⇦Chinese Physics Letters, 2023, Vol. 40, No. 7, Article code 076801 Synthesis of Chemically Sharp Interface in NdNiO$_{3}$/SrTiO$_{3}$ Heterostructures Yueying Li (李月莹)1,2†, Xiangbin Cai (蔡祥滨)3†, Wenjie Sun (孙文杰)1,2, Jiangfeng Yang (杨江枫)1,2, Wei Guo (郭维)1,2, Zhengbin Gu (顾正彬)1,2, Ye Zhu (朱叶)3*, and Yuefeng Nie (聂越峰)1,2* Affiliations 1National Laboratory of Solid State Microstructures, Jiangsu Key Laboratory of Artificial Functional Materials, College of Engineering and Applied Sciences, Nanjing University, Nanjing 210093, China 2Collaborative Innovation Center of Advanced Microstructures, Nanjing University, Nanjing 210093, China 3Department of Applied Physics, Research Institute for Smart Energy, The Hong Kong Polytechnic University, Hong Kong 999077, China Received 11 February 2023; accepted manuscript online 29 May 2023; published online 20 June 2023 †These authors contributed equally to this work.
*Corresponding authors. Email: yezhu@polyu.edu.hk; ynie@nju.edu.cn Citation Text: Li Y Y, Cai X B, Sun W J et al. 2023 Chin. Phys. Lett. 40 076801    Abstract The nickel-based superconductivity provides a fascinating new platform to explore high-$T_{\rm c}$ superconductivity. As the infinite-layer nickelates are obtained by removing the apical oxygens from the precursor perovskite phase, the crystalline quality of the perovskite phase is crucial in synthesizing high quality superconducting nickelates. Especially, cation-related defects, such as the Ruddlesden–Popper-type (RP-type) faults, are unlikely to disappear after the topotactic reduction process and should be avoided during the growth of the perovskite phase. Herein, using reactive molecular beam epitaxy, we report the atomic-scale engineering of the interface structure and demonstrate its impact in reducing crystalline defects in Nd-based nickelate/SrTiO$_{3}$ heterostructures. A simultaneous deposition of stoichiometric Nd and Ni directly on SrTiO$_{3}$ substrates results in prominent Nd vacancies and Ti diffusion at the interface and RP-type defects in nickelate films. In contrast, inserting an extra [NdO] monolayer before the simultaneous deposition of Nd and Ni forms a sharp interface and greatly eliminates RP-type defects in nickelate films. A possible explanation related to the polar discontinuity is also discussed. Our results provide an effective method to synthesize high-quality precursor perovskite phase for the investigation of the novel superconductivity in nickelates.
cpl-40-7-076801-fig1.png
cpl-40-7-076801-fig2.png
cpl-40-7-076801-fig3.png
cpl-40-7-076801-fig4.png
cpl-40-7-076801-fig5.png
 DOI:10.1088/0256-307X/40/7/076801 © 2023 Chinese Physics Society Article Text The recently discovered superconductivity in hole-doped infinite-layer nickelate Nd$_{1- x}$Sr$_{x}$NiO$_{2}$ films, as an analogue of cuprates, has attracted much research attention, providing new opportunities to investigate the underlying physics of high-$T_{\rm c}$ superconductivity (HTSC).[1-11] Due to the instability of $+$1 valence state of nickel, the infinite-layer nickelate is hard to grow directly. Instead, the precursor phase, perovskite nickelate (Nd,Sr)NiO$_{3}$, is synthesized first and soft-chemically reduced subsequently to the infinite-layer phase (Nd,Sr)NiO$_{2}$ with reductive agents, such as sodium hydride.[12,13] Though great advances have been made, it still remains challenging to eliminate lattice defects during the synthesis of superconducting nickelates.[14-16] In the soft-chemical reduction process, only the oxygen content in the oxide is changed while the cation structural framework remains unchanged. That is, the cation-related defects formed during the growth of perovskite phase can persist after the reduction, such as the Ruddlesden–Popper-type (RP-type) stacking fault, a common and influential defect in the superconducting nickelates.