Epitaxial Growth of LZO Film on NiW Substrate

Available online at www.sciencedirect.com

Physics Procedia 45 (2013) 161 – 164

ISS2012

Epitaxial growth of LZO film on NiW substrate C.S. Li, Z.M. Yu*, P. Odier, Y. Wang, L.H. Jin, J.Q. Feng, P.X. Zhang Northwest Institute for Nonferrous Metal Research

P.O. Box 51, Xi’an, Shaanxi, 710016, P.R. China

Abstract La2Zr2O7 (LZO) oxide film is a good choose for buffer layer of coated conductors, especially for Chemical Solution Deposition route to fabricate coated conductors. In this paper, the influences of process parameters on texture of LZO films deposited by chemical solution deposition process on NiW substrate were studied, and results indicated that the heat-treatment temperature, dwell time and the oxygen content in the heat-treatment atmosphere were important to obtain a cube textured LZO film on NiW substrate. Based on our results, it was believed that inhibition of crystallite growth of LZO grains on the top layer of LZO film was the key to understand the epitaxial growth of LZO film on NiW substrate.

© 2013 The Authors. Published by Elsevier B.V. © 2013 The Authors. Published by Elsevier B.V. Selection and/orCommittee. peer-review under responsibility of ISS Program Committee. Selection and/or peer-review under responsibilty of ISS Program Keywords: La2Zr2O7; chemical solution deposition (CSD); epitaxial growth

1. Introduction All fabrication techniques of coated conductors are based on the epitaxial growth techniques, either vacuum-based deposition techniques or chemical solution deposition (CSD) techniques. It is believed that the metalorganic deposition method using trifluoroacetates, one of CSD techniques, is an important and better method to deposited superconducting layer. But it is still developing for the deposition techniques of buffer layer. The main reason why it is difficult to optimize the depositing techniques of buffer layers is that the architecture of buffer layers is very complex in order to carry out the functions of buffer layers. So simplification of architecture of buffer layers is important to develop cost-effect deposition techniques of buffer layers. The architecture of CeO2/La2Zr2O7 buffer layers is a good choose, and it is interesting that the buffer layers of CeO2/La2Zr2O7 can be deposited by all CSD techniques [1, 2]. La2Zr2O7 (LZO) films can be deposited directly on NiW substrates by CSD techniques [3-6], and all films have a good cubic texture. However, the mechanism of epitaxial growth of LZO film on NiW substrate is still an opening question. In order to prevent the oxidization of NiW substrate, a reducing atmosphere is usually used during the heattreatment step. It will result in some residual carbon in the final crystalline film. The results in Ref. 7 indicated that residual carbon in grain boundaries influenced the texture of CeO2 film. Due to similar process between deposition of LZO film and deposition of CeO2 film, many researchers believed that development of LZO film’s texture resulted from the growth process of LZO film, and it was the key to explain the mechanism of epitaxial growth of LZO film on NiW substrate that residual carbon inhibited the growth of LZO grains [6, 8]. In this study, influence of process parameters on the texture of LZO film were investigated firstly, and then the mechanism of epitaxial growth of LZO film deposited by CSD process on NiW substrate was discussed. 2. Experiments

*Corresponding author. Tel.:+86-29-8631079; fax: +86-29-86224487 E-mail address: [email protected]

1875-3892 © 2013 The Authors. Published by Elsevier B.V. Selection and/or peer-review under responsibilty of ISS Program Committee. doi:10.1016/j.phpro.2013.04.077

