8/3/2019 BaC03 FTIR

1/6

Inorg. Chem. 1993,32 , 3093-3098 3093

The Interaction between CO and YBazCujO, As Studied by TG/DTA, FTIR, and XPSJ. Lin,,'vt K,G. Neoh,? N.Li,ts* T. C. TanJ A. T. S. Wee,g A. C. H. Huan,i and K. L. TansDepart ment s of Ch emica l Engineering and Physics, National University of Singapore, Singapore 0511Received December 2, I9 92

The interaction of CO with YBazCu30, ( x - 7) has been studied using differential thermal analysis (DTA)/thermogravimetry (TG), Fourier transform infrared spectroscopy (FT IR), and X-ray photoelectron spectroscopy(XPS ). The interaction at temperatures between 400 and 800 K proceeds as CO + O(1attice) = C02, as shownby the evolved gas analysis and DT A/T G calorimetric measurement. At 800-1200 K a large CO uptake is observed,resulting in the formation of Ba C0 3, which is characterized by FT IR, and the reduction of Cu(I1) and Y (III), asobserved by XPS. An app arent activation energy of 16 kcal/m ol a nd a heat of reaction of 60 kcal/mol are estimatedfor this high-temperature reaction from the D TA/T G studies. These results are compared with those obtained forYBa2Cu3OYO,- ) and the composite compounds, CuO, Y2O3, and BaC03. A detailed mechanism which mayexplain the higher catalytic activity on YBa2Cu30x han on CuO3 is suggested.Introduction

Cu-based oxides are presently the most widely used ca talystsfor meth anol synthesis and the water-gas shift reaction.' Theyare also viewed as prospective catalysts for removing nitric oxideand carbon monoxide from automative exhaust emission^.^.^Recen tly, the well-known supercondu cting Y-Ba cuprates ,including YBa2Cu307, showed high c atalytic activities,ss as wellas stron g affinity, toward som e small-molecule gases, including0 2 , H2, H20, C02 , CO, andN0.5-'0 In particular, theperovskite-type Y-Ba cuprates are found to be more active in catalyzin g thereaction N O + CO = '/2N2 + C0 2 than other types of coppercatalyst^.^ Therefore it is of practical significance to stu dy theinteractio n between CO and the perovskite-type Y-Ba cuprates .In this work two Y-Ba cuprates , YB a2Cu 30x nd YBa2Cu3OY,are prepared in the same way as reported in the literature,according o which x - nd y - .5.6 Th e interac tion betweenCO and these samples is studied using simu ltaneous thermalanalysis (STA), X-ray photoelectron spectroscopy (XPS), andFourier transform infrared spectroscopy (FT IR). ST A allowsfor the calorimetric measurement and kinetic study of the C Ointeractionby a combination of simu ltaneou sdifferential therma lanalysis (DTA), thermogravimetry (TG ), and evolved g as analysis(EGA). FTIR is used to characterize the reaction products inthe solid phase while XP S can provide inform ation on chan gesin the oxidation states due to the reaction. The results obtainedfor the YBa2Cu30, and YBa2Cu3O Yamples are com pared with

t Department of Chemical Engineering.t Present address: Department of Minerals Engineering and ExtractiveMetallurgy, Western Austra lian School of Mines, P.O. Box 597, Kalgoorlie6430, Westem A ustralia, Australia.8 Department of Physics.(1) Bridger, G. W.; Spencer, M. S. n Cutulyst Handbook, 2nd4.;wigg,

M. V.,Ed.; Wolfe Publishing Ltd.: London, 1989; p 446 and referen ctstherein.( 2 ) Ross, J. Appl. Cutul. 1992, E l , N18 and references therein.(3) Hala sz, I.; B renner, A.; Shelef, M.; Ng, K.Y. S.Cutul. Lett. 1 9 9 1 , l l ,327 and references therein.(4) Hansen, S.; tami ri, J.; Bovin, J.; Anderson, A. Nurure 1988,334, 143.(5) Mizuno,N.; amato, M .; Misono, M. J . Chem.Soc.,Chem. Commun.1988,887.(6) Gallagher, P. K.; O'Bryan, H. M.; Sunshine, S.A.; Murphy, D. W.Muter. Res. Bull. 1981,22, 995.(7) Haller, I.; Shafe r, M. W.; Figat, R.; Goland, D. B. In Chemistry ofOxide Superconductors; Rao, C. N. R., Ed.; Blackwell Scien tificF'ublicatioins: London, 1988; p 93.(8) Yan, M. F.; Barn, R. L.; O'Bryan, H. M.; Gallagher, P. K.; S h e r w d ,R. C.; Jin, S. Appl. Phys. Lert. 1987, 51 , 532.(9) Qiu, S.L.; Ruclunan, M. W.; Brookes,N. B.; Johnso n, D.; Chen, J.; L in,C. L.;Strongin , M.;Sinkovic, B.; Crow, J. E.; ec,C. Phys. Rev. B 1988,37, 3747.(10) McCallum, R. W. J. Met. 1989, 50.0020- 669/93J 1332-3093$04.00/0