[14,15] Therefore, the quality of the precursor phase has direct impacts on the quality of infinite-layer phase, and the cation-related defects should be avoided in advance during the growth process. To optimize the growth of perovskite-phase films, many factors should be taken into account precisely, such as the stoichiometry,[17] growth pressure,[18] and laser fluence.[14,19] In addition, the interface chemical composition also matters. The issue regarding interface effects deserves especial attention in nickelates, given that the superconductivity is hitherto only present in the infinite-layer nickelate films while absent in the bulk- and powdered-form samples without the existence of interface.[20-22] When observed in the [100]$_{\rm p}$ projection, perovskite nickelate NdNiO$_{3}$ consists of alternating (NdO)$^{+}$ and (NiO$_{2})^{-}$ layers [Fig. S1(a) in the Supplementary Materials], which is usually grown on the (001)-oriented SrTiO$_{3}$ (STO) substrates with (SrO)$^{0}$ and (TiO$_{2})^{0}$ layers, leading to the polar discontinuity at the polar/nonpolar interface. The resulting polar catastrophe will affect the quality of nickelate films on STO substrates, including the emergence of secondary phases with Ni$^{2+}$ valence state,[23,24] the rough surface morphology,[25] formation of oxygen vacancies,[26] and other defects.[27] The precise control of interface chemistry, especially the first atomic layer, has been demonstrated to contribute to the lattice structure in some other nonpolar/polar heterostructures such as the SrTiO$_{3}$/LaAlO$_{3}$,[28] SrVO$_{3}$/LaAlO$_{3}$,[29] and SrTiO$_{3}$/CeO$_{2}$.[30] Nevertheless, similar investigations on nickelate films are still lacking. In this work, utilizing the reactive molecular beam epitaxy (MBE), we optimize the crystalline quality of Nd-based nickelate/SrTiO$_{3}$ heterostructures by engineering the interface chemical composition with atomic precision. The defective interface with Nd vacancies, cation diffusion and RP-type stacking faults are observed in NdNiO$_{3}$ (NNO) films grown directly on TiO$_{2}$-terminated SrTiO$_{3}$ substrates. The deposition of one excess [NdO] atomic layer prior to the codeposition of Nd and Ni shows effective suppression of the Nd vacancies and Ti diffusion, which is also beneficial to reduce the lattice defects and to improve the crystalline quality. Further impacts of the interface engineering on the Sr-doped nickelate and the superconductivity are also revealed. The secondary phase with dense RP-type faults is dominated in (Nd,Sr)NiO$_{3}$ films without depositing excess [NdO] at the interface, resulting in a highly insulating behavior of the reduced phase instead of the superconductivity. Our work provides an effective route to modulate the interface structure between polar nickelate films and non-polar SrTiO$_{3}$ substrates, which significantly enhances the quality of nickelate films.
cpl-40-7-076801-fig1.png
Fig. 1. (a) Schematic diagrams of MBE growth and the interface engineering. The Nd and Ni beams are controlled separately by their shutters. Different doses of the inserted extra [NdO] before the codeposition of NdNiO$_{3}$ are shown and named by 0 ML, 0.5 ML, and 1 ML, respectively. (b) RHEED patterns of SrTiO$_{3}$ (STO) substrate (left panel) and NdNiO$_{3}$ (NNO) film (right panel) of the 1 ML sample. (c) RHEED intensity oscillations of 18-u.c.-NdNiO$_{3}$ films of different interface conditions, which are obtained by monitoring the 11 diffraction spots marked by the red dashed rectangles in (b) during the growth. The yellow areas indicate the growth of excess [NdO] monolayer where the Nd shutter is open while Ni shutter closed, and the green areas indicate the codeposition of the NdNiO$_{3}$ where the Nd and Ni shutters are both kept open simultaneously.
