Fe(Se,Te) coated conductors deposited on simple rolling-assisted biaxially textured substrate templates

In figure 1, we report the XRD analysis for a CeO₂ buffer layer grown on Ni–W. The intense (002) CeO₂ peak, the only reflection related to the film, indicates that the buffer layer develops a single orientation with the (001) direction perpendicular to the substrate. Even if a major part of the deposition process occurs in oxygen atmosphere, peaks related to the substrate oxidation are not detectable. In the insets, we report the ω-scans in the transverse direction with respect to the deformation process—which is the less favourable direction—through the (002) Ni–W and CeO₂ peaks. The FWHM are 7.5° and 6.5° respectively for the Ni–W and the CeO₂ confirming, as previously reported, that CeO₂ films are prone to improve the out-of-plane substrate texture [15]. The CeO₂ film texture analysed by pole figures in figure 2 reveals the in-plane epitaxial relationship: [100] Ni–W//[110] CeO₂. The FWHM value of the φ-scan across the (111) CeO₂ poles lie between 5.0° and 6.2°, similarly to those measured across the (111) Ni–W poles of the substrate which are between 5.2° and 6.7°, as it will be reported later.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Figure 3 shows SEM images of a CeO₂/Ni–W buffer layer surface after the annealing process in a temperature range relevant for 11 film deposition (>300 °C). The structural and morphological properties of the CeO₂ film are unchanged after the annealing, in particular the surface is uniform, smooth and dense. Above all, the CeO₂ film assures a good coverage of the grooves among the substrate–grain boundaries. This result shows that CeO₂ film on Ni–W is a robust template suitable for the 11 film deposition.

Zoom In Zoom Out Reset image size

Figure 4 reports a θ−2θ scan of an FST thin film deposited on the CeO₂/Ni–W template. Besides the (001) peaks corresponding to the 11 phase, the peaks coming from the template are clearly visible, but no other peaks are present relative to other orientations or phases, indicating an optimum c-axis alignment of the growth and a high purity of the phase. The ω-scan on the (002) peak, reported in the inset of figure 4, shows a FWHM value Δω of about 6.6° measured in the transverse direction, which is compatible with the value measured in the CeO₂ buffer layer (see inset of figure 1) and confirms a good out-of-plane texture.

Zoom In Zoom Out Reset image size

In order to establish the in-plane epitaxy of the growth, we performed φ scans of the (101) reflections of the 11 films. In figure 5(a), we show the polar figures of an FST thin film along the [101] and the [111] direction, while figure 5(b) reports the φ scans of the (101) peak from the thin film compared with those of the (111) peak of the CeO₂ and the Ni–W substrate. The lattice of the 11 thin films is therefore aligned cube-on-cube with the Ni–W substrate and rotated by 45° in the ab plane compared that of the CeO₂ buffer layer. Eleven thin films grown on CaF₂ single crystals and on RABiTS templates show the same behaviour [7, 16]. In fact, the diagonal of the lattice parameter of CeO₂ (and CaF₂) is about 3.82 Å (∼5.41/√2), which matches exactly with the a lattice parameter of FST. The average in-plane FWHM Δφ_FST is about 4.9°, comparable with Δφ_CeO2 which is about 4.7°. Therefore, we can conclude that the 11 thin films grown on CeO₂/Ni–W templates possess a good biaxial texturing determined by the texturing of the CeO₂ buffer layer. Such results are consistent with the data reported in the literature on IBAD templates [6, 11] and RABiTS with a CeO₂/YSZ/Y₂O₃ buffer layer [17].

