Reversible shift in the superconducting transition for La1.85Sr0.15CuO4 and BaFe1.8Co0.2As2 using piezoelectric substrates
Sascha Trommler, Ruben Hühne, Kazumasa Iida, Patrick Pahlke, Silvia Haindl, Ludwig Schultz, Bernhard Holzapfel
RReversible shift in the superconducting transition forLa . Sr . CuO and BaFe . Co . As using piezoelectric substrates S. Trommler, ∗ R. Hühne, K. Iida, P. Pahlke, S. Haindl, L. Schultz, ∗ and B. Holzapfel † IFW Dresden, P.O. Box 270116, 01171 Dresden,Germany, electronic mail: [email protected]
Abstract
The use of piezoelectric substrates enables a dynamic observation of strain dependent properties offunctional materials. Based on studies with La . Sr . CuO we extended this approach to the ironarsenic superconductors represented by BaFe − x Co x As to investigate strain driven changes in de-tail. We demonstrate that epitaxial thin films can be prepared on (001) Pb(Mg / Nb / ) . Ti . O substrates using pulsed laser deposition. The structural as well as the electric properties of the grownfilms were characterized in detail. A reversible shift of the superconducting transition of 0.44 K forLa . Sr . CuO and 0.2 K for BaFe . Co . As was observed applying a biaxial strain of 0.022%and 0.017% respectively. ∗ Dresden University of Technology, Department of Physics, Institute for Physics of Solids, 01062 Dresden † Dresden University of Technology, Department of Physics, Institute for Physics of Solids, 01062 Dresden,Germany a r X i v : . [ c ond - m a t . s up r- c on ] J u l he application of pressure has a significant influence on the physical properties of func-tional materials. Detailed experiments are required to enable a deeper understanding inphysics of these materials especially the sensitive interplay between structural parameterslike bonding length or angle and the electronic properties. Existing studies on bulk ma-terials predominantly cuprate high–temperature superconductors using hydrostatic pressuredemonstrate that compressive pressure increases the superconducting transition temperat-ure, T c , for most materials. Recent studies also reveal a strong influence of pressure onsuperconductivity for iron based superconductors resulting in a pressure dependent super-conducting dome in the electronic phase diagram similar to doping [1, 2]. It should be notedthat this shift in T c is highly anisotropic regarding strain along different crystallographicaxes. For some oxides like La − x Sr x CuO it is known that the resulting effect is partiallyneutralized for hydrostatic pressure [3–6].Therefore, in the last decade also the application of biaxial strain attracted increasingattention, especially for the model system La − x Sr x CuO [7, 8]. Similarly, recent studieson biaxial strained iron based superconductors like BaFe . Co . As and FeSe . Te . alsorevealed that compressive biaxial strain enhances T c [9, 10]. Typically, epitaxial thin filmsare prepared on various single crystalline substrates with a different lattice mismatch betweensubstrate and film inducing a biaxial tensile or compressive strain. However, this approachis often restricted to very thin films due to a limited layer thickness for coherent strainedgrowth. A large misfit typically leads to partial relaxation of the lattice and, therefore, to theimplementation of lattice defects. In this case it is difficult to correlate the applied strainwith superconducting properties directly, as the preparation conditions and the resultingmicrostructure may severely affect the latter.An alternative approach to static pressure experiments is the preparation of supercon-ducting films on single crystalline piezoelectric substrates. Using the inverse piezoelectriceffect the applied strain can be changed continuously and reversibly by an electric field.This approach offers the unique opportunity to investigate the strain dependent propertieson one and the same sample as shown already for ferromagnetic oxides [11–14]. Recentlywe reported on the epitaxial growth of superconducting YBa Cu O − δ and La . Sr . CuO (LSCO) thin films on pseudocubic (001) Pb(Mg / Nb / ) . Ti . O (PMN-PT) substrates[15, 16]. In this letter we extend this approach to BaFe . Co . As (Ba-122) thin films andreport on a reversible shift in T c with applied strain.2or the sample preparation on the PMN-PT we used a standard pulsed laser deposition(PLD) setup equipped with a Lambda Physiks LPX 305 KrF laser and stoichiometric targets.To reduce the lattice mismatch between PMN-PT ( a =4.02 Å) and the superconducting filmwe deposited smooth 20 nm thick buffer layers of either SrTiO (STO) ( a =3.905 Å) orCaTiO (CTO) ( a =3.82 Å) [17]. The buffer layers as well as the 300 nm thick LSCO filmsare prepared in 0.3 mbar oxygen atmosphere at substrate temperatures of 650 ◦ C–700 ◦ Cusing off–axis deposition [13, 18]. Films prepared under this condition typically exhibit avery smooth surface and droplet–free growth [18]. Subsequently, the films were cooled downin 0.4 bar oxygen atmosphere. A detailed description of the LSCO preparation as well asstructural analysis can be found in our previous publication [16]. A scheme of the filmarchitecture is given in fig.1.For the preparation of Ba-122 we used STO buffered PMN-PT prepared by off–axis–PLDas described above. Subsequently the substrate was transferred to an ultra high vacuumsystem with a base pressure of 10 − mbar where the Ba-122 was deposited at 650 ◦ C usingon–axis–PLD. A detailed description of the film preparation can be found in Iida et al.