Chemical Solution Deposition of YBCO Films with Gd Excess

Abstract

Chemical solution deposition of Gd-doped YBCO, Y₁Gd_yBa₂Cu₃O_7−δ, (YBCO-Gd), film was carried out following the metal-organic decomposition approach and in situ route. Two dopant concentrations, 5 and 10 mol %, were evaluated. The morphology and crystalline structure of the superconductor films were deeply investigated. In general, a homogeneous and well c-axis oriented film was observed by using scanning and transmission electron microscopy (SEM and TEM) and X-ray diffraction. However, compared to pure YBCO, YBCO-Gd samples showed an increased stacking faults concentration, as recognized by TEM. X-ray photoelectron spectroscopy allowed studying the Gd distribution in the films and gathered information about the Gd electronic environment. Superconducting properties were evaluated at different temperatures, magnetic field directions, and intensities. Higher zero-field critical current densities were measured with respect to undoped samples in the temperature range from 10 to 77 K with both Gd concentrations (i.e., 28, 27, and 13 MA·cm⁻², respectively, for YBCO-Gd 5%, YBCO-Gd 10%, and undoped YBCO at 10 K in self field condition). At low temperatures, this improvement was maintained up to 12 T, confirming the efficacy of Gd addition for the enhancement of transport properties of YBCO film.

1. Introduction

Several studies have been carried out on the substitution of Y³⁺ in YBa₂Cu₃O_7−δ (YBCO) by rare earth elements (RE), such as Er³⁺, Ho³⁺, and Gd³⁺, to form REBa₂Cu₃O_7−δ (REBCO) compounds. Their superconducting properties are generally improved with respect to YBCO due to the higher critical temperature (T_c) values and, consequently, higher irreversibility fields (H_irr) at 77 K [1]. Among REBCO systems, GdBa₂Cu₃O_7−δ (GdBCO) is one of the most investigated and extensively adopted for coating conductors manufacturing. GdBCO thin films can be successfully deposited through physical [2,3,4] and chemical methods [5,6], being the latter ones less expensive and, consequently, more appealing from an industrial point of view [7], which is a key issue in the perspective of application technology development. However, with respect to pure YBCO, chemical solution deposition (CSD) of GdBCO film requires more accurate control of the process parameters, especially during the crystallization step. In fact, GdBCO nucleation is favored by higher growth temperatures, faster ramp rates, and lower oxygen partial pressure (pO₂) than those used in the YBCO process [8,9,10,11].

Driven by the works on YBCO films, a great number of studies have been carried out on GdBCO doping with oxides, such as BaZrO₃, BaHfO₃, Gd₂O₃, aimed at introducing artificial pinning centers (APCs) to improve film performances when an external magnetic field is applied [5,6,10,12]. On the other hand, mixed compounds, such as Y_1−xGd_xBa₂Cu₃O_7−δ (0 < x < 1), have been proposed and studied as a suitable alternative route for the improvement of vortex pinning capabilities [13,14,15,16,17,18].

To our knowledge, the addition of Gd in stoichiometric excess with respect to Y, Y₁Gd_yBa₂Cu₃O_7−δ, has been reported only by two papers: Xu et al. [19] studied the Gd₂O₃ introduction in YBCO bulk, and Li et al. [20] grew, through a sol-gel process, YBCO films doped with Gd 10 and 50 mol %. These articles have shown promising results with enhanced critical current density (J_c) in both self-field and in-field conditions at 65 and 77 K, respectively. Therefore, in the present paper, we proposed a similar study applied to Gd doped YBCO film (YBCO-Gd) deposited by CSD and a low-fluorine metal-organic decomposition (MOD) approach. The low-fluorine formulation has already demonstrated to be suitable for GdBCO deposition, even better than the solution composed by all trifluoroacetate precursors, the so-called TFA-route [5,21]. Our work aimed at exploring whether effective defects could be introduced in YBCO film through the in situ low-fluorine MOD technique, i.e., using a single solution with YBCO and Gd precursor salts, and at determining the temperature and magnetic field regimes in which their contribution is more favorable. The introduction of zero-dimensional defects by CSD is generally carried out by the ex-situ route, based on the addition of preformed nanoparticles to the YBCO metal-organic precursor solution [22,23]. On the contrary, the in situ process has demonstrated to have limited control over the particle size of secondary phases that spontaneously segregate during YBCO film growth. Therefore, this system is normally employed for three-dimensional defects generation [24].

