Tailoring high temperature quasitwo-dimensional

Transcription

1 Tailoring high temperature quasitwo-dimensional superconductivity Dmitrii P. Pavlov 1, Rustem R. Zagidullin 1, Vladimir M. Mukhortov 2, Viktor V. Kabanov 1,3, Tadashi Adachi 4, Takayuki Kawamata 5, Yoji Koike 5, and Rinat F. Mamin 1 1 Zavoisky Physical-Technical Institute Federal Research` Center Kazan Scientific Center of RAS, Sibirskii trakt 10/7, Kazan, Russia 2 Southern Scientific Center of RAS, Chehova 41, Rostov-on-Don, Russia 3 Jozef Stefan Institute, Jamova 39, 1000 Ljubljana, Slovenia. 4 Department of Engineering and Applied Sciences, Sophia University, Tokyo, Japan 5 Department of Applied Physics, Tohoku University, Sendai, Japan Science should be fun, exciting and simple. Petr Kapitsa The creation of a high-tc quasi-two-dimensional superconductivity (HTq2DSC) 1-4 as well as a quasi-two-dimensional electron gas (q2deg) 5-11 at the interface and the ability to control such states by magnetic and electric fields were impossible without tailoring the atomically perfect interfaces Here we report superconductivity in heterostructure consisting of an insulating ferroelectric film (Ba 0.8Sr 0.2TiO 3) on an insulator single crystal in [001] orientation (La 2CuO 4), created by magnetron sputtering on a non-atomically-flat surface with inhomogeneities of the order of 1-2 nm. In this heterostructure the superconducting phase transition temperature T c gets as high as 30K. Our results open a new page in creating interfaces with q2deg and HTq2DSC, since it has been shown

2 experimentally that it is possible to create HTq2DSC by a relatively simple method at the boundary of the ferroelectric and parent compound of the high temperature superconductor (PCHTSC). On the one hand, this indicates that the idea of a polarization catastrophe is working. And on the other hand, it opens up new physics, when a polarization catastrophe is associated not with the polar oxides, but with the ferroelectric oxides. Therefore, it becomes possible to create q2deg and HTq2DSC on a non-atomically-flat surface and use more simple technologies and equipment. This highly robust phenomenon is confined within the interface area. If a weak magnetic field is applied perpendicularly to the interface of the heterostructure, a resistance appears, and Tc decreases. That confirms a quasi-two-dimensional nature of the superconductive state. The HTq2DSC arises from strongly increasing carrier density localised in the interface area in copper oxide while the polar discontinuity at the interface leads to the divergence of the electrostatic potential due to the polarization catastrophe 5,10. The proposed concept promises ferroelectrically controlled interface superconductivity which offers the possibility of novel design of electronic devices. The realization of high-t c superconductivity confined to nanometre-sized interface (HTq2DSC) area is a long-term goal because of potential applications 13,14 and the possibility to study quantum phenomena in two dimensions 7,15,16. Typical approaches to the realization of quasi-two-dimensional superconducting layer rely on creation of an ultrathin film of a known superconductor 13,14. Usually, the unique properties of functional materials are achieved due to the effects associated with the complex composition of the interface structure, thus another way is to use heterointerfaces Tailoring q2deg and HTq2DSC at the interface is impossible without the deep understanding of the nature of quasi-two-dimensional states. First, the q2deg has been created at the heterointerfaces between two insulating oxides, LaAlO 3/SrTiO 3, and unique transport properties were observed owing to strong electronic correlations In this case the system becomes superconducting below 300 mk 7. Then the superconductivity at 30 K in bilayers of an insulator (La 2CuO 4) and a metal (La 1.55Sr 0.45CuO 4), neither of which is superconducting in isolation, was reported 1,2. In the first case one of the components, LaAlO 3, was polar oxide where an alternation of the differently charged layers of (LaO) +1 2

