- Email: [email protected]
Synthesis and structure of copper- and yttrium-rich YBa2Cu3O7−x superconductors
PhysicaC 154 (1988) 201-207 H
SYNTHESIS AND STRUCTURE SUPEIKONDUcTORs D.J. LI, H. SHIBAHARA, Department
qf Materiab
Science
OF COPPER- AND YTTRIUM-RICH
YRa2Cu307-r
J.P. ZHANG and L.D. MARKS and Materials
Research
Center,
Northwestefq
University,
Evanston.
IL 60208,
USA
H.O. MARCY Department
of Electrical
Engineering
and Materials
Research
Center,
Northwestern
University,
Evanston,
IL 60208.
USA
S. SONG Department
of Physics
and Materials
Research
Center,
Nun’hwestern
University.
Evanston,
IL 50208,
USA
Received 1 June 1988 Revised manuscript received 4 August 1988
Copper- and yttrium-rich modifications of the YBa.&&O,- , superconductors have been synthesized using a liquid nitrogen quench followed by annealing. In both types of cation-rich materials the starting material is highly disordered but upon annealing the extra copper or yttrium precipitates out in the form of planar defects normal to the c axis which on annealing at higher temperatures start to coarsen from single plane defects to multiple plane defects and become better ordered normal to the c axis. In all cases X-ray diffraction gives an average onhorhombic structure. The existence of these defects appears to have no effect upon the sharpness of the superconducting transition for specimens annealed zo 8OO”C, leads to a small broadenits of the transitions in specimens annealed at 7OO”C, while the specimens annealed at 300°C show only a hint of superconducting behavior. This supports the view that these materials are one- or two-dimensional superconductors normal to the c axis with only a small contribution to the bulk superconductivity along the c axis
1. Introduction
From extensive phase diagram studies using primarily X-ray methods, e.g. [ 1,2], it is now well established that the superconductor YBa2Cu307-.T is a point compound. At the same time, there has also been a num’re; jf reports using high resolution electron microscopy of planar defects containing either an extra plane of copper [3-91 and/or rare earths [ 10,ll .I. In particular, in our own work on the gadolinium superconductors [ 8,9], thin film materials [ 121 and in one grain of liquid nitrogen quenched YBa$u~O,-,v [8] these defects can be very common, and by using a fast quench we have been able to produce trigonal phases with different metal atom ordering [ 8,131. Since X-ray diffraction techniques will as a rule be quite insensitive to the presence of planar defects, an unanswered question is whether such defects and cation non-stoichiometry in general
exists undetected in many samples, and whether it plays a role in degrading the superconducting properties either in the short or long term. For instance, additional copper or rare earth elements can be expected in time to precipitate out and if they decorate grain boundaries could lead to long term degradation of the current carrying capabilities or alter the mechanical properties making the boundaries more brittle. The exact source of the cation non-stoichiometry and planar defects is at present unclear. Van Tenderloo and Amelinckx [ 51 and Zandbergen et al. [ 61 have pointed out that copper-rich defects can occur in specimens left in air for several weeks which from our results [ 81 is due to leaching of the barium from the structure and similar defects can also occur due to mechanical deformation or ion beam damage [ 8 1. Although these processes are possible sources of the defect structures, it does not follow that the only way
202
D.Z Li et al. / Copper- and ytlrium-rich 123 superconductors
that they can be formed is by a damage process. For instance, our original work [ 14] on the superconductors indicated that order-disorder reactions were possible, and, reinforced by the observation o f trigonai phases [ 8,13], implies that cation disorder m~y have a more intrinsic source. In this paper we report a method o f preparing both copper-rich ztnd yttrium-rich supercond~.c~ors which explains the ,'ole of additional cations i~ the perovskite superconductors; they are not simply artifacts or due to damage but represen~ the early stages of precipitation of excess of either o f these cations,
In both cases the extra metal forms planar defects normal to the c axis which show indications o f coarsening and ord~,ring when annealed at higher temperatures, very similar to Guinier-Preston zones. The rr~aterials appear to follow what is in essence an agehardening process, with the additional cations precipitating out from a metastable solid solution as planar defects which can be several thousand ~ngstr6ms in length. Electrical measurements indicate that the superconducting transition is still very sharp even with a very high concentration o f these defects.