those for the composite compounds CuO, Y2O3, and BaC 03. Onthe basis of these studies, a detail ed mechanism of the in teractionis postulated.Experimental Section

The YBazCu 30, sample was prepared as described in the literature.5Powder mixtures of Y203 (Fluka, AG), BaCO3 (Merck, AG), and Cu O(Fluka, AG ) of the required stoichiometric ratio were pressed into pelletsand calcined at 1200 K overnight in air. They were then reground welland repressed into pellets and sintered at 1200 K overnight again. Thesample prepared was characterized on a JEOL JSM-T330A scanningelectron microscope (SEM), ith the atomic ratio of Y:Ba:Cu H 1:2:3as determined by the energy-dispersive analysis of X-ray s (EDA X). T hevalue of x in YBazCu30, thus prepared was reported to range from 6.5to 6.9.5 The Y BazCu30, sam ple was obtained by heating the YBaz-C U ~ O , ample to 1200 K in Nz. The T G measurement indicated thatYBazCu 30, had -1.4% weight loss, as compared with YBazCu30,.According to the li terature, the weight loss was due to the removal ofoxygen from the crystal: which would give a value close to 6 for y . Th ecrystal structur e of the YBazCu30, sample was determined by powderX-ray diffraction (XR D), which was performed on a Philips PW17 29X-ray diffractometer with Cu Ka! radiation. The X-ray powder diffractionspectrum of the YBazCu30, sample was found to be identical to thosereported in the literatur e, indicating tha t the sam ple prepared was indeeda single phase with the known perovskite-type structure.5.6The simultaneous thermal analysis was performed on a N etzsch ST A409. Approxim ately 100-200 mg of the sample prepared as describedabove was placed in an open crucible and heated from room tem peratureto - 200 K a t a linear rate of 10 OC/min by a temperatur e-program medheater. The same amoun t of kaolin, a thermally and chemically inertmaterial, was loaded in another identical and symmetrically locatedcrucible as the reference. Two thermocouples attached to the cruciblesmeasured the tem perature dependence (AT) between the sample and thereference mater ial, providing information on the he at flow during theexperiment. C O was passed through the sample at a flow rate of 100mL/m in for studying its interaction with the sample. The temperaturedifference (AT) was continuously recorded as a function of tempera ture ,giving the diff erential thermal analysis (DT A) curve, while the weightof the sample was recorded as a thermogravimetry (TG) plot. For aquantitative evaluation of the D TA signals, calibration was performedby melting standard materials such as Sn or KzCrO4. The calibrationfactor was determined by the calculation of the quotient of th e real heatof melting of the standard materials and the area of the D TA peak. Theweight change during the reaction was determined on the basis of thedepartureof theTG curvef rom thebaseline. Theevolvedgas wasanalyzedby gas chromatography (GC ) on a Shimadzu GC-14A using a PorapakQ (2 m ong) column with helium gas as the carrier gas (30 m L/mi n).The infrared spectra of the samples were recorded on a ShimadzuFTIR-8101 using the pellet technique, which involved mixing the finelyground samp le with potassium bromide powder an d pressing the mixture

0 1993 American Chemical Society

8/3/2019 BaC03 FTIR

2/6

3094 Inorganic Chem istry, Vol. 32 , No. 14 , 1993 Lin et al.- 1 I

25-2-

8 .5-

-1SI 4-2do 460 &I & 7bo 800 Qh lo0 0 l i0 0 1 2 boTemperature (19

Figure 1. TG data with a linear heating ra te of 10 OC/min for (1 ) YBaz-CulO, in CO, (2) YBazCu30, in Nz, (3 ) YBazCu30, in CO, and (4)CuO in CO. Weight change (%) is determined by the quotient of theweight change,AW, nd the initial weight of the sample, W.