162

C.S. Li et al. / Physics Procedia 45 (2013) 161 – 164

Lanthanum(III) 2,4-pentanedionate (La(acac)3) and zirconium(IV) 2,4-pentanedionate (Zr(acac)4) were used as the solutes. A stoichiometric mixture of La(acac)3 and Zr(acac)4 was solved into propionic acid. The solution was placed in an ultrasonic bath for 15 min, then stirred and kept at 60 for 30 min. Finally, a transparent yellow precursor solution was obtained [9]. The concentration of the precursor solution was 0.4 mol/L (with respect to La3+). Ni5at.%W substrates (size 10 mm x 10 mm) were cut from a long tape, and put into an ultrasonic bath with acetone to clean their surface. After that, a spin-coating technique was used to cover the surface of NiW substrates by LZO precursor solution (rotation speed: 2500 rpm, rotation time: 30 s, acceleration: 3000 rpm/min). Most of the samples were heated under Ar-5%H2 atmosphere, except that the sample was heated under a high content oxygen atmosphere in order to investigate the influence of oxygen partial pressure, in which an Ar-5%H2 mixture gas and an Ar200ppmO2 mixture gas were mixed to increase slightly the oxygen partial pressure during the heat-treatment. Based on our previous works, the gas flow was very important to obtain a cube-textured LZO film on NiW substrate [6]. So the gas flow was set the value of 0.5 L/h during all the samples’ heat-treatment in order to eliminate the influence of gas flow. The texture of LZO films was characterized by -2 scan. It should be mentioned that due to the high symmetry of crystalline structure of LZO phase, the result of -2 scan is sufficient to show preliminarily if a textured LZO film is obtained. Therefore, only the results of -2 scans have been presented in this paper, since we mainly focus on the type of texture. XPS has been used to probe the content of residual carbon in the top layer of LZO films. In order to eliminate the influence of absorption of other carbon-containing gases from keeping environment, all samples were sputtered by Ar ions for 10 s before characterized by XPS. 3. Results Figure 1 showed the influence of heat-treatment temperature on the texture of LZO films. It was clear that only LZO(400) peaks could be found obviously on the curves of -2 scans, and the intensity of LZO(400) peak increased with increasing the heat-treatment temperature. It indicated that the texture of LZO films was improved with increasing the heat-treatment temperature. The rates of I(400)LZO to I(200)NiW were used to demonstrate the increase of volume fraction of cube texture of LZO films in order to eliminate the error resulting from XRD equipment. The results showed that the rate was 1% if the sample was heat treated at 800 , and the rate increased to 1.28% for the sample heated at 900 , and the rate increased to 1.27% for the sample heated at 1000 . It indicated that almost all part of films were cube texture if the heat treatment temperature was higher than 900 . This conclusion has been proved by our previous results [10], and agreed with the results in Ref. 11. The results in Ref. 11 also proved that heattreatment temperature should be higher than 900 . Besides that the volume fraction of cube texture of LZO film was sensitivity with the heat-treatment temperature, the texture of LZO film was also very sensitivity with the dwell time because the rate of I(400)LZO to I(200)NiW was increased by 23% if the dwell time was increased from 10 min to 40 min (Fig. 2).

1500 1000 500 0

Intensity ( a.u. )

2000

o

1000 C o

900 C o

800 C

26

28

30

32

34

2 ( deg. )

Fig.1 Influence of heat-treatment temperature on the texture of LZO films Because the heat-treatment temperature during CSD process is lower than the melting point of LZO phase so much, the driving force for crystallization is significantly larger than the energy barriers for all nucleation events. In order to inhibit the occurrence of homogeneous nucleation, a fast heating process is a good choice [12]. The data shown in Fig. 3 proved this idea. It was clear that the intensity of (400)LZO peak increased with increasing the heating rate, and the rate of I(400)LZO to I(200)NiW increased by 12% when the heating rate accelerated from 300 /h to 600 /h. Based on above-mentioned data, it could be found that the influence of heating rate was not primary in our case comparing with the influence of heat-treatment temperature and of dwell time.

2000 1500 1000 500

Intensity ( a.u. )

C.S. Li et al. / Physics Procedia 45 (2013) 161 – 164

0 o

950 C x 40min o

950 C x 10min 26

28

30

32

34

2 ( deg. )

Fig.2 Influence of dwell time on the texture of LZO films

1500 1000 500

Intensity ( a.u. )

2000

0 o

heating rate: 600 C/h o

heating rate: 300 C/h

26

28

30

32

34

2 ( deg. )

Fig. 3 Influence of heating rates on the texture of LZO films

1500 1000 500

Intensity ( a.u. )

2000

0 Ar-5%H2 / Ar-200ppm O2 Ar-5%H2 26

28

30

32

34

2 ( deg. )

Fig.4 Influence of heat-treatment atmosphere on the texture of LZO films In order to prevent oxidization of NiW substrate, all the samples should be heat-treated under a reducing atmosphere. Ar-5%H2 mixture gas was usually used. However, a few results discussed the influence of oxygen content in the heat-treatment atmosphere. In order to increase slightly the oxygen content in the heat-treatment atmosphere, the two mixture gases were mixed, in which the one of two mixture gases was the mixture of Ar-5%H2, and the other was the mixture of Ar-200ppm O2. The influence of oxygen content in the heat-treatment atmosphere was showed in Fig. 4, and it was clear that the texture of LZO film was not a cube texture if the oxygen content was increased slightly. The rate of I(400) to I(222) should be 0.6 if all the part of LZO film was random texture, but the rate of I(400) to I(222) was 1.1 in Fig.4. It meant that the sample heat-treated under the high oxygen content atmosphere could be divided into two parts, a epitaxial part on the surface of NiW substrate and a random part on the top layer of LZO film. The carbon contents in the top layer of two samples in Fig.4 were characterized by XPS. In order to eliminate the adsorbed gases on the samples’ surface, all the samples were bombed by Ar ions for 10 s. Two zones of all the samples were checked, and average values were used to compare the influence of oxygen content on the content of residual carbon in the films. The results indicated that the content of residual carbon was decreased from 25.87 at.% to 17.03 at.% if the oxygen content in heat-treatment atmosphere was increased slightly. 4. Discussion