Fig. I. High resolution images of YBazCu~:z: (a) the as-qoench~l'material~ ( 6 ) aher annealing at 300°C; and ( c ) after annealing at 700°C.
D.J. Li et al. /Copter- and yttrium-rich 123 superconductors
2. Experimentalmethod The startingpoint of the synthesiswas to follow the standardfiring/grinding procedureusing compositionsYl.lBa2Su307-xand YBa2Cu3.20,-,with an annealingtemperatureof 940°C and an initial tiring of 20 hoursin air. (Wehavealsousedthe same procedurefor a compositionof YBa&u307+ but to date havenot beenable to producebarium-rich phases;insteadphaseseparationgiving barium cop per oxide wasobserved.)The materialswerethen liquid nitrogenquenchedand three different sampleswereannealedin oxygenat 300°C for a week, 700°C for 72 hours and 800°C for 48 hours, respectively. To characterizethe specimens,both X-ray diffractionandhighresolutionelectronmicroscopywere employed.For theX-ray work,a diffractometerwith a Cu-K, sourcewas used.For the high resolution electronmicroscopyspecimenswerepreparedby very gentlegrinding betweentwo glassplates (to minimize mechanicaldeformation), mounteddry onto holeyca n films and then examinedusinga Hitachi I%9000electronmicroscopeoperatedat 300 kV. In general,specimenswerefreshlypreparedbefore microscopyanalysisto avoid problemswith water vapor damageof old specimens.Imagesimulationswereperformedusingthe NUMIS suiteof programson an Apollo workstationring to cbaracterizethe materials. 3. Results
The as-quenched material (seefig. la) showeda highly defectivestructurewith disorderof the c axis periodicity. Basedupon our earlier studiesof fast quenchedmaterials[ 8,131we canunderstandthese resultsasdueto a disorderingof the metal atomsin the basicperovskite3.8A unit cell; at 940°C thetripleperovskitelatticeis notwell established but is only establishedon cooling.Theremay well be a solid solution rangeat highertemperatures whereeither(and p&ably both) additionalcopperandyttrium is dissolvedin the 123phase.After annealingto 3OO”C, {seefig. lb), the triple perovskiteorderingwas established. X-ray measurementsindicate that this materialhasan averageorthorhombiccomposition
203
Table I Average lattice parameters ds determined by X-ray masurements of the various annealed specimens. Note that from the Xray data alone there is no way of determining that the materials contain numerous planar defects, or &at the orthorhombic 300°C annealed material is not a bulk superconductor, it is in fact the “most” orthorhombic material. Annealing temperature
a (A)
b (A,
c 6)
Cu-rich
,300 700 800
3.82 I 3.833 3.835
3.885 3.880 3.885
11.675 11.644 11.659
Y-rich
800
3.828
3.884
Il.655
(seetableI). The high resolutionresultsshowthat thereis still substantialdisorderwith only limited orderedregions.As thematerialis annealedto higher temperatures(seefig, lc) the materialbecomesbetter ordered and planar defectsextendingseveral thousand5ngstriimsnormalto the c axisarepresent as shownin fig. 2. This, and all the other electron microscopeimageshereinshowthetypical structure of the material,ratherthanthan foundin only a few grains,and is not an artifact of mechanicaldeformation or otherdamageto the material.Analysisof the images(seefig. 3a) indicatesthat the copper planardefectshavethe samebasicstructureas reportedelsewhere[ 7-9JPwhich is shownin fig. 3b, with the two copperplanesoccurringpredominantly in the basalplanebetweentwo barium sites(rather than betweenbarium and yttrium) with a translation vectorbetweenthe two planesof eitheri [ 3011 or i [a311. The excessyttrium is accommodatedin a very similar fashionas shownin fig. 4a with the structureshown in fig. 4b. Although perhapsnot completelyapparentin theimagesshown,thereis no dangerof confusingthe copper-andyttrium-richdefects;thelattershowsubstantiallyhighercontrastand occur at different positionswith respectto the,adjacent perovskiteframework. For the 800°C annealedcopperspecimensthere are indicationsof coarseningleadingto defectscomposedof threecopper planesas shownin fig. 5. Althoughwe cannot rigorouslyprovethis interpretationfrom the image contrast,both the spacing,contrastlevelandtheabL senceof singleplanecopperdefectsin the regionof theselargerdefectstructuressupportthis interpre-
Fig, 2. L ~ ' mag~ifseatio~ imagg of~he cocker-rich material annealed to 800°C, The image shows a typical region containing numerous copper-rich defects running across the figure. The ¢lancs are continuous over at least several thousand tlngstr6ms.