Zoom In Zoom Out Reset image size

Figure 6 reports the surface of a patterned FST thin film grown on a CeO₂-buffered Ni–W substrate. The surface is uniform and shows an irregular hill-and-valley morphology. Average and rms roughness on a 1 μm² area are 0.7 nm and 0.9 nm, respectively, slightly larger than what observed for CeO₂ films (0.4 and 0.5 nm, for average and rms roughness, respectively). In order to correctly evaluate the thickness of the 11 films, the cross section of the fractured sample was analysed with SEM: as shown in figure 7, the thickness of the film is about 100 nm. Both the CeO₂ and 11 films appear compact and continuous along the whole thickness suggesting very high-density layers. No interdiffusion process is detectable at the Fe(Se, Te)/CeO₂ interface. The Fe(Se, Te) film shows some longitudinal features that suggest a layered growth mechanism.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

The microstructure of the 11 films was investigated by means of the EBSD technique.

Figure 8 shows the EBSD map (a) and the corresponding pole figures (b) of an FST film grown on a CeO₂/Ni–W template. In both the EBSD map and pole figures, points are displayed according to the local misorientation angle with respect to the ideal {001}〈010〉 orientation, as shown in the legend of figure 8(c). In addition, grain boundaries above 2° and 10° (thin and thick lines, respectively) are shown.

Zoom In Zoom Out Reset image size

Kikuchi patterns were well fitted using the P4/mmm tetragonal phase and no hexagonal component was detected. As can be seen, the 11 film shows a sharp {001}〈010〉 texture, with a fraction of oriented points of 98.9% within 10°. The film reproduces the microstructure of the underlying Ni–W substrate, as typically observed in RABiTS-based CCs [17]. In fact, the film shows a relative misorientation below 2° in regions of 35 ± 15 μm, which is the typical grain size of a Ni–W RABiT substrate [9, 16–18]. As can be seen, the vast majority of grain boundaries are below 10°. The microstructure is well connected, with the presence of several low-angle grain boundary percolation paths. The critical angle for the existence of a percolative path across the sampled area is 4.5°. This is in line with the value obtained in the case of YBCO films grown on a similar template [19].

The underlying CeO₂ buffer layer shows the same microstructure observed in the FST thin film. Figure 9 reports the EBSD map (a) and the corresponding pole figures (b) of the CeO₂ film analysed outside the patterned strip used for transport measurements. As can be seen, the CeO₂ film shows a strong {001}〈110〉 texture, with a fraction of oriented points of 99.2% within 10°.

Zoom In Zoom Out Reset image size

From EBSD microstructural investigation, we can conclude that an epitaxial relationship (001)[010] FST//(001)[110] CeO₂//(001)[100] Ni–W has been obtained.

In figure 10 we report the temperature dependence of the normalized resistivity of the 11 CCs up to 18 T in the two directions, perpendicular and parallel to the magnetic field. T_c,0 is about 16 K with an onset above 18 K, comparable with thin films on single crystals or on technological templates [4].

Zoom In Zoom Out Reset image size

All the plotted ρ(T) curves show low normal state resistivity values and an anomalous bump before the transition with a sharp increase in ρ followed by a steep reduction and then eventually reaching the zero resistance state. The maximum of the bump, T_m, shifts towards lower temperatures coherently with T_c as the applied magnetic field is increased. However, the presence of such a bump was not observed in unpatterned films grown on the same template studied in the present work.

The normal state temperature dependence of the resistivity exhibited by iron chalcogenides is, in general, not trivial. Semiconductor behaviour up to room temperature has been reported in both bulk and films, while a gradual crossover to metallic behaviour is usually observed for temperatures approaching the superconducting transition [20–22]. These features give rise to highly non-linear ρ(T) curves sometimes exhibiting local maxima. A similar bump in the resistivity has been observed in Ba-122/Fe bilayers grown on IBAD tapes with thin Y₂O₃ buffer layers and was ascribed to the effect of current shunt provided by the metallic template in the presence of a conductive buffer layer [23]. On the other hand, comparable behaviours of resistive transitions have been often reported for different superconducting films and ascribed to a phenomenon related to the current redistribution occurring in presence of sample inhomogeneity, i.e. regions characterized by slightly different T_c and multiple paths for the bias current [24]. Therefore, the present anomalous bump can be explained by invoking the occurrence of a metallic path, likely to be ascribed to a technological issue that arose during the film patterning process, in parallel with the superconducting path with some degree of film inhomogeneity. In particular, the resistive transition behaviour is thus reproduced by equation (1a) of [24], assuming R₁ of [24] = metallic path resistance of the present case.

