J Supercond Nov Magn(2012)25:811–816
DOI10.1007/s10948-011-1350-y
O R I G I NA L PA P E R
Texture Transfer Mechanism of Buffer Layer in Coated Conductors
Y.Wang·L.Zhou·C.S.Li·Z.M.Yu·J.S.Li·
L.H.Jin·P.F.Wang·Y.F.Lu
Received:9August2011/Accepted:21October2011/Published online:12November2011
©Springer Science+Business Media,LLC2011
salary什么意思Abstract We have studied the texture transfer process of buffer layers prepared by chemical solution deposition (CSD)methods on YSZ(00l)single crystal substrates and biaxial textured Ni–W substrates.The structure,texture,and surface morphology of the buffer layers were investigated by X-ray diffraction(XRD),atomic force microscopy(AFM), and scanning electron microscopy(SEM).Our re
sults show that the degree of texture and the surface morphology of the buffer layers vary with the changes of the substrate and the lattice mismatch of the top buffer layers with La2Zr2O7 (LZO)after crystallization in argon–hydrogen atmosphere. Moreover,the growth mode of multi-layerfilms and the type of the lattice strain have strong influence on the formation and the transfer of the bi-axial texture in multi-layer buffer architecture.It suggests that there exists a possible connec-tion between the strain relaxation and the texture transfer in buffer layer fabricated by CSD methods.Information on the texture transfer of buffer layer is important for optimizing the buffer layer architecture in coated conductors. Keywords Coated conductors·Buffer layer·Chemical solution deposition method
Y.Wang·L.Zhou·C.S.Li·Z.M.Yu·L.H.Jin·P.F.Wang·
Y.F.Lu( )我最高兴的一件事
拉萨美食Northwest Institute for Nonferrous Metal Rearch,Xi’an, 710016,P.R.China
e-mail:
Y.Wang·L.Zhou·J.S.Li
State Key Laboratory of Solidification Processing,
加湿器怎么清洗
Northwestern Polytechnical University,Xi’an,710072,P.R.China P.F.Wang
Shaanxi Normal University,Xi’an,710062,P.R.China 1Introduction
Coated conductors compod of metal substrate/buffer layer/superconducting layer/protective layer are very attrac-tive for electrical power application and superconducting magnet technology[1–6].Buffer layer,which prevents oxy-gen and metal atoms diffusion and also transfers the biax-ial texture from metal substrate to superconducting layer,is one of the key functional layers in coated conductors[7,8]. Thus the multi-layer architecture is generally adopted to realize the above-mentioned two main functions of buffer layer.However,the thicher buffer layers can lead to a more complicated fabrication procedure and an increa in prepa-ration cost.Therefore,the simplification of the buffer layer architecture is always the desired goal of buffer layer de-velopment.The knowledge of the texture transfer mecha-nism of buffer layer is very uful for the optimization of the buffer layer architecture in coated conductors.However, few studies about texture transfer mechanism of buffer layer, especially fabricated by chemical solution deposition(CSD) methods,have been reported.The study of the texture for-mation and the texture transfer of buffer layer are important for understanding the function of buffer layer and simplify-ing buffer layer architecture in coated conductors.
CSD methods offer the significant potential cost and pro-cessing advantages over physical vapor deposition(PVD) process[2,9].In addition,the preparation procedure of La2Zr2O7(LZO)thinfilm by CSD is readily controllable and rather mature.Although the LZO with pyrochlore struc-ture prents a relatively high lattice mismatch with YSZ (00l)single crystal substrates and Ni–W substrates,good epitaxial growth of LZOfilm is still possible[10–14].More-over,for the heteroepitaxy system,the lattice mismatch may lead to a ries of changes,including the growth mode,mi-crostructure and so on.Therefore,LZOfilms on YSZ(00l)
single crystal substrates and Ni–W substrates were lected as the model to further deposit CeO2,Nd2Zr2O7(NZO),and Nd2Mo2O7(NMO)films with different lattice mismatch to LZO by CSD methods,respectively.Then the texture trans-fer mechanism of buffer layer is explored by investigating the texture and surface morphology evolution of buffer lay-ers on YSZ(00l)single crystals and Ni–W tapes.