The NdNiO$_{3}$ and (Nd,Sr)NiO$_{3}$ films were epitaxially grown on the TiO$_{2}$-terminated (001)-oriented SrTiO$_{3}$ single-crystalline substrates using a DCA R450 MBE system equipped with the in situ reflection high-energy electron diffraction (RHEED). An electron beam of 15 keV energy was utilized during the RHEED measurements. SrTiO$_{3}$ substrates were etched in buffered HF acid for about 70 s and annealed in flowing pure oxygen at 1000 ℃ for 80 min before growth to obtain TiO$_{2}$-terminated step-and-terrace surface.[31] The films were grown at 550–650 ℃ (measured by a thermocouple) and under an oxidant (distilled ozone) background pressure of $\sim$ $4.0 \times 10^{-6}$ Torr. The doping level of Sr in (Nd,Sr)NiO$_{3}$ is 20%. The nominal beam flux ratio of Nd and Ni was measured by the quartz crystal microbalance roughly and optimized precisely to meet the stoichiometry.[15] In the MBE growth of NdNiO$_{3}$ films, the doses of Nd and Ni are controlled independently by their respective shutters, as shown in Fig. 1(a). As a result, the film could be grown via either shuttered method (Nd and Ni deposited alternatively) or co-deposition method (Nd and Ni deposited simultaneously), which is convenient for the atomically precise modulation of the interface structures. In this work, we utilize the co-deposition method to grow the NdNiO$_{3}$ layer and to modulate the interface with pre-depositing the excess [NdO] layer using the shuttered method. The film crystalline structure was examined by XRD using a Bruker D8 Discover diffractometer with Cu $K_\alpha$ source. The specimens for the cross-sectional scanning transmission electron microscopy (STEM) were prepared by focused ion beam (FIB) techniques. Atomic-resolution annular dark-field (ADF) images and energy-dispersive x-ray spectroscopy (EDS) maps were acquired on the JEOL JEM ARM 200F outfitted with an ASCOR fifth-order probe corrector. Figure 2 shows the analysis of atomic structure of the NdNiO$_{3}$ films directly grown on the TiO$_{2}$-terminated SrTiO$_{3}$ in stoichiometry without any interface engineering. The atomic-resolution EDS map reveals the chemical distributions of Nd (deep blue), Ni (green), Sr (light blue), and Ti (red) elements across the NdNiO$_{3}$/SrTiO$_{3}$ interface. The corresponding integrated concentration profiles are also shown in Fig. 2(a). The intensity of Nd has a prominent reduction at the interface layer compared to other layers within the film, that is, considerable Nd vacancies form at the interface. In addition, Ti atoms diffuse to the first NdNiO$_{3}$ layer and occupy part of Ni sites, as marked by the red arrow. The extent of Ti atom diffusion in our observation is comparable to the previous report.[32] Other elements show little diffusion across the interface. Meanwhile, the crystalline quality of NdNiO$_{3}$ films based on the rough interface is examined by the ADF-STEM image. There are some defective regions marked by the red rectangles in Fig. 2(b). Within those regions, the contrast between Nd and Ni atoms are reduced, implying the possible presence of the RP-type faults. To further explore the defects, Fourier filtering is employed, which is usually adopted to analyze the dislocation defects.[33] Figures 2(c) and 2(d) show the Fourier-filtered images from (001) and (100) reflections of the ADF-STEM images, respectively. Here, stacking faults with the half-atomic-layer displacement along $c$ axis are indicated by the white arrows in Fig. 2(c), which are consistent with the defective regions in Fig. 2(b). Edge dislocations with extra vertical atomic layers inserted above or below are also pointed in Fig. 2(d). Combining these two features, we can realize that the defective regions in the films are most likely to be RP-type stacking faults.
cpl-40-7-076801-fig2.png
Fig. 2. (a) EDS mapping and corresponding line profiles of the 0 ML sample. The intensities of all the spectra are normalized for clarity. The position of the NdNiO$_{3}$/SrTiO$_{3}$ interface is indicated by the dashed black line. The red arrow indicates the diffused Ti atoms. (b) Cross-sectional ADF-STEM image of the 0 ML sample. The yellow dashed lines represent the interface between the NdNiO$_{3}$ film and the SrTiO$_{3}$ substrate and that between the film and carbon. [(c), (d)] Fourier-filtered images using (001) (c) and (100) (d) reflections corresponding to the ADF image in (b). Scale bar: 2 nm.
