM image, is detected. The difference in contrast is not related to any compositional change, as confirmed by EDX analysis (data not shown), but is related to the different orientation of the YBCO lattice ([100] or [010]) [38]. The presence of twin boundaries, as observed in [39, 40], can be responsible for the anisotropic contribution to the pinning when the applied magnetic field is parallel to the c-axis.

In the case of YBCO grown on decorated substrates, the TEM image (figure 3(b)) shows that the superconducting film grows all around the BZO nanoparticle with a significant density of defects. These defects, visible as darker stripes in the YBCO matrix, can be found in the whole film thickness and are commonly referred to as stacking faults, consisting of an extra Cu-O chain between two Ba-O layers [38, 41, 42]. In these areas, the YBCO grows with a stoichiometry different from the usual one: instead of growing with the usual stoichiometry YBa₂Cu₃O_7−x or Y123, it grows as Y₂Ba₄Cu₈O₁₆ or Y248.

The formation of stacking faults is probably promoted in order to release the strain that accumulates at the YBCO/BZO or YBCO/ZO interface during YBCO growth [13]. Partial dislocations at whose position all the strain is accumulated delimit the stacking fault at both edges.

The effect of strain in the YBCO lattice can be observed in the Geometric Phase Analysis (GPA) map reported in figure 3(c). The variation of the structural parameter in the c-axis direction is calculated with reference to the area of YBCO free from defects (highlighted in green in figure 3(b)). This analysis shows a high amount of strain in YBCO localized in correspondence of the edges of the stacking faults.

The critical temperatures of the all samples (T_C), reported in table 1, are very similar within the limit of the experimental error, showing that the substrate decoration does not affect this parameter.

As regards transport properties, in figure 4(a), the critical current density (J_C) as a function of the applied magnetic field (B = μ₀H) in a configuration in which the magnetic field is parallel to the YBCO c-axis, at 77 K (filled symbols) and 65 K (empty symbols) is reported. As can be seen from figure 4(a), the values of the critical current density for the samples deposited on decorated substrates (red circles for BZO and black triangles for ZO decorated templates) are higher than pure YBCO (blue squares) in the whole field range and at both temperatures. More in detail, the values of J_C at 77 K for BZO and ZO decorated samples are 0.211 MA cm⁻² and 0.235 MA cm⁻² at 1 T and 0.0142 MA cm⁻² and 0.0175 MA cm⁻² at 5 T, respectively, representing a threefold improvement compared to the 0.107 MA cm⁻² at 1 T and the 0.0022 MA cm⁻² at 5 T for pristine YBCO. The same trend is observed at 65 K 1 T with J_C values of 0.393 MA cm⁻², 0.801 MA cm⁻² and 0.780 MA cm⁻² for pristine YBCO, BZO decorated and ZO decorated respectively. Furthermore, the self-field values (reported in table 1) are higher for decorated samples, and also the in-field behavior is improved, with the difference between decorated and pristine samples increasing with increasing applied magnetic field. The critical current density improvement, obtained for PAD–techniques, is larger than what reported in the case of MOD Ba₂YNbO₆ decorated substrate [43]. This different effect can be attributed to the higher density of nanoparticles obtained in these systems.

Zoom In Zoom Out Reset image size

Figure 5 shows the normalized pinning force f (where f = F_P/F_P^Max, F_P= J_C × μ₀H, and F_P^Max is the maximum value of the pinning force), as a function of the normalized magnetic field applied parallel to the c-axis, b (b = H/H^Max, where H^Max = H(F_p^max)) calculated for both 77 K and 65 K, for the three different samples. As can be seen, a perfect overlap of the curves is evident: this leads us to conclude that the mechanism responsible for pinning along the c-axis is similar in nature for both YBCO and decorated YBCO films.

Zoom In Zoom Out Reset image size

This improvement of the transport properties can also be analyzed through angular measurements. In figure 4(b), the critical current density is reported as a function of the angle between the magnetic field and the c-axis of the film for pure YBCO (blue squares) and for a BZO-decorated sample (red circles) at 77 K and three different applied magnetic fields, μ₀H= 0.5 T, μ₀H= 3 T and μ₀H= 5 T. Again, the decorated sample shows improved properties in the whole angular range, for both values of the applied magnetic field. This is confirmed for all the applied magnetic field values investigated.

As shown in figure 4(b), when the magnetic field is parallel to the ab-plane, a strong increase in J_C is obtained for both samples, in agreement with the YBCO structural anisotropy, with the flux lines accommodating in the insulating layers of the YBCO lattice that act as intrinsic pinning centers, as already reported in [44–46].

Under this condition, at θ = 90°, a valuable contribution is also provided by extrinsic planar defect (parallel to ab-plane) such as stacking faults or intergrowths [47–49]. Furthermore, when the magnetic field is applied close to the c-axis (θ= 180 °), a much weaker peak in J_C(θ) is also observed, already visible in the curves measured at 1 T (not shown), more evident as the field increases as in the case of the curves recorded at 5 T. This second peak is attributed to other defects present along this direction, such as twin boundaries, that act as strongly correlated pinning centers [39, 40].

To better identify the contribution of isotropic and anisotropic pinning in YBCO superconducting films, J_C(θ) data have been analyzed following the methodology based on the Blatter scaling function [50], recently developed and successfully applied to MOD YBCO films to sort out the anisotropic and isotropic contributions in different angular regimes [1].

