CHARACTERIZATION OF SUPERCONDUCTORS
BY ELECTRON BACKSCATTERED DIFFRACTION IN SEM
Z. Barkay1, E. Grunbaum2, A. Gholinia 3, S .Reich4
1Wolfson Applied Materials Research Center, Tel-Aviv University, Tel-Aviv, Israel
2Department of Physical Electronics, Faculty of Engineering, Tel-Aviv University, Tel-Aviv, Israel
3Oxford Instruments HKL A/S, Hobro, Denmark
4Department of Materials and Interfaces, Weizmann Institute of Science, Rehovot, Israel
Microtexture characterization by electron backscattered diffraction (EBSD) in the scanning electron microscope (SEM) provides a complementary method to conventional surface and element analysis methods in SEM. EBSD theoretical and experimental background is discussed with emphasis on phase identification and orientation mapping at high lateral resolution. We show an application of EBSD for electronic materials and particularly for low Tc and high Tc superconductors. The EBSD method is demonstrated on thermal vapour grown CsXWO3 0.005≤x≤0.3 crystals using HKL CHANNEL5 EBSD system in Quanta 200FEG ESEM. EBSD patterns showed phase transformation upon Cs doping to nominal concentrations of x=0.005, x=0.05 and x=0.3. In particular, the 2D superconducting crystals of x=0.005 nominal concentration, were of inhomogeneous crystallographic phase depending on the local Cs doping. The superconducting Cs-doped regions of the hexagonal phase were shown to be epitaxially grown on the WO3 monoclinic crystal surface with the (0001) planes parallel to the (001) planes in the WO3 monoclinic crystal and perpendicular to the sample surface. The utility of EBSD for high Tc superconductor texture characterization is demonstrated on YBa2Cu307 superconducting thin films.
Introduction
Electron backscatter diffraction (EBSD) in the scanning electron microscope (SEM) provides crystallographic information from the sample upper surface. In stationary mode, an electron beam strikes a tilted crystalline sample and the diffracted electrons form an EBSD pattern on a fluorescent screen. In mapping mode, EBSD patterns are accumulated from a scanned region for providing information on grain boundary misorientations, local crystalline perfection and preferred crystal orientations. The correlation between the microstructure and the crystallographic directions gives rise to microtexture information [1-2].
The mechanism by which the EBSD patterns are formed is based on scattering of electrons with a small energy loss close to the surface of the material forming a divergent source of electrons. Some electrons on either side of the divergent beam are arriving at Bragg angle to the crystallographic planes, producing two reflected cones for each crystallographic plane. Due to the short electron wavelength the diffracted electrons form large angle cones, which intersect the detector fluorescence screen as nearly straight parallel lines named as Kikuchi lines after its discoverer [3]. Hence the EBSD bridges the two traditional methods of studying crystalline materials: X-ray diffraction (XRD) and transmission electron diffraction (TED) in transmission electron microscope (TEM), providing the possibility of high spatial resolution in bulk specimen. The lateral and depth resolutions of the EBSD method are dependent on the sample composition, the e-beam diameter and interaction volume and are [4] both of the order of 10nm for field emission gun (FEG) SEM.
Low Tc and high Tc superconductor cryoelectronic devices both strongly depend on the development of reproducible high quality bulk and thin films. Most applications of YBa2Cu307 (YBCO) utilize conduction in the CuO2 layers i.e., conduction along the a-b crystal directions. This requires that films should be oriented with the c direction normal to the surface, while misorientation in grain boundaries influences detrimentally the transport critical current density [5].
Our work discusses the utility of EBSD for investigation of high Tc YBCO thin superconducting films and in particularly focuses on low Tc superconductors made of Cs doped WO3 crystals. It was shown [6-8] that CsxWO3, 0.3³x³0.19, is a 3D superconductor. Below x=0.19, a transition from a metal to an insulator occurs and no superconductivity is observed. For x£0.05 superconductivity reappears with 2D superconductivity at particularly low doping levels of x=0.005.
Local phase identification and texture orientation were studied in YBCO high Tc superconductors (HTSC) by other groups [9-10] using the EBSD method in SEM. Following our previous study [11-12], we will discuss texture analysis using the EBSD method in SEM for both families of low Tc and high Tc superconducting materials.
Experimental
Crystals of CsXWO3 0.005£x£0.3 were prepared by thermal vapor growth process as described elsewhere [13]. We will refer to crystals of nominal concentrations of x=0.3, x=0.05 and x=0.005. The process for YBCO thin film growth is described elsewhere [14].
EBSD analysis was performed at the conventional specimen configuration [15] with 70° tilt position relative to the main beam and with 15mm working distance. The EBSD system was HKL-Oxford Channel5 with Nordlys II detector mounted on Quanta 200FEG in the CsXWO3 study and on LEO Supra 55VP SEM in the YBCO study. All the analysis was carried out at 20kV acceleration voltage with a few nA beam current, without surface preparation. EBSD was locally correlated with element composition measurement using Energy Dispersive Spectroscopy (EDS).
