JOURNAL OF ELECTRON MICROSCOPY TECHNIQUE 8:263-272 (1988)
Structure of Superconducting Thin Films of YBa2C~307-x
Grown
on SrTi03 and Cubic Zirconia
L.A. TIETZ, B.C. DE COOMAN, C.B. CARTER, D.K. LATHROP,
S.E. RUSSEK, AND R.A. BUHRMAN
Department of Materials Science and Engineering (L.A.T,B. C.D.C., C.B.C.)
and School ofApplied and Engineering Physics (D.K.L..S.E.R.. R.A.B.).
Cornell Uniuersity, Ithacq New York 14853
KEY WORDS
High T, superconductors, Electron microscopy, HREM
image stimulations
ABSTRACT
Thin films of the superconductive oxide YBa2Cu307., have
been made by electron-beam coevaporation of the metals in a n oxygen atmosphere onto single-crystal { 001)-oriented SrTiO3 and yttria-stabilized zirconia
(YSZ) substrates. The oxide films were superconducting in the as-deposited
state (T, = 81-83K, J, = lo6 A/cm2 a t 4.2K). Bright-field imaging, selectedarea diffraction (SAD), and high-resolution imaging in the transmission electron microscope were used to characterize the microstructure of these films.
All of the films were polycrystalline. On SrTi03 the films were oriented, for
the most part, with (110) parallel to the substrate surface. On YSZ, two
microstructures were observed: one with smaller rectangular grains oriented
with (100) or (010) parallel to the substrate surface and the other with (001)
parallel to the surface (i.e., c-axis up).
INTRODUCTION
The discovery of the superconducting oxide
YBa2C~307.~
with a critical temperature
above 90°K (Chu et al., 1987; Wu et al., 1987)
has made the use of superconducting materials at liquid-nitrogen temperatures possible. Although the compound is easily formed
by a solid-state reaction of the component
oxides, many of the applications envisioned
will require other fabrication techniques to
obtain the required product specifications.
Some of these applications-for example, in
the area of microelectronics-will require
thin films of the superconducting oxide. Several groups have reported success in growing
such films on ceramic substrates by a variety
of techniques including electron-beam evaporation (Laibowitz et al., 1987; Naito et al.,
1987) and dc magnetron sputtering (Jin et
al., 1987; Moriwaki et al., 1987; Somekh et
al., 1987). YBa2C~307-x
thin films formed by
deposition and subsequent annealing in a n
oxygen atmosphere have had critical temperatures in excess of 80°K and critical current
densities better than lo5 A/cm2 at 77°K
(Chaudhari et al., 1987a,b).
The structure of the superconducting phase
has been shown to be a n orthorhombic layered perovskite structure with the c-axis ap-
0 1988 ALAN R. LISS, INC.
proximately three times the length of the aor b-axis (Beyers et al., 1987; Cava et al.,
1987). Recently, neutron scattering has been
used to refine this structure so that the correct designation for the compound is
YBa2Cq07., (space g o u p = Pmmm, a =
3.8231 A , b = 3.8864 A , c = 11.6807 A) (see
Fig. 1).The copper and oxygen ions are arranged a s Cu-0 chains along the b-direction
between the Ba-0 layers and as rumpled
CuO2 planes between the Ba-0 and Y layers
(Beno et al., 1987).
Studies of single crystals (Dinger et al.,
1987)have shown that the properties of these
materials are quite anisotropic. For example,
critical densities in the (001) plane are ten
times greater than that measured perpendicular to this plane. Because of this anisotropy,
high-quality thin films for device applications will require growth of correctly oriented YBa2C~307.~.
The choice of substrate
is expected to exert a strong influence on the
Received August 28, 1987; accepted November 17, 1987.
Address reprint requests to Prof. C. Barry Carter, Department
of Materials Science and Engineering, Bard Hall, Cornell University, Ithaca, NY 14853.
