Journal of Crystal Growth 88 (1988) 541—546 North-Holland, Amsterdam
PRIORITY COMMUNICATION SURFACE MORPHOLOGY OF FLUX-GROWN SINGLE CRYSTALS OF YBa2Cu3O78 K.N.R. TAYLOR, P.S. COOK
T. PUZZER, D.N. MATTHEWS, G.J. RUSSELL and P. GOODMAN
Advanced Electronic Materials Technology Group, School of Physics, University of New South Wales, Post Office Box 1, Kensington, NS W 2033, A usiralia
Received 10 December 1987; manuscript received in final form 12 February 1988
Single crystals of YBa2Cu3O7_5 have been grown in a CuO—BaCuO2 flux at a temperature of 9300 C and examined using SEM, XRD and electron diffraction techniques. The observed crystal habit is predominantly square planar with the c-axis normal to the plane. Surface growth spirals are observed in approximately 25% of the crystallites and in some cases extensive step decoration is exhibited. The growth surface usually involves two such spirals with Burgers vectors of the same sign. Opposed spirals have been observed in only two cases.
The growth and superconducting properties of single crystal YBa2Cu3O7_~ ((123) hereafter) material have been reported by a number of laboratories [1-.-3].This work clearly demonstrates the high anisotropy in the electrical and magnetic properties which may be anticipated for the superconducting state of the strongly axial distorted orthorhombic structure of the solid. This directional dependence of the current flow is believed to be partly responsible for the difference in the values of the critical current and critical field strength observed for randomly oriented poiycrystalline samples  compared with thin films of (123) . It is clear that commercial applications of bulk materials will benefit from the presence of texture in the polycrystalline conductor and it is consequently highly desirable to establish some understanding of the growth mechanisms which occur during material preparation. As part of an intensive research programme involving this new generation of superconducting cuprates, we have prepared and studied single crystal materials. The growth habits of these crystals are described in the following,
Division of Materials Science and Technology, CSIRO, Clayton, Victoria 3168, Australia.
Single crystal materials were grown in a platinum crucible using a flux technique from a starting composition (0.46CuO, 0.13Y2O3 and 0.42BaCO3 by weight). Heated to 1050°C, this provides a CuO—BaCuO2 flux in which is dispersed molten (123) material. After a prolonged stay (24 h) at this temperature to complete the formation of both (123) and BaCuO2 and ensure full homogeneity, the temperature was lowered to 930°C and held there for several days. At the end of this time, a continuous flow of oxygen was admitted to the furnace and the system cooled at approximately 1°C/mm to room temperature. On removal from the crucible, the growth systern appeared as a shell of solidified flux material lying above an extensively crystallized mass. While crystalline material could be observed throughout the mass, the largest and best quality crystals occurred in small groups in cavity regions between the top of the crystalline mass and the lower surface of the flux shell. The dominant morphology of the single crystals was square platelet in form, with edge dimensions of up to 1.5 mm and thicknesses in the range 1—lO j~sm.Long, square prism crystals with dimensions 2 x 0.1 X 0.1 mm were also observed. Energy dispersive analysis (EDAX) showed the compositions of the various components of the crucible contents
0022-0248/88/$03.50 © Elsevier Science Publishers B.V. (North-Holland Physics Publishing Division)
K.N.R. Taylor et al.
morphology ofJiux-grown single crystals of YBa ,Cu 307
to be BaCuO2 (shell). CuO (square prisms) and (123) (platelets). Despite the crucible being degraded during the growth, no evidence of platinum impurities was found in any of the components. Only the (123) crystals will be discussed in the following, One of the platelet crystals was crushed and mounted on a holed carbon film for electron diffraction. We were unable to find a crystal fragment in this material which was oriented in such a way that a c-axis projection could be obtained, This is due to a preferred cleavage direction in
these crystals lying parallel to the c-axis. Consequently, it was not possible to distinguish between the known tetragonal and orthorhombic forms of the (123) compound. Two platelet crystals were extracted from the batch for X-ray diffraction, using a Philips automatic diffractometer operating with Cu Ka X-rays. The surfaces of both plates were found to be normal to the  direction and therefore paralid to the superconducting planes. The unit cell dimensions obtained were a 3.859 A, b 3.859 A and c 11.812 A for the first crystal and a =
Fig. 1. Scanning electron micrographs of YBa,Cu ~O., cr\stals and (inset) the bottom area. Specific features. labelled A and B, are discussed in the text. The plate scale given as 017*3 nm is equivalent to 17 X 10~nm, i.e. 17 ~sm.
K.N.R. Taylor et al.