However, when added in excess with respect to Y, Gd can play two different roles: it can replace Y to form GdBCO with the exceeding Y, possibly forming other oxides, such as Y₂O₃. Otherwise, Gd can form oxides as Gd₂O₃. Both situations should contribute to the generation of a complex defects landscape, namely, the formation of several oxides and enhancement of the lattice strain due to the accommodation of two comparable but not identical crystalline cells (i.e., YBCO and GdBCO) [25,26,27] (

Table 1. Lattice parameters in orthorhombic REBCO ¹ and cubic C-type RE₂O₃ ² phase. RE = Y, Gd.

Lattice Parameter	YBCO	GdBCO	Y2O3	Gd2O3
a (Å)	3.8230 (5)	3.8416 (4)	10.6021	10.8122
b (Å)	3.8857 (5)	3.8984 (4)	-	-
c (Å)	11.680 (2)	11.707 (1)	-	-
V (Å3)	173.5	175.32	1192	1264

¹ Rietveld refinements done by [25]. ² Data from [26,27].

). These nanometric defects, various for chemical nature and size, can act in a synergetic way, determining an overall improvement of the superconducting transport properties [17,28].

YBCO-Gd films were prepared through the low-fluorine MOD approach and in situ route with two dopant concentrations, 5 and 10 mol %, of Gd with respect to Y. Experiments were carried out using the thermal treatment optimized for YBCO. The purpose was to investigate the fate of Gd when process conditions are unfavorable to GdBCO growth, i.e., high pO₂ and slow ramp rate [8,10]. The morphology and crystalline structure of superconductor films were deeply investigated. In general, a homogeneous and well c-axis oriented film was observed by using scanning and transmission electron microscopy (SEM and TEM) and X-ray diffraction (XRD). X-ray photoelectron spectroscopy (XPS) allowed studying the Gd distribution in the films and gather information about the Gd electronic environment.

Results suggested that we could maintain the advantages related to the presence of Gd despite the non-optimum experimental growth conditions. Gd was finely and uniformly dispersed within the YBCO film matrix, likely forming GdBCO. The formation of a large number of stacking faults was also reported. The improved J_c(H) performances of YBCO-Gd films indicated that the rich defect landscape recognized in these samples could act as vortex pinning centers, particularly effective at low-temperature regimes.

2. Materials and Methods

YBCO and YBCO-Gd films were deposited following the low-fluorine MOD approach. The YBCO precursor solution was prepared by dissolving in propionic acid (Sigma Aldrich, St. Louis, MO, USA, 99.5%) stoichiometric amounts of yttrium (III) acetate hydrate (Sigma Aldrich, 99.9%), barium (II) trifluoroacetate hydrate (Alfa Aesar, Haverhill, MA, United States), and copper (II) acetate (Alfa Aesar 99.9%). The Y:Ba:Cu ratio was equal to 1:2:3, and the final solution concentration corresponded to [YBCO] = 0.2 M. Further details on the solution preparation can be found elsewhere [29]. YBCO-Gd solutions were prepared through the in situ approach, using a single solution with YBCO and Gd precursor salts and comparing two dopant concentrations, 5 and 10 mol %, of Gd (added as gadolinium (III) acetate hydrate, Sigma Aldrich, 99.9%) with respect to Y. Precursor solutions were spin-coated for 60 s at a spinning rate of 3000 rpm on (001) strontium titanate, SrTiO₃ (STO), single crystals (supplied by Wollemi).