264
L.A. TIETZ ET AL.
polished, (001)-oriented, single-crystal substrates of both SrTiO3 and YSZ were used in
this study. Following deposition, the chamA d a , along
[loo] [OlO] %[001] CX(OK-')~ ber was backfilled with 100 mtorrs oxygen
while the film cooled. The as-deposited oxide
Substrate
films were black, tough, and exhibited excelSrTi0s3
0.021 0.0048 0.0030 0 . 8 2 ~ 1 0 - ~lent adhesion to the substrates. The films
YSZ4
0.294 0.278 0.276
1.04x
were superconducting in the as-deposited
state with T, = 8 1 4 3 ° K and a transition
'All values calculated for 25°C.
width ATc = 4°K. Films on both substrates
~ , = 2 . 5 ~ 1 0 (Yukino
-~
'For YBa2CU307-o a1=a2= 1 . 1 ~ 1 0 -a3
et al., 1987).
had critical current densities of lo6 Mcm2 at
3JCPDS powder diffraction file; Krishnan, Srinivasan, and
4.2"K (Lathrop et al., 1987).
Devanaravanan. 1979.
4JCPDS p"owder'diffraction file; Adams, Nakamura, Ingel, and
Plan-view sections of the films were preRice, 1985.
pared for examination by transmission electron microscopy. Discs, 3 mm in diameter,
were cut with a n ultrasonic drill. The discs
quality of the films through epitaxy and in- were then ground, polished, and dimpled
teraction or reaction with the film. The two from the substrate side to less than 20 pm.
substrates chosen for this study have very All of the cutting and polishing steps were
different lattice mismatches with the super- performed in nonaqueous media (e.g. ethylconducting phase. SrTiO3(ST) has a perov- ene glycol). In order to preserve the informaskite-type structure which closely matches tion on the crystallographic relationship
the YBazCu307., phase in lattice parameter between the film and the substrate, it was
in thermal expansion. Misfit calculations necessary to thin the specimens by ion-milllisted in Table 1 suggest that the (100) plane ing. The specimens were ion-milled to perfoof the superconducting phase is the favorable ration from the substrate side only with 5interface with a (001)ST surface; they also kV Ar+ ions. The specimen was not cooled
suggest that the c-axis can be oriented during ion-milling; this procedure may cause
equally well with either [100]s~or [olO]sT in a n increase in lattice damage over that prothis plane. On the other hand, yttria-stabi- duced by milling while cooling the sample,
lized cubic zirconia (YSZ) has a fluorite-type but the different crystallographic orientastructure and a large lattice parameter mis- tions could still be easily identified by each
match with respect to the superconducting technique used. The foils did not require carphase. It is less clear for this substrate what bon coating for examination in the electron
orientation would be expected for the film.
microscope. Bright-field imaging and SAD
In this study, the microstructure YBa2- studies were performed on a JEOL 1200EX
C U ~ Ofilms
~ . ~has been characterized by us- 120-kV or a Siemens Elmiskop 102 125-kV
ing bright-field imaging, selected-area dif- transmission electron microscope. High-resofraction (SAD), and high-resolution electron lution lattice-imaging was performed on a
microscopy (HREM) to determine the influ- JEOk 4000EX a t 400 kV with better than
ence of the substrate on their epitactic 1.8-A point-to-point resolution.
growth. Computer stimulation of the highresolution images has been used to confirm
Computer simulation of high-resolution
that HREM is able to verify this structure by
images
locating the oxygen structural vacancies and
lattice distortions and can thus supplement
An extensive analysis of simulated images
the information obtained by SAD.
of the superconducting phase and related
structures has been carried out to assess the
MATERIALS AND METHODS
suitability of HREM for direct study of the
Thin films of the superconducting Y-Ba- structure described by Beno et al. (1987). The
Cu-0 phase 1 pm thick were deposited on simulated images were all obtained by using
single-crystal substrates by electron beam the SHRLI program (M.A. O'Keefe, private
coevaporation of the metals, Y, Ba, and Cu communication) and show that HREM can
from three separate sources. During deposi- provide two important pieces of information
tion, the oxygen partial pressure in the in this study. First, by using either a 125-kV
chamber was approximately 1 mtorr and the or a 400-kV machine, the orientation of the
substrates were held at 700°C. Mechanically c-axis can be identified in grains which are
TABLE 1. Misfit calculations and linear coefficients of
thermal expansion for YBaZCu307-x on various
substrates.'