....
b
a b
O~ @v
Fig. 3. High resolution image (a) and structure model (b) of the copper-rich defects taken from the 800°C annealed specimen with a calculated image inset. The image contrast correlates well with that reported elsewhere [7-9].
D.J. Li et al. /Copper- and yttriwn-rich
I?:..
A W-L
#a\ with imae &&tjmw :, ,:,
Fig. 5. An
I23 superconductors
(inset4 . : and struature model (b) for the yttrium-rich defects. :,;
planes.
206
D.J. Li et al. /Copper- and yttrium-rich 123 superconductors
s°F----
a
tation. The density of the copper-or yttrium-rich planeswas consistentwith the initial startingcomposition, and the total excesscation concentration appearedto remain essentiallyconstantduring the annealing.However,we cannotcompletelyrule out thepossibilitythat a small fraction of the excesscationshasphaseseparated,but wecandefmitely state that themajority of the excesscationshavenot. For instance,in the caseof the yttrium-rich material, substantial quantities of the green phase ( Yb2BaCu0,)werenot apparentin either X-ray or electrondiffraction patterns,and the material was primarily “single phase”Y, .,Ba2Cu307--x. In manyrespectsthemostsurprisingresultwasthe width of the superconductingtransition using four point resistance measurements on freshpelletswhich is shownin f;,g.6. For the 300°Cannealthe material is almost completelysemiconducting,with only a small indication of a superconductingtransition around90 K, while the 700°C samplesshowa small tail andin the 800°C samplesthe transitionis very sharp.Experimentsto analyzein moredetail the superconductingpropertiesfrom Meissnermeasurements are in progresswhich may reveal more information. 4. Discussion
TEMPERATJJRE NI)
TEMPERA TURE CKJ
Fig. 6. Resistivity meawrements: (a) 300°C annealed specimen, (b) 700°C annealed specimen, and (c) 800°C annealed specimen.
Oneclearconclusionfrom this work is thatcopper and/oryttrium planardefectsmay be very common in the YBa2Cu30,-, superconductingsystem,and may well bea sourceof systematicerrorsin property measurements.For instance,we feel that the conclusionsof experimentswhereother elementshave beensubstitutedinto the superconductors, e.g.ref. [ 151,haveto be consideredassuspectsinceit is not clear that theseelementshave in fact replacedthe copper;theymaybepresentasplanardefectsor even precipitatesandit doesnot appearthat this hasbeen tested.The factthat annealingat 800°Cfor 48hours doesnot eliminate (by phaseseparation)the excess metal explainstheir quite common observationby high resolutionelectronmicroscopy.The surprising resultthat we seeno changein the superconducting transitionwidth for the 800°Cspecimensimpliesthat, the superconductionis quite stronglylocalizednormal to the c axis,andis not effectedby theexistence
D.J. Li et ai. /Copper- and yttrium-rich I23 superconductors
of disruptions to the lattice 10 or 20 A away. The quenching of the transition in the 3OW.Z samples indicates that disrupting the ordering normal to the c axis strongly effects the bulk superconductivity, although there may be small superconducting domains which are electrically disconnected. Fnrther expeniments to explore this and, among other properties, the mechanical characteristics of these materials are in progress. Acknowledgements This work was supported by the National Science Foundation through Northwestern University Materials Research Center on Grant number DMR 85 20280 and the Airforce Offtce of Scientific Research on grant number AFOSR 86-0344 DEE One of us (HOM) would :ike to acknowledge support from the ONR-SRO program. References [ 1] G. Wang, S.J. Hwu, S.N. Song, J.B. Ketterson, L.D. Marks, K.R. Poeppelmeier and T.O. Mason, Adv. Ceram. Mater. 2 (38) (1987) 313.