The irreversibility field and the upper critical field values, H_irr and H_c2, are usually evaluated from the R(T, H) curves using, respectively, the 10% and 90% of the superconductive transition criterion. This method is more difficult to be applied to the measurements presented in this work since the reported anomalous behaviour, though accounted for by the presence of multiple paths for the bias current, makes any evaluation of the normal state resistivity value quite speculative. The alternative criterion is thus adopted, considering the 10% of the peak resistance at T_m and T_m(H) dependences as the H_irr and the H_c2 curves, respectively. In the first case, the values will be close to those evaluated with the standard 10% criterion. In the upper critical field case, a qualitative picture of the H_c2(T) dependence can be obtained, even though quantitatively underestimated. The H_c2(T) uncertainty associated with this criterion can be estimated as the temperature interval between T_m and the temperature at which the 50% of the resistance bump is achieved: this means about 0.3 K in self-field increasing to about 0.8 K at 18 T for H perpendicular to the ab planes, while is 0.3 K for any value of H parallel to ab plane.

In figure 11 we report the H versus T diagram, with the upper critical field H_c2 and the irreversibility field H_irr calculated as reported above in both the field directions. H_c2 versus T is very steep near T_c, with a downward curvature and slopes of about 63 T K⁻¹ and 7 T K⁻¹ in the parallel and perpendicular directions, respectively. Such values are quite high, even though not as high as those observed in thin films deposited on single crystals which can reach up to 500 T K⁻¹ and 30 T K⁻¹ in the two directions [25]. The extremely high values reported on our strained film deposited on LaAlO₃ and CaF₂ were explained in terms of an extreme Pauli-limited H_c2(T), indicative of the Fulde–Ferrell–Larkin–Ovchinnikov state [25]. The slopes near T_c are significantly higher than those reported on IBAD-LaMnO₃ (LMO) buffered metal tapes [18]. Trying to estimate H_c2 values at zero temperature, we applied the Werthamer–Helfand–Hohenberg relationship H_c2(0) = −0.693T_c(dH_c2/dT)_∣Tc and obtained 790 T and 90 T for H//ab and H//c respectively. The anisotropy is not high: ${{H}_{{\rm{c}}2}}^{//{\rm{ab}}}/{{H}_{{\rm{c}}2}}^{//{\rm{c}}}$ is about 2.3 at 17 K while ${{H}_{{\rm{irr}}}}^{//{\rm{ab}}}/{{H}_{{\rm{irr}}}}^{//{\rm{c}}}$ is about 1.8 at 15 K, both higher though than the values reported in thin films grown on single crystals (where H_c2 anisotropy can be as low as 1.2 [16]) and on IBAD-LMO buffered metal tapes [18].

Zoom In Zoom Out Reset image size

In figure 12(a), we report the critical current density of the CC as a function of the magnetic field measured up to 18 T at 4.2 K and 9 K. The self-field J_c value at 4.2 K is above 1.7 × 10⁵ A cm⁻² and above 3.0 × 10⁴ A cm⁻² in fields up to 18 T, while the self-field J_c value at 9.0 K is above 7.0 × 10⁴ A cm⁻². The J_c curves measured at 4.2 K with H parallel and perpendicular to the tape surface overlap in the whole magnetic field range up to 18 T, indicating an isotropic behaviour in these regimes. In figure 12(b), the pinning force densities calculated as F_p = J_c · B are reported: at 4.2 K we are still far from its maximum at 18 T, while we observe a plateau starting from 14 T at 9 K. The relatively low J_c values with respect to the previously reported FST CC architectures [7, 18] can be explained by the presence of some inhomogeneity in the 11 film. However, this effect can be mitigated through the optimization of growth conditions and texturing of both the buffer layer and FST.

Zoom In Zoom Out Reset image size