3 and (AlO 2) -1 exists and (LaO) +1 /(TiO2) 0 interface displays n-type conductivity 5-8. In the second case the components from family La 2-xSr xcuo 4 (LSCO) were used. It was considered 1,11, that the interface must be atomically perfect to obtain the superconductivity at the interface in copper oxides, because the coherence length is very short (ξ=1-3 nm) 17. Thus the price for result in the both cases was that the interface should be atomically perfect 1,7,12. We believe that, for the case of polar oxides, continuous polarization at the interface occurs only at atomically-flat interface boundary. On the other hand, in the case of a ferroelectric oxide deposited on the copper oxide, the conditions are not so stringent for the appearance of the effect: inhomogeneities of the order of ξ are possible if their envelope is much greater than ξ. In our investigation, a La 2CuO 4 single crystal was grown using a travelling-solvent-floating-zone technique and was characterized by magnetic susceptibility and resistivity measurements. And then a Ba 0.8Sr 0.2TiO 3 ferroelectric oxide was deposited on ab surface of the single crystal by reactive sputtering of stoichiometric targets using RF plasma (RF-sputtering) method 18 at 650 C. Therefore, we tried to combine the advantages of both approaches described above in order to get the maximum effect in the easy way. We have used a La 2CuO 4 (LCO) single crystal as a substrate in order to obtain a high T c. We have used ferroelectric oxide instead of polar oxide to deposit the film, because in this case the polarization at the interface will occur independently of the interface quality (see Fig.1a, 1b). Therefore, we can use a relatively simple method of creating the interface. Typical surface roughness of LCO single crystal determined from atomic force microscopy data before deposition is about 2 nm (see Fig. 1c, 1d), which is slightly more than one unit cell in c direction (1.3 nm in LCO). Heteroepitaxial BSTO ferroelectric thin film (thickness of 200 nm from atomic force microscopy data) was deposited on LCO single crystal (001) substrate. BSTO belongs to ferroelectric perovskites. In the ferroelectric phase below the Curie temperature of ferroelectric phase transition (T c = 353 K) 19 it has a tetragonal unit cell. The as-grown film shows built-in polarization in the [001] crystallographic direction 19. The size of domains is about 200 nm from atomic force microscopy data (see also Supplementary Materials A. and B.). 3

4 The temperature dependences of magnetic susceptibility χ(t) and resistivity ρ(t) of the La 2CuO 4 (LCO) single crystal are shown in Fig. 2a and 2b. The peak in χ(t) clearly observed around 306 K corresponds to the Neel temperature below which a long-range antiferromagnetic order is formed. The temperature dependence of resistivity ρ(t) is usual for LCO 20,21. Both of these results indicate a good quality of the crystal. The interface between the ferroelectric and insulating oxides shows superconducting behaviour with high Tc about 30 K (Fig. 2c and 2d). The beginning of a transition to superconducting state occurs around 40 K, similar to what is observed in bulk LSCO single crystals at optimal doping 20,21. When a weak magnetic field is applied to the heterostructure in the direction perpendicular to the surface of the interface, the finite resistance of the interface appears and it increases with the increasing of the field (see Fig. 3) as it was predicted 22. We did not perform the measurements in higher magnetic fields intentionally, since we know from the previous experiment 23 (see Supplementary Materials C.) that the effects of magnetostriction in relatively small magnetic fields can lead to partial peeling of the film from the substrate resulting in partial or complete disappearance of the observed effect. On the other hand, we believe that in our case, by analogy with LSCO, H C2 is of the order of 39 T, which is inaccessible for us. That confirms a quasi-two-dimensional nature of the superconductive state. The most common mechanism for q2deg is the polarization catastrophe (PC) model 5,10, which was also discussed for the case of ferroelectric/dielectric interface The polar discontinuity at the interface leads to the divergence of the electrostatic potential. In addition to this possibility, the occurrence of HTq2DSC is possible due to the impact of cation interdiffusion (primarily Sr from BSTO to LCO) and oxygen non-stoichiometry. Strontium diffusion is unlikely due to low diffusion coefficient at 650 C. Reduction of oxygen during deposition of the film is also unlikely, since the process is carried out at elevated oxygen pressure. For that matter, an introduction of additional oxygen in this process would be more likely. The following three experimental facts argue against this. The first is that q2deg was created at the interface of the Ba 0.8Sr 0.2TiO 3/LaMnO 3 heterostructure, which was obtained by a 4