207
[2] R.S. Roth, K.L. Davis and J.R. Dennis, Adv. Ceram. Mater. 2 (1987) 303. [3] H.W. Zandbergen, G. Van Tenderloo, T. Okabe and S. Amelinckx, Phys. Status Solidi, 102 (1987) 45. [4] G. Van Tenderloo, H.W. Zandbergen, T. Okabe and S. Amelinckx, Solid State Commun., 63 (15187) 969. [ 51 G. Van Tenderloo and S. Amelinckx, J. Electron Microsc. Techn. 8 (1988) 285. [6] H.W. Zandbergen, R. Gronsky and G. Thomas, Phys. Status Solidi a, preprint. [ 7) H.W. Zandbergen, R. Gronsky, K. Wang and G. Thomas, Nature 33 1 (1988) 596. [8] L.D. Marks, D.J. Li, H. Shibahara and J.P. Zhang, J. Electron Microsc. Techn. 8 (1988) 297. [9] H. Shibahara, S.J. Hwu, KR. Poeppelmeier and L.D. Marks, J. Solid State Chem., submitted. [ lo] A. Ourmazd, J.C. Spence, J.M. Zuo and C.H. Li, 3. Electron Microsc. Techn. 8 (1988) ?5![ I1 ] H.W. Zandbergen, G.F. Holland, P. Tejedor, R. Gronsky and A.M. Stacey, Adv. Ceram. Mater. 2 (1987) 688. [ 121 H. Shibahara, D.J. Li, J.P. Zhang and L.D. Marks, in preparation. [13] D.J. Li, J.P. Zhang, J.P. Theil and L.D. Marks, in preparation. [ 141 L.D. Marks, J.P. Zhang, S.-J. Hwuand K.R. Poeppelmeier, J. Solid. State Chem. 69 (1987) 189. [ 151 G. Xiao, M.Z. Cieplak, A. Gavin, F.H. Streitz, A. Bakhshai andC.L. Chien, Phys. Rev. Lett. 60 (1988) 1446.
SYNTHESIS AND STRUCTURE SUPEIKONDUcTORs D.J. LI, H. SHIBAHARA, Department
qf Materiab
Science
OF COPPER- AND YTTRIUM-RICH
YRa2Cu307-r
J.P. ZHANG and L.D. MARKS and Materials
Research
Center,
Northwestefq
University,
Evanston.
IL 60208,
USA
H.O. MARCY Department
of Electrical
Engineering
and Materials
Research
Center,
Northwestern
University,
Evanston,
IL 60208.
USA
S. SONG Department
of Physics
and Materials
Research
Center,
Nun’hwestern
University.
Evanston,
IL 50208,
USA
Received 1 June 1988 Revised manuscript received 4 August 1988
Copper- and yttrium-rich modifications of the YBa.&&O,- , superconductors have been synthesized using a liquid nitrogen quench followed by annealing. In both types of cation-rich materials the starting material is highly disordered but upon annealing the extra copper or yttrium precipitates out in the form of planar defects normal to the c axis which on annealing at higher temperatures start to coarsen from single plane defects to multiple plane defects and become better ordered normal to the c axis. In all cases X-ray diffraction gives an average onhorhombic structure. The existence of these defects appears to have no effect upon the sharpness of the superconducting transition for specimens annealed zo 8OO”C, leads to a small broadenits of the transitions in specimens annealed at 7OO”C, while the specimens annealed at 300°C show only a hint of superconducting behavior. This supports the view that these materials are one- or two-dimensional superconductors normal to the c axis with only a small contribution to the bulk superconductivity along the c axis
1. Introduction
From extensive phase diagram studies using primarily X-ray methods, e.g. [ 1,2], it is now well established that the superconductor YBa2Cu307-.T is a point compound. At the same time, there has also been a num’re; jf reports using high resolution electron microscopy of planar defects containing either an extra plane of copper [3-91 and/or rare earths [ 10,ll .I. In particular, in our own work on the gadolinium superconductors [ 8,9], thin film materials [ 121 and in one grain of liquid nitrogen quenched YBa$u~O,-,v [8] these defects can be very common, and by using a fast quench we have been able to produce trigonal phases with different metal atom ordering [ 8,131. Since X-ray diffraction techniques will as a rule be quite insensitive to the presence of planar defects, an unanswered question is whether such defects and cation non-stoichiometry in general
