34 Growth Parameter Control of Structures and Properties Of

Growth Parameter Control of Structures and Properties of Perovskite Thin Films Skye Tackkett Physics, Missouri University of Science and Technology REU Program: 2017 Cornell NanoScale Science & Technology Facility International Research Experience for Undergraduates (CNF iREU) Program at the National Institute of Material Science (NIMS), Tsukuba, Ibaraki, Japan CNF iREU Principal Investigator: Dr. Naoki Ohashi, Electroceramics Group, National Institute for Materials Science REU Final Reports 2017 CNF iREU Mentor: Dr. Takeo Ohsawa, Electroceramics Group, National Institute for Materials Science Primary Source of CNF iREU Funding: National Science Foundation under Grant No. OISE #1559368 Contact: [email protected], [email protected], [email protected] Website: http://www.cnf.cornell.edu/cnf_2017reu.html Abstract: Application of growth parameters to control thin film structures and properties is an important method for improving film quality for applications in advanced electronic devices. Lattice engineering of thin films is key to improve the crystallinity and quality of films and the devices they comprise. In this study, as a model system we seek to control the growth of a perovskite oxide film, strontium titanate (SrTiO3), through magnetron sputtering. Film growth of SrTiO3 was investigated on lattice-matched lanthanum aluminate- strontium aluminum tantalite ((LaAlO3)0.3(Sr2AlTaO6)0.7) substrates, where we focused on the distance between substrates and sputtering target. We find that alteration of the distance leads to good Sr and Ti control, high crystallinity, and smooth surfaces, but oxygen deficiency in the films is still an obstacle to improving the electrical properties. Introduction: important in the development of many devices,Stacked thinsuch filmas heterostructuresmultilayer ceramic are capacitors, batteries, and integrated interfaces are essential in determining the circuits [1]. Highly crystalline films and Table 1: Typical deposition growth parameters for STO thin films. performance. quality of thin films and improving device The most common technique for producing practical Experimental Procedure: the advantages of low-cost deposition and relative 3) (Sr2 6) simplicitystacked thin of thefilms system is magnetron [2]. To obtainsputtering, high-quality due to Thin films of STO were grown on lattice-matched 0.3 3 0.7 frequency(LaAlO magnetronAlTaO (100)sputtering. (LSAT) In-plane and lanthanum lattice aluminate, LaAlO (LAO) substrates using radio- selectionthin films, of lattice substrates and orientationwith larger matching or smaller between lattice respectively, resulting in (aSTO-aLSAT a substrate and film are important [3]. For instance, Typicalconstants growth of STO conditions and LSAT areare summarized0.390 and 0.387 in Table nm, LSAT 1. These conditions were held constant)/ for=0.0103. each ofconstants substrates than andthose growth of the filmsparameter results optimizationin tensile or sample, but the sample holder height was changed to duringcompressive sputtering strain depositionin the films, are respectively. crucial to Selectionimprove distance (ds-t) between substrate and sputtering target Our goal in this project is to alter growth parameters to 130 mm, 150 mm, or 170 mm, which correspond to a film properties. In this study, we used a compact sputtering gun with of 53 mm, 73 mm, and 93 mm, respectively. After the ancontrol adjustable perovskite head andthin controlled film structures the distance and properties. between sputtering depositions, STO films were characterized the substrate and sputtering target during deposition via profilometry, atomic force microscopy (AFM), x-ray to fabricate SrTiO usingdiffraction vacuum (XRD), evaporation and x-ray to fluorescence form Ohmic (XRF). contacts Gold/ for 3 resistivitytitanium electrodesand Hall-effect were measurements. deposited onto the films (STO) thin films. 34 REU Final Reports 2017 We were able to conclude that our sputtering method grew with smooth surfaces. Cation stoichiometryhighly crystalline for SrSTO and thin Ti filmswas controlled through the changing of ds-t. Electric properties remained poor due to oxygen deficiency in the FutureSTO films. Work: Oxygen control is necessary in improving the properties of Figure 1, left: XRD wide scan of STO/LSAT samples. Figure 2, right: XRD (200) peak of achieve this control, we can next STO/LSAT samples. usesputtered oxygen thingas exposurefilms. In in order addition to to argon gas during the sputtering Results and Conclusions: we can test the effect of changing the angle of the head deposition procedure. Additionally, samples deposited under different ds-t. It is clear the of our sputtering gun and tune the angle for property Figure 1 showsc-axis XRDoriented wide and scans single-phase of the STO/LSAT without films were Acknowledgements:control of the film growth. impurity peaks. A closer look at the 200 peak of the angles than the bulk STO peak, represented by the I would like to thank my PI, Dr. Naoki Ohashi, my dashedfilms shows line in all Figure three 2. filmThe lowerpeaks peak occurred angles at indicate lower Rathbun and Nancy Healy, and the CNF iREU Program. Thismentor, work Dr. Takeowas supportedOhsawa, the by iREU the site National staff, Drs. Science Lynn the films are compressively strained, as the lattice constant of LSAT is lower than that of STO, which led to the lengthening of the c-axis parameters. Additionally, Foundation Award OISE: 1559368. oscillations visible on the 200 peak are an indication References: the films are highly crystalline, which was confirmed by [1] Kambara, H. et al. Chemical Solution Deposition of Functional rocking curve scans showing peak widths around 0.1°. The AFM images (Figure 3) of the STO/LSAT samples Oxide Thin Films. Vienna (Austria): Springer. Chapter 22: Thin revealed the thin films were significantly more smooth, Film Multilayer Capacitors; p. 547-570(2013). [2] Harp, G.R. et al. Thin Solid Films, 288 (1-2):315-324 (1996). diagonalwith a root-mean-square lines visible in Figureroughness 3 indicated of 0.177 stepped- nm and 0.271 nm for samples A and C, respectively. The faint [3] Hecht, J. Understanding Lasers: An Entry-Level Guide. New and-terraced surface, implying the surface of sample York (NY): Wiley-IEEE Press. Chapter 8: Semiconductor Lasers; p.246-294 (2001). [4] Verma, A. et al. AP Letters, 107(19):192908 (2015). C was very flat. We performed XRF measurements on measurement.the STO/LAO samplesThe XRF rather analysis than of the stoichiometrySTO/LSAT, as LAO does not contain Sr which would convolute the resultsrevealed imply a Srthat deficiency the enlargement in sample of c-axis A, parameters but near resultedstoichiometric from compressive ratios of Sr:Ti strain in samples and amount B and of C. oxygen These rather than cation (non)stoichiometry. Electric properties collected from Hall-effect cm-3) in sample C were severalmeasurements orders ofshowed magnitude resistivity 19different (1.1 than Ωcm) the otherand samples.carrier concentration In fact, sample (2.8*10 C looked light gray, indicating an oxygen-deficient film as is usually seen in n-type oxide semiconductors. Although former literatures C,reported could not that show compressively these effects. strained STO films show a Figure 3: 2000 2000 nm 0.98 Hz AFM scan of 170 mm STO/LSAT ferroelectric transition [4], our STO films, even sample thin film. × 35.

Load more