5 similar technology using the same equipment 23. It would be unlikely that a change in the oxygen concentration in LaMnO3 could lead to the appearance of q2deg. The second is that the application of the magnetic field, which partially destroys the contact at the interface between Ba 0.8Sr 0.2TiO 3 and LaMnO 3, leads to q2deg disappearance (see Supplementary Materials). From this fact it was concluded that the occurrence of q2deg is related to the proximity effect, rather than to diffusion processes. Therefore, we expect that in our system the effect is also related to the proximity to the ferroelectric layer, rather than to the change in oxygen stoichiometry. And the third fact is illustrated in Fig. 4. Here we applied electrodes for resistance measurements on the back side of the heterostructure (from single crystal side, see Supplementary Materials D) superconducting state is not observed directly. The resistance decreases below a certain temperature but not below 4 Ohm. Here we obtain the results from all heterostructures and superconducting state gives the impact in these results (details in Supplementary Materials). Therefore, we can conclude that this is not a case of surface superconductivity. It means that the oxygen does not penetrate the surface layer during the sputtering of the film. The possibility of the reduction of oxygen in the interface area during deposition had been also discussed for the case of bilayers La 2CuO 4/La 1.55Sr 0.45CuO 1,2 4. It was concluded that Interstitial oxygen in La2CuO4+δ is mobile and, in particular in very thin films, it diffuses out of the sample on the scale of hours or days 1. The experimental results suggest that the polarization catastrophe (PC) model 5,10 can be considered as the main explanation of the observed phenomena. Therefore, we believe that the polar discontinuity at the interface due to ferroelectric polarization leads to the divergence of the electrostatic potential and an atomically-flat surface of the interface is not necessary for this case. As a result, q2deg is formed at the interface, which, when the temperature is lowered below ~ 30 K, becomes HTq2DSC state. This allows us to control the interface superconductivity by applying an electric field, as it was done in the case of the ionic liquid 26. In this way we can switch on and off the 5

6 superconducting properties of the interface and use these phenomena in a novel design of electronic devices. Acknowledgments. We wish to thank I.A. Garifullin, N.N. Garif yanov, R. Khasanov (PSI), A.V.Leontyev, Yu.I. Talanov and R.V. Yusupov for help and advices. R.F.M. is grateful to his wife Aigyul for stimulation of this research. The work at ZPhTI FRC KazSC RAS was funded by Russian Scientific Foundation, research project No V.V.K. acknowledges financial support from Slovenian Research Agency Program P Gozar, A. et. al. High-temperature interface superconductivity between metallic and insulating copper oxides, Nature 455, (2008). 2. Gozar, A. & Bozovic, I. High-temperature interface superconductivity. Physica C: Superconductivity and its applications , (2016). 3. Jian-Feng Ge et. al. Supercon- ductivity above 100 K in a single layer FeSe films on doped SrTiO, Nat. Mater 14, (2014). 4. Qing Lin He et. al. Two-dimensional superconductivity at the interface of a Bi 2Te 3/FeTe heterostructure. Nature Communications 5, 4247 (2014). 5. Ohtomo, A., & Hwang H.Y. Nature 427, 423 (2004). 6. Thiel, S., Hammerl, G., Schmehl, A., Schneider, C.W. & Mannhart J. Tunable quasi-twodimensional electron gases in oxide heterostructures, Science 313, 1942 (2006). 7. Reyren, N. et. al. Superconducting interfaces between insulating oxides., Science 317, 1196 (2007). 6

7 8. Brinkman, A. et. al. Magnetic effects at the interface between non-magnetic oxides. Nature Mater. 6, 493 (2007). 9. Kalabukhov, A. et. al. Effect of oxygen vacancies in the SrTiO 3 substrate on the electrical properties of the LaAlO 3 SrTiO 3 interface. Phys. Rev. B 75, (2007). 10. Biscaras, J. et. al. Two-dimensional superconductivity at a Mott insulator/band insulator interface LaTiO 3/SrTiO 3. Nature Communications 1, 89 (2010). 11. Moetakef, P. et. al. Electrostatic carrier doping of GdTiO 3/SrTiO 3 interfaces. Appl. Phys. Lett. 99, (2011). 12. Bozovic, I. et. al. Giant proximity effect in cuprate superconductors. Phys. Rev. Lett. 93, (2004). 13. Ahn, C. H., Triscone, J.-M. & Mannhart, J. Electric field effect in correlated oxide systems. Nature 424, (2003). 14. Ahn, C. H. et al. Electrostatic modulation of superconductivity in ultrathin GdBa 2Cu 3O 7-d. Science 284, (1999). 15. Berezinskii, V. L. Destruction of long-range order in one-dimensional and 2-dimensional systems having a continuous symmetry group 2. Quantum systems. Sov. Phys. JETP 34, (1972). 16. Kosterlitz, J. M. & Thouless, D. J. Ordering, metastability and phase transitions in twodimensional systems. J. Phys. C 6, (1973). 17. Mourachkine, M. Determination of the coherence length and the Cooper-pair size in unconventional superconductors by tunnelling spectroscopy. Journal of Superconductivity 17, 711 (2004). 18. Mukhortov, V. M., Golovko, Y. I., Tolmachev, G. N. & Klevtzov, A. N. The synthesis mechanism of complex oxide films formed in dense RF plasma by reactive sputtering of stoichiometric targets. Ferroelectrics 247, (2000). 7