exists undetected in many samples, and whether it plays a role in degrading the superconducting properties either in the short or long term. For instance, additional copper or rare earth elements can be expected in time to precipitate out and if they decorate grain boundaries could lead to long term degradation of the current carrying capabilities or alter the mechanical properties making the boundaries more brittle. The exact source of the cation non-stoichiometry and planar defects is at present unclear. Van Tenderloo and Amelinckx [ 51 and Zandbergen et al. [ 61 have pointed out that copper-rich defects can occur in specimens left in air for several weeks which from our results [ 81 is due to leaching of the barium from the structure and similar defects can also occur due to mechanical deformation or ion beam damage [ 8 1. Although these processes are possible sources of the defect structures, it does not follow that the only way
202
D.Z Li et al. / Copper- and ytlrium-rich 123 superconductors
that they can be formed is by a damage process. For instance, our original work [ 14] on the superconductors indicated that order-disorder reactions were possible, and, reinforced by the observation o f trigonai phases [ 8,13], implies that cation disorder m~y have a more intrinsic source. In this paper we report a method o f preparing both copper-rich ztnd yttrium-rich supercond~.c~ors which explains the ,'ole of additional cations i~ the perovskite superconductors; they are not simply artifacts or due to damage but represen~ the early stages of precipitation of excess of either o f these cations,
In both cases the extra metal forms planar defects normal to the c axis which show indications o f coarsening and ord~,ring when annealed at higher temperatures, very similar to Guinier-Preston zones. The rr~aterials appear to follow what is in essence an agehardening process, with the additional cations precipitating out from a metastable solid solution as planar defects which can be several thousand ~ngstr6ms in length. Electrical measurements indicate that the superconducting transition is still very sharp even with a very high concentration o f these defects.
Fig. I. High resolution images of YBazCu~:z: (a) the as-qoench~l'material~ ( 6 ) aher annealing at 300°C; and ( c ) after annealing at 700°C.
D.J. Li et al. /Copter- and yttrium-rich 123 superconductors
2. Experimentalmethod The startingpoint of the synthesiswas to follow the standardfiring/grinding procedureusing compositionsYl.lBa2Su307-xand YBa2Cu3.20,-,with an annealingtemperatureof 940°C and an initial tiring of 20 hoursin air. (Wehavealsousedthe same procedurefor a compositionof YBa&u307+ but to date havenot beenable to producebarium-rich phases;insteadphaseseparationgiving barium cop per oxide wasobserved.)The materialswerethen liquid nitrogenquenchedand three different sampleswereannealedin oxygenat 300°C for a week, 700°C for 72 hours and 800°C for 48 hours, respectively. To characterizethe specimens,both X-ray diffractionandhighresolutionelectronmicroscopywere employed.For theX-ray work,a diffractometerwith a Cu-K, sourcewas used.For the high resolution electronmicroscopyspecimenswerepreparedby very gentlegrinding betweentwo glassplates (to minimize mechanicaldeformation), mounteddry onto holeyca n films and then examinedusinga Hitachi I%9000electronmicroscopeoperatedat 300 kV. In general,specimenswerefreshlypreparedbefore microscopyanalysisto avoid problemswith water vapor damageof old specimens.Imagesimulationswereperformedusingthe NUMIS suiteof programson an Apollo workstationring to cbaracterizethe materials. 3. Results
The as-quenched material (seefig. la) showeda highly defectivestructurewith disorderof the c axis periodicity. Basedupon our earlier studiesof fast quenchedmaterials[ 8,131we canunderstandthese resultsasdueto a disorderingof the metal atomsin the basicperovskite3.8A unit cell; at 940°C thetripleperovskitelatticeis notwell established but is only establishedon cooling.Theremay well be a solid solution rangeat highertemperatures whereeither(and p&ably both) additionalcopperandyttrium is dissolvedin the 123phase.After annealingto 3OO”C, {seefig. lb), the triple perovskiteorderingwas established. X-ray measurementsindicate that this materialhasan averageorthorhombiccomposition
203