The polar discontinuity between NdNiO$_{3}$ and SrTiO$_{3}$ is commonly considered as the driving force of the interface reconstruction reported both theoretically and experimentally.[32,34] Likewise, the results of our experiment indicate that the alternating [NdO]$^{+}$ and [NiO$_{2}$]$^{-}$ layers with polarity are less inclined to stack spontaneously on the nonpolar layers of the SrTiO$_{3}$ substrate. The formation of Nd vacancies is a possible way of interface reconstruction, which could reduce the polarity of the interfacial [NdO] layer. Therefore, excess Nd may be in demand to form the complete [NdO] layer and to enhance the interface quality. Herein, the [NdO] monolayer (ML) is deposited before the codeposition of Nd and Ni to complete the interfacial [NdO] layer, forming a polar terminated surface on the nonpolar substrate prior to NdNiO$_{3}$ growth. As shown in Fig. 1(a), various dosages of pre-deposited [NdO] monolayer are referred as 0 ML, 0.5 ML, and 1 ML, respectively. The 1 ML represents the insertion of one full layer of [NdO] on the substrate surface and 0.5 ML represents half layer. It is noted that the films were grown under the identical growth condition except for the interface and on the same day to minimize the possible variation in growth parameters, such as beam fluxes and background pressure. The pre-deposition of [NdO] monolayer can be controlled based on the in situ RHEED, which is commonly employed to in situ monitor the growth process and provide the accurate deposition time. As shown in Fig. 1(c), continuous RHEED intensity oscillations of the (11) diffractions marked in Fig. 1(b) are observed during the NdNiO$_{3}$ growth in each case. During the deposition of [NdO] monolayer on the TiO$_{2}$-terminated SrTiO$_{3}$, the RHEED intensity decreases first and then increases to the maximum, the emergence of which indicates that one full [NdO] monolayer has been deposited. Note that the incident angle of the electron beam is related to the mean inner potential and consequently influences the trend of RHEED intensity.[35] Hence, it is necessary to find the appropriate incident angle at which the intensity maximum emerges at the end of [NdO] monolayer growth.[36] After the [NdO] monolayer growth, the Nd and Ni shutters were opened together to grow the NdNiO$_{3}$ film subsequently. The thickness of the co-deposited NdNiO$_{3}$ films is controlled to be nearly identical [18 unit cells (u.c.)] by RHEED intensity oscillations, the period of which corresponds to the growth of one-unit-cell NdNiO$_{3}$.
cpl-40-7-076801-fig3.png
Fig. 3. [(a), (b)] EDS mapping with corresponding line profiles of the 0.5 ML (a) and 1 ML (b) samples. (c) The evolution of intensities of the interface Nd layer and the diffused Ti at the interface under different interface conditions. The error bars represent the fitting error of the line profiles. (d)–(f) Cross-sectional ADF-STEM image (d) and corresponding Fourier-filtered images using (001) (e) and (100) (f) reflections of the 1 ML sample. Scale bar: 2 nm.
The modulated interface structure of the samples with the pre-deposition of [NdO] layer is also revealed by atomic-resolution EDS. As shown in Figs. 3(a) and 3(b), the Nd vacancies have been reduced after interface engineering reflected by the increasing intensity of interface Nd. From 0.5 ML to 1 ML, the interface [NdO] atomic layer is completed gradually. Simultaneously, the extent of Ti diffusion remains virtually unchanged in the 0.5 ML sample but is substantially suppressed in the 1 ML sample, consistent with the minimized Nd vacancies. For clarity, the evolution of Nd intensity of the interface layer along with the diffused Ti intensity on the interface conditions has been extracted from the peaks of first-unit-cell NdNiO$_{3}$ in their respective line profiles and plotted in Fig. 3(c). The error bars denote the fitting error of the intensity peaks. In the perovskite oxide $AB$O$_{3}$, the migration of smaller $B$-site atoms through the larger $A$-site vacancies is under expectation, such as in SrTiO$_{3}$ as reported previously.[37,38] It can be inferred that the coupled existence and disappearance of Nd vacancies and Ti diffusion in our experiments are expected because the diffused Ti atoms can migrate through the Nd-vacancy channel, implying that the suppression of the Nd vacancies is key to rehabilitating the interface structure. Hence, the chemically sharp and abrupt interface with the atomic structure of NdO/TiO$_{2}$ could be produced forcedly when excess Nd is provided at the interface. Moreover, the gradually filled Nd vacancies in the 0.5 ML and 1 ML samples suggest that the dosage of excess Nd is related to the degree of Nd vacancies, which can also be interpreted as the different degree of compensation for polar discontinuity by the Nd vacancies. Additionally, the crystalline quality of the entire film is also enhanced as confirmed by STEM. A sharp contrast between Nd and Ni atoms is observed as shown in Fig. 3(d). In Figs. 3(e) and 3(f), the Fourier-filtered images show well-ordered atomic layers in the film without obvious dislocations. That is, the RP-type faults are less likely to form in films with the sharp interface.
cpl-40-7-076801-fig4.png
Fig. 4. (a) XRD 2$\theta$–$\omega$ patterns of NdNiO$_{3}$ films under different interface conditions along [00$l$] direction. The asterisks denote the peaks of the STO substrate. The curves are offset vertically for clarity. (b) Corresponding RSM images around the STO (103) peak for NdNiO$_{3}$ films. The dashed lines indicate the peak positions. (c) Corresponding temperature-dependent resistivity of the NdNiO$_{3}$ films. The solid and dashed curves represent the cooling and warming processes, respectively. Inset: the $\rho (T)$ before the transition normalized by the resistivity at room temperature ($\rho_{\scriptscriptstyle{\rm RT}}$) shown in linear scale.