THIN FILMS OF YBazCu@-, GROWN ON SrTiO3 AND YSZ
O Y
0 Ba
0
cu
.O
Structural
vacancy
"YBaZCu307"
B
Fig. 1. A "Perfect" YBazCu309 structure. B: YBaz
Cu307 structure described by Beno et al. (1987). Lattice
parameters: a = 3.823 if, b = 3.8886 if, c = 11.681 A .
Not drawn to scale.
much too small for SAD analysis. Second, a t
the higher voltage, the simulations show
that both the presence of oxygen structural
vacancies at [%,O,O] and [O,O,%]sites and the
distortion ("rumpling") of the Cu02 planes
should be detectable by the observation of
image symmetry with the electron beam directed parallel to either [loo] or [OlO]. In the
present paper, only a limited summary of the
results of the computer simulations will be
presented to illustrate these uses and to show
how high-resolution imaging can supplement the SAD information regarding the epitactic alignment of the epilayer and the
substrate. Projected-potential diagrams and
simulated images have been calculated for
both the "perfect"
YBazCu309 and
YBazCu307 structures and are shown in Figure 1A and B, respectively. YBa2Cu309 contains no oxygen structural vacancies; all of
the possible oxygen sites are filled. Although
the lattice is orthorhombic, it can be considered to be essentially cubic since the lattice
parameters a-b-cI3 and all of the cubic
subcells are identical except for the central
atom, which can be either Ba or Y. The [loo]
projected-potential diagram for this structure is shown in Figure 2A. The YBazCu307
structure, however, contains oxygen structural vacancies arranged in rows along the
[loo] and [OlO] directions and distortions in
the C u 0 2 planes, so the structure cannot be
considered to be nearly cubic. Viewed along
[1001, only those vacancies in the Y-plane are
visible, while vacancies in both the Cu-0and Y-planes are visible by viewing the
structure along [OlO]. This is shown by the
projected potentials in Figures 3A and 4A
respectively.
265
A series of three simulated images, all calculated at Scherzer defocus for the 400-kV
machine, is shown for each of the three
cases-namely,
[100]-YBa2Cu309, [loo]YBazCu307, and [010]-YBa2Cu307(see captions to Figs. 2B-D, 3B-D, and 4B-D for
more details). In each of the series of image
simulations, the image contrast varies dramatically with crystal thickness, but the
overall symmetry of the image remains the
same. The thin crystal images (Figs. 2,3,4B)
show the least amount of structural detail;
the detail increases with crystal thickne%s
and contrast reversal occurs between 30 A
and 66 A. All three simulated images of the
"perfect" structure exhibit a pseudofourfold
symmetry as would be expected from the essentially cubic symmetry of the YBa2Cu309
structure. In the YBa2Cu307 structure, the
pseudofourfold symmetry is reduced to twofold symmetry by the presence of the structural vacancies and lattice distortions. This
change is reflected in the simulated images
in Figures 3 and 4. Image simulation has
also been performed for the YBazCu309 and
YBa2Cu307 structures for a [110] beam direction; examples are shown, for the same set of
imaging conditions, in Figures 5 and 6, respectively. Notice that none of the rows of
oxygen structural vacancies is visible in
YBa2Cu307 when viewed along [110], although the distortions in the Cu02 plane are
visible, as can be seen in the projected-potential diagram (Fig. 6A). A comparison of these
two cases shows that the contrast in the images is very different particularly in the
thicker crystals; however, the overall symmetry of the two serigs is the same and the
presence of the 11.7-A periodicity can be directly recognized. The latter result. also applies to images obtained by using the 125-kV
machine. In order to distinguish between the
YBa2Cu309 and YBa2Cu307 structures,
then, the crystal thickness and defocus values for the image must be known so that the
position of the structural features can be located. However, if a [loo] or [OlO] beam direction is used, the presence of the structural
vacancies and lattice distortions can be determined simply from changes in the symmetry of the 400-kV images; it is not
necessary to know the thickness of the crystal. HREM can therefore be used to identify
the composition of a particular grain by using the more precise structural information
available from x-ray studies. When the grains
are large, SAD is adequate for identifying
266
L.A. TIETZ ET AL.
THIN FILMS OF YBazCu@-, GROWN ON SrTiO3 AND YSZ
267
the orientation of the c-axis; for small grains,
HREM provides this information.
