Brought to you by:
Paper

A water-free metal organic deposition method for YBa2Cu3O7−δ thin film fabrication

, , , , , , , , , and

Published 14 October 2013 © 2013 IOP Publishing Ltd
, , Citation Rongxia Huang et al 2013 Supercond. Sci. Technol. 26 115010 DOI 10.1088/0953-2048/26/11/115010

0953-2048/26/11/115010

Abstract

During the fabrication of YBa2Cu3O7−δ (YBCO) thin films via the metal organic deposition (MOD) method, the water content of the precursor solution should be reduced to prevent precipitation and other detriment. In most of the reported MOD-YBCO studies, water was used as a solvent, and vacuum distillation was utilized to reduce the water content of the precursor solution. In this study, we propose a water-free MOD method using propionic anhydride and methanol as the solvents. The crystal water of raw precursor salts was removed by heating, and the water content in the final precursor solution was as low as 0.2 wt%. Using this solution with low fluorine content and the optimized maximum temperature of pyrolysis, the final YBCO thin films about 250 nm thick had the critical current density (Jc) of 3.8 MA cm−2 (77 K, self-field). The water content of the precursor solution was further reduced to less than 10−2 wt% by vacuum distillation. According to the Jc performance, thermal analysis, x-ray diffraction investigations, and morphology observations, no significant improvement was found for the final YBCO thin films fabricated by the distilled solution, indicating that vacuum distillation is not necessary in the water-free MOD method.

Export citation and abstract BibTeX RIS

1. Introduction

Owing to the excellent superconducting properties and promising application prospects, YBa2Cu3O7−δ (YBCO) coated conductor fabrication has become a focused area of research [13]. Fabrication with low cost is required for large-scale applications of YBCO coated conductors. Therefore chemical solution deposition (CSD), which requires no vacuum apparatus and permits a high deposition rate [4, 5], is attracting great interest worldwide. Among the CSD techniques, the trifluoroacetate metal organic deposition (TFA-MOD) method is very popular, since this approach can avoid the formation of BaCO3 and produce high-performance YBCO films [68].

During preparation of the precursor solution, it is important to reduce the water content. Although the microscopic influences of water molecules on the drying and decomposition process still remain unclear [3], it has been noticed that when the precursor solution contains too much water the formation of uniform gel film could be hindered [7], and the superconducting properties of the final YBCO thin films would be deteriorated [9]. However, water is generally used as a solvent in the preparation of conventional MOD solution. To remove water molecules from the solution is difficult because of the hydrogen bonds between fluorine in trifluoroacetic acid and hydrogen in water [9]. Thus the precursor solution needs to be refined several times by vacuum distillation to reduce the water content. However, the refining process using vacuum distillation would increase the production cost, because it is time consuming and needs a corresponding device. In this study, we try to abandon the addition of water as a solvent, and eliminate the vacuum distillation process.

Besides, in the conventional TFA-MOD process, a major drawback is the long duration of the pyrolysis step (usually 10–16 h) [9]. Such a long pyrolysis step is designed because strong internal stresses accompany the decomposition reactions, and buckling or cracking might occur such that the final YBCO films are degraded [10, 11]. In order to shorten the pyrolysis time, various improvements of the precursor solution preparation have been proposed in the past decade, such as lowering the fluorine content [1214], using organic additives [1517], and using trifluoroacetic anhydride (TFAA) [18]. Low-fluorine solution has been proven efficient by many groups, because the generation of fluorine-related gases could be reduced and the stress could be alleviated [12]. In this study, we also prepare precursor solutions with low fluorine content to shorten the pyrolysis time.

In this work, we propose a water-free MOD method, in which, instead of water, propionic anhydride and methanol are used as solvents. Therefore, the main source of the water content in the solution is the crystal water of raw precursor salts. By directly heating the raw salts in advance of dissolution, the water molecules could be removed before they participate in the formation of hydrogen bonds in the solution. A low-fluorine precursor solution was prepared without vacuum distillation. We demonstrate that high-performance YBCO thin films can be fabricated from this solution with a short pyrolysis step. Thermal analysis of the precursor solution, as well as the microstructure and superconducting properties of the YBCO films are discussed in detail. The influence of the maximum temperature of pyrolysis is investigated.