0 4-

-2300 400 500 600 700 800 900 loo0 1100 1200Temperature (K)Figure 2. DTA data with a linear heating rate of 10 OC/min for (a)YBazCu30, in CO, (b ) YBazCu30, in CO, and (c) CuO in CO. Heatflow is determinedby the measurement of the differential temperature( A T ) n mV/unit.in an evacuable die at 8 tons of pressure to produce a transparent disk.Forty scans were taken for each spectrum o obtain a good signal-to-noiseratio.The XP S experiments were performed on a VG ESCALAB MKIIspectrometer using a Mg K a X-ray source (1253.6 eV, 120 W) at aconstant analyzer pass energy of 20 eV. With an ultimate base pressureof - X l0-l0mbar, henormaloperating pressurein heanalysischamberwas < 2 XResults and Analysis1. CO Interaction at Temperatures below 800 K. Th edifferential thermal analysis (DTA ) and thermogravimetry (TG)results for the interaction of C O with YBa2Cu30x,YBa2Cu30y,and their com posite compou nds are illustrated in Figu res 1 and2, respectively. Curv e 1 in Figure 1 is the T G dat a for the YBa2-C u3 0x ample which was heated to 1200 K in C O in the NetzschSTA 409. Curve 2 in Figure 1 was obtained by heating theYBa2Cu30x ample under N2 in the STA 409, which resulted inthe formation of YBa~C u30,. After cooling to room temperature,the YBazCusO, sample was heated to 1200 K in CO at a linearrate of 10 OC/min, giving curve 3. Th e composite compoundsCuO , Y2O3, and BaC 03 were also studied by S TA. Cu O reactswith C O strongly as shown in curve 4 of Figure 1. The data forthe Y2O3 and BaCO3 samples are not presented, since little changein the sample weight was observed for the Y2O3 and BaCO3samples heated in C O, even if they were pretreated, prior to th einteraction with C O, by the s ame procedure as in the preparationof the Y Ba2Cu30xsam ple. This indicates tha t these two compositecompounds of YBazCu30, are fairly stablewith respect to heating

mbar during the course of th e measurements.

so0 400 500 600 70 0 800 900 lo00 1100 1200T

Figure3. Evolved COz analysisby GC for the YBazCu30, sample heated(at 10 "C/min) in CO (with a flow rate of 100mL/min).an d CO interaction at temperatures below 1200 K. The DTAresults are illustrated in Figure 2 with curves a-c referring to theYBa2Cu30x, YBa2Cu3OY, nd C uO samples heated in C O a t alinear rat e of 10 OC/min, respectively. As shown in curve 1 ofFigure 1 and curve a of Figu re 2, YBazCusO, has exhibitedsimultaneousweight loss (- 1%) and heat release at emperaturesbetween 40 0 and 800 K. Th e evolved gas analysis (EGA ) by G Cindicates that COz is the unique product in the evolved gas andits concentration has a maximum a t temperatu res between 700and 800 K (see Figure 3). The EGA results are in a goodagreement with the T G data shown in curve 1 of Figure 1, eadingto th e conclusion tha t th e loss in the sam ple weight under theseconditions is totally du e to the formation a nd evolution of COz.Thus the interaction between CO and YBazCu30, a t 400-800K can be expressed as

C O ( g ) + O(1attice) = CO,(g) (1)On he basis of this observation, a quantitative evaluation of thetherm al effect of the reaction can be performed. A reaction heat,AH, f 38 kcal/mol is derived from the DT A/T G results accordingto t he following formula:

AH = Q / ( A W / l 6 )where Q is the heat release measured from the DT A study andAW the w eight loss (due to t he depletion of the latt ice oxygenas shown in reaction 1) obtained from the TG curve. This valueis almost the sam e as that reported in the literature," givingweight to the conclusion derived from th e EG A study. No evidentinteraction is observed between CO and YBa~C u30, n thistemp erature region (see curve 3 in Figu re 1 and curve b in Figure2). As shown in curve 4 of Figure 1 and curve c of Figure 2, C uOreacts strongly with C O in this temperature range. The EGAand calorimetric measurem ents of the heat of reaction give resultsvery sim ilar to those observed for the YBa2 Cu30 , sample.

