The role of the substrate temperature on superconducting properties of

Thin Solid Films 611 (2016) 33–38

Contents lists available at ScienceDirect

Thin Solid Films journal homepage: www.elsevier.com/locate/tsf

The role of the substrate temperature on superconducting properties of sputtered Nb films T.C. de Freitas, J.L. Gonzalez, V.P. Nascimento, E.C. Passamani ⁎ Departamento de Física, Universidade Federal do Espírito Santo, Vitória, 29075-910, ES, Brazil

a r t i c l e

i n f o

Article history: Received 10 December 2015 Received in revised form 4 May 2016 Accepted 6 May 2016 Available online 11 May 2016 Keywords: Superconductivity Nb films Critical current Microstructure

a b s t r a c t The influence of the substrate temperature TS on the superconducting properties of 100 nm sputtered Nb films, prepared directly onto orientated Si (100) substrate, was systematically investigated by measuring their structural, morphological and magnetic properties. Within the TS interval 293–373 K no significant change is observed either in the Nb lattice parameter or in the superconducting transition temperature TC. For TS N 373 K, a degradation of the Nb superconducting properties was observed concomitantly with an increase of the Nb lattice parameter. This effect was attributed to an interdiffusion at the Si/Nb interface and/or an enhancement of the internal stress caused by high temperature deposition. The temperature dependence of the critical current density JC(T) was estimated from magnetization measurements and its behavior is explained based on the granular morphology of Nb films. This work provides some insights on the optimum TS responsible for the highest JC value. It also brings information on how TS affect the superconducting properties of Nb films sputtered directly on Si (100) by DC magnetron sputtering. © 2016 Elsevier B.V. All rights reserved.

1. Introduction The production of niobium (Nb) superconducting thin films is important both from the fundamental viewpoint and for the development of superconducting-based technological devices, i.e., nano-squids [1]; cryogenic circuits for switches operating at high frequency [2]; single-photon detectors [3] and so on. Furthermore, the development of these superconducting systems allows studying different superconducting properties, which are crucial for a better understanding of the entire superconducting phenomenon [4–8]. As an example, it has recently been reported that many hetero-structures are built with the niobium being the active superconducting element [9–12]. In this sense, these hybrid systems provide a unique opportunity to explore interesting phenomena in nanoscale regime; for instance, the correlation between superconductivity and ferromagnetism. Nb films, with superconducting properties, can be obtained with different preparation procedures [10–21]. According to results reported in the literature [8,10–21], it can be inferred that sputtering seems to be the most suitable method to produce superconducting Nb films, i.e., a nanostructured material with optimal critical transition temperature TC and other appropriate physical properties. However, during a sputtering process, there are several deposition parameters that may determine the final physical properties of the sputtered films. The most relevant deposition parameters are, in general, the substrate material, the substrate temperature TS, the residual chamber pressure ⁎ Corresponding author. E-mail address: [email protected] (E.C. Passamani).

http://dx.doi.org/10.1016/j.tsf.2016.05.012 0040-6090/© 2016 Elsevier B.V. All rights reserved.

(the value and its constituent) and, obviously, the deposition rate. As a general rule, finding out the optimal parameters is a challenging task, since they will depend on the particularities of the experimental setup used in the deposition process. The necessary knowhow to obtain good Nb films comes from a robust and clear comprehension about the correlation between superconducting (TC, critical current density JC, etc.) and structural (lattice parameters, grain sizes, roughness, etc.) properties. It has been shown that the TC value of sputtered Nb films does not only depend on the lattice parameter [17,19] and the grain size [17, 22], but it can also be modified by the material used as substrate [17, 23]. Moreover, the TC value can either decrease when the Nb film is annealed ex-situ to temperatures close to 573 K [16] or it may enhance when hydrostatic pressure is applied [13]. It has been reported that Nb films deposited on Si substrate show a larger lattice parameter and reduction of both, thickness and grain size [20]. These two effects were discussed in terms of the O2 content at Nb grain boundaries [20]. In thin films, the TC value can also be affected by other mechanisms as, for example, the proximity effect [8,15,18]. The references cited above show that the deposition by DC sputtering allows to prepare Nb films with different structural parameters, where the superconducting properties can clearly be correlated with the structural disorder effect [15]. Optimal superconducting properties have been reported for 50 nm thick Nb films grown on sapphire substrates [15]. In spite of the efforts devoted to understand this problem, some particular questions are still open, deserving a particular attention. One point that has not extensively explored in the literature is how the substrate temperature TS can modified the superconducting