Films were pyrolyzed in a quartz tube furnace with flowing oxygen (0.94 L min⁻¹, humid for T >100 °C) up to 480 °C using the following heating ramp rate: 10 °C min⁻¹ (up to 90 °C), 2.5 °C min⁻¹ (in the range 90–200 °C), 1.2 °C min⁻¹ (in the range 200–300 °C), 10 °C min⁻¹ (in the range 300–480 °C), 10 °C min⁻¹ (during the cooling step). A second treatment, named “Firing”, was then performed to promote YBCO crystallization and oxygenation: firstly, the sample was heated with a rate of 10 °C·min⁻¹ to the dwell temperature of 830 °C, kept for 50 min in a humid mixture of nitrogen and oxygen (flow rate N₂: 2.83 L min⁻¹, O₂: 1.15·10⁻³ L min⁻¹) and further 10 min in the same dry mixture (crystallization step). Subsequently, the film was cooled down to 450 °C with a rate of 10 °C·min⁻¹, kept at this temperature for 15 min in dry oxygen (0.94 L·min⁻¹) for the oxygenation step and then cooled to room temperature with the same ramp rate. The aforementioned procedure was defined as the “standard” process. Several experiments were carried out varying some parameters of the thermal treatment in order to estimate if the YBCO-Gd system needs, as Y_1−xGd_xBa₂Cu₃O_7−δ, 0< x ≤1, compounds, a precise optimization of the processing conditions. In particular, a faster heating/cooling ramp rate of 30 °C min⁻¹ (fast process) was studied in comparison with the standard process (10 °C min⁻¹). A sample of pure YBCO film, used as a reference, was always deposited together with YBCO-Gd films. A final thickness of 75 nm was obtained for both YBCO and YBCO-Gd films. Structural characterizations were carried out employing XRD. The θ-2θ and ω-scans were performed by a Rigaku Geigerflex diffractometer with Cu Kα radiation in Bragg Brentano configuration. Film morphology was evaluated by SEM images, obtained using a Gemini LEO 1525 field emission high-resolution SEM with 20 kV operating voltage and in-lens high-resolution annular detector. XPS data were acquired in an ultra-high vacuum system operating at 2 × 10⁻⁸ Pa base pressure and equipped with a CLAM2 hemispherical analyzer working at constant pass energy mode and calibrated on the Ag 3d_5/2 and Ag M4NN lines according to [30] and by setting at 285 eV the binding energy (BE) of the adventitious carbon C1s core-level measured on the as-inserted sample. For quantification, data were normalized to the photoelectron cross-section, analyzer transmission function, and electron mean free path. The estimated BE maximum error was ±0.1 eV. Ar⁺ etching was used for depth profiling. For sample etching, a PSP Vacuum Technology ISIS 3000 system was used as the Ar⁺ ion source. The samples were bombarded with a defocused beam of 1000 eV Ar⁺ ions at about 18 μA total sample current. The sputtering rate was calibrated against a sample of a known thickness. The YBCO-Gd sample was inserted into the analysis chamber “as received”. The Gd₂O₃ sample (PLD target, 99.95% supplied by Franco Corradi) used for reference was prepared in a glove box in the N₂ atmosphere and then loaded into the XPS sample load-lock system without exposure to the atmosphere.

For TEM investigations, a cross-sectional lamella was produced by a focused ion beam (FIB) technique, employing an FEI Dual Beam Helios NanoLab system (FEI, Thermo Fisher, OR, USA). The annular dark-field and high-angle annular dark-field scanning transmission electron microscopy (ADF and HAADF-STEM) imaging and energy-dispersive X-ray spectroscopy (EDX) were performed using an FEI Titan electron microscope, equipped with a Cs-aberration corrector for the probe-forming lens and a “Super-X” wide solid angle EDX detector operated at 200/300kV acceleration voltages [31,32].

The zero-resistance critical temperature, T_c, of the films was assessed by d.c. electric measurements in the four-probe configuration. Magnetization measurements were performed using a vibrating sample magnetometer (VSM) equipped with a 12 T magnet and a controlled He gas flow system. Isothermal magnetization loops, M(H) up to 12 T, were recorded at different temperatures. The field swept at a rate of 0.5 T min⁻¹ (B = μ₀H//YBCO c-axis). The critical current density values were extracted from the magnetization data through the Bean model for a rectangular thin film [33].

In order to perform I-V transport measurements, the films were patterned, using standard UV photolithography and wet etching to obtain 1 mm long strips with a width of 20 and 50 μm. The patterned samples were mounted on a sample holder and loaded in a He gas flow cryostat provided with a 12 T superconducting magnet. Critical current values as a function of the applied magnetic field were obtained from the I-V characteristics using 1 μV·cm⁻¹ criterion.