OBSERVATIONS
SrTiO3
Figure 7 shows a lattice image of the asdeposited superconducting Y B ~ ~ C Ufilm.
~ O ~ - ~
The planes have a n 11.5-A spacing, which
corresponds well with that measured by Beno
et al. (1987) for the (001) planes in
YBazCu307,. Thus, in this grain, the c-axis
is parallel to the (001) substrate surface.
The film is polycrystalline with small areas
of different orientation lying within the
larger single-crystal regions. The apparent
porosity of this film is believed to be due to
surface roughness which would lead to the
formation of holes when the specimen was
ion-milled from the substrate side only.
The major portion of the film was oriented
with respect to the substrate as shown by the
SAD pattern in Figure 8a which was recorded from a n area similar to that shown in
Fig. 5. “Perfect” YBazCu309 structure, [ l l O ] beam
Figure 7. It shows that the YBa2Cu307., film direction. A Projected potential (unit cell is outlined):
has grown with (110) parallel to ( 0 0 1 ) s ~ ~heavy circle = BdO, light circle = Y/O, square = Cu,
although it was not possible to determine closed circle = 0. Image simulations (Scherzer defocus)
from the pattern whether the plane is (€10) at: (B)lO.9O-A, (C) 32.71-A, (D)65.42-A thickness.
Fig. 6. YBazCu307 structure, [110]beam direction. A:
or (110). This pattern also confirms that the
potential (unit cell is outlined): heavy circle =
grain size was -1 pm, the diameter of the Projected
BdO, light circle = Y/O-vacancy, square = Cu, closed
area included in the SAD aperture. These circle = O/O-vacancy. Image simulations (Scherzer defoobservations differ from those of Naito et al. cus) at: (B) lO.9O-A, (C) 32.71-A, 0)
65.42-A thickness.
(1987), who reported seeing randomly oriented grains, grains with a-axes perpendicu- pendicular to that of the larger surrounding
lar to the substrate, and a mixture of grains area, but still parallel to the substrate surwith a- and c-axes perpendicular to the sur- face. This 90” microstructure is to be exface in their films depending on the film’s pected for { 100)-oriented cubic substrates
composition and the substrate temperature because of the fourfold rotational symmetry
during deposition.
about < 100> as discussed previously.
Figure 8b is a SAD pattern taken on a n
High-resolution imaging was also perarea such as the small area arrowed in Fig- formed on these films a t 400 kV. The
ure 7. It shows that these small areas are YBazCu307, film appeared to be heavily
oriented with their c-axes approximately per- faulted parallel to the (001)planes; these defects may have been caused by damage durFig. 2. “Perfect” YBazCu309 structure, [loo] beam ing observation in the transmission electron
direction. A Projected potential (unit cell is outlined): microscope or by the ion-milling rather than
heavy circle = Ba, light circle = Y, square = CdO,
closed circle = 0. Image simulations (Scherzer defocus) be intrinsic to the structure.
at (B)7.65-A, (C) 30.58-A,(D) 61.17-A thickness.
Fig. 3. YBazCu307 structure [loo] direction: A Projected potential (unit cell is outlined): heavy circle = Ba,
light circle = Y, square = C d O or CdO-vacancy, closed
circle = 0, star = missing row of 0. Image simulations
(Scherzer defocus) at (B) 7.65-A, (C) 30.58-A, (D) 61.17A thickness.
Fig, 4. YBazCu307 structure [OlO] beam direction.
A: Projected potential (unit cell is outlined): heavy circle
= Ba, light circle = Y, square = CdO, closed circle =
0, star = missing row of 0. Image simulations (Scherzer
defocus) a t (B) 7.77-A, (C) 31.09-A, (D) 62.18-A
thickness.
Yttria-stabilized zirconia
The microstructure of the thin film deposited on YSZ was different from those grown
on SrTiO3. Figure 9 is a general view of one
of the as-deposited films. Two distinct types
of microstructure are visible: (1) mosaic
structures (Ml, M2) and (2) nonmosaic structures (N). The mosaic regions consist of rectangular grains 200-300 nm in length.
Individual grains within these regions are
oriented with either (100) or (010) parallel to
268
L.A. TIETZ ET AL.