2Experimental Procedures
All of the MOD precursor solutions were prepared in am-bient atmosphere.The stable yellow-colored LZO starting solution was produced by mixing a stoichiometric amount of reagents lanthanum(III)2,4-
pentanedionate(La(acac)3) and zirconium(IV)2,4-pentanedionate(Zr(acac)4)(re-ceived from Strem)with propionic acid along with con-tinuous stirring at75◦C for25min.Similarly,cerium (III)2,4-pentanedionate(Ce(acac)3)(received from Strem) was dissolved in propionic acid by heating with continu-ously stirring to synthesize a stable brown-colored CeO2 precursor solution.However,the NZO and NMO precur-sor solutions were prepared from neodymium(III)acetyl-acetonate(Nd(acac)3),Zr(acac)4and molybdenum(IV) dioxo acetylacetonate(MoO2(acac)2)in methanol and pro-pionic acid.Details of the solution preparation process can be found elwhere[15–20].Spin coating was ud to deposit LZO on YSZ(00l)single crystal substrates and textured Ni–W substrates at a spin rate of2500rpm for30s.The spin coatedfilms were then heat-treated at 900◦C for40min in aflowing mixture of Ar-4%H2gas. Then the CeO2precursor solution was spin-coated onto the above-obtained crystallized LZOfilms,followed by heat treatment at1000◦C for15min in Ar-4%H2to fab-ricate CeO2/LZO/YSZ and CeO2/LZO/NiW.On the other hand,the NZO and NMO precursor solutions were re-spectively spin coated onto the as-grown LZO/YSZ and LZO/Ni–W templates.Subquently,the coatings were heated at550◦C for1h and then annealed at1000◦C for2h in a reducing forming gas atmosphere of Ar-4%H2.At the end of the heat-treatment cycles,the samples were quenched to room temperature in the same atmosphere to fabricate NZO/LZO/YSZ,NMO/LZO/YSZ,NZO/LZO/Ni–W,and NMO/LZO/Ni–W.
All the bufferfilms were characterized by X-ray diffrac-tion(XRD),which was performed to carry outθ–2θscan using CuKαradiation at40kV and50mA,for pha purity and texture.In addition,the out-of-plane and in-plane scans were also confirmed by XRD for the degree of texture of buffer layer.Thefilm thickness of all buffer layers were calibrated and determined byα-step apparatus.The surface homogeneity and microstructure analysis of the samples were performed by JSM-6700field emission scanning elec-tron microscopy(SEM)and a digital instruments nanoscope SPI3800-SPA-400atomic force microscopy(AFM)in con-tact mode.
3Results and Discussion
洗头的最佳时间
The typicalθ–2θscans for LZO/YSZ,CeO2/LZO/YSZ, NZO/LZO/YSZ,and NMO/LZO/YSZ samples are shown in Fig.1.All of the LZO,CeO2,NZO,and NMOfilms prepared in Ar-4%H2have been completely crystallized to form the c-axis orientation becau only highly(00l) diffraction peaks of buffer layers and YSZ(00l)single crys-tal substrates are obrved,and neither(111)nor(222)peak is prent in the patterns.
In addition,Fig.1also shows the development of the FWHM values of the out-of-plane and in-plane texture of the buffer layers as a function of the lattice mismatch of the top buffer layers with LZO.The
variation of the FWHM values of the top buffer layers reprents mainly a texture transfer effect of the bottom buffer layer.Therefore,the relative FWHM from the omega and phi scans(derived from the specific value FWHM buffer(00l)/FWHM LZO(004), and FWHM buffer(111)/FWHM LZO(222)or FWHMLZO(111)/ FWHM LZO(222))are plotted in Fig.1,which reveals the tex-ture transfer of the buffer layer with the lattice mismatch of the top buffer layer to LZO.The relative FWHM value from omega scans in0.3%of lattice mismatch between
the Fig.1XRD patterns for the as-grown LZO/YSZ,CeO2/LZO/YSZ, NZO/LZO/YSZ,and NMO/LZO/YSZ samples in Ar-4%H2.The de-velopment of the relative FWHM from omega scans and phi scans of buffer layer as a function of the lattice mismatch of the top buffer layer with LZO
top buffer layer and LZO reaches a maximum value and then decreas as the lattice mismatch increas;finally,it slightly increas when the lattice mismatch between the top buffer layer and LZO increas from 1.3%to 3.3%.How-ever,the relative FWHM value from phi scans drops moder-ately as the lattice mismatch increas and reaches a min-imum value in 1.3%of the lattice mismatch between the top buffer layer and LZO,and then increas markedly as the lattice mismatch increas to 3.3%.Compared with the LZO film on YSZ substrate,the degraded out-of-plane tex-ture of CeO 2layer on the LZO/YSZ template may be related to its especial surface morphology prented below.How-ever,the larger lattice mismatch between NZO and LZO may increa the influence of surface energy on the ori-entation of buffer layer,which leads to a smaller FWHM value determined by rocking curve measurement [21].For NMO/LZO/YSZ sample,the wor out-of-plane and in-plane texture could be explained by the strain relaxation in NMO film with increasing the thickness of the buffer layer [17].It should be pointed out that the thickness of the top and bottom buffer layers in the different architectures is almost the same value of around 50nm.
