⇦Chinese Physics Letters, 2019, Vol. 36, No. 7, Article code 078101 Van der Waals Epitaxy of Anatase TiO$_{2}$ on mica and Its Application as Buffer Layer * Han Xu (徐晗)1, Zhen-Lin Luo (罗震林)1**, Chang-Gan Zeng (曾长淦)1, Chen Gao (高琛)1,2** Affiliations 1National Synchrotron Radiation Laboratory & CAS Key Laboratory of Materials for Energy Conversion, Department of Physics, University of Science and Technology of China, Hefei 230026 2Beijing Advanced Sciences and Innovation Center of Chinese Academy of Sciences, Beijing 101407 Received 5 March 2019, online 20 June 2019 *Supported by the National Key Research and Development Program of China under Grant No 2016YFA0300102, the National Natural Science Foundation of China under Grant Nos 11675179, 11434009 and 11374010, and the Fundamental Research Funds for the Central Universities under Grant No WK2340000065.
**Corresponding author. Email: zlluo@ustc.edu.cn; cgao@ustc.edu.cn Citation Text: Xu H, Luo Z L, Ceng Z G and Gao C 2019 Chin. Phys. Lett. 36 078101    Abstract MICAtronics, based on the functional oxide/mica heterostructures, has recently attracted much attention due to its potential applications in transparent, flexible electronics and devices. However, the weak van der Waals interaction decreases the tolerable lattice mismatch and thus limits the species of function oxides that are able to be epitaxially grown on mica. We successfully fabricate relatively high-quality epitaxial anatase TiO$_{2}$ thin films on mica substrates. Structural analyses reveal that the carefully chosen growth temperature (650$^\circ\!$C) and suitable crystalline phase (anatase phase) of TiO$_{2}$ are the key issues for this van der Waals epitaxy. Moreover, as a buffer layer, the TiO$_{2}$ layer successfully suppresses the decomposition of BiFeO$_{3}$ and the difficulty of epitaxial growth of BiFeO$_{3}$ is decreased. Therefore, relatively high-quality anatase TiO$_{2}$ is proved to be an effective buffer layer for fabricating more functional oxides on mica. DOI:10.1088/0256-307X/36/7/078101 PACS:81.15.-z, 68.55.-a, 61.05.cj © 2019 Chinese Physics Society Article Text MICAtronics,[1] which is based on the functional oxides epitaxially grown on mica substrates, has strongly attracted the interest of researchers. As the most abundant subgroup of phyllosilicates,[2] mica is flexible, transparent and could be fabricated to be ultrathin two-dimensional (2D) flakes.[3–5] Furthermore, the low-cost, light-weight and biocompatibility of mica makes it promising for potential applications in flexible and wearable electronics.[6] Benefiting from these characteristics of mica substrates, intriguing functionalities have been achieved in epitaxial oxide films on mica substrates, i.e., transparent conducting oxides,[7–9] spintronics,[10,11] magnetostrictive oxides,[12] thermochromic devices,[13] as well as ferroelectric memories.[14] However, the interaction between the as-grown functional oxide thin films and the top surface of mica substrates is normally van der Waals interaction.[1,6,15] The weak bonding energy (40–70 meV)[16] of van der Waals interaction restricts the epitaxial growth of oxides with large lattice-misfit, hindering the further development of the MICAtronics. For example, only limited kinds of perovskite oxides, such as SrRuO$_{3}$,[17,18] LaNiO$_{3}$,[19] PZT[20] and SrTiO$_{3}$,[6] have been reported for epitaxial growth on mica substrates directly. An alternative pathway is to use a buffer layer (for example, CoFe$_{2}$O$_{4}$,[20] MoO$_{2}$,[21] Al-doped ZnO,[8,9] LaNiO$_{3 }$[19]) between the desired perovskite oxide films and the mica substrates, which could be able to provide effective epitaxial surface for desired perovskite oxide. In this work, we report the epitaxial growth of high-quality anatase TiO$_{2}$ (TO) thin films on mica substrates (001 plane, muscovite). The van der Waals epitaxial structure and the growth mechanism of TO are revealed using x-ray diffraction (XRD) and x-ray absorption spectroscopy (XAS). In addition, BiFeO$_{3}$/TiO$_{2}$/mica (BFO/TO/mica) heterostructure was successfully prepared, which verifies the potential applications of such an anatase TO layer as a buffer layer for growth of functional perovskite oxides on mica. The muscovite mica (001 plane) single crystal substrates (purchased from the MTI corporation) were firstly cleaved to obtain fresh atomically flat surfaces. Then, TO films were fabricated by pulsed laser deposition (PLD) (KrF excimer laser, $\lambda=248$ nm), using a stoichiometric ceramic target of TiO$_{2}$. The target is commercially available from the MTI corporation. To optimize the growth parameters, the deposition temperature was varied in the range 600–700$^\circ\!