Fig. 7. Plan-view bright-field images of a YB?&u307., film deposited on SrTiO,. A small
region of different orientation arrow and - 11.5-Aspacing of the (001)planes.
Fig. 8. SAD patterns of YBazCu307., film on SrTiO3. A Main orientation of the film:
or [110]pole. B: Two grains with c-axes rotated 90" to each other.
-
[%lo]
THIN FILMS OF YBazCu307-, GROWN ON SrTiO3 AND YSZ
269
be the result of damage to the Y B a z C ~ 3 0 7 . ~
structure from ion-milling or from the electron beam in the microscope during observation. Reports of planar defects in specimens
prepared by ion-thinning have been made by
Bentley et al. (19871, who reported that such
defects could by avoided by ion-thinning with
the benefit of LN2 cooling of the specimen.
Furthermore, Kelly et al. (1987)reported that
Lal.8Sro,2Cu04 foils experienced oxygen loss
when heated in vacuum (e.g., by ion or electron beams) and Ar implantation during ionmilling. Clark et al. (1987) have observed
that YBazCu307., phase become amorphous
during irradiation with 300-keV electrons in
the microscope, although in the present
study, a stable image could be obtained by
using a 400-kV microscope.
Damage by ion and electron irradiation
which leads to planar defects has been obDISCUSSION
served in other related materials. For examSpecimen preparation and its
ple, in p"'-alumina, it has been shown
relationship to defects
(h4orrissey et al., 1984)that cations (e.g., Na)
The planar stacking defects observed by can be removed from a n area of interest durhigh-resolution microscopy in this study may ing electron-beam irradiation. This displace-
the ( 0 0 1 ) substrate
~ ~ ~
surface so that the caxis is parallel to the interface as shown by
the SAD pattern in Figure 10a and b taken
from a similar area of the same specimen.
Within a mosaic region (e.g., Ml), adjoining
grains are oriented with either their c-axes
parallel or rotated by 90°C to each other.
Adjacent mosaic structures that are rotated
45" with respect to each other (e.g., M1 and
M2 in Fig. 9) are often found in these films.
Large areas of the films exhibit the nonmosaic-type structure such as region N indicated in Figure 9. These areas consist of
larger, irregularly shaped grains up to 1 pm
in diameter. The grains have their c-axes
perpendicular to the substrate surface as
shown in the SAD pattern in Figure 1Oc and
d taken from a similar region of the same
specimen.
Fig. 9. Plan-view bright-field image of YBaZCu307., film deposited on YSZ. General view
shows mosaic structure (MI,
Mz) and nonmosaic structure (N).
270
L.A. TIETZ ET AL.
ment can lead to the collapse of the
conduction planes which appear as planar
stacking defects in the high-resolution image. Ar+ ions can also replace alkali-metal
cations during ion-thinning, as shown by
EDS. These effects are ascribed to the very
open structure of the P"' phase. YBa2Cu307.,
also has a n open structure due to the oxygen
structural vacancies. Thus, YBa2Cu307.,
may be a candidate for similar damage.
In thin film studies such as that presented
in this paper, it is necessary to use ion-thinning in order to preserve the information on
the fildsubstrate interface. Therefore, care
should be used in ascribing the defects observed here to the deposition process, as they
may simply be artifacts of the specimen preparation technique and observation in the
electron microscope.
Comparison of observations on thin film
microstructure with those reported
in the literature
Based on the lattice mismatch calculation
listed in Table 1, certain orientation relationships were expected for epitactic film growth.
In the case of SrTi03 substrate, the c-axis
was found to lie parallel to the (001)srr SUface, as expected from the very small mismatch for this direction. Also, 90" rotation of
the c-axes in different grains was observed
as expected from the substrate symmetry.
Fig. 10. SAD patterns of YBazCu307., film on YSZ. a: SAD on grain mosaic region showing
orientation with substrate. b: Schematic of a showing pole of grain is [loo] or [OlO]. Stars
indicate spots from c-axis of a n adjoining grain. c: SAD on grain in nonmosaic region. d:
Schematic of c showing pole of grain is [OOl].