3. Results and Discussion

3.1. Structural Properties and Morphology

Good epitaxial YBCO films were obtained with the addition of both 5% and 10% Gd using both standard and fast-firing conditions. XRD patterns of films obtained with different doping levels are reported in Figure 1. The spectra were characterized by the presence of peaks ascribable to 00ℓ reflections of the YBCO phase and to the substrate, with few less intense features ascribable to the substrate contribution (e.g., broad halos at 2θ = 12° and 32° and K_β reflections) and no clear pieces of evidence of the presence of crystalline secondary phases. Going further into details, from a detailed analysis of the 001 reflections (see inset in the figure), it can be observed that all samples with Gd doping were characterized by the presence of a broad peak at a lower angle with respect to the YBCO reflection. This feature might be ascribed to the presence of Y₂Ba₄Cu₇O₁₅ (Y247) phases or Y₁Ba₂Cu₄O₈ (Y124) (e.g., PCPDF card # 473–410), suggesting, therefore, a higher amount of stacking faults in these samples. However, this feature was observed only on the 001 reflection, suggesting that the stacking faults were widely dispersed, without the formation of large coherent scattering domains of these phases, and thus limiting the order of observable Bragg reflections.

To evaluate differences in the microstructure due to the chemical composition, the c lattice parameter was calculated and is reported in

Table 2. Results of the XRD analysis for the different samples deposited by the standard and fast process, with lattice parameter c, grain size, and microstrain ε calculated along the 00ℓ directions:

Sample	Process	Lattice Parameter c (Å) 1	Grain Size (nm)	ε (%)
YBCO	Standard	11.67 ± 0.01	77 ± 7	0.05 ± 0.01
YBCO-Gd 5%		11.67 ± 0.01	77 ± 3	0.07 ± 0.01
YBCO-Gd 10%		11.67 ± 0.01	70 ± 5	0.07 ± 0.01
YBCO		11.67	100	0.08
YBCO-Gd 5%	Fast	11.68	80	0.06
YBCO-Gd 10%		11.67	80	0.09

¹ For the standard process, the mean and standard deviation of results obtained for several samples prepared in the same laboratory and with the same procedure are provided.

, as well as the grain size and the microstrain along the c axis were evaluated by means of the Williamson–Hall method [35]. It can be observed that all samples were characterized by similar c axis values, without a significant effect due to Gd addition. Regarding the grain size, the values were well in agreement with the thickness of the films, suggesting a coherent growth of the YBCO phase from the substrate interface to the film surface. Few differences could be observed also regarding the microstrain values, with results ranging in the 0.05%–0.1% range. Besides, in this case, a clear trend with the doping amount could not be evidenced, suggesting that stress accumulation due to the lattice mismatch between YBCO and GdBCO phases could be relieved through the stacking faults formation.

The XPS depth profile of a YBCO-Gd 5% sample grown on an STO substrate is shown in Figure 2. The depth profile consisted of seven sputtering cycles for a 70 min total sputtering time. The interface was reached at about 57 min, resulting in a sputtering rate value of about 1.3 nm·min⁻¹. The Gd normalized intensity is shown on the left axis, while the Y, Sr, and Ti intensities are on the right axis. The depth profile showed the Gd intensity increasing with the etched depth, while the Y intensity remained fairly constant up to the interface region. Sr and Ti were also detected all over the thickness, with their intensities increasing in the proximity of the interface region. Diffusion from the STO substrate has been already observed in previous works, and it has been explained as due to the action of the propionic acid used as the solvent [34,36]. Figure 3 shows the Gd/Y distribution along with the −5% sample thickness. The Gd/Y behavior was quite homogeneous in the bulk film, i.e., far from the surface (30 min sputtering time) and up to the interface (57 min sputtering time). The Gd/Y stoichiometric ratio resulted in about 0.06 inside the film and about 0.06 in the total thickness, in good agreement with the nominal Gd 5 mol % content. The stoichiometric behavior of the YBCO-Gd 10% sample resulted similarly to the 5% doped film. In this case, the Gd/Y intensity ratio resulted in about 0.11 in the bulk film and about 0.13 in the total thickness. Figure 4 shows the binding energies of the Gd 3d_5/2 (Figure 4a) and O1s (Figure 4b), measured on both the 5% and 10% YBCO-Gd samples as a function of the film depth. The binding energy values far from the surface (where they were not influenced by contamination) and before the interface (at about 57 min) were fairly constant. The red dot-dashed lines in Figure 4 indicate the binding energy of the Gd₂O₃ sample used as a reference. The corresponding binding energy value measured for the Gd 3d_5/2 and O1s were 1187.0 eV and 530.1 eV, respectively, in agreement with the literature data on Gd₂O₃ oxide [37,38].