Additionally, the XRD patterns and transport properties are studied to investigate the macroscopic impact of the interface engineering. As shown in Fig. 4(a), the 2$\theta$–$\omega$ scans with clear (00$l$) peaks of 1 ML samples further demonstrate the higher macroscopic crystalline quality, while the weaker (003) peak of 0 ML and 0.5 ML samples indicates the lower crystalline quality. It is noted that the RP-type faults only exist in nanoscale regions in our films, so it cannot be identified in XRD scans for the lack of long-range order. The peak position of NdNiO$_{3}$ in the 2$\theta$–$\omega$ scans exhibits an obvious shift along with the interface engineering of excess [NdO] monolayer deposition, indicating that the corresponding out-of-plane lattice constant $c$ decreases gradually. The epitaxial growth without lattice relaxation is confirmed by reciprocal space mapping (RSM) images in all three samples, as shown in Fig. 4(b). Thus, the larger $c$ can reflect the possible lattice expansion due to the defects in the samples without interface engineering. The trend of out-of-plane lattice constant $c$ is more intuitive in the RSM images. The XRD results demonstrate the effective crystallinity improvement of NdNiO$_{3}$ films after the interface engineering, in agreement with the microscopic observations. The intrinsic metal-insulator transition (MIT) behavior resulting from the charge disproportionation together with the hysteresis loop is prototypical in NdNiO$_{3}$ films.[39] Figure 4(c) shows the transport properties of the corresponding NdNiO$_{3}$ films under different interface conditions. In the 0 ML sample, the insulating behavior starting from room temperature to the lowest temperature is observed, consistent with the inferior crystalline quality. With the interface engineering, the insulating behavior becomes weaker with lower resistivity in the 0.5 ML sample and the MIT behavior revives finally in the 1 ML sample. The increase in conductivity together with the gradually enhanced thermal hysteresis verifies the improvement of the overall crystallinity influenced by the interface condition. It is known that the perovskite NdNiO$_{3}$ is the precursor phase of the infinite-layer nickelate. Since the infinite-layer phase is obtained via removing apical oxygen atoms of the NiO$_{6}$ octahedra, the lattice quality of the precursor phase is crucial for the topotactic reduction. As shown in Fig. S2 in the Supplementary Materials, the sample reduced from the 0 ML precursor phase with no inserted [NdO] shows little sign of infinite-layer phase in 2$\theta$–$\omega$ scan, which is highly insulating with resistivity beyond the measurement limit. Instead, the (00$l$) diffraction peaks indicating prototypical infinite-layer phase can be detected in the samples with insertion of the interface [NdO] layer. However, only the NdNiO$_{2}$ sample with one [NdO] monolayer shows the metallic behavior which is corresponding to the self-doping effect in NdNiO$_{2}$, whereas the 0.5 ML sample is still insulating. It is thus clear that the interface engineering is also important to synthesize high-quality infinite-layer nickelate films. Furthermore, we conducted the interface engineering in Sr-doped nickelate films (Fig. 5). XRD pattern of the (Nd,Sr)NiO$_{3}$ film without interfacial [NdO] monolayer (0 ML) shows a strong peak near STO (002) whose position is below 48$^{\circ}$ and a very weak peak near STO (001), demonstrating the formation of a secondary phase consisting of dense RP-type faults.[14] Pre-deposition of one [NdO] monolayer significantly enhances the crystallinity of (Nd,Sr)NiO$_{3}$ phase. Sharp (00$l$) diffraction peaks are detected with correct position of (002) peak at 48.317$^{\circ}$. The metallic behavior is also consistent with previous reports. After the topotactic reduction, similar results for the reduced phase were obtained as the undoped compound. The infinite-layer phase exhibiting superconductivity can be obtained from the 1 ML samples, while the 0 ML sample shows neither infinite-layer phase nor superconductivity owing to the dense defects. The resistivity of the 0 ML (Nd,Sr)NiO$_{2}$ sample runs out of the measurement limit. Hence, the interface engineering via pre-deposition of one full [NdO] monolayer plays a critical role on the crystallinity of Sr-doped NdNiO$_{3}$ and thus crucial to achieving superconductivity in the infinite-layer nickelate.
