Multiferroic properties of sputtered BiFeO3 thin films - NTU.edu

32 downloads 109 Views 384KB Size Report
experiment. 1C. Michel, J.-M. Moreau, G. D. Achenbach, R. Gerson, and W. J. James, ... D. Bucci, B. K. Robertson, and W. J. James, J. Appl. Crystallogr. 5, 187.
APPLIED PHYSICS LETTERS 92, 132908 共2008兲

Multiferroic properties of sputtered BiFeO3 thin films Yibin Li,1,2,a兲 Thirumany Sritharan,1 Sam Zhang,3 Xiaodong He,2 Yang Liu,4 and Tupei Chen4 1

School of Materials Science and Engineering, Nanyang Technological University, 50 Nanyang Avenue, Singapore 639798, Singapore 2 Center for Composite Materials, School of Astronautics, Harbin Institute of Technology, P.O. Box 3010, Harbin 150001, People’s Republic of China 3 School of Mechanical and Aerospace Engineering, Nanyang Technological University, 50 Nanyang Avenue, Singapore 639798, Singapore 4 School of Electrical and Electronic Engineering, Nanyang Technological University, Nanyang Avenue 639798, Singapore

共Received 13 December 2007; accepted 25 February 2008; published online 2 April 2008兲 A cosputtering method was used to deposit BiFeO3 thin films on Pt/ Ti/ SiO2 / Si substrates. It was confirmed as a polycrystalline film with a tetragonal crystal structure in the annealed state. Both Fe2+ and Fe3+ ions were found to coexist in the film. The leakage current density is as low as 10−3 A / cm2 at 120 kV/ cm. This sputtered film shows multiferroic properties exhibiting a saturated ferroelectric loop with a large remnant polarization of 37 ␮C / cm2 and a saturated ferromagnetic loop with saturation magnetization of 21 emu/ cm3 at room temperature. © 2008 American Institute of Physics. 关DOI: 10.1063/1.2901871兴 BiFeO3 共BFO兲 as one of a few multiferroic materials has attracted increasing interest due to its high ferroelectric Curie temperature 共TC ⬃ 1100 K兲 and high antiferromagnetic Néel temperature 共TN ⬃ 650 K兲, which appears to have the best potential applications such as information storage, spintronics, and sensors.1 In theory, the spontaneous polarization of bulk BFO could be up to 91.5 ␮C / cm2 because of its large atomic displacement and high Curie temperature.2 Unfortunately, experimental studies3,4 showed unexpectedly weak remnant polarization 共Pr兲 of 3.5 ␮C / cm2 along the 共001兲 direction and 6.1 ␮C / cm2 along the 共111兲 orientation for bulk BFO due to high leakage current at a high electric field. Excitingly, Wang et al. have reported well-saturated ferroelectric hysteresis loops with a large Pr of 50– 150 ␮C / cm2 and large coercivity 共Hc兲 of 120– 200 kV/ cm in epitaxial BFO films on single crystal SrTiO3 substrates.5,6 Subsequently, much effort has been made to study the epitaxial BFO thin films deposited by sol-gel,7 magnetron sputtering,8,9 and chemical solution deposition.10 Nevertheless, the selection of single crystal substrates inevitably leads to high product cost and relatively large resistivity, which will hamper the commercialization process. These weaknesses could be overcome by depositing polycrystalline BFO films on commercial Pt/ Ti/ SiO2 / Si substrates. Magnetron sputtering was also used to deposit polycrystalline BFO films with a competitive Pr of 30– 50 ␮C / cm2 but the saturation magnetization 共M s兲 was weak 共3 – 5 emu/ cm3兲.11,12 Furthermore, the film composition during sputtering single BFO target is difficult to control. In this work, we used a cosputtering technique where an additional Fe target was utilized to precisely control the film composition as Fe deficiency was commonly found if only one BFO target was sputtered. Both sintered Bi1.1FeO3 共99.9% in purity兲 and metallic Fe 共99.95% in purity兲 targets were used to cosputter BFO films on Pt/ Ti/ SiO2 / Si substrates. The power densities for Bi1.1FeO3 and Fe targets are 1.8 and 1.06 W / cm2, respectively. The process pressure was set to 1.5 Pa with an Ar: O2 a兲

Electronic mail: [email protected].