2. CO Interaction at High Temperatures (800-1200 K).Furt her enhancement in the reaction temperature for the CO/YBazCu30, system results in the rapid increase in the weight ofthe Y Ba2Cu30, sample and n the simultaneous strong heat release(see curve 1 of Figure 1 and curve a of Figure 2). The exothermicheat effect measured corresponds to a heat of reaction of 60kcal/mol, if it is postulated th at th e increase in the weight is dueto the uptake of C O through the reaction

CO + OMO(1att ice) = M C O , (2)Th e heat of reaction thus m easured agrees well with the knownvalue reported in t he literature," supporting the above assumption.The formation of th e carbonate in the solid phase as the resultof the CO interaction has been confirmed by th e FTIR tudy. In(11) Stone, F.S. du. C a r d . 1962, 23, 1.

8/3/2019 BaC03 FTIR

3/6

The Interaction between CO and YBa2Cu30,

-8-

-8.2

-8.4-Y -8.6-

Inorganic Chemistry, Vol. 32,No. 14, 1993 3095

- 9TRANSYISSI0N

1 I 11500 1doo 560Wavenumber(dl)

Figure 4. Comparison of the FTIR spectra obtained from (a) a mixtureof BaCO3, Y2O3, and CuO and (b) the YBa2Cu30, sample afterinteraction with CO at temperatures up to 1200 K.Figure 4 the infra red spectra are displayed for a m ixtureof BaCO3,Y2O3, and CuO (spectrum a) and for the YBa2Cu30, sampleafter the interaction with CO at high temperatures (spectrum b).In the w avenumber region between 650 and 1600 cm-l the twospectra are basically the sam e. This is strong evidence of theformation of Ba C0 3 resulting from the CO interaction, since,accord ing to the literature12 as well as our F TI R results on pureBaC 03, he absorption bands at 1420, 1060,855, and 695 cm-lare due to BaC0 3with the aragonite structure and can beassignedto u3(E), ul(Al ), uz(A2), and u&) for the C 0 3group. In thelower wavenumber region ( 35 06 50 cm-I) the infrared bands at500 cm-1, which ar e found in spectr um a but not in spectrum b,are due to the Cu (II )-O stretching vibrations, while the bandsat 560, 420, an d 400 cm-1, which are observed in both spectr uma and spectrum b, are due to yttrium oxides, on the basis ofexperim ents on pure Cu O and Y2O3. The sim ilarity betweenspectra a and b in the bands due to yttrium oxides appears toindicate that, in addition to barium carbonate, yttrium oxidemay also be a product of the C O interact ion of YBaz Cu3 0x.Onthe other hand, the fact that no evident signals are observed inspectrum b at 500 and 620 cm-1, where the Cu-0 stretchingvibratio n is normally observed with the presence of pure CuO(500 cm-1) or Cu 20 (6 20 cm-I), seems to mean that copper oxidesmay not be the pro ducts of the C O interac tion of YBazCu30,.This conclusion is in a good agreement with the XPS esults,which show that most of the Cu atoms in the crystal are reducedto Cu(0) after the CO interaction (vide infra).It is possible to evalua te the activation energy of the rea ctionfrom the TG data. In reaction 2 the concentration of CO canberegarded as a constant (e.g. 1)sothat the rea ction rate appears(12) Nakamoto, K. In Infrored and Romon Spectro of Inorganic ondCoordination Compounds, 4t h 4.;ohn Wiley & Sons: New York,1986; 8ec also references therein.(13) Coats, A. W.; Rcdfcm, J. P. Nature 1964,201, 6 8 .

-8.81

8/3/2019 BaC03 FTIR

4/6

3096 Inorganic Chemistry, Vol. 32, No. 14, 1993 Lin et al.

T I

: , .., **. '3-926.2 938.2 950.2Binding energy (eV)

Figure 6. Comparison of the Cu 28312 X-ray photoelectron spectraobtained from (a) the YBazCu30, sample before the CO interaction, (b )the YBazCu30, sample after the CO interaction, (c) the YBazCu30,sample before the CO interaction, and (d) the YBazCu30, sample afterthe CO interaction. The spectra were recorded by using a Mg Ka X-raysource (1253.6 eV, 120 W) at a constant analyzer pass energy of 50 eV.The charging shifts of the samples have been corrected by using thebinding energyof the contaminant C 1s= 284.6 eV below the Fermi levelas reference. The peaksof broken lines were produced by a peak-synthesisprocedure that fixed the peak positions and widths and varied the heightsto achieve a best fit of the experimental (dotted) spec tra. i and ii denotethe copper species with different oxidation states.