34

T.C. de Freitas et al. / Thin Solid Films 611 (2016) 33–38

properties of sputtered Nb films grown directly on the Si substrates [17]. So, in the present work, it was done a systematic study to describe the influence of TS on the quality of superconducting properties of Nb films prepared by DC sputtering onto Si (100) substrates. To reach this goal, several depositions were performed at different TS values, while the other parameters [deposition rate, gas pressure, residual gas atmosphere (measured and “controlled” by residual gas analyzer RGA), Si substrate wafer, etc.] were fixed during each process. Our results clearly demonstrate a dependence of the superconducting properties of Nb films on their morphologies; the latter governed by the TS value. In addition, the results are reproducible, since we have prepared three similar batches of samples and they have shown quite similar TC and JC values. For example, the TC value changes roughly 0.1 K in the reproducibly tests. 2. Experimental 100 nm thick Nb films were prepared onto Si (100) substrates, kept at different TS, using the AJA Orion-8 DC magnetron sputtering setup. The TS values were measured using a K-type thermocouple sensor coupled to the rotating sample holder. TS was stabilized (during 5 min) before starting the Nb film depositions. Thus, we can assume that the film is in thermal equilibrium with the sample holder (film temperature TF = TS). A 5 in. silicon wafer was cut in squared shapes, with typical areas covering about 10−2 × 10− 2 m2. These Si pieces were preliminary cleaned in a bath with Extran neutral detergent by an ultrasonic machine. Subsequently, they were washed in consecutives and repeated baths of water and acetone. Finally, they were dried with a flux of inert gas. 2 in. Nb disk, with 99.995% purity, was used as the target material. The residual chamber pressure was better than 5 × 10−6 Pa before filling the deposition chamber with 3 × 10− 1 Pa argon gas (99.999% purity). The residual gas atmosphere was measured by a Residual Gas Analyzer (RGA) device. The film thickness was calibrated by measuring X-ray reflectivity of a single film. The deposition rate of about 0.72 Å/s (maximum rate of Nb obtained with our power supply), obtained by X-ray, was monitored using the quartz crystal balance, which is positioned close to the sample holder position. A pre-sputtering process was performed before the film depositions in order to remove 10 nm from the Nb target and reducing possible natural oxide layers during the deposition. The 100 nm films were prepared at TS equal to 293, 323, 373, 393 and 423 K. No buffer and capping layers of other materials were used, which means that the Nb films were directly deposited onto the Si (100) substrate. In principle, since the sample characterizations were done ex-situ, a natural oxide Nb phase is formed before the sample measurements. This natural Nb oxide seems to passivate the metallic core Nb and keeps the properties of the samples for a long time interval. Measurements performed two weeks after the sample preparation, keeping the film in vacuum setup, show similar physical features and absence of aging effect (no significant change in the TC value − ΔTC ~ 0.1 K). The grazing incidence X-ray diffraction (GIXRD) patterns were recorded, using a Cu Kα-radiation (λ = 1.5418 Å), in an X-ray diffractometer, Rigaku Ultima IV, where the incident angle was kept fix (the sample holder is fix with αi = 1°), while the detector scanned the film surface in steps of 0.05° and count rate of 3 s/pass. The GIXRD patterns were fitted by the Maud software [24], which already considers the small 2θ displacements that occur due to the refraction effects [25]. The relevant extracted parameters were: the lattice parameter (a) and the crystalline grain size (τ). The morphology of the films surface was studied using an atomic force microscopy (SPM 9600 Shimadzu). The data were acquired in a non-contact method, where parameters like the root-mean-square roughness (σ), the lateral correlation length (ξ), the fractal dimension (δ) of the surface and the effective area A of the particles, were measured on typical areas of 1 × 1 μm2. The magnetic properties of the Nb films were obtained by using a Physical Property Measurement System (PPMS) from Quantum Design