8 19. Anokhin, A. S., Razumnaya, A. G., Yuzyuk, Y. I., Golovko, Y. I. & Mukhortov, V. M. Phase transitions in barium-strontium titanate films on MgO substrates with various orientations. Physics of the Solid State 58, (2016). 20. Kastner, M. A. & Birgeneau, R. J. Magnetic, transport and optical properties of monolayer copper oxides. Rev. Mod. Phys. 70, (1998). 21. Takagi, H. et al. Superconductor-to-nonsuperconductor transition in (La 1-xSr x) 2CuO 4 as investigated by transport and magnetic measurements. Phys. Rev. B 40, (1989). 22. Kabanov, V.V., Piyanzina, I.I., Tayurskii, D.A. & Mamin, R.F. Toward a High Temperature Quasi-two-dimensional Superconductivity. ArXiv: (2018) 23. Pavlov, D. P. et al. Two-dimensional electron gas at the interface of Ba 0.8Sr 0.2TiO 3 ferroelectric and LaMnO 3 antiferomagnet. JETP Letters 106, (2017). 24. Niranjan, M.K., Wang, Y., Jaswal, S.S. & Tsymbal, E.Y. Prediction of a switchable twodimensional electron gas at ferroelectric oxide interfaces. Phys. Rev. Lett. 103, (2009). 25. Wang, Y. et al. Prediction of a spin-polarized two-dimensional electron gas at the LaAlO3/EuO (001) interface. Phys. Rev. B 79, (2009). 26. Xiang Leng et al. Electrostatic control of the evolution from a superconducting phase to an insulating phase in ultrathin YBa 2Cu 3O 7-x films. Phys. Rev. Lett. 107, (2011). 8

9 Figures Figure 1. The structures of Ba 0.8Sr 0.2TiO 3/La 2CuO 4 and LaAlO 3/SrTiO 3. The schematic structures of Ba 0.8Sr 0.2TiO 3/La 2CuO 4 (q2deg is marked by red curve on the blue insert) (A) and structure of LaAlO 3/SrTiO 3 with q2deg (shown by transparent red) (B) are shown. AFM image of the La 2CuO 4 single crystal surface without the film (C) illustrates the inhomogeneity of the interface. Its cross-section (D, see also Supplementary Materials A) exhibits a characteristic roughness up to 2 nm. In the case of LaAlO 3/SrTiO 3 heterostructe a perfect interface is necessary for a q2deg creation, where as in the Ba 0.8Sr 0.2TiO 3/La 2CuO 4 heterostructure case smooth interface is not important. 9

10 A ( emu/g) C 0,06 0,05 0,04 0,03 0, T(K) (Ohm cm) B , T (K) D 200 R (Ohm) 100 R (Ohm) T (K) T (K) Figure 2. The magnetic properties and the resistivity of La 2CuO 4 single crystal and the resistance of LaAlO 3/SrTiO 3 heterostructure. The temperature dependence of the magnetic susceptibility (A), and the temperature dependence of the resistivity (B) of La 2CuO 4 single crystal (without ferroelectric film) are typical for this system 20,21 and indicate a good quality of the La 2CuO 4 crystal. The temperature dependence of the resistance of Ba 0.8Sr 0.2TiO 3/La 2CuO 4 heterostructure in the wide temperature range (C) shows that above ~40 K the resistance has usual semiconducting behavior. At the low temperatures (D) the resistance drops very rapidly and superconducting behavior is observed at low temperatures. 10

11 0,20 0,15 R(Ohm) 0,10 0,05 T=22.3 K 0, H (Oe) Figure 3. The magnetic field dependence of the resistance of Ba 0.8Sr 0.2TiO 3/La 2CuO 4 heterostructure. The magnetic field was applied perpendicular to the surface and parallel to c axis of the La 2CuO 4 substrate at T=22.3 K. The magnetic field dependence of the heterostructure resistance shows that non-zero resistance appears at a very low field. The Hc1 for the thin layer of superconductor is very small and magnetic field penetrates in the superconducting layer. 11