cpl-40-7-076801-fig5.png
Fig. 5. XRD 2$\theta$–$\omega$ patterns along [00$l$] direction (a) and temperature-dependent resistivity (b) of (Nd,Sr)NiO$_{3}$ and (Nd,Sr)NiO$_{2}$ films obtained under different interface conditions. The resistivity of both 0 ML and 0 ML samples runs out of the measurement limit. Insets: schematic crystal structures of perovskite and infinite-layer phases.
Our work provides a specific case to show the critical role of an interfacial [NdO] layer at the beginning of the film growth for the quality of nickelate films. It should also be noted that the exact initial interfacial composition of nickelate films could be sensitive to the growth techniques and the substrate surface termination. As such, different amounts of intentional deposition of initial interfacial [NdO] layer in different cases could be drastically different. In summary, we provide an effective interface engineering strategy in this work to obtain chemically sharp interface between NdNiO$_{3}$ and SrTiO$_{3}$ and avoid resultant RP-type stacking faults in NdNiO$_{3}$ films. The critical impacts of the interface structure on the overall nickelate film quality and the superconductivity are demonstrated. Our findings are instructive for growth of high-quality nickelate films with sharp interface on nonpolar SrTiO$_{3}$ substrates, aiding in syntheses of high-quality superconducting nickelates. Acknowledgements. This work was supported by the National Key Projects for Research and Development of China (Grant No. 2022YFA1402502 and 2021YFA1400400), the National Natural Science Foundation of China (Grant No. 11861161004), the Fundamental Research Funds for the Central Universities (Grant No. 0213–14380221), the Research Grants Council of Hong Kong (Grant No. N_PolyU531/18), and the Hong Kong Polytechnic University (Grant No. ZVRP). References Superconductivity in an infinite-layer nickelateSuperconductivity in nickel-based 112 systemsA Nickelate RenaissanceSuperconducting Dome in Nd 1 − x Sr x NiO 2 Infinite Layer FilmsPhase Diagram and Superconducting Dome of Infinite-Layer Nd 1 − x Sr x NiO 2 Thin FilmsSuperconductivity in a quintuple-layer square-planar nickelateCharge density waves in infinite-layer NdNiO2 nickelatesMagnetic excitations in infinite-layer nickelatesSingle particle tunneling spectrum of superconducting Nd1-xSrxNiO2 thin filmsIsotropic Pauli-limited superconductivity in the infinite-layer nickelate Nd0.775Sr0.225NiO2The discovery and research progress of the nickelate superconductorsSodium Hydride as a Powerful Reducing Agent for Topotactic Oxide Deintercalation: Synthesis and Characterization of the Nickel(I) Oxide LaNiO2Reversible changes of epitaxial thin films from perovskite LaNiO3 to infinite-layer structure LaNiO2Aspects of the synthesis of thin film superconducting infinite-layer nickelatesImpact of Cation Stoichiometry on the Crystalline Structure and Superconductivity in NickelatesNickelate Superconductivity without Rare‐Earth Magnetism: (La,Sr)NiO 2Effects of Off-Stoichiometry in the Epitaxial NdNiO3 Film on the Suppression of Its Metal-Insulator-Transition PropertiesPressure dependence of the metal-insulator