Binding energy (eV)Figure 7. Comparison of the Cu LVV Auger spectra obtained from (a)the YBazCu30, sample before the CO interaction, (b) the YBazCu30,sample after the CO interaction, (c) the YBazCu30, sample before theCO interaction, and (d) the YBazCu3O, sample after the CO interaction.The spectra were obtained under the same conditions as for Figure 6.Broken lines are shown in spectra b and d for convenient comparison ofspectra a with b and spectra c with d.i) still retain the high oxidation state. In order to determinewhether t he reduction product (species ii) is Cu(0 ) or Cu( I), itis necessary to measure t he A uger parameter,14 since Cu(1) an dCu(0) are known to have the sa me binding energy for their Cu

Y3d. - . .....*. a. * . ,

S I

Binding energy (eV)Figure 8. Comparison of the Y 3d spectra obtained from (a ) the YBaz-CU ~ O ,ample before the CO interaction and (b) the YBazCu30, sampleafter the CO interaction under the same conditions as for Figure 6. Thepeaks of the broken lines wereobtained for the 3 d ~ pnd 3d3p componentsby the same peak-synthesis procedure as in Figure 6.2p peaks but different values for their Auger C u LVV peaks. Bycomparison of Auger spectra a an d b in Figure 7, the interactionof CO with YBazC u30x is found to result in the sh ift of some ofthe A uger electron intensity to lower binding energy by abo ut 3eV. This gives an Auger param eter of 1851 eV for the CO -reduced copper species (species ii in Fig ure 6), indicating t hatmost of the copper atoms in th e YB azCu30 x ample are reducedto Cu(0) after the CO interaction. Spectra a and b in Figure 8represent the Y 3d peaks obtained for the YBa2Cu3OX amplesbefore and after th e interaction with CO . Obviously the Y 3dpeaks are shifted to lower binding energy by 1.2 eV due to th eC O nteraction, indicating that th e yttrium atoms in YBa2Cu3OXhave been reduced by CO. Littl e change is observed in the Ba3d spectrum for th e YBa2Cu30x sample upon th e C O nteraction.

Spectrumc in Figure 6 is obtained from the Y Ba2Cu30y amplebefore the interaction with CO . A comparison of spectra c an da shows that when YBa2Cu30x s heated in Nz,bout one-thirdof the Cu 2~312ntensity shifts to lower binding energy by 1.6 eV.On th e other hand, in comparison with spectrum a in F igure 7,the Auge r spectrum of the YBazCu30, sample (spectrum c inFigure 7) is found to peak in the sa me energy region, but withless intensity in the higher binding energy region. These arecharacteristic of Cu(I).15 This indicates that heatingYB azCu30,in Nz a t elevated temperat ures would result in th e reduction ofsome Cu(I1) to Cu(1)becauseof th e removal of oxygen as reportedin the literature.16 Spectrum d in Figu re6 is the C u 2~312 pectrumrecorded for the YBazCu30, samp le after th e interaction withCO. In comparison with spectrum c in Figure 6, m ore intensityis shifted to lower binding energy in spectrum d, showing tha t th eC O interaction results in the furthe r reduction of the Cu(I1)(14) Wagner, C.D.; Riggs,W. M.; Davis, L. E.; Moulder,J. F.; Mullenberg,G. E. In Handbook of X-ray Photoelectron Spectroscopy; Perkin-Elmer: Eden Prairie, MN, 1979.(15) Ramaker, D. E.; Tuener, N. H.; Murday, J. S.; Toth, L. E.; Osofsky,M.; Hutson, F. L. Phys. Rev.B 1987, 36, 5672.(16) Goodenough, J. B.; Manthiram, A. In Chemistry of Oxide Supercon-ducrors;Rao,C .N. R.,Ed.; lackwell Sc ientificPublications: London,1988; p 101.