As it can be seen, the position of the O1s and Gd 3d_5/2 in Gd₂O₃ were at higher binding energy and well separated from the average binding energy of O1s and Gd 3d_5/2 measured in the YBCO-Gd films. This feature indicated that the electronic environment of Gd in the YBCO-Gd system was not similar to the one of Gd in Gd₂O₃. Preliminary results on the GdBCO film suggested that the nature of the Gd bond in YBCO-Gd was closer to Gd in GdBCO [39]. Even though detailed work is still in progress and defined conclusions cannot be drawn, the scenario resulting from XPS investigations was then more consistent with the presence of a GdBCO phase within the YBCO-Gd films rather than Gd oxide formation.

The highly c-axis oriented grain structure, revealed by XRD (Figure 1 and Table 2), corresponded to dense, smooth, and well-connected film surfaces observed with SEM (Figure 5). The good morphology of YBCO-Gd films, comparable with those of pure YBCO, evidenced that even 10% of Gd introduction did not produce any film degradation. The YBCO-Gd films obtained by using the fast ramp showed as well a similar morphology with respect to pristine YBCO deposited in the same conditions. As can be seen in Figure 5d–f, some porosity was evident in pure YBCO, but it was almost absent in the Gd-doped films. In general, among YBCO-Gd samples, there were no significant differences due to dopants concentration or ramp rate.

Figure 6a shows the HAADF STEM overview of the YBCO-Gd 5% layer on the STO substrate. No precipitate particles or secondary phases ascribable to RE₂O₃ were present, in contrast with the previously reported study by [20]. The vertical brighter/darker contrast corresponded to YBCO twin grains with an average size of 18.5 nm in [100]/[010]YBCO projection (giving ~13 nm in [110]YBCO projection).

Figure 6b,c show the presence of Y124 intergrowths (in other words, ‘stacking faults’) in the YBCO matrix. The edges of these intergrowths produced strain in the YBCO matrix [40]. This strain was visualized by diffraction contrast in ADF STEM images as brighter (Figure 6b) or warmer (Figure 6c) contrast areas. The longer Y124 intergrowths contained Cu/O-deficient areas (see the highlighted area in Figure 6c), which appeared to be 1D ferromagnetic clusters [41]. The visualized features observed in the sample (twins, strains, and clusters in Y124) created a rich pinning landscape.

EDX analysis, Figure S1 and Figure 7, confirmed a uniform distribution of Gd in the film being in a good agreement with XPS studies on a larger area (5 mm × 5 mm). Therefore, TEM and XPS results demonstrated that a compositional homogeneity of the YBCO-Gd films was clearly evidenced at different scales. In addition, this analysis showed that the film was perfectly epitaxial to the STO substrate. The integrated profile over the Gd map confirmed that Gd occupied some of the Y positions in YBCO perovskite structure, in agreement with XPS indications.

Finally, no RE₂O₃ was evidenced by XPS or TEM. A limited number of (Y,Gd)₂Cu₂O₅ nanoparticles were observed on the surface of the film by TEM. Therefore, we can suppose that the stoichiometric balance could be guaranteed by the presence of other phases like YGdCu₂O₅ or Gd₂BaCuO₅, being the latter one the GdBCO decomposition product at high pO₂ conditions [8]. This point is still not completely elucidated, and it will be studied in more detail.