transition in the charge-transfer oxides R NiO 3 ( R =Pr,Nd, Nd 0.7 La 0.3 )Effects of Nonequilibrium Growth, Nonstoichiometry, and Film Orientation on the Metal-to-Insulator Transition in NdNiO3 Thin FilmsSynthesis and characterization of bulk Nd 1 − x Sr x Ni O 2 and Nd 1 − x Sr x Ni O 3 Absence of superconductivity in bulk Nd1−xSrxNiO2Topotactic transformation of single crystals: From perovskite to infinite-layer nickelatesPolarity-driven nickel oxide precipitation in LaNiO 3 -LaAlO 3 superlatticesPolarity compensation in ultra-thin films of complex oxides: The case of a perovskite nickelateEffect of polar discontinuity on the growth of LaNiO3/LaAlO3 superlatticesPolarity-driven oxygen vacancy formation in ultrathin LaNiO 3 films on SrTiO 3 Roles of the first atomic layers in growth of SrTiO3 films on LaAlO3 substratesInterface chemistry and electrical properties of SrVO3/LaAlO3 heterostructuresRole of the First Atomic Layers in Epitaxial Relationship and Interface Characteristics of SrTiO 3 Films on CeO 2 /YSZ/Si(001)Atomic Control of the SrTiO3 Crystal SurfaceResolving the polar interface of infinite-layer nickelate thin filmsEpitaxial Growth of LiMn2 O4 Thin Films by Chemical Solution Deposition for Multilayer Lithium-Ion BatteriesPolarity-induced electronic and atomic reconstruction at NdNiO 2 / SrTiO 3 interfacesChemically specific termination control of oxide interfaces via layer-by-layer mean inner potential engineeringFirst-principles study on migration mechanism in SrTiO3Ti diffusion in La-doped SrTiO3 single crystalsMetal-insulator transitions in NdNiO 3 thin films
[1] Li D F, Lee K, Wang B Y, Osada M, Crossley S, Lee H R, Cui Y, Hikita Y, and Hwang H Y 2019 Nature 572 624
[2] Gu Q Q and Wen H H 2022 Innovation 3 100202
[3] Mitchell J F 2021 Front. Phys. 9 813483
[4] Li D F, Wang B Y, Lee K, Harvey S P, Osada M, Goodge B H, Kourkoutis L F, and Hwang H Y 2020 Phys. Rev. Lett. 125 027001
[5] Zeng S W, Tang C S, Yin X, Li C, Li M, Huang Z, Hu J, Liu W, Omar G J, Jani H, Lim Z S, Han K, Wan D, Yang P, Pennycook S J, Wee A T S, and Ariando A 2020 Phys. Rev. Lett. 125 147003
[6] Pan G A, Ferenc S D, LaBollita H, Song Q, Nica E M, Goodge B H, Pierce A T, Doyle S, Novakov S, Córdova C D, N'Diaye A T, Shafer P, Paik H, Heron J T, Mason J A, Yacoby A, Kourkoutis L F, Erten O, Brooks C M, Botana A S, and Mundy J A 2022 Nat. Mater. 21 160
[7] Tam C C, Choi J, Ding X, Agrestini S, Nag A, Wu M, Huang B, Luo H, Gao P, García-Fernández M, Qiao L, and Zhou K J 2022 Nat. Mater. 21 1116
[8] Lu H, Rossi M, Nag A, Osada M, Li D F, Lee K, Wang B Y, Garcia-Fernandez M, Agrestini S, Shen Z X, Been E M, Moritz B, Devereaux T P, Zaanen J, Hwang H Y, Zhou K J, and Lee W S 2021 Science 373 213
[9] Gu Q Q, Li Y Y, Wan S Y, Li H Z, Guo W, Yang H, Li Q, Zhu X Y, Pan X Q, Nie Y F, and Wen H H 2020 Nat. Commun. 11 6027
[10] Wang B Y, Li D F, Goodge B H, Lee K, Osada M, Harvey S P, Kourkoutis L F, Beasley M R, and Hwang H Y 2021 Nat. Phys. 17 473
[11] Li D F 2021 Sci. Sin.-Phys. Mech. Astron. 51 047405 (in Chinese)
[12] Hayward M A, Green M A, Rosseinsky M J, and Sloan J 1999 J. Am. Chem. Soc. 121 8843