ratio of 4:1. The substrate temperature was 450 ° C. The asdeposited BFO films with 320 nm thickness were annealed in a rapid thermal processor at 600 ° C for 5 min. A top Au electrode was sputtered onto the annealed BFO films through a shadow mask to produce a circular diode of 0.2 mm diameter for polarization measurement. Grazing incidence x-ray diffraction 共GIXRD兲 pattern in Fig. 1 reveals the polycrystalline nature with a perovskite structure and no impurities were detectable. The inset 共a兲 magnifies the peak in the vicinity of 2␪ = 32° where the 共104兲 and 共110兲 reflections overlap into a single broad peak indicating an orthorhombic or tetragonal structure. This is different from a previous report where the 共104兲 and 共110兲 reflections were clearly separated, suggesting that the BFO film in their study was rhombohedrally distorted.13 Lattice parameters obtained by Rietveld refinement using TOPAS Version 3.0 are a = 3.984 Å and c = 4.006 Å 共c / a = 1.019兲. This further confirms the tetragonal structure of the film. The field emission scanning electron microscope image in Fig. 1 inset 共b兲 demonstrates a dense surface morphology with a grain size of about 100 nm. Normalized time of flight secondary

FIG. 1. GIXRD pattern and surface morphology of annealed BiFeO3 thin film.

0003-6951/2008/92共13兲/132908/3/$23.00 92, 132908-1 © 2008 American Institute of Physics Downloaded 05 Apr 2008 to 155.69.4.4. Redistribution subject to AIP license or copyright; see http://apl.aip.org/apl/copyright.jsp

132908-2

Li et al.

Appl. Phys. Lett. 92, 132908 共2008兲

FIG. 4. 共Color online兲 共a兲 Typical J-E characteristics of Au/ BiFeO3 / Pt capacitor and various fittings of these data are shown to help determine the leakage mechanism: 共b兲 SCLC and 共c兲 FN tunneling.

FIG. 2. 共Color online兲 SIMS profile of annealed BiFeO3 / Pt/ Ti/ SiO2 / Si multilayers.

ion mass spectrometry 共SIMS兲 analysis in Fig. 2 confirms a clearly defined multilayered structure with sharp interfaces. Aside from this, SIMS depth profiling also shows compositional uniformity of Bi, Fe, and O depthwise indicating that cosputtering could produce chemically homogeneous films. The binding energy of Fe 2p measured by x-ray photoelectron spectroscope 共XPS兲 is plotted in Fig. 3, where peak deconvolution and fitting to data show that the experimental Fe 2p peak could be fitted well by using two peaks corresponding to the chemical states Fe3+ at 710.8 eV and Fe2+ at 709.4 eV of 2P3/2. Thus, both charge states 共3+ and 2+兲 are simultaneously present but 3+ state appears to be dominant. The presence of Fe2+ ions signifies oxygen vacancies to meet electrical neutrality. It must be noted that the metallic Fe is not present proving the absence of Fe precipitates. The leakage current density versus electric field 共J-E兲 characteristics of an Au/BFO/Pt capacitor measured at 300 K is shown in Fig. 4共a兲. At the applied field of ⫾120 kV/ cm, J is around 10−3 A / cm2, comparable to or lower than the previous reports.14–16 The asymmetry of J-E characteristics im-