that operates with an evercool type-II system. The critical temperature TC of the Nb superconducting films was obtained from the onset of the diamagnetic signal in the M(T) curve recorded for a probe field of 3979 A/m. Magnetic loops [M(H)] were recorded with the scan field applied perpendicular to the sample plane for different temperatures, after cooling the sample from T N TC in zero-field mode (ZFC), for each M(H) case. 3. Results and discussions Room temperature GIXRD patterns of the Nb films, prepared at different TS, are plotted in Fig. 1. The most relevant peaks (those clearly observable in the figure) of these GIXRD profiles were indexed, using the Maud software [24], to the body centered cubic (bcc) structure of bulk Nb (space group: I m–3 m, 229). Furthermore, three low intense broad Bragg lines [100 times lower than the most intense peak, (110)] are peaked between 2θ = 40° and 52°. The intensity is too low that the peaks are diluted in the background, being not detected in two of the five films investigated here (investigations done in the same samples after 14 days still show similar features). These additional peaks are related to the thin natural Nb oxide formed at the film surface during ex-situ experiments. The very low intensities of these lines (relatively to the main peaks of the metallic Nb) inform that only few monolayers were naturally oxidized and according to above results, this process was enough to passivate and protect the metallic core of Nb films against deeper oxidation of Nb layer. Another aspect to be noted is the fact that in the GIXRD method (using a fix sample holder), the diffraction lines [different (h k l)] can only appear if the film is not textured and the crystallographic grains have random orientations, like in a polycrystalline powder [26]. In other words, from the structural viewpoint, all these Nb films are polycrystalline, have a bcc structure and keep their pure metal properties, since it is previewed that the small amount of oxide does not influence the magnetization results, as will be shown later. Table 1 and Fig. 2 display the dependence of the a- and τ-parameters on the TS; values that were obtained by fitting the GIXRD patterns with the Maud software [24]. The uncertainties were calculated, from the fitting procedure, to be equal to 0.002 Å and 2 nm for the a- and τ-quantities, respectively. As can be noted, for TS ≤ 373 K, the a-quantity remains nearly constant. Further increase of TS markedly shift the lattice parameter of the film to a value well above the one found in the bulk phase (3.303 Å) [17], and according to the literature, this would lead to a considerable decrease of the TC value [17,19]. This increase of the a-parameter for the films prepared at TS N 373 K may be attributed to two main effects: Si interdiffusion to Nb layer and/or an enhancement of the internal stress caused by high temperature deposition. Thus, these two effects would degrade the superconducting properties. The AFM images of the Nb films are shown in Fig. 3. The σ [1.4(2)– 1.6(2) nm] and δ [2.03(2)–2.05(2)] parameters are roughly constant for all films, when their uncertainties are considered. The δ parameter describes (by a single parameter) a scale-independent measure of the surface roughness (for lengths lower than ξ), i.e., it represents the roughness irregularities taken in many scales of length (bξ). Here, the value around 2 for the δ quantity informs that the film roughness is relatively smooth (bξ), pictured as sand dunes. On the other hand, it should be mentioned that the ξ quantity is a statistical parameter that gives a measure of the roughness lateral length scale (it can be roughly understood as an average of roughness wavelengths taken over all directions). Therefore, the ξ parameter works as an effective cut-off length, where the height correlations vanish, i.e., if the distance between two points is within of the ξ value, their heights are correlated and the roughness can be described by the δ parameter [27,28]. Thus, our results suggest that the ξ parameter varies from 2(1) nm (TS = 293 K, 323 K) to 3(1) nm (TS = 373 K, 393 K), which indicates that the roughness oscillations have predominantly short wavelengths. Consequently, it can be said that the films roughness is similar and it is not dependent