[13] Kawai M, Inoue S, Mizumaki M, Kawamura N, Ichikawa N, and Shimakawa Y 2009 Appl. Phys. Lett. 94 082102
[14] Lee K, Goodge B H, Li D F, Osada M, Wang B Y, Cui Y, Kourkoutis L F, and Hwang H Y 2020 APL Mater. 8 041107
[15] Li Y Y, Sun W J, Yang J F, Cai X B, Guo W, Gu Z B, Zhu Y, and Nie Y F 2021 Front. Phys. 9 719534
[16] Osada M, Wang B Y, Goodge B H, Harvey S P, Lee K, Li D F, Kourkoutis L F, and Hwang H Y 2021 Adv. Mater. 33 2104083
[17] Yamanaka T, Hattori A N, Pamasi L N, Takemoto S, Hattori K, Daimon H, Sato K, and Tanaka H 2019 ACS Appl. Electron. Mater. 1 2678
[18] Obradors X, Paulius L M, Maple M B, Torrance J B, Nazzal A I, Fontcuberta J, and Granados X 1993 Phys. Rev. B 47 12353
[19] Breckenfeld E, Chen Z, Damodaran A R, and Martin L W 2014 ACS Appl. Mater. & Interfaces 6 22436
[20] Wang B X, Zheng H, Krivyakina E, Chmaissem O, Lopes P P, Lynn J W, Gallington L C, Ren Y, Rosenkranz S, Mitchell J F, and Phelan D 2020 Phys. Rev. Mater. 4 084409
[21] Li Q, He C P, Si J, Zhu X Y, Zhang Y, and Wen H H 2020 Commun. Mater. 1 16
[22] Puphal P, Wu Y M, Fürsich K, Lee H, Pakdaman M, Bruin J A N, Nuss J, Suyolcu Y E, van Aken P A, Keimer B, Isobe M, and Hepting M 2021 Sci. Adv. 7 eabl8091
[23] Detemple E, Ramasse Q M, Sigle W, Cristiani G, Habermeier H U, Benckiser E, Boris A V, Frano A, Wochner P, Wu M, Keimer B, and van Aken P A 2011 Appl. Phys. Lett. 99 211903
[24] Middey S, Rivero P, Meyers D, Kareev M, Liu X, Cao Y, Freeland J W, Barraza-Lopez S, and Chakhalian J 2014 Sci. Rep. 4 6819
[25] Liu J, Kareev M, Prosandeev S, Gray B, Ryan P, Freeland J W, and Chakhalian J 2010 Appl. Phys. Lett. 96 133111
[26] Tung I C, Luo G, Lee J H, Chang S H, Moyer J, Hong H, Bedzyk M J, Zhou H, Morgan D, Fong D D, and Freeland J W 2017 Phys. Rev. Mater. 1 053404
[27]Yang H F, Liu Z T, Fan C C, Yao Q, Xiang P, Zhang K L, Li M Y, Liu J S, and Shen D W 2016 AIP Adv. 6 085115
[28] Kim D W, Kim D H, Kang B S, Noh T W, Lee D R, and Lee K B 1999 Appl. Phys. Lett. 74 2176
[29] Kim D W, Kim D H, Noh T W, Char K, Park J H, Lee K B, and Kim H D 2000 J. Appl. Phys. 88 7056
[30] Yamada T, Kiguchi T, Wakiya N, Shinozaki K, and Mizutani N 2003 MRS Online Proc. Libr. 747 17
[31] Kawasaki M, Takahashi K, Maeda T, Tsuchiya R, Shinohara M, Ishiyama O, Yonezawa T, Yoshimoto M, and Koinuma H 1994 Science 266 1540
[32] Goodge B H, Geisler B, Lee K, Osada M, Wang B Y, Li D, Hwang H Y, Pentcheva R, and Kourkoutis L F 2023 Nat. Mater. 22 466
[33] Ikuhara Y H, Gao X P, Huang R, Fisher C A J, Kuwabara A, Moriwake H, and Kohama K 2014 J. Phys. Chem. C 118 19540
[34] He R, Jiang P H, Lu Y, Song Y D, Chen M X, Jin M L, Shui L L, and Zhong Z C 2020 Phys. Rev. B 102 035118
[35] Sun H Y, Mao Z W, Zhang T W, Han L, Zhang T T, Cai X B, Guo X, Li Y F, Zang Y P, Guo W, Song J H, Ji D X, Gu C Y, Tang C, Gu Z B, Wang N, Zhu Y, Schlom D G, Nie Y F, and Pan X Q 2018 Nat. Commun. 9 2965
[36] Haeni J H, Theis C D, and Schlom D G 2000 J. Electroceram. 4 385
[37] Mizoguchi T, Takahashi N, and Lee H S 2011 Appl. Phys. Lett. 98 091909
[38] Gömann K, Borchardt G, Gunhold A, Maus-Friedrichs W, and Baumann H 2004 Phys. Chem. Chem. Phys. 6 3639
[39] Catalan G, Bowman R M, and Gregg J M 2000 Phys. Rev. B 62 7892