plies different mechanisms operating at negative and positive biases. To investigate the leakage mechanism, the experimental data were plotted in four manners corresponding to the four conduction mechanism models: Schottky emission; space-charge-limited conduction 共SCLC兲; Fowler–Nordheim 共FN兲 tunneling, and Poole–Frenkel 共PF兲 emission.17–19 The Schottky and PF emissions were excluded as possible mechanisms because no linearity could be obtained in both ln共J / T2兲 ⬃ E1/2 and ln共J / T2兲 ⬃ E1/2 plots 共not shown兲 in either positive or negative bias. Linearity was however, obtained in the log共J兲-log共E兲 plot with a slope of 1 in Fig. 4共b兲 when E 艋 −1 kV/ cm, indicating the normal Ohmic contact behavior. When −190艋 E 艋 −1 kV/ cm, the plot is still linear but with a slope of 2.9 indicating a power law relation 共J ⬀ E2.9兲. This matches with the characteristic of SCLC mechanism with deep traps where the space charge originates from traps in the band gap induced by oxygen vacancies, similar to previous reports.14,20 As shown earlier, oxygen vacancies are indeed present in this BFO film. These oxygen vacancies could then generate deep-trap energy levels in the band gap for activated electrons to become mobile. Consequently, the free carriers increase with increase in oxygen vacancies. Linearity was also identified in ln共J / E2兲 versus 1 / E plot when E 艌 110 kV/ cm, as shown in Fig. 4共c兲, which means FN tunneling effect governs the leakage mechanism. In the FN mechanism, charge carriers are injected into an insulator layer 共BFO兲 from electrodes 共Au兲 by tunneling effect. It is worth mentioning that at positive bias lower than 110 kV/ cm, experimental data fail to fit any of these four leakage mechanisms suggesting none of them dominates the leakage. The ferroelectric nature of BFO film measured at 300 K is shown in Figure 5共a兲, where Pr = 37 ␮C / cm2 and Hc = 110 kV/ cm are yielded. Although the Pr value is smaller than that reported for epitaxial films grown on single crystal SrTiO3 substrates 共55 ␮C / cm2 on 关100兴 SrTiO3 and 95 ␮C / cm2 on 关111兴 SrTiO3兲,21 it is much higher than those of sol-gel derived films22 and the bulk material3,4 共less than 2 ␮C / cm2兲. Thus, the distortion in our films should be larger on polarization compared to that in sol-gel derived films. Magnetization measurement performed on a Lakeshore 736 vibrating sample magnetometer at room temperature demonstrates that a ferromagnetic characteristic with a good saturation magnetization 共M s = 21 emu/ cm3兲 is evident in Fig. 5共b兲. The remnance 共1 emu/ cm3兲 and coercivity 共70 Oe兲 were, however, low. Here, M s value is much enhanced in comparison with experimental measurements in the bulk ceramic23 and theoretical simulations.24 This enhancement could be attributed to oxygen vacancies as a perfectly stoichiometric BFO crystal should not show any net magnetic moment. Some models have been proposed to explain this enhancement in terms of oxygen vacancies. Coey et al.25

FIG. 3. 共Color online兲 X-ray photoemission spectrum and deconvolution of Fe 2p in BiFeO3 film. Downloaded 05 Apr 2008 to 155.69.4.4. Redistribution subject to AIP license or copyright; see http://apl.aip.org/apl/copyright.jsp

132908-3

Appl. Phys. Lett. 92, 132908 共2008兲

Li et al.

dominant leakage mechanism in negative bias is spacecharge limited conduction while the FN tunneling dominates in positive bias, resulting in asymmetric J-E curve. The film shows multiferroic properties with a remnant polarization of 37 ␮C / cm2 and a saturation magnetization of 21 emu/ cm3 which are of good magnitude for commercial exploitation. This work was done under the A*STAR-SERC-PSF Grant No. 052 101 0019. The authors would like to thank Lai Mei Ying 共IMRE, A*STAR, Singapore兲 for SIMS experiment. 1

FIG. 5. 共a兲 P-E hysteresis loop and 共b兲 M-H hysteresis loop, measured at room temperature for BiFeO3 thin film.

proposed an F-center exchange mechanism model where spin-polarized electrons trapped at oxygen vacancy sites cause the high magnetic moment in Fe-doped SnO2 thin films. This F-center mechanism requires clustering of the magnetic ions around vacancies. No evidence for such clustering of Fe was forthcoming in this study. Particularly, there was no metallic Fe in the film as evidenced by XPS investigations. The second possibility is a ferrimagnetic arrangement in which the moments of the Fe2+ ions are aligned opposite to those of the Fe3+ ions, leading to a net magnetic moment.26 The third possible mechanism is an increase in the canting angle of Fe ions which could be driven by the oxygen vacancy generation.26 The latter two mechanisms are the possible reasons for observed high magnetization but further study is still required to clarify this. In conclusion, the sputtered BiFeO3 thin film is polycrystalline with tetragonal structure. Both Fe2+ and Fe3+ ions coexist in BFO film which implies the formation of oxygen vacancies. The Au/ BiFeO3 / Pt capacitor gives a leakage current density of about 10−3 A / cm2 at 120 kV/ cm. The pre-