35

Fig. 1. X-ray diffraction patterns of the sputtered Nb films deposited at different substrate temperature TS.

on TS, at least, in the range used in this work. In this sense, it should be noted that the argon pressure seems to be one of the most important parameter that determines the surface roughness of the sputtered Nb films [29]. This finding (the surface roughness of the films is not dependent on the TS value) is also in agreement with the fact that a similar pressure of 3 × 10− 1 Pa was used in the preparation of the studied films. Thus, the AFM images show surfaces characteristic of polycrystalline particulate films, corroborating with the GIXRD results. For TS N 323 K, AFM data suggest that there is an in-plane particle size

increase from 42 to 58 nm (Table 1), which is slightly larger than the coherence length of bulk Nb; a value of about 38 nm [30]. Also, the 2D effective area S (defined as the in-plane area of the particles times the

Table 1 Morphological and structural parameters obtained from X-ray and AFM measurements of the Nb films. Substrate temperature TS (K)

Lattice parameter (a) (Å)

Grain size (τ) DRX (nm)

Effective surface S (nm2)

Grain size (D) AFM (nm)

293 323 373 393 423

3.303 3.303 3.307 3.317 3.347

13.6 13.3 12.4 12.5 12.7

2023 1503 2709 3209 –

42 38 48 56 –

Fig. 2. The behaviors of the lattice parameter a and the grain size τ as a function of the substrate temperature TS for the Nb films.

36


Fig. 3. 2D AFM images of Nb films prepared at different TS: (a) 293 K, (b) 323 K, (c) 373 K, (d) 393 K. At the right of each image, the z-profile of the film surface is plotted.

number of particles) abruptly enhances (~60%) when TS value increases to higher temperatures. In this sense, Thornton et al. have reported that the sputtered film microstructures are strongly determined by external parameters, like as: working pressure and TS [31]. In addition, the microstructure is strongly dependent on stress effects, which, in our case, can be caused by two sources: the intrinsic stress due to the lattice mismatch between the film/substrate interface and the thermal stress originated from the TS and measuring temperature T [32]. Considering that we used 3 × 10−1 Pa for film preparations, the only source, according to Thornton et al. [31], that affects the film morphology should be the TS quantity. As they have also demonstrated, an enhancement of the TS value provokes more compacted films and less granular (more flat layer); as we experimentally observed if one takes the data plotted in Fig. 3. Fig. 4 displays the dependence of TC of Nb thin films on TS. As can be noted, the TC value is roughly constant for TS b 373 K, but it abruptly decreases for higher TS value (the film prepared at 423 K does not show a superconducting transition down to 2 K). Thus, considering that the lattice parameter substantially increases for TS N 373 K and that TC significantly reduces, it seems that TC scales inversely with the a-parameter and the reduction of TC can be assumed to be due to an enhancement of the internal stress in Nb layers, as also reported in the literature [17,19]. Fig. 2 also shows that the crystalline grain size (according to XRD data) remains mostly constant, consequently would not influence on the variation of TC. However, it should be, once more, stressed that there are two possible main sources for the TC vs TS behavior: (i) the internal stress caused by high substrate deposition temperature of the Nb layers and/or (ii) the Si interdiffusion effect, which is dominated for TS N 373 K.

ZFC M(H) loops obtained at different temperatures (T ≤ TC) were recorded for all samples. A typical set of M(H) is plotted in Fig. 5 for the Nb film prepared at TS = 293 K. In general, these M(H) loops reproduce the typical diamagnetic response expected for type-II superconductors. Thus, considering the Bean's model [33], intergranular critical current densities JC were estimated from the M(H) data. This model suggests a

Fig. 4. Behavior of the critical superconducting temperature TC as a function of the substrate temperature TS for the 100 nm Nb films. Inset figure shows the M(T) curves of the Nb film prepared at TS = 323 K (the onset temperature is indicated by vertical dashed line or arrow).