[10, 19].The superconducting properties were characterized in a Quantum Design Physical Prop-erties Measurement System (PPMS) using a four probe technique. For the evaluation ofthe transition temperature a 50% resistance criterion is used. To confirm epitaxial growthand to study the structural properties standard x-ray diffraction (XRD) in Bragg–Brentanogeometry, pole figure measurements and reciprocal space mapping (RSM) were performedusing a Phillips XPert MRD Diffractometer with Cu K α radiation. X–ray reflectivity wasused to determine the layer thickness and the roughness of the buffer layers.We achieved a c –axis oriented growth and cube on cube epitaxy for both, buffer layerand LSCO [16]. Also for Ba-122 the pole figure of the (103) peak, given in fig.2(a), provesperfect cube on cube epitaxy without any misorientation since the BaFe . Co . As peaks areoriented parallel the substrate [100] directions. The superconducting transition of LSCO at17.5 K on CTO buffered PMN-PT is slightly smaller compared to 18.5 K on STO bufferedPMN-PT. However, the CTO buffered system was used for further investigations due toreduced affinity to crack during dynamic strain measurements.In the case of BaFe . Co . As the STO buffered films exhibit a T c of 14 K which issignificantly reduced compared to 23 K for films prepared on bare STO [10]. Part of this3eduction is attributed to the poorer crystalline quality of the PMN-PT substrate comparedto STO and the much larger transition width as we use a 50% criterion.In the first step it is necessary to ensure the transfer of strain into the superconductinglayer. Detailed investigations by Bilani et al. showed that the strain is transferred from thePMN-PT to the STO buffer [17]. We performed additional high resolution XRD and RSMto verify the strain transfer to the superconducting layer. An example is given in fig.2(b)for the Ba-122 (008) peak without and with 16.6 kV/cm applied field.The change of the lattice parameters in PMN-PT single crystals at room temperaturein dependence of the applied electric field is well investigated [20], however, there is nodata available for lower temperatures. Nevertheless, the knowledge of the low temperaturebehavior is essential to correlate the strain with the change in the superconducting properties.To gauge the magnitude of strain at lower temperatures we deposited a thin meander-shaped platinum wire at room temperature directly on CTO buffered PMN-PT. The res-istivity of the wire correlates to the biaxial strain due to a change of the wire geometry. Thechange of the relative resistivity with the applied electric field is given in fig.3(a) for threedifferent temperatures. This data reveals a strong reduction of the strain at constant electricfield with decreasing temperature. Compared to room temperature ( (cid:15) a =0.12% at 10 kV/cm)we achieve half the value at 90 K and less than 20% at 20 K. The biaxial in–plane strain, (cid:15) a , is defined as ( a - a strained ) / a , where a is the unstrained in–plane lattice parameter.To check the suitability of our approach we used the well known model system LSCO.Applying an electric field of E =10 kV/cm at 20 K which corresponds to (cid:15) a =0.022% weachieved a reversible shift of the superconducting transition temperature of 0.4 K (fig.3(b)).We compared this shift to available literature data using a simple equation of the strain de-pendent T c for an orthorhombic unit cell, where T c (0) denotes the superconducting transitiontemperature of the unstrained film: T c = T c (0) + δT c δ(cid:15) a (cid:15) a + δT c δ(cid:15) b (cid:15) b + δT c δ(cid:15) c (cid:15) c (1)The values for the derivatives are well investigated for LSCO [4]. Due to the biaxial strain (cid:15) a equals (cid:15) b . Taking the correlation of (cid:15) c and (cid:15) a from statically strained LSCO films [8] wecan replace (cid:15) c in the the out–of–plane term and finally summarize equation(1) to: T c = T c (0) + β(cid:15) a , (2)4here β =2000 K. This results in a theoretical change of the transition temperature of0.44 K for (cid:15) a =0.022%, which is in good agreement with our experimental results (fig.3(b)).We checked the reversibility of the applied strain and the relaxation time of the PMN-PTat low temperatures, which is the time a piezoelectric material needs to reach the equilibriumstrained state. We measured the resistivity depending on the applied electric field at afixed temperature within the superconducting transition at 18 K. At