3.2. Superconducting Properties

The J_c behavior by dc measurements of the YBCO-Gd films was investigated in the 10–77 K temperature range as a function of the magnetic field direction and strength. In Figure 8a, the J_c(H) curves of both 5 and 10 mol % Gd samples were plotted. In the same graph, the curves of a typical MOD YBCO films grown in the same process conditions were also shown for a direct comparison. As can be seen, at 77 K, all J_c curves overlapped with some differences only in the mid-field region, approaching the irreversibility field (about 6 T). Reducing the temperature, J_c curves of the YBCO-Gd samples exhibited better in-field behavior than pristine YBCO film in the whole investigated magnetic field region. In

Table 3. Critical current density (J_c) was measured at different temperatures and applied magnetic fields. Magnetic and electric transport analyses were compared¹. Italics is used for I-V data. Films were deposited using the standard process.

Temperature (K)	Applied Field (T)	Jc	YBCO 2 (MA·cm−2)	YBCO-Gd 5% (MA·cm−2)	YBCO-Gd 10% (MA·cm−2)
77	0	VSM I-V	1.80 ± 0.59 2.26 ± 0.93	3.23 4.57	2.71 4.27
	1	VSM I-V	0.10 ± 0.03 0.15 ± 0.04	0.11 0.16	0.10 0.16
	3	VSM I-V	0.02 ± 0.01 0.04 ± 0.01	0.03 0.04	0.01 0.03
65	0	VSM I-V	3.81 ± 1.31 4.37 ± 1.76	6.69 9.22	6.05 8.94
	1	VSM I-V	0.40 ± 0.12 0.55 ± 0.15	0.53 0.77	0.47 0.70
	3	VSM I-V	0.17 ± 0.05 0.25 ± 0.07	0.23 0.35	0.21 0.31
	9	VSM I-V	0.04 ± 0.02 0.08 ± 0.03	0.05 0.08	0.03 0.06
10	0	I-V	13.3 ± 1.2	28.4	27.2
	1	I-V	3.27 ± 0.42	5.47	5.73
	3	I-V	1.91 ± 0.44	2.97	3.04
	9	I-V	1.20 ± 0.32	1.70	1.65

¹ The abbreviation VSM and I-V are used for magnetic and electric transport analyses, respectively. ² YBCO average value represents the mean and standard deviation of results obtained for several YBCO samples prepared in the same laboratory and with the same procedure.

, the J_c values, recorded by magnetic and dc transport measurements, of YBCO-Gd samples, at selected temperatures and magnetic fields, were compared to the corresponding averaged values of YBCO samples deposited with the standard process conditions. This comparison demonstrated that the improvement induced by Gd addition was a more general and reproducible behavior. Moreover, the J_c values, calculated from VSM magnetization data, were in good agreement with transport data, both at 77 and 65 K, in a broad range of fields. Such positive evidence strengthened the reliability of the following discussion.

It is worth noticing that a clear trend with the content of Gd excess in YBCO films did not emerge from these measurements. In fact, both Gd 5 and 10 mol % samples exhibited very subtle differences in J_c values and field dependences. This feature could be more consistently evaluated by investigating samples with a higher Gd doping level, and it will be addressed in future works.

Besides the in-field improvement, a remarkable effect of Gd excess was also revealed in self-field J_c values, J_c(0), that were larger than those of YBCO films at all temperatures. In Figure 8b, the J_c(H) curves measured at 77 and 10 K were plotted in a double logarithmic scale, in order to more easily appreciate the low-field region behaviors. In the same Figure 8b, the J_c(H) curves of the YBCO-Gd processed with fast ramp were also reported. As can be seen, the fast ramp process promoted a similar J_c behavior with a slight increase in the low-field region at both temperatures. At high temperature, J_c(0) = 4–5 MA/cm² was achieved for YBCO-Gd films, corresponding to values more than double with respect to those obtained with the pristine YBCO (see also Table 3 for details). Interestingly, the low-field plateau of J_c, representing the field region of a single vortex pinning regime, was not affected by Gd introduction in YBCO films. The cross-over field from single to collective vortex pinning regimes was usually identified by μ₀H*, determined at J_c = 0.9 J_c(0). The calculated values at 77 K were very close to each other to μ₀H* ≈ 15 mT (

Table 4. Critical temperature (T_c), cross over fields (H*), α parameter of the J_c power-law dependence (J_c(H) = J_c(0) × H^−α), maximum of the pinning force densities (F_p^max), and irreversibility field (H_irr) for YBCO and YBCO-Gd films deposited with the standard process.