37

T/TC)n. Fig. 6 inset displays that the n-exponents are always around the value of 2. This value signalizes that the non-superconducting barriers, between the grains, are normal metals. The JC(0) values, obtained from the fit (also shown in Fig. 6 inset), can be interpreted according to the superconducting-normal-superconducting (S-N-S) granular theory (proximity-effect junctions) for dirty superconductors, and therefore JC(0) ∝l/L [37]. Here, l is the mean free path of the electrons in the normal metal and L is the length of the junction [37]. It should be noted that JC(0) increases with TS and, in principle, this behavior could be associated with an increase of the l/L-parameter. In other words, a systematic increase of TS reinforces the metallic nature of the intergranular regions during the sputtering process, since the l/L parameter also increases. 4. Conclusions

Fig. 5. M(H) loops measured at different temperatures for the Nb film prepared at TS = 293 K (the vertical dashed line corresponds to the field (0.1 T) used to calculate JC according to Bean's model). Inset figure shows the entire scan field range that the M(H) loops were obtained.

relation between JC and a magnetization variation ΔM at a specific field given by JC = ΔM/D, where the ΔM quantity, given in A/m, was taken as the difference between the magnetization curves for decreasing and increasing branches at μoH = 0.1 T (vertical dashed line in Fig. 5) and the D quantity was assumed to be the width of the films (D) in millimeter. The JC values are therefore reported in A/m2. Following the ΔM criterion defined by Bean's model, the temperature dependence of the JC quantity was calculated for all superconducting films and the results are displayed in Fig. 6. It should be first said that: (i) the calculated JC values are in the same order of magnitude to those reported in the literature for granular superconducting materials [34]; (ii) the JC (T) decreases as the temperature approaches TC and (iii) the highest JC value is observed for the film prepared at TS = 373 K (above this temperature, internal stress and/or interface interdiffusion degrade the superconducting properties of the Nb films). On the other hand, given the granular nature of the Nb films studied herein (see AFM images in Fig. 3), a granular model [35] was used to analyze the JC data. Following the procedure describe in the literature [36], the calculated JC(T) data of Nb films, with the nonzero TC value, were fitted with the following relation JC(T) = JC(0)(1 −

Fig. 6. Critical current densities JC versus temperature for the Nb films. Inset figures displays data that were fitted with JC = JC(0) (1 − t)n, being n the fitting parameter and t = T / TC (reduced temperature).