Table I Average lattice parameters ds determined by X-ray masurements of the various annealed specimens. Note that from the Xray data alone there is no way of determining that the materials contain numerous planar defects, or &at the orthorhombic 300°C annealed material is not a bulk superconductor, it is in fact the “most” orthorhombic material. Annealing temperature
a (A)
b (A,
c 6)
Cu-rich
,300 700 800
3.82 I 3.833 3.835
3.885 3.880 3.885
11.675 11.644 11.659
Y-rich
800
3.828
3.884
Il.655
(seetableI). The high resolutionresultsshowthat thereis still substantialdisorderwith only limited orderedregions.As thematerialis annealedto higher temperatures(seefig, lc) the materialbecomesbetter ordered and planar defectsextendingseveral thousand5ngstriimsnormalto the c axisarepresent as shownin fig. 2. This, and all the other electron microscopeimageshereinshowthetypical structure of the material,ratherthanthan foundin only a few grains,and is not an artifact of mechanicaldeformation or otherdamageto the material.Analysisof the images(seefig. 3a) indicatesthat the copper planardefectshavethe samebasicstructureas reportedelsewhere[ 7-9JPwhich is shownin fig. 3b, with the two copperplanesoccurringpredominantly in the basalplanebetweentwo barium sites(rather than betweenbarium and yttrium) with a translation vectorbetweenthe two planesof eitheri [ 3011 or i [a311. The excessyttrium is accommodatedin a very similar fashionas shownin fig. 4a with the structureshown in fig. 4b. Although perhapsnot completelyapparentin theimagesshown,thereis no dangerof confusingthe copper-andyttrium-richdefects;thelattershowsubstantiallyhighercontrastand occur at different positionswith respectto the,adjacent perovskiteframework. For the 800°C annealedcopperspecimensthere are indicationsof coarseningleadingto defectscomposedof threecopper planesas shownin fig. 5. Althoughwe cannot rigorouslyprovethis interpretationfrom the image contrast,both the spacing,contrastlevelandtheabL senceof singleplanecopperdefectsin the regionof theselargerdefectstructuressupportthis interpre-
Fig, 2. L ~ ' mag~ifseatio~ imagg of~he cocker-rich material annealed to 800°C, The image shows a typical region containing numerous copper-rich defects running across the figure. The ¢lancs are continuous over at least several thousand tlngstr6ms.
....
b
a b
O~ @v
Fig. 3. High resolution image (a) and structure model (b) of the copper-rich defects taken from the 800°C annealed specimen with a calculated image inset. The image contrast correlates well with that reported elsewhere [7-9].
D.J. Li et al. /Copper- and yttriwn-rich
I?:..
A W-L
#a\ with imae &&tjmw :, ,:,
Fig. 5. An
I23 superconductors
(inset4 . : and struature model (b) for the yttrium-rich defects. :,;
planes.
206
D.J. Li et al. /Copper- and yttrium-rich 123 superconductors
s°F----
a
tation. The density of the copper-or yttrium-rich planeswas consistentwith the initial startingcomposition, and the total excesscation concentration appearedto remain essentiallyconstantduring the annealing.However,we cannotcompletelyrule out thepossibilitythat a small fraction of the excesscationshasphaseseparated,but wecandefmitely state that themajority of the excesscationshavenot. For instance,in the caseof the yttrium-rich material, substantial quantities of the green phase ( Yb2BaCu0,)werenot apparentin either X-ray or electrondiffraction patterns,and the material was primarily “single phase”Y, .,Ba2Cu307--x. In manyrespectsthemostsurprisingresultwasthe width of the superconductingtransition using four point resistance measurements on freshpelletswhich is shownin f;,g.6. For the 300°Cannealthe material is almost completelysemiconducting,with only a small indication of a superconductingtransition around90 K, while the 700°C samplesshowa small tail andin the 800°C samplesthe transitionis very sharp.Experimentsto analyzein moredetail the superconductingpropertiesfrom Meissnermeasurements are in progresswhich may reveal more information. 4. Discussion
TEMPERATJJRE NI)
TEMPERA TURE CKJ
Fig. 6. Resistivity meawrements: (a) 300°C annealed specimen, (b) 700°C annealed specimen, and (c) 800°C annealed specimen.