2. Experimental details

Conventional TFA-MOD precursor solution was prepared by the route described by Araki et al [9]. The low-fluorine precursor solution proposed in this study was prepared starting from Y(CH3COO)3⋅4H2O,Ba(CF3COO)2⋅1.91H2O and Cu(CH3COO)2, corresponding to the 1:2:3 stoichiometry. The crystal water of Y(CH3COO)3⋅4H2O and Ba(CF3COO)2⋅1.91H2O was removed by heating at 150 °C for 30 min, then the remaining and Cu(CH3COO)2 were mixed in the proper amounts of methanol and propionic anhydride. Full dissolution of the mixture could be achieved by magnetic stirring and heating at 70 °C to form a uniform and transparent solution. The molar concentration of metal ions in the precursor solution was adjusted to 1.5 mol l−1 by adding methanol. The utilization of propionic anhydride as a solvent in this study was different from the acid solvents in the conventional TFA-MOD studies, thus this solution was named PAn. To investigate whether or not further reducing the water content of PAn could improve the properties of final YBCO films, PAn was refined through vacuum distillation and re-dissolved in methanol three times. The distilled solution, named PAnD, also had a metal ion molar concentration of 1.5 mol l−1. The preparation of precursor solutions was carried out in an ultra-clean chamber, where the atmosphere was dry. The water contents of PAn and PAnD evaluated by the Karl Fischer method were 0.2 wt% and below 10−2 wt%, respectively.

Lanthanum aluminate (LAO) single crystal substrates with (001) orientation were annealed at 670 °C for 10 h in a dry oxygen atmosphere. PAn and PAnD were spin-coated on the LAO substrates at a rotation rate of 6000 rpm for 60 s. The spin coating was conducted in a sealed glove box, into which pure nitrogen continually flowed to control the relative humidity below 50%. Following this, the heat treatment of the coated films consist of three steps: pyrolysis, crystallization and oxygenation. The heating profiles are shown in figure 1. In the pyrolysis step, the samples were decomposed by heating to the maximum temperature of pyrolysis (MTP) of 400, 500 or 600 °C with a heating rate of 10 °C min−1 in an atmosphere of O2. When the furnace temperature was higher than 110 °C, the gas flowed through water (25 °C) to become humid prior to following into the furnace. In the crystallization step, the samples were treated using a heating rate of 50 °C min−1 up to 800 °C with a dwell time of 60 min in a humid 100 ppm O2/N2 atmosphere, and a further dwell time at 800 °C of 10 min in a dry 100 ppm O2/N2 atmosphere. In the oxygenation step, the samples were cooled to 525 °C in a dry 100 ppm O2/N2 atmosphere, then further cooled in a dry oxygen atmosphere. Using the fabrication procedure described above, the thickness values of all the YBCO films by PAn and PAnD were about 250 nm.

Figure 1.

Figure 1. The heat treatment profiles with different maximum temperatures of pyrolysis (MTPs) for YBCO superconducting thin film fabrication using PAn and PAnD.

Standard image High-resolution image

X-ray diffraction (XRD) patterns of the pyrolyzed films and the final YBCO films were obtained using a Rigaku D/MAX-2500 diffractometer with a Cu Kα1 source (λ = 0.154 06 nm). Microscopic morphology of the sample surface was examined by scanning electron microscopy (SEM) using LEO-1530. SEM was also used to measure the thickness of YBCO thin films by observing their cross sections. Differential thermal analysis (DTA) and thermogravimetric analysis (TGA) of the gel powder dried from the precursor solutions were measured with a NETZSCH STA 409 PC/PG thermal analyzer. The critical temperature (Tc) of the YBCO films was measured using the four-probe method with the current fixed at 1 mA. The magnetic field dependence of the critical current density (Jc) of the YBCO films at 77 K was measured by a magnetic property measurement system (MPMS) XL-7 of Quantum Design. Jc in self-field at 77 K was measured by a home-made AC susceptometer, which consisted of a lock-in amplifier, a power amplifier and four Helmholtz coils. Two of the Helmholtz coils served as magnetizers; the other two, the measuring and compensating coils, were mounted concentric with the sample [19]. The AC susceptibility (χ = χ' + jχ'') of the YBCO films as a function of the magnetic field amplitude (Hm) could be measured. When the imaginary part (χ'') reached its maximum, the field amplitude was denoted as ${H}_{\mathrm{m}}({\chi }_{\mathrm{m}}^{\prime\prime})$. In this study, the width and length of YBCO films were approximately equal (about 5 mm), and the thickness (d) was measured by SEM cross sectional observation. Thus Jc could be calculated through equation (1), which was derived based on the Bean model by Chen [20].