Polycrystalline Nb films, sputtered onto single crystal Si (100) substrates, have shown in-plane average particle sizes independent of the TS value. In general, the cubic lattice parameter (a) seems to scale with the superconducting transition temperature TC, being roughly invariant for films prepared with TS ≤ 373 K. Further increase of TS has reduced TC due to an increase of the a-parameter, which could be explained assuming either due to a Si/Nb interface interdiffusion effect or an enhancement of the internal stress. Finite-size effect added to interface interdiffusion can be responsible for the relatively smaller TC value (~ 6.5 K) of the Nb films when compared with the bulk Nb (9.2 K). Critical current density JC measurements suggest a metallic character of the intergranular regions. This metallic behavior of the weaklinks seems to be reinforced as the TS value is increased. Finally, it has to be noted that we have prepared, at least, three sets of similar films in different times and the results are reproducible (the TC value of similar samples changes roughly 0.1 K). Also, the films have been kept with similar structural and physical properties during, at least, 14 days. In brief, our study demonstrates that the substrate temperature TS plays a crucial role on the structural, morphologic and superconducting properties of Nb thin films deposited directly onto Si (100) substrates. Acknowledgments This work was supported by the Brazilian agencies CNPq, FAPES and CAPES. References [1] X. Liu, X. Liu, H. Wang, L. Chen, Z. Wang, The fabrication and characterization of nano-SQUIDs based on Nb thin films, Phys. C 515 (2015) 36. [2] N. Alcheikh, P. Xavier, J.M. Duchamp, K.F. Schuster, C. Malhaire, B. Remaki, C. Boucher, X. Mescot, Temperature dependence of the electromechanical characteristics of superconducting RF-MEMS switches, Microsyst. Technol. 21 (2015) 301. [3] A.J. Annunziata, D.F. Santavicca, J.D. Chudow, L. Frunzio, M.J. Rooks, A. Frydman, D.E. Prober, Niobium superconducting nanowire single-photon detectors, IEEE Trans. Appl. Supercond. 19 (2009) 327. [4] D. Carmo, F. Colauto, A.M.H. De Andrade, A.A.M. Oliveira, W.A. Ortiz, T.H. Johansen, Trapping flux avalanches in Nb films by circular stop-holes of different size, IEEE Trans. Appl. Supercond. 25 (2015) 7501004. [5] O. Crauste, F. Couëdo, L. Bergé, C.A. Marrache-Kikuchi, L. Dumoulin, Destruction of superconductivity in disordered materials: a dimensional crossover, Phys. Rev. B 90 (2014) 060203. [6] D. Bothner, R. Seidl, V.R. Misko, R. Kleiner, D. Koelle, M. Kemmler, Unusual commensurability effects in quasiperiodic pinning arrays induced by local inhomogeneities of the pinning site density, Supercond. Sci. Technol. 27 (2014) 065002. [7] T. Matsuda, K. Harada, H. Kasai, O. Kamimura, A. Tonomura, Observation of dynamic interaction of vortices with pinning centers by Lorentz microscopy, Science 271 (1996) 1393. [8] A.I. Gubin, K.S. in, S.A. Vitusevich, M. Siegel, N. Klein, Dependence of magnetic penetration depth on the thickness of superconducting Nb thin films, Phys. Rev. B 72 (2005) 064503. [9] V. Zdravkov, J. Kehrle, G. Obermeier, D. Lenk, H.A. Krug von Nidda, C. Müller, M. Kupriyanov, A. Sidorenko, S. Horn, R. Tidecks, L. Tagirov, Experimental observation of the triplet spin-valve effect in a superconductor-ferromagnet heterostructure, Phys. Rev. B 87 (2013) 144507. [10] O.V. Dobrovolskiy, M. Huth, V.A. Shklovskij, Anisotropic magnetoresistive response in thin Nb films decorated by an array of Co stripes, Supercond. Sci. Technol. 23 (2010) 125014.

38


[11] S.F. Lee, Y. Liou, Y.D. Yao, W.T. Shih, C. Yu, Magnetic flux penetration depth study in Nb/Co system, J. Appl. Phys. 87 (2000) 5564. [12] M. Velez, C. Martınez, A. Cebollada, F. Briones, J. Vicent, Magnetic proximity effect in Fe/Pt/Nb multilayers, J. Magn. Magn. Mater. 240 (2002) 580. [13] G. Pristás, S. Gabáni, E. Gazo, V. Komanický, M. Orendác, H. You, Influence of hydrostatic pressure on superconducting properties of niobium thin film, Thin Solid Films 556 (2014) 470. [14] M. David Henry, S. Wolfley, T. Monson, B.G. Clark, E. Shaner, R. Jarecki, Stress dependent oxidation of sputtered niobium and effects on superconductivity, J. Appl. Phys. 115 (2014) 083903. [15] O.V. Dobrovolskiy, M. Huth, Crossover from dirty to clean superconducting limit in dc magnetron-sputtered thin Nb films, Thin Solid Films 520 (2012) 5985. [16] Y.W. Kim, S.G. Lee, J.H. Choi, Ex situ annealing effect on Nb thin films prepared by DC magnetron sputtering, Phys. C 471 (2011) 1193. [17] J. Liu, J. Li, T. Li, T. Li, W. Wu, W. Chen, Study of stress and morphology of superconducting niobium thin films, IEEE Trans. Appl. Supercond. 19 (2009) 245. [18] J.H. Kim, Y.W. Kahng, Y.H. Choi, Critical properties of Submicrometer-patterned Nb thin film, IEEE Trans. Appl. Supercond. 19 (2009) 2649. [19] D. Hazra, M. Mondal, A.K. Gupta, Phys. C 469 (2009) 268. [20] D. Hazra, S. Datta, M. Mondal, J. Ghatak, P.V. Satyam, A.K. Gupta, J. Appl. Phys. 103 (2008) 103535. [21] M. Meepagala, S.J.T. Chen, L.E. Wenger, Thickness dependence on the superconducting properties of thin Nb films, Phys. Rev. B 49 (1994) 15235. [22] S. Bose, R. Banerjee, A. Genc, P. Raychaudhuri, H.L. Fraser, P. Ayyub, Size induced metal–insulator transition in nanostructured niobium thin films: intra-granular and inter-granular contributions, J. Phys. Condens. Matter 18 (2006) 4553. [23] U.D. Chacón Hernandez, M.A. Sousa, F.J. Litterst, V.P. Nascimento, E. BaggioSaitovitch, Effects of Nb buffer layer on superconducting and magnetic behavior of IrMn/NiFe/Nb/NiFe spin-valves, J. Magn. Magn. Mater. 390 (2015) 114. [24] M. Ferrari, L. Lutterotti, Method for the simultaneous determination of anisotropic residual stresses and texture by X-ray diffraction, J. Appl. Phys. 76 (1994) 7246.