THIN FILMS OF YBazCu307-, GROWN ON SrTiO3 AND YSZ
271
However, it is unclear why the films adopted (ECS-82-00312).Additional support was proa < 100> growth direction. With the plan- vided by the Office of Naval Research under
view specimens it was not possible to deter- grant No. N00014-K-0296.
mine which directions were aligned in the
REFERENCES
filmhubstrate interface. Naito et al. (1987)
have suggested that the occurrence of orien- Adams, J.W., Nakamura, H.H., Ingel, R.P., and Rice
tations not favored by lattice mismatch may
R.W. (1985)Thermal expansion behavior of single crystal zirconia. J. Am. Ceram. Soc.,68:C228-C231.
be stabilized by the presence of a second
Beno, M.A., Soderholm, L., Capone D.W. 11, Hinks, D.G.,
phase.
Jorgensen, J.D., Segre, C.U., and Zhang, K. (1987)
In other studies of Y B a 2 C ~ 3 0 7 .films
~
Structure of the single-phase high-temperature superformed by electron-beam coevaporation
conductor Y B a z C ~ 3 0 7Appl.
. ~ Phys. Lett., 51:57-59.
(Chaudhari et al., 1987a; Naito et al., 19871, Bentley, J., Fisher, A.T., Haltom, C.P., Brynestad, J.,
Lauf, R.J., Williams, R.K., and Kroeger, D.M. (1987)
the authors report that the as-deposited films
Electron microscopy of high temperature oxide superare not superconducting and exhibit highly
conductors. Presented at the annual meeting of Electron Microscopy Society of America, Baltimore, MD,
disordered or amorphouos structures and that
August 2-7, 1987.
subsequent annealing is required to obtain
R.J., Batlogg, B., Van Dover, R.B., Murphy, D.W.,
the superconducting phase. In this study, all Cava,
Sunshine, S., Siegrist, T., Remeika, J.P., Reitman, E.A.,
of the films discussed showed highly crystalZahurak, S., and Espinosa, G.P. (1987) Bulk superconlized microstructures and were superconductivity at 91°K in a single phase oxygen deficient
perovskite Ba2YCu309.,. Phys. Rev. Lett., 58:1676ducting a t 80°K in the as-deposited state.
1679.
Finally, it should be noted that the results Chaudhari,
P., Koch, R.H., Laibowitz, R.B., McGuire,
presented here for plan-view specimens give
T.R., and Gambino, R.J. (1987a) Criticalcurrent measurements in epitaxial films of YBa2Cu307.x comno information on the distribution of the mipound. Phys. Rev. Lett., 58:2684-2686.
crostructure in the direction parallel to the
Chaudhari, P., Collins R.T., Freitas, P., Gambino, R.J.,
electron beam. Naito et al. (1987) have reKirtley, J.R., Koch, R.H., Laibowitz, R.B., LeGoues,
ported that the YBa2C~307.~
grains in their
F.K., McGuire, T.R., Penney, T., Schelsinger, Z., and
Segmiiller, A.P. (1987) Properties of epitaxial films of
thin films exhibit a columnar microstrucYBa2C~307.p
Preprint.
ture, while Chaudhari et al. (198710) have
C.W., Hor, P.H., Meng, R.L., Gao, L., Huang, Z.J.,
described their films as having a “bilayer” Chu,
Wang, Y.Q., Wu, M.K., Ashburn, J.R., and Huang, C.Y.
microstructure. In the latter study, single(1978) Superconductivity at 98K in the Y-Ba-Cu-0crystal growth occurs near the substratelepicompound system at ambient pressure. Phys. Rev.
Lett., 58:911.
layer interface with the c-axis perpendicular
G.J., LeGoues, F.K., Marwick, A.D., Laibowitz,
to the interface, and a polycrystalline micro- Clark,
R.B.. and Koch. R. (1987). Ion beam amorDhization of
structure of grains with c-axes in the plane
YBazCu30,. Preprint.
of the film lies on top of this such that the c- Dinger, T.R., Worthington, T.K., Gallagher, W.J., and
Sandstrom, R.L. (1987)Direct observation of electronic
axes align with the a- or b-axis of the underanisotropy in single-crystal Y 1BazCu307., Phys. Rev.
lying Y B a 2 C ~ 3 0 7 .film.