The EDS system for element composition measurement was nitrogen-cooled Oxford INCA with 133eV resolution. The EDS analysis was performed at the system optimum configuration of horizontal sample position at 10mm working distance. The EBSD lateral resolution (mentioned above) is about order of magnitude better than the EDS micro-scale resolution, which provides local and comparable information.
Results and Discussions
A typical secondary electron (SE) image of the thermal vapor grown CsXWO3 crystals is shown at fig. 1. The crystals are typically of 100mm size and tend to agglomerate within mm size volume with their crystal facets inclined in various directions. The sample was a challenge for EBSD analysis due to the inhomogenueity in Cs concentration on the surface, and due to the requirement of selecting a facet with the appropriate inclination to the EBSD detector.
In an enlarged SE image (fig. 2a) of CsXWO3 crystal with nominally x=0.005, two locations have been chosen for EBSD analysis. Figure 2b shows the EBSD pattern at point 1, which fits to the Cs-doped hexagonal phase (6/mmm Laue group, a=7.42Å, b=7.42Å, c=7.57Å). Figure 2c shows the EBSD pattern at point 2, which fits the WO3 monoclinic phase (2/m Laue group, a=5.28Å, b=5.16Å, c=7.66Å); this corresponds also to undoped WO3 crystal samples. The comparison of EBSD patterns for the other Cs doped samples showed the hexagonal phase structure for x=0.05, and the trigonal phase (-3m Laue group with a=7.26Å, b=7.26Å, c=110.58Å) for x=0.3. Table 1 summarizes the surface crystallographic phases at the Cs rich regions versus the nominal doping of the CsXWO3 crystals.
Nominal doping (x) / Crystallographic phase0 / Monoclinic
0.005 / Hexagonal
0.05 / Hexagonal
0.3 / Trigonal
Table 1: Fig. 1.
Surface crystallographic phase at the Typical SE image of an agglomeration
Cs-rich regions versus the nominal of CsXWO3 crystals.
doping of the CsXWO3 crystals.
EBSD mapping in areas 1 and 2 showed texture orientation of each phase. The analysis in fig. 2a includes 40% patterns corresponding to the monoclinic phase (fig. 2c) and the other 60% corresponding to the hexagonal phase (fig. 2b). The location of each phase is shown in the phase-map (fig. 2d), where the left side (in red) corresponds to the Cs-doped hexagonal phase and the right side (in blue) corresponds to the WO3 monoclinic phase. Correlation with the SE image at fig. 2a shows that each of the two phases in the adjacent planes occupies about 10mm2. Missing points within the mapping frame region actually constitute unresolved EBSD patterns due to the topographic shadow effect at the slope of the plane and to incomplete flatness of the surface.
EDS analysis of x=0.005 crystals at the Cs-rich regions yielded concentrations which are order of magnitude higher than the nominal value of 0.5 wt%. In particular, EDS analysis from the left side of fig. 2d showed Cs enhancement to 8.5 wt% in correlation with the EBSD phase-map. Similar concentration enhancement was obtained by EDS analysis for crystals of x=0.05 nominal concentration, whilst CsXWO3 crystals of x=0.3 yielded values of 15-20 wt% at the Cs-doped regions.
Based on the EBSD patterns, which were collected for crystal phase mapping (fig. 2d), pole figures are derived for crystal orientation mapping of both the Cs-doped hexagonal phase (fig. 3a) and the monoclinic WO3 phase (fig. 3b) regions. The intensity in the pole figures is presented in multiples of uniform density (MUD) units. We use a 10º half width plane spread angle and 5º clustering of data points for increased MUD weighting. The pole figures are presented in X-Y sample coordinates while the Z direction is perpendicular to the surface.
Fig. 2, CsXWO3 crystal of x=0.005 nominal concentration: (a) SE image (b) EBSD pattern of crystal in fig. 2a at location 1 - Cs-doped hexagonal phase (c) EBSD pattern of crystal in fig. 2a at location 2 - WO3 monoclinic phase (d) EBSD phase-map corresponding to the framed region of fig. 2a: red – Cs-doped hexagonal phase, blue - WO3 monoclinic phase.
Fig. 3. Pole figures of CsXWO3 crystal of x=0.005 nominal concentration: (a) (0001) plane orientation of the hexagonal Cs-doped phase (b) (001) plane orientation of the monoclinic WO3 phase.
Alignment within each phase is clearly derived from the corresponding pole figures for each of the crystallographic planes. In particular, the (0001) crystallographic plane of the Cs-doped hexagonal phase lies within the X-Y sample plane, and is parallel to the (001) crystallographic plane of the monoclinic WO3 phase. Texture alignment was also observed within the Cs-doped regions of x=0.3 and x=0.05 CsXWO3 crystals.