Figure 2shows the XRD patterns obtained from the LZO/Ni–W,CeO 2/LZO/Ni–W,NZO/LZO/Ni–W,and NMO/LZO/Ni–W multilayer architectures.Besides the ma-jor (002)diffraction peaks of the Ni–W substrates,only the (00l)reflections of the buffer layers appear,which indicates the good c-axis texture of the buffer
films.
Fig.2Typical θ–2θscans of the as-prepared LZO/Ni–W,CeO 2/LZO/Ni–W,NZO/LZO/Ni–W,and NMO/LZO/Ni–W samples in Ar-4%H 2.Dependence of the relative FWHM from omega scans and phi scans of buffer layer on the lattice mismatch of the top buffer layer with LZO
Figure 2also prents the relationship between the rela-tive FWHM value of the omega scans as well as phi scans and the lattice mismatch of the top buffer layer with LZO.The relative FWHM value from omega scans slightly de-creas when the lattice mismatch between the top buffer layer and LZO increas to 0.31%.Furthermore,the rela-tive FWHM suddenly increas as the lattice mismatch in-creas.It should be pointed out that the variation trend in the relative FWHM value from phi scans is very similar to that of omega scans as the lattice mismatch between the top buffer layer and LZO increas but only the fluctuation range of relative FWHM value is wider in omega scans than that in phi scans.We found that the FWHM values of out-of-plane and in-plane alignments of CeO 2film has improved compared with that of the bottom LZO buffer layer.This tex-ture improvement was generally obrved in many samples,which was already reported by other rearchers [22–24].This result is related to the lattice mismatch of the top buffer layer with LZO and the growth mode of the top buffer layer on LZO/Ni–W templates.
For LZO/YSZ,the round-shape grains have an average grain size of about 40nm and the root means square rough-ness value (R rms )of 0.63nm is calculated over an area of 5µm ×5µm,as en in Fig.3(a).It indicates that this very smooth LZO surface is especially beneficial for the sub-quent growth of the top buffer layers.Figure 3(b)–(d)shows the surface morphology of CeO 2,NZO,and NMO films on LZO/YSZ template,and their R rms values are 2.10,0.51,and 1.42nm with a scan area of 5µm ×5µm,respectively.The CeO 2/LZO/YSZ grown in Ar-4%H 2indicates a volcanic-vent shaped surface structure,which results in a larger R rms value and is related to the large FWHM values from its omega and phi scans.In addition,NZO film with a flat crystalline plain has been obtained on LZO/YSZ template.In contrast to the surface morphology of NZO/LZO/YSZ,NMO layer deposited on LZO/YSZ prents a relatively rough topography along with large crystalline growing out of the coating volume.Thus,it can be considered that the surface and interface roughness have little effect on the epi-taxial growth of the buffer layers on LZO/YSZ templates due to the very smooth surface of the bottom buffer layer.This suggests that the surface morphology evolution is re-lated to the change of the epitaxial growth and texture trans-formation behavior of buffer layers.
SEM micrographs for LZO film on Ni–W substrate,CeO 2,NZO,and NMO films on LZO/Ni–W templates are shown in Fig.4.The grain boundary grooves of Ni–W sub-strates were found to be com
pletely covered.Figure 4shows that all of the film surfaces are homogenous as well as crack free.For CeO 2/LZO/Ni–W and NZO/LZO/Ni–W,there is no significant difference with respect to the surface mor-phology compared with that of the LZO/Ni–W sample ex-cept that the average grain size increas to a certain ex-
Fig.3AFM images for as-grown(a)LZO/YSZ,(b)CeO2/LZO/YSZ,(c)NZO/LZO/YSZ,and(d)NMO/LZO/YSZ samples in Ar-4%H2,re-spectively
tent as the layer amount of the buffer layer increas.In ad-dition,we have noticed that the average grain size for the CeO2film is larger than that for NZOfilm on LZO/Ni–W. It suggests that the growth of NZO grains under the given conditions is slower than that of CeO2,which leads to a higher grain boundary concentration in NZOfilm than that in CeO2film.However,the larger grain size may result in the lager residual stress infilms.It indicates that the part of strain relaxation takes place in NZO/LZO/Ni–W sam-ple becau of the larger lattice mismatch of NZO with LZO than that of CeO2.This result is assigned to the larger FWHM values from omega and phi scans of NZOfilm than that of CeO2film on LZO/Ni–W.It is expected that a larger strain relaxation in the NZOfilm is not in fa-vor of the texture transfer in NZO/LZO/Ni–W architecture. However,some large grains bulge on the surface of the NMO/LZO/Ni–W,which is very similar to the surface mor-phology of NMO/LZO/YSZ,ca
n be related to its especial epitaxial growth mode on LZOfilm with a large lattice mis-match of3.3%.The strain energy in strainedfilm increas as the lattice mismatch of epitaxialfilm with substrate in-creas.The result reveals that the growth of CSD-NMO films on LZO/Ni–W and LZO/YSZ templates does not take a two-dimensional layer-by-layer mode.It may result in the decrea of texture degree of NMO layers,which was concluded from the largest FWHM values in out-of-plane scans and in-plane scans of NMOfilms on LZO/Ni–W and LZO/YSZ.Therefore,we believe that their surface morphol-ogy evolution is associated with the texture transfer of the buffer layers.