Sample	Tc (K)1	H* (mT)	α	Fpmax (GN·m−3)	Hmax (T)	Hirr (T)	H* (mT)	α
		77 K					10 K
YBCO	90.2 ± 0.4	17.7	0.6	2.0	1.6	6.7	39	0.5
YBCO-Gd 5%	90.6 ± 0.2	13.7	0.6	1.8	1.5	6.5	31	0.5
YBCO-Gd 10%	90.3 ± 0.1	14.7	0.7	1.6	1.3	6.4	46	0.6

¹ The mean and standard deviation of results obtained for several samples prepared with the same procedure are reported.

), in line with the typical values reported for MOD YBCO films [42,43]. The similarity in μ₀H* could be explained considering that the introduction of Gd did not significantly alter the nature, i.e., neither the density nor the strength, of the defects acting in the pristine YBCO as vortex pinning centers at high temperatures, such as 77 K. On the other hand, the introduction of Gd promoted a more compact and dense film microstructure, reflected in a better current percolation (increase in J_c(0)). The discrepancies at higher magnetic field values indicated, however, that some minor effects were acting in the collective regime.

At lower temperatures, similar features could be recognized in the low-field behavior. At 10 K, the introduction of Gd led to an increase in J_c(0) to about 30 MA·cm⁻² for both 5 and 10 mol % films (Figure 9a), but, as observed at 77 K, not for μ₀H* (Figure 8b and Table 4). The obtained values in the range 30–50 mT agreed with the reported results for MOD YBCO films at low temperatures. It was observed that the low-temperature value of μ₀H* was directly related to the microstrain generated by the incorporation of nanoparticles in nanocomposite YBCO films. On this basis, the reported μ₀H* values in YBCO-Gd films were consistent with the relatively low effect of Gd addition on the film microstrain (see Table 2) [44].

On the other hand, the improved in-field values of J_c indicated that a more efficient low-temperature vortex pinning was present in YBCO-Gd films. In Figure 9b–d, the J_c data at different temperatures for selected magnetic fields are reported for a better comparison. As shown by microstructural characterizations, Gd addition promoted a rich defect landscape. Gd was uniformly and finely distributed within the YBCO film matrix, and a large number of stacking faults, i.e., Y124/Y247 phases, was generated. Since this was a typical mechanism of stress release, the presence of stacking faults suggested that during the film growth stage, more stress was accumulated, likely for the accommodation of excess of Gd. As reported by TEM, lattice strain was localized at the edges of stacking faults, acting as efficient pinning centers. Moreover, additional defects recognized within the stacking fault, such as Cu/O vacancies, are proposed as effective pinning centers [41]. In agreement with TEM and XPS results, Gd was expected to be incorporated in YBCO by locally substituting Y in the YBCO crystal structure, thus forming unit cells of the superconducting GdBCO dispersed within the YBCO film matrix. Due to the slight difference in lattice parameters, lattice strain should locally emerge at the interfaces between YBCO and GdBCO phases. Since the local oxygen deficiency is an efficient c-axis strain relief mechanism in YBCO films, a strain-induced local oxygen nonstoichiometry could likely emerge in the YBCO lattice close to the YBCO-GdBCO interfaces. This local oxygen nonstoichiometry could produce a point-like vortex pinning center, as already proposed in previous studies on pulsed laser deposited mixed REBCO films [45]. It has to be considered that this kind of defect is particularly efficient in pinning vortices at lower temperature conditions, thus providing an additional isotropic contribution to the increase of the pinning strength at low temperatures.