$C, while the oxygen pressure was kept at 100 mTorr. For the growth of BFO/TO/mica heterostructure, BFO layers were deposited on the TO buffer layers at 675$^\circ\!$C using a stoichiometric ceramic target of Bi$_{1.1}$FeO$_{3}$ (MTI corporation). After the deposition, the films were cooled down to room temperature with steps of 5$^\circ\!$C/min in 760 mTorr of oxygen. The crystalline qualities of the films were carefully checked by XRD performed using a conventional diffractometer ($\lambda=1.54$ Å, Smartlab, Riguka) and BL14B beamline of Shanghai Synchrotron Radiation Facility ($\lambda=1.024$ Å). Atomic force microscopy (AFM) was utilized (Bruker Dimension Icon microscope) for real-space surface analysis. XAS experiments were performed at the Ti $L$-edge (453–465 eV) at beamline BL12B in the National Synchrotron Radiation Laboratory. High-resolution XRD was employed to analyze the crystalline structure of the TO/mica heterostructure. A typical out-of-plane $\theta$–$2\theta$ scan is shown in Fig. 1(a). For the TO layer, only four diffraction peaks (002, 004, 006 and 008) were detected beside diffraction peaks belonging to the mica substrate, indicating the epitaxy relationship in the out-of-plane direction. Based on the 2$\theta$ value of TO 004 diffraction peak (38.13$^\circ$) and Bragg equation $d=k\lambda /2\sin (\theta)$, the out-of-plane layer spacing of the TO film was calculated to be 9.43 Å, consistent with $d_{001}$ of bulk anatase TO (9.37 Å). Here the slight increase (0.6%) of out-of-plane lattice suggests that the TO layer is underneath small in-plane compressive stress. Subsequently, the crystalline quality of the TO film was further analyzed by measuring the x-ray reflection (XRR). As shown in Fig. 1(b), the pronounced Kiessig fringes observed in the XRR curve indicate a flat surface of TO. Based on the reflection oscillation number and angles, the TO film thickness was calculated to be 12.93 nm. Moreover, benefiting from the merits of mica illustrated above, the obtained TO/mica heterostructure is transparent and flexible, as definitely demonstrated in Fig. 1(c). The size of as-prepared TO/mica is 13 mm$\times$18 mm$\times$0.12 mm, and no observable cracks are created even when the bending angle reaches $\sim $$180^{\circ}$. An FWHM value of 0.25$^{\circ}$ is found in the rocking curve (Fig. 1(d)), which implies a relatively high quality of TO/mica in comparison with the oxides films grown on mica directly.[7,10,21–23] In addition, the RSM around the TO 004 diffraction spot, as exhibited in Fig. 1(e), indicates that no impurity phases present in the film. The surface morphology of the TO thin film was analyzed by AFM, and the typical results are shown in Fig. 1(f). Relatively flat surface with roughness below 1.3 nm is seen, consistent with the XRR result. The clear domain morphology implies an island growth mode for the TO thin film, which stems from the weak bonding nature of van der Waals force, and is common for films grown on mica surface.[21]
cpl-36-7-078101-fig1.png
Fig. 1. (a) The out-of-plane $\theta$–$2\theta$ scan of TO/mica. (b) The x-ray reflection (XRR) of TO/mica. (c) The transparency and flexibility of the TO/mica sample. (d) The rocking curve of the TO 002 diffraction peak. FWHM of 0.25$^\circ\!$ is observed. (e) The reciprocal space mapping (RSM) result around the TO 004 diffraction spot. (f) The AFM images of TO thin film surface with 5$\times $5 µm$^{2}$ (left) and 1$\times $1 µm$^{2}$(right).
cpl-36-7-078101-fig2.png
Fig. 2. (a) The $\varphi$-scan of TO 116 and mica 202 diffraction peaks (displayed in polar coordinate form). (b) The complete diffraction diagrams of TO/mica heterostructure with different growth temperatures. (c) The enlarged representations of TO 004 and mica 004 diffraction peaks. (d) The XAS results of TO thin films at different growth temperatures (630$^\circ\!$C and 640$^\circ\!$C).