this point the slope isvery steep enabling the detection of minor changes in resistivity, when the transition curveis shifted. Starting at E=8.66 kV/cm we determined the resistivity by varying the fieldin steps of 0.66 kV/cm. By successive change of the electric field a reversible change inresistivity was obtained (fig.4(a)). The minor deviation from linear behavior we attributeto the fact, that the time between field change and data acquisition was too less to reachthe equilibrium strain state. To characterize the relaxation time we reverse the polarity ofthe electric field starting at 13.3 kV/cm within 5 seconds and subsequently measured theresistivity depending on the time. Choosing a criterion for the equilibrium of less than 1%resistance change per hour, the equlibrium is reached after 30 minutes.Applying an electric field to an STO buffered Ba-122 film we observed a shift of thesuperconducting transition of 0.2 K for (cid:15) a =0.017% (fig.4(b)) corresponding to β =1700 K.We compared the data with the results on statically epitaxial strained Ba-122 thin filmswhere compressive strain results in different c / a ratios. There a strain of (cid:15) a =1.2% wasachieved resulting in a shift of the critical temperature of about 8 K [10]. The corresponding β =670 K is less than half the value, we achieve with the dynamic approach. Analyzing thisdifference one has to take into account that equation(1) is only valid for small strain effects.In addition hydrostatic pressure experiments reveal a non linear change in T c with pressure[1]. we expect a similar behavior for biaxial strain.In summary, we demonstrated the suitability of the inverse piezoelectric effect for thedynamical investigation of stain dependent superconducting properties. We observed a sig-nificant change in the superconducting transition temperature for both, LSCO and Ba-122thin films. We conclude that compressive biaxial strain enhances the critical temperaturefor Ba-122 similar to cuprates like LSCO. 5 CKNOWLEDGMENTS
This work was partially supported by the German Research Foundation. [1] S. A. J. Kimber et al. Nature Mat. , 471 (2009).[2] H. Okada et al. J. Phys. Soc. Jpn. , 113712 (2008).[3] S. L. Bud’ko, J. Guimpel, O. Nakamura, M. B. Maple, and I. K. Schuller, Phys. Rev. B ,1257 (1992).[4] F. Gugenberger, C. Meingast, G. Roth, K. Grube, V. Breit, T. Weber, H. Wühl, S. Uchida,and Y. Nakamura, Phys. Rev. B , 13137 (1994).[5] X. Chen, G. X. Tessema, and M. J. Skove, Physica C , 340 (1991).[6] M. Nohara, T. Suzuki, Y. Maeno, T. Fujita, I. Tanaka, and H. Kojima, Phys. Rev. B , 570(1995).[7] H. Sato, A. Tsukada, and M. Naito, Physica C , 848 (2004).[8] H. Sato, Physica C , 991(2008).[9] E. Bellingeri et al., Appl. Phys. Lett. , 102512 (2010).[10] K. Iida, J. Hänisch, R. Hühne, F. Kurth, M. Kidszun, S. Haindl, J. Werner, L. Schultz, andB. Holzapfel, Appl. Phys. Lett. , 192501 (2009).[11] C. Thiele, K. Dörr, O. Bilani, J. Rodel, and L. Schultz, Phys. Rev. B, , 054408 (2007).[12] K. Dörr, C. Thiele, J.-W. Kim, O. Bilani, K. Nenkov, and L. Schultz, Philos. Mag. Lett. ,269 (2007).[13] A. D. Rata, A. Herklotz, K. Nenkov, L. Schultz, and K. Dörr, Phys. Rev. Lett. , 076401(2008).[14] A. Herklotz, A. D. Rata, L. Schultz, and K. Dörr, Phys. Rev. B , 092409 (2009).[15] R. Hühne, D. Okai, K. Dörr, S. Trommler, A. Herklotz, B. Holzapfel, and L. Schultz, Super-cond. Sci. Technol. , 075020 (2008).[16] S. Trommler, P. Pahlke, R. Hühne, L. Schultz, and B. Holzapfel, J. Phys. Conf. Ser., ,012045 (2010).[17] O. Bilani-Zeneli, A. D. Rata, A. Herklotz, O. Mieth, L. M. Eng, L. Schultz, M. D. Biegalski,H. M. Christen, and K. Dörr. Appl. Phys. , 054108 (2008).
18] B. Holzapfel, B. Roas, L. Schultz, P. Bauer, and G. Saemann-Ischenko, Appl. Phys. Lett. ,3178 (1992).[19] K. Iida, J. Hänisch, T. Thersleff, F. Kurth, M. Kidszun, S. Haindl, R. Hühne, L. Schultz, andB. Holzapfel, Phys. Rev. B , 100507(R), (2010).[20] S.-E. Park and T.R. Shrout, J. Appl. Phys. , 1804 (1997). igure 1. Schematic film architecture. The bottom contacts are sputtered NiCr/Au, whereas thetop electrodes are deposited gold. igure 2. (a) Pole figure of the Ba-122 (103) reflection. The intensity is scaled quadratic, wheremaximum intensity corresponds to 650 cps; (b) the BaFe . Co . As (008) peak is shifted to lowerangles when compressive in–plane strain is applied. igure 3. (a) Relative change of the resistivity of a platinum strain gauge in dependence of the fieldfor three different temperatures; (b) shift of transition temperature with applied field for LSCO forE=-10 kV/cm (black), E=0 kV/cm (red) and E=10 kV/cm (blue). igure 4. (a) Reversible change of the resistivity over the applied field at a fixed temperature of18 K; (b) superconducting transition temperature of Ba-122 at E=-6.6 kV/cm (black) and E=6.6kV/cm (red).igure 4. (a) Reversible change of the resistivity over the applied field at a fixed temperature of18 K; (b) superconducting transition temperature of Ba-122 at E=-6.6 kV/cm (black) and E=6.6kV/cm (red).