Angular measurements supported the aforementioned scenario and provided some additional hint on the role of Gd addition on J_c behavior in YBCO films. In Figure 10, the J_c(θ) angular curves measured at 77 and 30 K and μ₀H = 3 T for YBCO-Gd and pristine YBCO samples were plotted. At 77 K, the pristine J_c curve exhibited a different dependence with the magnetic field angle. In fact, J_c values were higher than the J_c measured for Gd samples in a wide angular range between about −45°< θ < 45° (in agreement with the already discussed J_c(H) plotted in Figure 8a), whereas the opposite could be observed in other magnetic field orientations and, in particular, for θ = ±90°. These improvements in YBCO-Gd films suggested that ab-plane intrinsic pinning was particularly strengthened in such samples. This could be ascribed to the increased density of stacking faults recognized by TEM investigations with respect to the pristine YBCO film. In fact, it is well known that effective ab-aligned planar defects often extend their effect on a wider angular range because, due to energetic reasons, it is more convenient for magnetic vortices, within a certain angle spread centered at θ = ±90°, to be accommodated along the defect length [46]. More interestingly, it has to be pointed out that in that region, the 10 mol % Gd film had the highest J_c values. J_c(θ) curves recorded at both 77 and 30 K for YBCO-Gd 10 mol % with faster temperature ramp appeared very similar to the standard ramp sample, indicating that the increase from 10 to 30 °C·min⁻¹ was not enough to promote significant changes in YBCO-Gd films, in agreement with previous analyses.

Looking at the J_c(θ) curves at 30 K, since similar ab-plane peak broadening was observed, it was inferred that this mechanism was less effective at low temperature, likely due to the increased stiffness of vortex lines at lower temperatures. In addition, the J_c(θ) curves measured at 30 K revealed a very similar dependence between Gd-added and pristine YBCO films, with YBCO-Gd samples having J_c values higher in the whole angular range. The observed behavior indicated that defects related to Gd addition had an isotropic character and strengthened their pinning capability as the temperature was reduced. These features were consistent with the proposed local strain accumulation at GdBCO/YBCO interfaces.

The broad bump centered at θ = 0° in the J_c(θ) could be attributed to the twin boundary pinning contribution, which resulted particularly effective when these boundaries, extending over the entire film thickness (see Figure 6a), acted as directional pinning centers, as often recognized in MOD YBCO samples. At 30 K, the broad bump was weakened, as often reported for the twin boundary pinning contribution [47].

It is now worth drawing some remarks, even taking into account the results reported by previous work on a similar system, i.e., the addition of 10 mol % Gd to YBCO film by the sol-gel method [20]. Li et al. observed an increase in both T_c and J_c in the whole field range at 77 K. These authors ascribed the enhancement to the significant presence of RE₂O₃ phases in their films. In our work, instead, we observed a different behavior: T_c was not affected by Gd addition (Table 4), whereas J_c behaviors were more markedly improved at lower temperatures rather than at 77 K. This result, in agreement with the structural and morphological characterizations, evidenced the effect of the different nature of the introduced defects (no RE₂O₃ phases were recognized in our samples), likely due to the different growth methods [20].

In this sense, the improvement of J_c observed in the present study was consistent with what was reported in the literature, relatively to Y replacement by Gd in YBCO-GdBCO systems deposited by MOD [6,16]. Actually, the present XPS and TEM investigations suggested that also in our samples, Y substitution by Gd occurred despite the unfavorable conditions for GdBCO nucleation (pO₂ too high and thermal treatment ramp rates too slow for GdBCO film [8,9]). The findings of our work were in agreement with previous results obtained by Hayashi et al. on (Y,Pr)BCO films [48]. The authors of that work assumed that both YBCO and PrBCO formed, although PrBCO nucleation requires a pO₂ at least 1000 times lower than that necessary for YBCO formation. This inconsistency can be explained considering that since the YBCO growth temperature is lower than GdBCO one, the YBCO lattice starts to nucleate first and then acts as a suitable template for the successive GdBCO nucleation. This feature deserves more attention, and more detailed studies are in progress based on thermodynamic and kinetic approaches [49].

4. Conclusions

Highly epitaxial Gd-doped YBCO films were deposited through the process optimized for pure YBCO. A uniform Gd distribution, revealed by both XPS and TEM, corresponded to improved superconducting properties with respect to pristine YBCO. It seemed to be confirmed that the presence of different defects, originated from Y and Gd competition during REBCO phase formation, had a beneficial effect on transport properties, particularly significant at low temperatures. These results suggested that homogeneously dispersed nano-sized defects could be originated even by an in situ YBCO with Gd excess approach, avoiding the complex use of nanoparticles required by the ex-situ route.

Further investigations will focus on the tuning of Gd and Y phase formation through experimental condition optimization. Higher dopant concentrations will be also evaluated.