12 100 R(Ohm) T(K) Figure 4. The temperature dependence of the resistance of Ba 0.8Sr 0.2TiO 3/La 2CuO 4 heterostructure from the substrate side. The temperature dependence of the heterostructure resistance is measured by electrodes deposited on the LCO surface opposite to the surface with the film. Below 50 K we have impact of the interface and of the substrate in the measured resistance (details in Supplementary Materials D, Fig 4). Below 50 K the temperature dependence is similar to the temperature dependence of the interface (Fig. 2D) but the superconductivity is not observed directly. It means that oxygen does not penetrate to the surface of the LCO sample. 12

13 Figure captions. Figure 1. The structures of Ba 0.8Sr 0.2TiO 3/La 2CuO 4 and LaAlO 3/SrTiO 3. The schematic structures of Ba 0.8Sr 0.2TiO 3/La 2CuO 4 (q2deg is marked by red curve on the blue insert) (A) and structure of LaAlO 3/SrTiO 3 with q2deg (shown by transparent red) (B) are shown. AFM image of the La 2CuO 4 single crystal surface without the film (C) illustrates the inhomogeneity of the interface. Its cross-section (D, see also Supplementary Materials A) exhibits a characteristic roughness up to 2 nm. In the case of LaAlO 3/SrTiO 3 heterostructe a perfect interface is necessary for a q2deg creation, where as in the Ba 0.8Sr 0.2TiO 3/La 2CuO 4 heterostructure case smooth interface is not important. Figure 2. The magnetic properties and the resistivity of La 2CuO 4 single crystal and the resistance of LaAlO 3/SrTiO 3 heterostructure. The temperature dependence of the magnetic susceptibility (A), and the temperature dependence of the resistivity (B) of La 2CuO 4 single crystal (without ferroelectric film) are typical for this system 20,21 and indicate a good quality of the La 2CuO 4 crystal. The temperature dependence of the resistance of Ba 0.8Sr 0.2TiO 3/La 2CuO 4 heterostructure in the wide temperature range (C) shows that above ~40 K the resistance has usual semiconducting behavior. At the low temperatures (D) the resistance drops very rapidly and superconducting behavior is observed at low temperatures. Figure 3. The magnetic field dependence of the resistance of Ba 0.8Sr 0.2TiO 3/La 2CuO 4 heterostructure. The magnetic field was applied perpendicular to the surface and parallel to c axis of the La 2CuO 4 substrate at T=22.3 K. The magnetic field dependence of the heterostructure resistance shows that non-zero resistance appears at a very low field. The Hc1 for the thin layer of superconductor is very small and magnetic field penetrates in the superconducting layer. Figure 4. The temperature dependence of the resistance of Ba 0.8Sr 0.2TiO 3/La 2CuO 4 heterostructure from the substrate side. The temperature dependence of the heterostructure resistance is measured by electrodes deposited on the LCO surface opposite to the surface with the 13

14 film. Below 50 K we have impact of the interface and of the substrate in the measured resistance (details in Supplementary Materials D, Fig 4). Below 50 K the temperature dependence is similar to the temperature dependence of the interface (Fig. 2D) but the superconductivity is not observed directly. It means that oxygen does not penetrate to the surface of the LCO sample. Author Contributions D.P.P., R.R.Z. and R.F.M. designed an experimental setup and carried out the main experiments. D.P.P., V.V.K. and R.F.M. analyzed these experimental data and prepared the figures. V.M.M., T.A., T.K. and Y.K. prepared and characterized the samples. R.R.Z. carried out and analyzed atomic force microscopy experiments and analyzed these results. R.F.M. conceived, designed and implemented the project. V.V.K. and R.F.M. performed all data analysis and developed phenomenological description. R.F.M., V.V.K. wrote the paper with important contributions to all its parts from T.A. and V.M.M. All authors discussed the experimental results and worked out the final concept of the paper. Competing financial interests The authors declare no competing financial interests. Author Information Correspondence and requests for materials should be addressed to R.F.M. (mamin@kfti.knc.ru). 14