For the heteroepitaxy system,the lattice distortion gener-ates the tensile stress or compressive stress in thefilm by the short range attractive interactions.The lattices distor-tion may be enhanced as the lattice mismatch of buffer layer
劳动最光荣演讲稿Fig.4SEM micrographs of as-prepared(a)LZO/Ni–W,
(b)CeO2/LZO/Ni–W,
(c)NZO/LZO/Ni–W,and
吞拿鱼(d)NMO/LZO/Ni–W samples in Ar-4%H2,
respectively
with LZO increas,which may be in favor of the coher-ent epitaxial growth of buffer layer during the high tempera-ture heat-treatment process.However,the strain relaxation can occur when the stress in thefilms overstep a limita-tion value,which may disfavor the epitaxial growth of buffer layer and the texture transfer from the substrate to the buffer layer.The larger lattice mismatch of buffer layer with sub-strate leads to the smaller the critical thickness of epitaxial film,at which the lattice relaxation takes place.After the buffer layer deposited on LZOfilm,the lattice distortion of LZO layer may be relaxed becau of an opposite effect in the misfit strain between LZO and the underlying substrate, which results in extending the strain to the top buffer layer during the subquent epitaxial growth of buffer layer[25]. In addition,surface energy of precursorfilm,which may be related to the growth mode of buffer layer,could influ-ence the orientation growth of buffer layer on template layer. Therefore,the texture transfer in multi-layer architecture of buffer layer is not only influenced by the lattice mismatch between the buffer layer and LZO,but also affected by the growth mode of buffer layer,especially for the buffer layer on LZO/Ni–W template.In other words,compared with the epitaxial growth of buffer layer on LZO/YSZ,the mismatch between the buffer layer and LZO is a minor contribution to the texture transfer from Ni–W substrate to the top buffer layer,which is mainly determined by growth mode of the top buffer layer.4Conclusions
We have obtained LZO,CeO2,NZO,and NMOfilms with strong(00l)orientation for further fabrication of coated con-ductors by the CSD route and studied the texture transfer mechanism of buffer layer in the CSD process.The XRD patterns,AFM images,and SEM micrographs have shown that the texture and surface morphology are dependent on the lattice mismatch of the top buffer layers with LZO and their growth mode.The lattice mismatch of the top buffer layer with LZO influences the strain relaxation process, which could be related to the texture formation and texture transfer of buffer layer in coated conductors. Acknowledgements This work wasfinancially supported by the Na-tional Science Fund Program and the National High Technology Re-arch and Development Program China(Grant No.50872115and 2009AA03Z203).
References
1.Arda,L.,Ataoglu,S.,Sezer,S.,Abdulaliyev,Z.:Surf.Coat.Tech-
nol.202,439(2007)法国谚语
2.Paranthaman,M.P.,Sathyamurthy,S.,Bhuiyan,M.S.,Martin,
P.M.,Aytug,T.,Kim,K.,Fayek,M.,Leonard,K.J.,Li,J.,Goyal,
A.,Kodenkandath,T.,Li,X.,Zhang,W.,Rupich,M.W.:IEEE
Trans.Appl.Supercond.17,3332(2007)
3.Shiohara,K.,Sakai,S.,Ohki,S.,Yamada,Y.,Tachikawa,K.,
Koizumi,T.,Aoki,Y.,Hikichi,Y.,Nishioka,J.,Hagawa,T.: Physica C469,1870(2009)