/ Surface morphology offlux-grown
3861 A b=3863 A and c=11 746 A for the second with errors of ±0 003 A The equality of a and b within experimental error suggests that both crystals are tetragonal and the values of c also lie in the range observed for this structure  An extended oxygen treatment or post-growth annealing in an oxygen atmosphere is needed to create the desired orthorhombic phase. Crystals were examined by scanning electron microscopy (Hitachi S-405A operating at 25 keV, and JOEL JXA84O at 15 key) and several micrographs illustrating typical features are presented. In the upper half of fig. 1, a pair of large plates is seen edge on. Such overlapping growth occurred frequently and may be of significance for the production of oriented superconductors. The coarse material adhering to the mid-line of the upper plate has been shown to be of the same composition as the flux, using calibrated energydispersive analysis. In other cases, similar edge deposits were identifiable as the compound BaCuO2. A number of growth spirals are clearly visible on the face of the large crystal which appears in the lower portion of fig. 1. Such spiral growth was observed on about one in four of the crystals examined. The inset shows this section at higher magnification viewed normal to the surface; a slight change in the step contrast may be discerned due to the change of viewing angle. Several platelet fractures are apparent which evidently occurred after crystal growth had ceased since the spirals are continuous across the fracture boundaries, Fig. 2 shows a growth spiral on a different platelet at higher magnification. In this case we see a double spiral which nucleates from a pair of cooperating screw dislocations. In fact, no conciusive evidence for single spirals was found in this study. It thus appears to be favourable for screw dislocations possessing Burgers vectors of the same sign to form close together in this material. It is interesting to note a finite decoration of the spiral in fig. 2, which consists of birth-contrast particles whose diameter is of order 0.1 .tm. It is particularly evident around its outer sections (as indicated in fig. :2), where the steps advance more rapidly , and appears to segregate at both the
single crystals of YBa,Cu
1-1g. 2. A double gro~~th cpiral showing an Interesting area of step decoration.
upper and lower edges of the steps. Investigation of compositional variations in these regions is in progress. Fig. 3 shows micrographs of several other crystals. In one of these, fig. 3a, the nucleation region of a double interlaced spiral is shown, with clearly developed fourfold symmetry. This core region is very reminiscent of the double spiral in fig. 2, but has not developed to the same extent nor has it established the decoration of that exampie. The axis separation of approximately 2 p~mis roughly half of the growth step width and the line joining the two centres is noticeably parallel to the square sides of the growth pattern, which in themselves are likely to be parallel to close packed planes in the structure . Fig. 3b shows an intersection region between two interlacing spiral growth patterns of the same sign. Again, extensive decoration is visible between the growth steps and analysis is now in progress to establish the composition of this material. Finally, fig. 3c displays a fully estab-
K.N.R. Taylor et aL
Surface morphology offlux-grown single crystals of YBa,Cu
Fig. 3 Further examples of growth spirals illustrating (a) nucleation region of a double spiral, (b) competition between steps of neighbounng spirals and (c) octagonal symmetry.
K. N. R. Taylor et a!.
Surface morphology of flux-grown single crystals of YBa
Fig. 3 (continued).
lished interlaced double spiral with evidence of octagonal symmetry. Again the line joining the nucleation centres is parallel to one of the edges. Polygonal growth habits of the type shown in figs. 3a and 3c are normally found only in crystals grown from a liquid phase . Since all these crystals were found in the space between the crystalline mass and the solidified flux cap, it is clear that during cooling of the growth volume, the flux separates from the crystalline (123) mass on solidification. It was very noticeable that, despite the large number and variety of double spiral growth patterns which we observed in this study, only two examples of opposite spirals were observed. Both of these had internuclear distances of a few microns and neither showed any evidence of polygonal growth habit well away from the source. In conclusion the flux route to single crystal growth appears capable of providing extremely high quality crystals; however, to obtain the superconducting orthorhombic phase requires exten-
sive oxygen treatment. The dominance of double growth spirals in the crystals was unexpected but may be related to the tendency of these materials to twin. The observation of preferential cleavage parallel to the c-axis is undesirable from the point of view of establishing texture, and efforts are necessary to establish a means of selectively inducing basal plane cleavage. We thank Dr. P. de Munk for performing some of the SEM measurements and Dr. B. Gatehouse for the XRD analyses.
References  Y. lye, T. Tamegai, H. Takeya and H. Takei. Japan. J. AppI. Phys. 26 (1987) L1057.  SW. Tozer, A.W. Kleinsasser, T. Penney, D. Kaiser and F. Holtzberg, Phys. Rev. Letters 59 (1987) 1768.  R.N. Shelton, R.W. McCallum, M.A Damento and K.A. Gschneidner, Jr., Intern. J. Mod. Phys. BI (1987) 235.  J.W. Ekin, Advan. Ceram. Mater. 2 (1987) 586.
K.N.R. Taylor eta!.
morphology of flux-grown single crystals of YBa
 P. Chaudhari, RH. Koch, RB. Laibowitz, T.R. McGuire
and R.T. Gambino, preprint.  K. Nakamura, T. Hatano, A. Matsushita, T. Oguchi, T. Matsumoto and K. Ogawa, Japan. J. AppI. Phys. 26 (1987) L791.
 H. Frey and H.J. Meyer, J. Crystal Growth 82 (1987) 435.  F.C. Frank, Advan. Phys. 1(1952) 91.