The normally weak van der Waals bonding between the film and mica 001 surface will make researchers doubt the epitaxial nature of the TO thin films. To reply to this concern, $\varphi$-scan diffraction was further performed for the TO/mica heterostructure. As shown in Fig. 2(a), the diffraction peaks belonging to anatase TO 116 and the mica 202 were found to locate at the same $\varphi$ angles with 120$^\circ\!$ intervals, which verify the good in-plane orientation and the epitaxy nature of the TO films. The revealed in-plane orientations indicate that the epitaxial relationship is [110]$_{\rm TO}$//[100]$_{\rm mica}$. The lattice constant of bulk anatase TO along [110] is 0.528 nm, while $d_{100}$ of mica is 0.519 nm. The small lattice mismatch ($\sim $1.7%) may be the reason for the above-found van der Waals epitaxy of relatively high-quality TO thin films. During the optimization of growth parameters for TO/mica heterostructure, we find that the growth temperature window is quite asymmetric according to the XRD results. As shown in Figs. 2(b) and 2(c), the diffraction peaks of TO do not appear at temperatures below 630$^\circ\!$C. However, the crystallinity of TO is enhanced significantly at 650$^\circ\!$C, and is then suppressed as the temperature further increases up to 700$^\circ\!$C. The XAS experiments performed at Ti $L$-edge (Fig. 2(d)) reveal that a phase transformation occurs in the TO film at a temperature between 630$^\circ\!$C and 640$^\circ\!$C, as the 630$^\circ\!$C-grown TO is rutile phase while the 640$^\circ\!$C-grown one is anatase phase.[24,25] For oxide/mica with small lattice mismatch, the combined island growth and the following lateral growth mode could be able to form a continuous epitaxial film.[21] Thus, for TO/mica, low temperature growth with high-intensity small island[26,27] could increases the sample quality in principle. However, as revealed by the XAS results, the crystalline phase of TO is rutile phase while grown below $\sim $$630^\circ\!$C, with a larger lattice mismatch compared to that of mica. Therefore, the best crystallinity appears after the rutile-anatase phase transition of TO, i.e., at 650$^\circ\!$C. Further increasing the growth temperature, however, may induce possible large island growth[26] and thus decrease the quality. That is to say, the growth temperature not only influences the van der Waals bonding but also modulates the crystalline phase and growth mode of TO. As mentioned above, the weak van der Waals bonding between mica and as-grown thin film greatly limits the epitaxial growth of functional perovskite oxides. An adaptive route is choosing a suitable material as the buffer layer. To demonstrate this, BFO/mica and BFO/TO/mica heterostructures are fabricated. BFO is a prototype simple-phase perovskite multiferroic material, and the most suitable epitaxial planes grown on mica are BFO 111 plane (2$\times d_{\rm BFO 110}=0.558$ nm) and mica 001 plane ($d=0.519$ nm), respectively.[1] Even though, the resulting in-plane lattice mismatch is still relatively larger, offering a good platform to investigate the potential of anatase TO as a buffer layer. The $\theta$–$2\theta$ scan of BFO/mica is shown in Fig. 3(a). The diffraction peaks at 38.07$^\circ\!$ and 57.73$^\circ\!$ belong to BFO 111 and Bi$_{2}$O$_{2.5}$ 402 peaks, which reveal the existence of Bi$_{2}$O$_{2.5}$ and the decomposition of the epitaxial BFO film. In contrast, for BFO/TO/mica only the 111 diffraction peak of BFO and the 004 diffraction peak of TO are observed (Fig. 3(b)). The AFM image (Fig. 3(c)) shows the domain morphology of BFO/TO/mica, which implies that the BFO layer follows the island-growth mode of the TO layer, and the $\varphi$-scan (Fig. 3(d)) reveals that the in-plane epitaxial relationship of BFO/TO/mica is [110]$_{\rm BFO}$//[110]$_{\rm TO}$//[100]$_{\rm mica}$ direction. Thus the application of TO buffer suppresses the decomposition of BFO and the difficulty of epitaxial growth of BFO is decreased. This experimental example demonstrates that high-quality anatase TO thin films can be used as an adapter to transfer the epitaxial growth mode from weak van der Waals bonding to conventional chemical bonding.
cpl-36-7-078101-fig3.png
Fig. 3. (a) The $\theta$–$2\theta$ scan of BFO/mica. (b) The $\theta$–$2\theta$ scan of BFO/TO/mica heterostructure. (c) The AFM image of BFO/TO/mica thin film surface. (d) The $\varphi$-scan of BFO 112 and mica 202 diffraction peaks (displayed in polar coordinate form).