C. Michel, J.-M. Moreau, G. D. Achenbach, R. Gerson, and W. J. James, Solid State Commun. 7, 701 共1969兲. 2 J. B. Neaton, C. Ederer, U. V. Waghmare, N. A. Spaldin, and K. M. Rabe, Phys. Rev. B 71, 014113 共2005兲. 3 J. D. Bucci, B. K. Robertson, and W. J. James, J. Appl. Crystallogr. 5, 187 共1972兲. 4 J. R. Teague, R. Gerson, and W. J. James, Solid State Commun. 8, 1073 共1970兲. 5 J. Wang, J. B. Neaton, H. Zheng, V. Nagarajan, S. B. Ogale, B. Liu, D. Viehland, V. Vaithyanathan, D. G. Schlom, U. V. Waghmare, N. A. Spaldin, K. M. Rabe, M. Wuttig, and R. Ramesh, Science 299, 1719 共2003兲. 6 J. Wang, H. Zheng, Z. Ma, S. Prasertchoung, M. Wuttig, R. Droopad, J. Yu, K. Eisenbeiser, and R. Ramesh, Appl. Phys. Lett. 85, 2574 共2004兲. 7 Y. Wang, Q. H. Jiang, H. C. He, and G. W. Nan, Appl. Phys. Lett. 88, 142503 共2006兲. 8 R. R. Das, D. M. Kim, S. H. Baek, C. B. Eom, F. Zavaliche, S. Y. Yang, R. Ramesh, Y. B. Chen, X. Q. Pan, X. Ke, M. S. Rzchowski, and S. K. Streiffer, Appl. Phys. Lett. 88, 242904 共2006兲. 9 Y.-H. Lee, C.-C. Lee, Z.-X. Liu, C.-S. Liang, and J.-M. Wu, Electrochem. Solid-State Lett. 9, F38 共2006兲. 10 H. Uchid, R. Ueno, H. Funakubo, and S. Koda, J. Appl. Phys. 100, 014106 共2006兲. 11 R. Y. Zheng, X. S. Gao, Z. H. Zhou, and J. Wang, J. Appl. Phys. 101, 054104 共2007兲. 12 Y.-H. Lee, C.-S. Liang, and J.-M. Wu, Electrochem. Solid-State Lett. 8, F55 共2005兲. 13 S. K. Singha, H. Ishiwara, K. Sato, and K. Maruyama, J. Appl. Phys. 102, 094109 共2007兲. 14 X. Qi, J. Dho, R. Tomov, M. G. Blamire, and J. L. MacManus-Driscoll, Appl. Phys. Lett. 86, 062903 共2005兲. 15 G. W. Pabst, L. W. Martin, Y. H. Chu, and R. Ramesh, Appl. Phys. Lett. 90, 072902 共2007兲. 16 Y. Wang and C. W. Nan, Appl. Phys. Lett. 89, 052903 共2006兲. 17 G. W. Dietz, W. Antpohler, M. Klee, and R. Waser, J. Appl. Phys. 78, 6113 共1995兲. 18 K. H. Ahn, S. S. Kim, and S. G. Baik, J. Appl. Phys. 92, 421 共2002兲. 19 C. J. Peng and S. B. Krupanidhi, J. Mater. Res. 10, 708 共1995兲. 20 J. F. Scott, C. A. Araujo, B. M. Melnick, L. D. McMillan, and R. Zuleeg, J. Appl. Phys. 70, 382 共1991兲. 21 J. F. Li, J. Wang, M. Wuttig, R. Ramesh, N. Wang, B. Ruette, A. P. Pyatakov, A. K. Zvezdin, and D. Viehland, Appl. Phys. Lett. 84, 5261 共2004兲. 22 Y. Wang, Q. H. Jiang, H. C. He, and C. W. Nan, Appl. Phys. Lett. 88, 142503 共2006兲. 23 S. T. Zhang, M. H. Lu, D. Wu, Y. F. Chen, and N. B. Ming, Appl. Phys. Lett. 87, 262907 共2005兲. 24 C. Ederer and N. A. Spaldin, Phys. Rev. B 71, 060401共R兲 共2005兲. 25 J. M. D. Coey, A. P. Douvalis, C. B. Fitzgerald, and M. Venkatesan, Appl. Phys. Lett. 84, 1332 共2004兲. 26 J. Wang, A. Scholl, H. Zheng, S. B. Ogale, D. Viehland, D. G. Schlom, N. A. Spaldin, K. M. Rabe, M. Wuttig, L. Mohaddes, J. Neaton, U. Waghmare, T. Zhao, and R. Ramesh, Science 307, 1203b 共2005兲.

Downloaded 05 Apr 2008 to 155.69.4.4. Redistribution subject to AIP license or copyright; see http://apl.aip.org/apl/copyright.jsp