163

164

C.S. Li et al. / Physics Procedia 45 (2013) 161 – 164

The process of solution-derived textured oxide film is a nucleation and growth process, so the texture of films deposited by CSD processes is decided by nucleation events and the modes of growth of grains. The larger driving force that governs the transformation from precursor film to crystalline film is enough large to overcome the energy barriers for all nucleation events if the heat-treatment temperature is low, so a high heat-treatment temperature is helpful to inhibit homogeneous nucleation event in the bulk of precursor film, and beneficial to obtain a cube-textured LZO film. Base on the results in Fig. 2, it was believed that the top layer of LZO film kept an amorphous state for a long time. Otherwise, it is impossible that no LZO(222) peak can be found. That was reason why we believed the unhomogeneous distribution of residual carbon was important in our previous works [6], and more residual carbon in top layer of LZO film inhibited the homogenous nucleation and growth events, and the epitaxial layer could develop from the interface between LZO film and NiW substrate to the top surface of LZO film. After a precursor LZO film was heat-treated under a reducing atmosphere, residual carbon should be found in the LZO film [6, 8, 13]. In our previous works, it has been shown that the reason resulting in the residual carbon in the LZO film was that some carbon-containing gases which came from the pyrolysis of precursor film were reduced [6]. So if the oxygen content in the heat-treatment atmosphere was changed, the distribution of residual carbon in the LZO film would change. It was believed that an unhomogeneous distribution of residual carbon in the LZO film resulted in the difference between epitaxial growth of LZO grains and the growth of LZO grains which came from the homogeneous nucleation in the top of LZO film [6]. Due to the increase of oxygen content in the heat-treatment atmosphere, the content of residual carbon in the top of LZO film decreased, and there was no enough residual carbon to inhibit the growth of LZO grains which resulted from the homogeneous nucleation, so the random textured LZO grains could be found in the final crystalline LZO film. 5. Conclusions The influence of four main parameters on the texture of LZO films was investigated. The results indicated that, comparing with the heat-treatment temperature and dwell time, the heating rate was not a primary parameter in order to obtain a cube-textured LZO film. The oxygen content in heat-treatment atmosphere was very important to deposit a cube-textured LZO film on NiW substrate. Increasing the oxygen content in heat-treatment atmosphere would decrease the residual carbon in the top layer of LZO film, and it resulted in that there was no enough residual carbon to inhibit the occurrence of homogenous nucleation and growth of LZO grains, which resulted in the random textured LZO grains in the top layer of LZO film. Acknowledgements This work was supported by the Program of International S&T Cooperation (Contract No. 2012DFA50780) and the National Natural Science Fund of China (Contract No. 5120201). It was also supported by the International Laboratory for the Applications of Superconductor and Magnetic Materials (LAS2M-CNRS-NIN). References [1] M. Paranthaman, S. Sathyamurthy, L. Heatherly, P.M. Martin, A. Goyal, T. Kodenkandath, et al., Physica C 445-448 (2006) 529–532. [2] S. Engel, K. Knoth, R. Huhne, L. Schultz, B. Holzapfel, Supercond. Sci. Technol. 18 (2005) 1385–1390. [3] K. Knoth, R. Huhne, S. Oswald, L. Schultz, B. Holzapfel, Supercond. Sci. Technol. 18 (2005) 334–339. [4] S. Sathyamurthy, M. Paranthaman, H.Y. Zhai, S. Kang, T. Aytug, C. Cantoni, et al., J. Mater. Res. 19 (2004) 2117–2123. [5] T. Caroff, S. Morlens, A. Abrutis, M. Decroux, P. Chaudouët, L. Porcar, et al., Supercond. Sci. Technol. 21 (2008) 075007. [6] Z.M. Yu, P. Odier, S. Morlens, P. Chaudouet, M. Bacia, L. Zhou, et al., J. Sol-Gel Sci. Technol. 54 (2010) 363–370. [7] A. Cavallaro, F. Sandiumenge, J. Gazquez, T. Puig, X. Obradors, J. Arbiol, et al., Adv. Funct. Mater. 16 (2006) 1363–1372. [8] S. Engel, R. Huhne, K. Knoth, A. Chopra, N.H. Kumar, V.S. Sarma, et al., J. Crys. Growth 310 (2008) 4295–4300. [9] Z.M. Yu, P. Odier, L. Ortega, L. Zhou, P.X. Zhang, A. Girard, Mater. Sci. Eng. B 130 (2006) 126–131. [10] Z.M. Yu, L. Zhou, P.X. Zhang, X.M. Xiong, L.H. Jin, C.S. Li, et al., Rare metal materials and engineering 37 (2008) 119–122. [11] K. Knoth, R. Huhne, S. Oswald, L. Schultz, B. Holzapfel, Acta Materialia 55 (2007) 517–529. [12] R.W. Schwartz, T. Schneller, R. Waser, C.R. Chimie 7 (2004) 433–461.