15 Supplementary Materials: Tailoring high temperature quasi-twodimensional superconductivity Dmitrii P. Pavlov 1, Rustem R. Zagidullin 1, Vladimir M. Mukhortov 2, Viktor V. Kabanov 1,3, Tadashi Adachi 4, Takayuki Kawamata 5, Yoji Koike 5, and Rinat F. Mamin 1 1 Zavoisky Physical-Technical Institute Federal Research` Center Kazan Scientific Center of RAS, Sibirskii trakt 10/7, Kazan, Russia 2 Southern Scientific Center of RAS, Chehova 41, Rostov-on-Don, Russia 3 Jozef Stefan Institute, Jamova 39, 1000 Ljubljana, Slovenia. 4 Department of Engineering and Applied Sciences, Sophia University, Tokyo, Japan 5 Department of Applied Physics, Tohoku University, Sendai, Japan A. Sample preparation. The La 2CuO 4 single crystal (LCO) was grown using a travelling-solvent-floating-zone technique and was characterized by magnetic susceptibility and resistivity measurements. The surface of the single crystal was polished using diamond paste up to roughness of 1-2 nm and a size of the roughness in the plane was nm (Fig. 1 and Fig. 1c, 1d of the article). Heteroepitaxial Ba 0.8Sr 0.2TiO 3 (BSTO) thin film was deposited on LCO (001) substrate by RF-sputtering of stoichiometric polycrystalline target at 650 C. Transparent BSTO film was realized by means of layer-by-layer growth (Frank-van der Merwe mechanism). Details of the growth conditions have been previously reported in ref. 1. The 15

16 sample was examined by atomic force microscopy method after the deposition of the film. The inhomogeneity of a film surface was nm with the size in the plane of approximately nm (Fig. 1b of the article). The measured thickness was 200 nm (Fig. 2a, 2b). The as-grown film consists of 180-degrees domains which are not compensated due to an interfacial strain and show built-in polarization in the [001] crystallographic direction 1, 2. The size of domains was about 200 nm (Fig. 2a). B. Measurements of resistance and superconducting state at the interface. Resistance measurement on the interface of the heteroctructure was performed by four contacts method as shown on Fig. 1a of the article. The electrodes were applied on the LCO surface at the boundary with film. The electrodes were in contact with the interface. The distance between potential electrodes was different in the different experiments. The current flows by different routes at different temperatures and depends on relation between a substrate conductance and an interface conductance. At high temperature the main current flows through substrate. Below 50 K the main current flows on interface area. The measurement of resistance at magnetic field was performed by two different ways. In one case the heterostructure has been cooled without magnetic field applied and then the resistance was measured in magnetic field at certain temperature changing the magnetic field step by step from 0 to 4000 Oe. In the other case the measurement of resistance was carried out during cooling at the different magnetic fields (Fig. 3 of the article). C. Magnetic field dependence of resistance of Ba0.8Sr0.2TiO3/LaMnO3 heterostructure obtained by the same method. Ba 0.8Sr 0.2TiO 3/LaMnO 3 heterostructure (BSTO/LMO) was grown by the same method as Ba 0.8Sr 0.2TiO 3/La 2CuO 4 heterostructure, thus the results on BSTO/LMO heterostructure give us the information about the features of the method. Temperature dependence of the resistance of BSTO/LMO heterostructure on different magnetic field applied in the plane shows strong and 16

17 irreversible change (Fig. 3). It could be connected with strong magnetostriction effect. Due to that the contact of film with the single crystal could be destroyed partially since it was obtained by the same method. D. Measurements of resistance from the substrate side. Resistance measurement from the back side of heteroctructure was performed as shown on Fig. 4a. The electrodes were deposited on the LCO surface opposite to the surface with film. And electrodes were not in contact with the interface. We believe that the current line distributions are strongly different at different temperatures (Fig. 4b and 4c) and depend on relation between a substrate conductance and an interface conductance. At high temperature the main current flows through substrate. Below 50 K the main current flows on interface area. And when the resistance is measured from the side of substrate the temperature dependence of the heterostructure is the same as that for the interface (Fig. 4 of the article). But superconductivity is not observed directly because the surface of substrate is not superconducting and the resistance is observed from superconducting state (for T<30 K) and from part of substrate (Fig. 4d). The result of the resistance measurement is presented on Fig. 4 of the article. The conclusion is that there is no superconductivity of the substrate surface. It means that the interstitial oxygen does not penetrate in the LCO surface in the interface area during the sputtering of the film. The possibility of the reduction of oxygen in interface area during deposition had been also discussed for bilayers La 2CuO 4/La 1.55Sr 0.45CuO 3,4 4. They conclude that Interstitial oxygen in La2CuO4+δ is mobile and, in particular in very thin films, it diffuses out of the sample on the scale of hours or days Mukhortov, V. M., Golovko, Y. I., Tolmachev, G. N. & Klevtzov, A. N. The synthesis mechanism of complex oxide films formed in dense RF plasma by reactive sputtering of stoichiometric targets. Ferroelectrics 247, (2000). 17