Equation (1)

3. Results and discussion

By heating the raw precursor salts to remove the crystal water, the obtained PAn solution had water content as low as 0.2 wt%, according to measurement by the Karl Fischer method. 0.2 wt% was similar to the water content of the conventional TFA solution reported by Araki et al [9, 21]. With such a low water content, the precursor solution could be long-term storable. In our study, a vial of PAn was stored for three months and remained clear without any precipitation.

Using PAn and the fabrication procedure previously described in the experimental details section, YBCO film with the optimized performance could be obtained when the maximum temperature of pyrolysis (MTP) was 600 °C. Its Jc was as high as 3.8 MA cm−2 (77 K, self-field, about 250 nm thick), which is comparable with those of YBCO films reportedly fabricated by the conventional all TFA solution [22] and low-fluorine solution [23].

As shown in figure 1, the pyrolysis duration was approximately 40–60 min, which was much shorter than that of the conventional TFA-MOD method [9]. To investigate the decomposition reactions during the pyrolysis step, thermal analyses were conducted using the gel powder samples dried from PAn, PAnD and conventional TFA solution under the same conditions (air and 5 °C min−1). As presented in figure 2, DTA of the conventional TFA solution sample displayed several major exothermic peaks between 230 and 310 °C, which was consistent with the decompositions of the TFA salts [18, 2426]. PAn and PAnD samples both exhibited three exothermic peaks in the temperature range from 200 to 340 °C, which was broader than that of the conventional TFA solution sample. For the TGA curves (inset in figure 2), the mass loss of PAn or PAnD sample initialized at about 200 °C, and completed at about 320 °C. Thus the decomposition temperature range of PAn and PAnD samples could be estimated as about 120 °C, which was also broader than that of the conventional TFA solution sample (about 80 °C). The broadening of the decomposition temperature range was also found in the study of terpineol additive by Ding et al [16]. They suggested that this phenomenon could be related to the alleviation of stress generated during the decomposition reactions, thus smooth films could be obtained after a rapid pyrolysis step. The achievement of a short pyrolysis time (approximately 40–60 min) in our study might also be attributed to such effects of the decomposition temperature range broadening.

Figure 2.

Figure 2. Differential thermal analysis (DTA) and thermogravimetric analysis (TGA) curves of the gel powder samples prepared by PAn, PAnD and conventional TFA solution.

Standard image High-resolution image

The relationship between Jc and MTP is illustrated in figure 3. For the films fabricated by PAn and PAnD, Jc increased along with MTP. In addition, Jc of the films by PAn was slightly higher than that of the films by PAnD. For the two YBCO films prepared by PAn and PAnD with MTP of 600 °C, the dependence of Jc on magnetic field (B) and the relationship of resistance against temperature were measured, as shown in figures 4 and 5 respectively. It could be noticed that their Jc values both decreased rapidly along with the increase of B. Tc of the YBCO films prepared by PAn and PAnD were 89.6 K and 88.9 K, respectively.

Figure 3.

Figure 3. Relationships between critical current density (Jc) of the final YBCO films and the maximum temperature of pyrolysis (MTP).

Standard image High-resolution image
Figure 4.

Figure 4. The dependence of Jc on the magnetic field (B) of the YBCO films fabricated by PAn and PAnD, with the maximum temperature of pyrolysis (MTP) of 600 °C.

Standard image High-resolution image
Figure 5.

Figure 5. The relationships between resistance (R) and temperature (T) of the YBCO films fabricated by PAn and PAnD, with the maximum temperature of pyrolysis (MTP) of 600 °C.

Standard image High-resolution image

To investigate the influence of MTP, phase characterization of the pyrolyzed films and final YBCO films was carried out by XRD measurements. XRD θ/2θ patterns of the pyrolyzed film samples are shown in figure 6. These samples were quenched at an MTP of 400, 500 or 600 °C. All the pyrolyzed films had smooth surface morphologies. As could be observed in figure 6, all the XRD patterns presented the same phases including Ba1−xYxF2+x (BYF) and CuO, which were also found in the pyrolyzed films prepared by conventional TFA solution [28, 29]. Apart from BYF and CuO, YOF(009) peak could be observed in all the XRD patterns, which were also detected in the pyrolyzed films deposited by the low-fluoride metal propionate-based solution [27]. The main influence of MTP increasing on the XRD patterns in figure 6 was that the peaks of BYF shifted toward lower 2θ angles. The peak shift was also found during the heat treatment using conventional TFA solution [28, 29]. In these studies, it was believed that the peak shift indicated the development of a BaF2-related superstructure and the Y release.