This method assumes that angular J_C measurements at fixed H can be plotted as J_C(H_eff) where H_eff = H× (γ, θ), being ²(γ, θ) = cos²(θ) + γ⁻²sin²(θ), with γ = $\frac{{{m_{ab}}}}{{{m_c}}}$ the electron mass anisotropy of the superconductor film. In the presence of pinning centers which do not affect the electronic anisotropy in YBCO, such as point (or 0D) defects, J_C(H_eff) data should collapse in a single curve. Failure of this scaling can be considered as an indication of the presence of anisotropic pinning contribution.

Applying this data processing for pristine YBCO sample at 77 K with γ = 5, typical value for a pristine YBCO film [51, 52], most of the experimental points collapse on a single curve as reported in figure 6(a). The dashed line represents the J_C^iso(H_eff) contribution obtained by fitting with the analytical expression f(H) = (1 + H/H₀)^−α (1−H/H_irr)^β, being H₀, H_irr and α and β fitting parameters. The failure of Blatter scaling indicates the presence of a dominant anisotropic pinning contribution for some θ intervals. By making the deconvolution of the J_C(H× (γ, θ)), it is possible to plot the isotropic behavior J_C^iso(θ). The same procedure was applied to YBCO on decorated substrates assuming the same γ= 5. This is justified considering that the nature of the defects introduced by NPs, as shown by TEM analysis, does not significantly affect the intrinsic electronic mass anisotropy [53]. The as-obtained isotropic curves and the data for the YBCO and YBCO samples on BZO, at 77 K and μ₀H = 0.5 T, are shown in figure 6(b), where the solid lines represent the as calculated isotropic J_C(θ), J_C^iso(θ).

Zoom In Zoom Out Reset image size

Notice that there is a large discrepancy between J_C^iso(θ) and experimental data in a large angular range close to ab-planes and c-axis indicating the occurrence of relevant anisotropic pinning mechanisms in those conditions. As already described, the origin of these anisotropic contributions can be mainly ascribed to the intrinsic ab-plane and twin boundaries pinning.

To quantify the anisotropic contribution of pinning in our samples, we calculated for the case of c-axis,

$$\begin{equation*}c\left( {\text{% }} \right) = \frac{{\Delta {J_{\text{C}}}\left( {H||c} \right)}}{{{J_{\text{C}}}\left( {H||c} \right)}}\end{equation*}$$

Where ΔJ_C(H||c) = J_C(H||c)—J_C^iso(H||c), J_C(H||c) is the critical current density measured along the c-axis and J_C^iso(H||c) is, instead, the critical current density along the c-axis, calculated from the isotropic contribution previously estimated.

In the case of ab-plane, we similarly evaluate

$$\begin{equation*}ab\left( {\text{% }} \right) = \frac{{\Delta {J_{\text{C}}}\left( {H||ab} \right)}}{{{J_{\text{C}}}\left( {H||ab} \right)}}\end{equation*}$$

where ΔJ_C(H||ab) = J_C(H||ab)—J_C^iso(H||ab), here J_C(H||ab) is the critical current density measured along the ab-planes and J_C^iso(H||ab) is, instead, the critical current density along the ab-planes, calculated from the isotropic contribution.

Thus, we can evaluate the trend of both ab(%) and c(%), that represent the percentage of the anisotropic contribution along both directions, for the BZO decorated, ZO decorated and pristine YBCO samples, as a function of the applied magnetic field. In detail, from figure 7(a) it is possible to notice that in the case of c(%), the three samples show similar values in the whole range of applied magnetic field at 77 K, while at 65 K the values are slightly different but still very close to each other. This means that the trend of the anisotropic contribution does not change significantly among the three samples, and that the improvement in the pinning properties along the c-axis, observed in the decorated sample, is most likely due only to an isotropic contribution. The isotropic nature of this pinning mechanism causes a shift of the J_C(θ) curves measured on the decorated samples to higher values in the whole angular as observed in figure 4(b).

Zoom In Zoom Out Reset image size

On the contrary, from figure 7(b), higher values for the ab(%) parameter in the samples deposited on decorated substrates are obtained at 77 K, indicating a more valuable strengthening of the anisotropic vortex pinning contribution in these samples, when the magnetic field is approaching the ab-plane. On lowering the temperature to 65 K, the effectiveness of the additional ab-plane correlated pinning centers, induced in the decorated samples, is reduced.

The origin of the different types of pinning mechanisms, most likely isotropic along c-axis and anisotropic along ab-planes, is consistent with the TEM analyses, previously reported.

The twin boundaries present in the bare YBCO sample, reported in figure 3(a), are a typical example of preferential pinning along the c-axis.

Moreover, in the case of the YBCO grow on decorated substrates, the edges of the stacking faults, highlighted in the GPA map reported in figure 3(c), act as point defects with a 3D isotropic nature. providing a contribution in the vortex pinning landscape. We think that the presence of this point defects can be the cause of the increase of the isotropic contribution in the YBCO film deposited on decorated substrate with respect to the bare YBCO, contributing to the shift to higher values of the critical current density in the whole range of fields and angles.

On the contrary, looking at the situation when the magnetic field is applied along the ab-plane, the stacking faults, showed in figure 3(b), can act as pinning centers themselves. The additional Cu-O chains could serve as preferential accommodation sites for flux lines along this direction, behaving as anisotropic pinning centers. Therefore, this effect can explain the increase in the anisotropic contribution along the ab-plane in the decorated samples at 77 K.