[25] E.I. Alessandrini, Morphology of thin superconducting Nb films, J. Vac. Sci. Technol. 18 (1981) 318. [26] B.D. Cullity, S.R. Stock, Elements of X-ray Diffraction, third ed. Prentice Hall Inc., Upper Saddle River, New Jersey, 2001. [27] S.K. Sinha, E.B. Sirota, S. Garoff, X-ray and neutron scattering from rough surfaces, Phys. Rev. B 38 (1988) 2297. [28] V. Holy, J. Kubena, I. Ohlidal, K. Lischka, W. Plotz, X-ray reflection from rough layered systems, Phys. Rev. B 47 (1993) 15896. [29] J. Du, A.D. Charles, K.D. Petersson, Study of the surface morphology of Nb films and the microstructure of Nb/AlOx-Al/Nb trilayers, IEEE Trans. Appl. Supercond. 17 (2007) 3520. [30] G. Deutscher, New Superconductors: From Granular to High Tc, 1o ed. World Scientific Publishing, 5 Toh Tuck Link, Singapore, 2006. [31] J.A. Thornton, D.W. Hoffman, Stress-related effects in thin films, Thin Solid Films 171 (1989) 5. [32] T.P. Bertelli, E.C. Passamani, C. Larica, V.P. Nascimento, A.Y. Takeuchi, M.S. Pessoa, Ferromagnetic properties of fcc Gd thin films, J. Appl. Phys. 117 (2015) 203904. [33] C.P. Bean, Magnetization of hard superconductors, Phys. Rev. Lett. 8 (1962) 250. [34] H. Wang, H. Chang, C. Wang, Study on the critical state of high-Tc superconducting junctions, Physica C 242 (1995) 135. [35] V.Y. Irkhin, M.I. Katsnelson, On the reconstruction of the conduction electron spectrum in metal-oxide superconductors owing to the intention with coherent atomic displacements, Phys. C 160 (1989) 397. [36] G.M. Stollman, M.A.M. Gijs, J.W.C. De Vries, Analysis of the critical current density in high-Tc superconducting films, Physica C 157 (1989) 406. [37] M.I. Petrov, D.A. Balaev, D.M. Gokhfel'd, K.a. Sha, S.V. Ospishchev, K.S. Aleksandrov, Crossover from the “clean” limit to the “dirty” limit in a network of S–N–S weak links in Y3/4Lu1/4Ba2Cu3O7 + BaPb1 − xSnxO3 (0 ≤ x ≤ 0.25) composites, Phys. Solid State 46 (2004) 1792.