Figure 6.

Figure 6. XRD θ/2θ patterns of the pyrolyzed films prepared by PAn or PAnD with different maximum temperatures of pyrolysis (MTPs).

Standard image High-resolution image

XRD θ/2θ patterns of the final YBCO films fabricated by PAn and PAnD are very similar to each other, as presented in figure 7. All the samples are mainly c-axis oriented, while the weak (200) peaks indicate the presence of a small portion of a/b-axis oriented YBCO grains. The detrimental BaCO3 is not detected in the XRD patterns. In addition, the secondary phase Y2Cu2O5 could be observed in all the films. Y2Cu2O5 is one of the equilibrium phases involved in the YBCO formation process, and also related to the CuO grains in the pyrolyzed films in the conventional TFA-MOD method [6, 30]. In our future study, the elimination of Y2Cu2O5 from the final YBCO thin films will be the aim of the crystallization process improvement.

Figure 7.

Figure 7. XRD θ/2θ patterns of the final YBCO films prepared by PAn or PAnD with different maximum temperatures of pyrolysis (MTPs).

Standard image High-resolution image

The surface morphologies of final YBCO films fabricated by PAn and PAnD were observed by SEM, as shown in figure 8. For all the films, the surfaces mainly consisted of c-axis orientated plate-shaped YBCO grains, which was consistent with the XRD patterns in figure 7. For the films fabricated either by PAn or by PAnD, with the increase of MTP from 400 to 600 °C, the size of pores became smaller, and the number of pores was also reduced. In the XRD patterns of the final YBCO films, no obvious difference could be found with the increase of MTP. Therefore, the enhancement of Jc with increasing MTP shown in figure 3 could be ascribed to the densification phenomenon indicated by SEM.

Figure 8.

Figure 8. SEM images of the final YBCO films prepared by PAn or PAnD with different maximum temperatures of pyrolysis (MTPs).

Standard image High-resolution image

In addition, as revealed in the phase characterizations by XRD and the morphology observations by SEM (figures 68), the pyrolyzed films or the final YBCO films prepared by PAn and PAnD with the same MTP were very similar. This was consistent with the similar behavior of their Jc with the increase of MTP as shown in figure 3. After the further reduction of water content from PAn (0.2 wt%) to PAnD (below 10−2 wt%) by vacuum distillation, we could not find any significant improvement in the properties of final YBCO thin films. Therefore, the preparation process of PAn without vacuum distillation proposed in this study was applicable for high-performance YBCO film fabrication. In the future simplification of this method, the heating of raw precursor salts could be omitted by using salts without crystal water.

4. Conclusions

In this study, a water-free MOD method is proposed. Using propionic anhydride and methanol as the solvents, a low-fluorine precursor solution PAn containing only 0.2 wt% water content could be prepared. YBCO thin films about 250 nm thick with Jc as high as 3.8 MA cm−2 (77 K, self-field) could be obtained by PAn with a short pyrolysis step of approximately 60 min. Thermal analysis results indicate that the decomposition temperature range of PAn is broader than that of conventional TFA solution. By increasing MTP from 400 to 600 °C, Jc could be significantly enhanced, which is due to the reductions of pore size and number according to SEM observations. No significant improvement was observed by further reducing water content from PAn to PAnD using vacuum distillation, as revealed by the Jc, XRD and SEM measurements. Thus vacuum distillation is not necessary for the water-free MOD method. Therefore, the water-free MOD method can be applied to avoid the utilization of distillation apparatuses, and shorten the solution preparation time. It is a promising technique for the low-cost manufacturing of YBCO coated conductors.

Acknowledgments

This work was supported by the National Natural Science Foundation of China (51202124), China Postdoctoral Science Foundation (2013M530615), Beijing Natural Science Foundation (2122026), Fundamental Research Program of Shenzhen (JCYJ20120614193005764), and State Key Laboratory of New Ceramic and Fine Processing Tsinghua University.

Please wait… references are loading.
10.1088/0953-2048/26/11/115010