Oneclearconclusionfrom this work is thatcopper and/oryttrium planardefectsmay be very common in the YBa2Cu30,-, superconductingsystem,and may well bea sourceof systematicerrorsin property measurements.For instance,we feel that the conclusionsof experimentswhereother elementshave beensubstitutedinto the superconductors, e.g.ref. [ 151,haveto be consideredassuspectsinceit is not clear that theseelementshave in fact replacedthe copper;theymaybepresentasplanardefectsor even precipitatesandit doesnot appearthat this hasbeen tested.The factthat annealingat 800°Cfor 48hours doesnot eliminate (by phaseseparation)the excess metal explainstheir quite common observationby high resolutionelectronmicroscopy.The surprising resultthat we seeno changein the superconducting transitionwidth for the 800°Cspecimensimpliesthat, the superconductionis quite stronglylocalizednormal to the c axis,andis not effectedby theexistence
D.J. Li et ai. /Copper- and yttrium-rich I23 superconductors
of disruptions to the lattice 10 or 20 A away. The quenching of the transition in the 3OW.Z samples indicates that disrupting the ordering normal to the c axis strongly effects the bulk superconductivity, although there may be small superconducting domains which are electrically disconnected. Fnrther expeniments to explore this and, among other properties, the mechanical characteristics of these materials are in progress. Acknowledgements This work was supported by the National Science Foundation through Northwestern University Materials Research Center on Grant number DMR 85 20280 and the Airforce Offtce of Scientific Research on grant number AFOSR 86-0344 DEE One of us (HOM) would :ike to acknowledge support from the ONR-SRO program. References [ 1] G. Wang, S.J. Hwu, S.N. Song, J.B. Ketterson, L.D. Marks, K.R. Poeppelmeier and T.O. Mason, Adv. Ceram. Mater. 2 (38) (1987) 313.
207
[2] R.S. Roth, K.L. Davis and J.R. Dennis, Adv. Ceram. Mater. 2 (1987) 303. [3] H.W. Zandbergen, G. Van Tenderloo, T. Okabe and S. Amelinckx, Phys. Status Solidi, 102 (1987) 45. [4] G. Van Tenderloo, H.W. Zandbergen, T. Okabe and S. Amelinckx, Solid State Commun., 63 (15187) 969. [ 51 G. Van Tenderloo and S. Amelinckx, J. Electron Microsc. Techn. 8 (1988) 285. [6] H.W. Zandbergen, R. Gronsky and G. Thomas, Phys. Status Solidi a, preprint. [ 7) H.W. Zandbergen, R. Gronsky, K. Wang and G. Thomas, Nature 33 1 (1988) 596. [8] L.D. Marks, D.J. Li, H. Shibahara and J.P. Zhang, J. Electron Microsc. Techn. 8 (1988) 297. [9] H. Shibahara, S.J. Hwu, KR. Poeppelmeier and L.D. Marks, J. Solid State Chem., submitted. [ lo] A. Ourmazd, J.C. Spence, J.M. Zuo and C.H. Li, 3. Electron Microsc. Techn. 8 (1988) ?5![ I1 ] H.W. Zandbergen, G.F. Holland, P. Tejedor, R. Gronsky and A.M. Stacey, Adv. Ceram. Mater. 2 (1987) 688. [ 121 H. Shibahara, D.J. Li, J.P. Zhang and L.D. Marks, in preparation. [13] D.J. Li, J.P. Zhang, J.P. Theil and L.D. Marks, in preparation. [ 141 L.D. Marks, J.P. Zhang, S.-J. Hwuand K.R. Poeppelmeier, J. Solid. State Chem. 69 (1987) 189. [ 151 G. Xiao, M.Z. Cieplak, A. Gavin, F.H. Streitz, A. Bakhshai andC.L. Chien, Phys. Rev. Lett. 60 (1988) 1446.