18 2. Anokhin, A. S., Razumnaya, A. G., Yuzyuk, Y. I., Golovko, Y. I. & Mukhortov, V. M. Phase transitions in barium-strontium titanate films on MgO substrates with various orientations. Physics of the Solid State 58, (2016). 3. Gozar, A. et. al. High-temperature interface superconductivity between metallic and insulating copper oxides, Nature 455, (2008). 4. Gozar, A. & Bozovic, I. High-temperature interface superconductivity. Physica C: Superconductivity and its applications , (2016). 18

19 A µm µm B Figure 1. The surface of the La 2CuO 4 single crystal. The cross-section of surface profile of the single crystal without the film at good polish (A) and bad polish (B) areas are shown. 19

20 A 1,2 h(nm) 0,8 0,4 0,0 0,2 0,4 0,6 0,8 1,0 1,2 1,4 1,6 L( m) B h(nm) 1,6 1,2 0,8 0,4 0,0 0,0 0,5 1,0 1,5 L( m) Figure 2. The film thickness measurement. AFM profile (A) and cross-section (B) of surface of the Ba 0.8Sr 0.2TiO 3/La 2CuO 4 heterostructure in the area of film boundary show the thickness of about 200 nm. 20

21 R(Ohm) Ba 0.8 Sr 0.2 TiO 3 /LaMnO 3 H(T) 0 0,2 0,4 0,6 0, T(K) Figure 3. The temperature dependence of the resistance of Ba 0.8Sr 0.2TiO 3/LaMnO 3 heterostructure at different magnetic fields. The magnetic field was applied parallel to the surface and perpendicular to c axis of the LaMnO 3. The sequence of application of the magnetic field is shown in the inset. We measured the resistance in a zero magnetic field before and after application of the magnetic field as shown. Thus the strong and irreversible increasing of resistance is shown. 21

22 A q2deg B Ba 0.8 Sr 0.2 TiO 3 J La 2 CuO 4 J J J V V C D J J R2D Rs V V Figure 4. Schematic illustration of the resistance measurement from the back side of Ba 0.8Sr 0.2TiO 3/La 2CuO 4 heterostructure: contacts layout (A), the schematic current distribution is shown by lines with arrows for T<50 K (B) and T>50 K (C) the lines of potential are shown by the dash lines, and scheme of the effective resistances between potential electrodes (R 2D is resistance of interface, R S is resistance of substrate, an additional resistance of substrate in the vicinity of interface is also shown) (D). Below 50 K the main current flows in the interface area (B). 22

23 Figure captions for Supplementary Materials. Figure 1. The surface of the La 2CuO 4 single crystal. The cross-section of surface profile of the single crystal without the film at good polish (A) and bad polish (B) areas are shown. Figure 2. The film thickness measurement. AFM profile (A) and cross-section (B) of surface of the Ba 0.8Sr 0.2TiO 3/La 2CuO 4 heterostructure in the area of film boundary show the thickness of about 200 nm. Figure 3. The temperature dependence of the resistance of Ba 0.8Sr 0.2TiO 3/LaMnO 3 heterostructure at different magnetic fields. The magnetic field was applied parallel to the surface and perpendicular to c axis of the LaMnO 3. The sequence of application of the magnetic field is shown in the inset. We measured the resistance in a zero magnetic field before and after application of the magnetic field as shown. Thus the strong and irreversible increasing of resistance is shown. Figure 4. Schematic illustration of the resistance measurement from the back side of Ba 0.8Sr 0.2TiO 3/La 2CuO 4 heterostructure: contacts layout (A), the schematic current distribution is shown by lines with arrows for T<50 K (B) and T>50 K (C) the lines of potential are shown by the dash lines, and scheme of the effective resistances between potential electrodes (R 2D is resistance of interface, R S is resistance of substrate, an additional resistance of substrate in the vicinity of interface is also shown) (D). Below 50 K the main current flows in the interface area (B). 23

24 24