In summary, we have successfully prepared high-quality epitaxial anatase TO thin films on mica substrates (001 muscovite). Systemat structural characterizations reveal that the carefully chosen growth parameters and suitable crystalline phase of TO are the keys for the realization of van der Waals epitaxy. Moreover, the successful preparation of epitaxial BFO/TO/mica heterostructure implies that the high-quality anatase TO could be used as an effective buffer layer for the growth of other functional oxides on mica, and thus to contribute to the further development of MICAtronics, especially the fabrication of transparent and flexible electronic devices. This work was partially carried out at the USTC center for Micro and Nanoscale Research and Fabrication. We thank the beamtime provided by Shanghai Radiation Facility (SSRF) beamline BL14B, BL08U and National Synchrotron Radiation Laboratory (NSRL) beamline BL12B. References MICAtronics: A new platform for flexible X-tronicsQuantifying the Mechanisms of Site-Specific Ion Exchange at an Inhomogeneously Charged Surface: Case of Cs + /K + on Hydrated Muscovite Mica2D Mica Crystal as Electret in Organic Field-Effect Transistors for Multistate MemoryMechanical properties of freely suspended atomically thin dielectric layers of micaLayer-by-layer removal of insulating few-layer mica flakes for asymmetric ultra-thin nanopore fabricationVan der Waals oxide heteroepitaxyOxide Heteroepitaxy for Flexible OptoelectronicsAZO/Ag/AZO transparent flexible electrodes on mica substrates for high temperature applicationHigh-Quality AZO/Au/AZO Sandwich Film with Ultralow Optical Loss and Resistivity for Transparent Flexible ElectrodesHeteroepitaxy of Fe 3 O 4 /Muscovite: A New Perspective for Flexible SpintronicsStrain-mediated magnetic properties of epitaxial cobalt ferrite thin films on flexible muscoviteFlexible Heteroepitaxy of CoFe 2 O 4 /Muscovite Bimorph with Large MagnetostrictionFree-standing SWNTs/VO 2 /Mica hierarchical films for high-performance thermochromic devicesFlexible, Fatigue-Free, and Large-Scale Bi 3.25 La 0.75 Ti 3 O 12 Ferroelectric MemoriesAtomic-Scale Distribution of Water Molecules at the Mica-Water Interface Visualized by Three-Dimensional Scanning Force MicroscopyProgress, Challenges, and Opportunities in Two-Dimensional Materials Beyond GrapheneMechanically Tunable Magnetic Properties of Flexible SrRuO 3 Epitaxial Thin Films on Mica SubstratesFlexible PbZr 0.52 Ti 0.48 O 3 Capacitors with Giant Piezoelectric Response and Dielectric TunabilityToward artificial intelligent self-cooling electronic skins: Large electrocaloric effect in all-inorganic flexible thin films at room temperatureAll-inorganic flexible piezoelectric energy harvester enabled by two-dimensional micaVan der Waals epitaxy of functional MoO 2 film on mica for flexible electronicsTowards the development of flexible non-volatile memoriesMechanical Strain-Tunable Microwave Magnetism in Flexible CuFe 2 O 4 Epitaxial Thin Film for Wearable SensorsSolvothermal Synthesis and Photoreactivity of Anatase TiO 2 Nanosheets with Dominant 001 FacetsComparison of the electronic structure of anatase and rutile TiO 2 single-crystal surfaces using resonant photoemission and x-ray absorption spectroscopySize control of self-assembled InP/GaInP quantum islandsCompressive Stress in Polycrystalline Volmer-Weber Films