~
However, in that
Lett., 58:2687-2690.
case, the reported structure was only ob- Jin, B:Y., Lee, S.J., Song, S.N., Hwu, S.-J., Thiel, J.,
Poeppelmeier, K.R., and Ketterson, J.B. (1987)High T,
tained after postdeposition annealing in oxyY-Ba-Cu-0thin films prepared by DC magnetron sputgen. The as-deposited YBa2Cu307., films
tering. Adv. Ceram. Mater., 2(Spec. Issue):436-443.
grown on YSZ and described in this paper Kelly,
T.F., Lee, P.J., Hellstrom, E.E., and Larbalestier,
could also have this “bilayer” microstrucD.C. (1987) Study of segregation in Lal,$ro.zCu04
superconductors by STEM-EDS microanalysis. Preture. It has been observed that the grains in
sented at the annual meeting of the Electron Microsthe mosaic regions are often aligned with
copy Society of America, Baltimore, MD, Aug. 2-7,
their c-axes parallel to the a- or b-axis of the
1987.
grains in the adjacent nonmosaic region. Krishnan, R.J., Srinivasan, R., and Devanarayanan, S.
(1979)Thermal Exuansion of Crvstals. Perzamon
Cross-section TEM samples are being pre- Press.
New York, p. 142.
pared in order to clarify this point.
Laibowitz, R.B., Koch, R.H., Chaudhari, P., and GamACKNOWLEDGMENTS
The authors would like to thank Mr. R.
Coles for maintenance of the TEM facility
and Ms. M. Fabrizio for photographic work.
The Materials Science Facility for Electron
Microscopy is supported, in part, by NSF
(DMR-85-16616). This research was supported by NSF under grant No. DMR-8521834 and through use of the National Nanofabrication Facility a t Cornell University
bino, R.J. (1987) Thin superconducting oxide films.
Phys. Rev. B, 358821-8823.
LathroD. D.K.. Russek, S.E., and Burhman, R.A. (1987)
Production of YBazCu307., superconducting thin films
in-situ by high pressure reactive evaporation in rapid
thermal annealing. Preprint.
Moriwaki, K., Suzuki, M., Enomoto, Y., and Murakami,
T. (1987)Electrical properties of superconducting Gal.,
SR,)2Cu04 and BaZYC~3074
thin films. In: MRS Extended Abstracts. High Temperature Superconductors.
D.U. Gubser and M. Schluter, eds. Proceedings of Symposium S, 1987 Spring Meeting of the Materials Research Society, April 23-24, Anaheim, CA.:85-87.
272
L.A. TIETZ ET AL.
Morrissey, K.J., Elgat, Z., Kouth, Y., and Carter, C.B.
(1984) High resolution TEM studies of 0-alumina type
structures. Mater. Res. SOC.Symp. Proc., 31:331-336.
Naito, M., Hammamond, R.H., Oh, B., Hahn, M.R., Hsu,
J.W.P., Rosenthal, P., Marshall, A.F., Beasley, M.R.,
Geballe, T.H., and Kapitulnik, A. (1987) Thin film synthesis of the high-T, oxide superconductor YBazCu307
by electron beam co-deposition. Preprint.
Powder diffraction file (1977) No. 30-1468. International
Center for Diffraction; Swarthmore, PA.
Powder diffraction file (1985) No. 35-734. International
Center for Diffraction, Swarthmore, PA.
Somekh, R.E., Blamire, M.G., Barber, S.H., Butler, K.,
James, J.H., Morris, G.W., Tomlinson, E.J., Schwarzenberger, A.P., Stobbs, W.M., and Evetts, J.E. (1987) High
superconducting transition temperatures in sputterdeposited Y Ba Cu 0 thin films. Nature, 326:857.
Wu, M.K., Ashburn, J.R., Torng, G.J., Hor, P.H., Meng,
R.L., Gao, L., Huang, Z.J., Wang, Y.Q., and Chu, C.W.
(1987) Superconductivity at 93K in a new mixed-phase
Y-Ba-Cu-0 compound system at ambient pressures.
Phys. Rev. Lett., 58:908-910.
Yukino, K., Sato, T., Ooba, S., Ohta, M., Okamura, F.P.,
and Ono, A. (1987) Studies on the termal behavior of
Ba2YCu307., by x-ray powder diffraction method. Jpn.
J. Appl. Phys. Lett., 26:L869-L870.