The superconducting YBCO has an orthorhombic structure, which is built from three perovskite unit cells, where the difference between a, b and c axes is about 1%. Such small differences give rise to pseudo-symmetry problems in the orientation analysis. The EBSD pattern was indexed using the YBCO-tetragonal crystal structure to distinguish between the c and a axis (4/mmm Laue group a=b=3.88Å, c=11.79Å). Figure 4a shows the band contrast image provided by EBSD mapping. The (001) plane alignment with respect to the x, y and z directions of the sample is shown by inverse pole figure (IPF) map (fig. 4b). The pole-figure for the (001) plane is shown (fig. 4c) with a schematic diagram (fig. 4d) referring to the alignment of the unit cell. The microtexture analysis in the YBCO thin film shows that most of the sample exhibit c-axis orientation, with tiny elongated grains of a-axis and b-axis orientations.
Fig. 4. The EBSD data shown in a) band-contrast map, b) IPF map, c) (001) pole figure and d) color key showing the alignment of the YBCO crystal structure, where the x-axis of the sample is rotated 45° with respect to the horizontal of the map.
Conclusions
EBSD provided the possibility to reveal phase and texture information on both low Tc crystal superconductors and high Tc thin superconducting film with sub-micron lateral resolution and high surface sensitivity. The EBSD results on CsXWO3 crystals support previous observations by scanning tunneling microscopy (STM) indicating the existence of a local surface superconducting phase within the x=0.005 CsXWO3 crystals. In addition, it confirmed the existence of a small amount of hexagonal phase previously observed by XRD. Apart from showing the correlation between the crystallographic phases and doping levels, we showed a high texture alignment for the 2D superconducting phase of x=0.005 nominal concentration. In particular, the Cs-doped WO3 (0001) plane was in alignment with the (001) WO3 crystal plane, which indicated an epitaxial growth of the superconducting phase due to Cs doping.
As complimentary information we introduced the analysis of high Tc superconductors by EBSD. The orientation of a and c axis growth directions in the YBCO film were shown at sub-micron resolution. The band-contrast and IPF images both show that the YBCO grain shapes are either flat or of needle shape aligned in the horizontal or vertical directions. In the flat grains, the YBCO structure was found to have c-axis perpendicular to the sample surface, where as in the needle shaped grains the c-axis was in the sample plane.
Microtexture analysis could be further correlated with macroscopic measurements of critical current density in case of cryoelectronic device applications. The applications of EBSD in SEM could be extended to the enlarged family of alkali doped WO3 crystals as well as to other HTSC materials.
Acknowledgement
Acknowledgement is given to G. Leibovitch and Prof. G. Deutscher from the physics department in Tel-Aviv University for providing the thin YBCO film for EBSD measurement.
References
1. Randle V and Engler O: Introduction to Texture Analysis, CRC Press, 2000.
2. Randle V: Micotexture determination and its applications, London, MANEY, 2003.
3. Kikuchi S: 'Diffraction of Kathode Rays by Mica'. Imp. Acad. Tokyo, Proc. 1928 (4) 271-78.
4. Humphreys FJ: 'Characterisation of fine-scale microstructures by electron backscatter diffraction (EBSD)'. Scripta Materialia 2004 (51) 771-76.
5. Fairhurst RJ, Zhai HY, Meen JK, Elthon D: 'Crystallographic study of superconducting YBCO thin films by EBSD'. Physica C 2000 (341-348) 2019-20.
6. Reich S and Tsabba Y: 'Possible nucleation of a 2D superconducting phase on WO3 single crystals surface doped with Na+'. Eur. Phys. J. 1999 (B9) 1.
7. Skokan MR, Moulton WG, and Morris RC: 'Normal and superconducting properties of CsxWO3' Phys. Rev. B 1979 (20) 3670-77.
8. Reich S, Leitus G, Tssaba Y, Levi Y, Sharoni A, and Millo O: 'Localized High-Tc Superconductivity on the Surface of Na-Doped WO3'. J. Superconductivity 2000 (13) 855-61.
9. Grossin D, Henrist C, Mathieu J-Ph, Meslin S, Harnois C, Noudem J-G, Cloots R, and Chateigner D: 'EBSD study on YBCO textured bulk samples: correlation between crystal growth and 'microtexture''. Supercond. Sci. Technol. 2006 (19) 190.
10. Fairhurst R. J, Zhai H, Meen JK, and Elthon D: 'Crystallographic study of superconducting YBCO thin films by EBSD'. Physica C 2000 (341-348) 2019-20.
11. Barkay Z, Grunbaum E, Leitus G and Reich S: 'Study of superconducting Cs-doped WO3 crystal surface by electron backscattered diffraction'. J Supercond. Nov. Magn. 2008 (21) 145-50.
12. Gholinia A: 'EBSD applications catalogue'. 2005 HKL technology App. 9.
13. Leitus G, Cohen H and Reich S: 'Interplay of Cs concentration, dimensionality and superconductivity in CsWO3'. Physica C 2002 (371) 321-29.
14. Leibovitch G, Beck R and Deutscher G: 'High quality a-axis outgrowth on c-axis Y1-xCaxBa2Cu3O7-δ '. Superconductor Science and Technology 2004 (17) 1065-68.
15. Deal A, Tao X, and Eades A: 'EBSD geometry in the SEM: simulation and representation'. Surf. Interface Anal. 2005 (37) 1017-20.
8-3