[1] Bitla Y and Chu Y H 2017 FlatChem 3 26
[2] Loganathan N and Kalinichev A G 2017 J. Phys. Chem. C 121 7829
[3] Zhang X, He Y, Li R, Dong H and Hu W 2016 Adv. Mater. 28 3755
[4] Castellanos G A, Poot M, Amor-Amorós A, Gary A S, Herre S J, Nicolás A and Gabino R B 2012 Nano Res. 5 550
[5] Gao J, Guo W, Geng H, Hou X, Shuai Z and Jiang L 2012 Nano Res. 5 99
[6] Chu Y H 2017 npj Quantum Mater. 2 67
[7] Bitla Y, Chen C, Lee H C, Do T H, Ma C, Qui L V, Huang C, Wu W, Chang Li, Chiu H and Chu Y 2016 ACS Appl. Mater. Interfaces 8 32401
[8] Li M, Wang Y and Wei X 2017 Ceram. Int. 43 15442
[9] Zhou H, Xie J, Mai M, Wang J, Shen X, Wang S, Zhang L, Kisslinger K, Wang H, Zhang J, Li Y, Deng J, Ke S and Zeng X 2018 ACS Appl. Mater. Interfaces 10 16160
[10] Wu P C, Chen P F, Do T H, Hsieh Y, Ma C, Ha T, Wu K, Wang Y, Li H, Chen Y, Juang J, Yu P, Eng L, Chang C, Chiu P, Tjeng L and Chu Y 2016 ACS Appl. Mater. Interfaces 8 33794
[11] Gao G Q, Jin C, Zheng W C, Zheng D X and Bai H 2018 Europhys. Lett. 123 17002
[12] Liu H J, Wang C K, Su D, Amrillah T, Hsieh Y, Wu K, Chen Y, Juang J, Eng L, Jen S and Chu Y 2017 ACS Appl. Mater. Interfaces 9 7297
[13] Chen Y, Fan L, Fang Q, Xu W, Chen S, Zan G, Ren H, Song L and Zou C 2017 Nano Energy 31 144
[14] Su L, Lu X, Chen L, Wang Y, Yuan G and Liu J 2018 ACS Appl. Mater. Interfaces 10 21428
[15] Fukuma T, Ueda Y, Yoshioka S and Asakawa H 2010 Phys. Rev. Lett. 104 016101
[16] Butler S Z, Hollen S M, Cao L, Cui Y, Gupta J A, Gutierrez H, Heinz T F, Hong S S, Huang J, Lsmach A F, Johnston-Halperin E, Kuno M, Plashnitsa V V, Robinson R D, Ruoff R S, Salahuddin S, Shan J, Shi L, Spencer M G, Terrones M, Windl W and Goldberger J E 2013 ACS Nano 7 2898
[17] Liu J, Feng Y, Tang R, Zhao R, Guo J, Shi D and Yang H 2018 Adv. Electron. Mater. 4 1700522
[18] Gao W, You L, Wang Y, Yuan G, Chu Y, Liu Z and Liu J 2017 Adv. Electron. Mater. 3 1600542
[19] Wang D, Chen X, Yuan G, Jia Y, Wang Y, Mumtaz A, Wang Y and Liu J 2019 J. Materiomics 5 66
[20] Wang D, Yuan G, Hao G and Wang Y 2018 Nano Energy 43 351
[21] Ma C H, Lin J C, Liu H J, Do T, Zhu Y, Ha T, Zhan Q, Juang J, He Q, Arenholz E, Chiu P and Chu Y 2016 Appl. Phys. Lett. 108 253104
[22] Jiang J, Bitla Y, Huang C W, Do T, Liu H, Hsieh Y, Ma C, Jang C, Lai Y, Chiu P, Wy W, Chen Y, Zhou Y and Chu Y 2017 Sci. Adv. 3 e1700121
[23] Liu W, Liu M, Ma R, Zhang R, Zhang W, Yu D, Wang Q, Wang J and Wang H 2018 Adv. Funct. Mater. 28 1705928
[24] Yang H G, Liu G, Qiao S Z, Sun C, Jin Y G, Smith S C, Zou J, Cheng H M and Lu G Q 2009 J. Am. Chem. Soc. 131 4078
[25] Thomas A G, Flavell W R, Mallick A K, Kumarasinghe A R, Tsoutsou D, Khan N, Chatwin C, Rayner S, Smith G C, Stockbauer R L, Warren S, Johal T K, Patel S, Holl, D, Taleb A and Wiame F 2007 Phys. Rev. B 75 035105
[26] Porsche J, Ruf A, Geiger M and Scholz F 1998 J. Cryst. Growth 195 591
[27] Koch R, Hu D and Das A K 2005 Phys. Rev. Lett. 94 146101