8/3/2019 BaC03 FTIR

5/6

The Interaction between C O and YB a2Cu30x Inorganic Chemistry, Vol. 32, No. 14 , 1993 3097toward yttrium . It is well establis hed that the oxygen atoms inthe Cu02 planes are strongly bound and cannot be removedwitho ut phase decomposition, while the oxygen atoms in the C uOcha ins are loosely bound an d can be easily removed.16.17 Th erefo reheating YBa2C u30, in N2 above 700K easily removes the oxygenatom s in the C uO planes, resulting in the w eight loss (as observedfrom curve 2 in Figure 1)and th e formation of the oxygen-deficientsamp le, YBazCu30,. In the presence of C O, the removal of theoxygen atom s becomes even easier and can occu r at much lowertemp eratures (400-800 K) (as shown in curve 1, Figu re 1). Thislow-te mpera ture nteraction esults in the formation and evolutionof C0 2 with the creatio n of new oxygen vacancies in the C uOplane (as schematica lly llustrated in block 2 in Figure 9). In theYBa2Cu3OY amp le most of the loosely bound oxygen atoms havebeen removed upon heatin g so that not evident reaction with COwas observed at tem peratures below 800 K. Note that the DTA/TG results observed for the C O interaction with the YBazCu30,sample at 400 -800 K arevery similar to those for CuO. It appearsthat the reaction CO + O(1attice) = C0 2 occurring on YBa2-Cu 30 , may follow a m echanism somewhat similar to that whichhas been suggested for CO ads orption on transition meta l oxidessuch as MnO2, CuO, and N i0.11J8J9 According to this mech-anism , the gene ral adsorbed state of CO on an oxide surface isthe surfac e carbonate, which m ay then decom pose or react withCO, forming C02.

Previous studies using EXA FS, MGssbauer spectroscopy, andother structure-sensitivemethods have shown that the interacti onof halogen with Y Ba2C u30, proceeds by in tercalation of halogeninto the vac ant anion positions in the C uO layers of the YBa2-Cu 30 , structure.20 In the present study the removal of the oxygenatoms from the unit cell of YBa2Cu30, by C O at the temperaturesof 400-800 K appears to result in breaking certain interionicbonds, softening the structure, and creating m ore oxygen vacanciesin the Cu O chains. These oxygen vacancies may serve as theactive centers for the CO intercalation. The intercalated YBa2-Cu3O, appears to be a metastable compound and can undergochem ical ransform ation at elevated tempe ratures, leading to thedegradation of the matrix and to the formation of morethermodynamically stable products. Note th at the B aO layerfalls between the C u02 and the CuO planes so that the Ba atomsmay share its coordinated oxygen atoms as well as oxygenvacancies with the copper atom s in the C uO planes. In otherwords the Ba atoms are in the neare st positions to the interca latedCO (

8/3/2019 BaC03 FTIR

6/6

3098 Inorganic Chemistry, Vol. 32, No. 14 , 1993The above-postulated mechanism has shown the importantrole that the perovskite structure may play in CO uptake byY-Ba cuprates: it can offer the mobility of oxygen ions, favoringthe oxidation of adsorbed molecules with the creation of activecenters for further CO uptake; it can offer structure stabilizerssuch as Ba(I1) ions, which are well distributed a t coordinationsites nearest to the active centers, aiding the ab sorption of COthrough the formation of carbonate. Surface carbonate has beensuggested o be an important intermed iate n the oxidation reactionof C O over transition oxides.~1J8J9 Therefore the earlier

observations of the hig her catalytic activity for the reaction CO+ N O = l /2N2+ CO zon the perovskite-type barium cupratesthan on CuO and other non-perovskite Cu-containing catalystsmay become understandable.3 In addition, the perovskite-typestructure of yttrium-barium cuprates has shown an ability toretain some copper atoms in high oxidation states even undervigorous reduction conditions (in the presence of CO at hightemperatures). This may be significantly important for the

Lin et al.methanol synthesis from syngas over copper-based catalysts sincecopper atoms in higher oxidation states m ay play a different rolein CO activation comp ared to the metallic copper atoms.zl.22Furthermore, from the S EM X-ray m ap studies we also observedthe stabilization of active metals including Cu, a, and Y n highdispersion against the CO interaction at elevated temperatures.All of these are valu able properties for catalysis and thereforeoffer the perovskite-type yttrium-barium cuprates the potentialto serve as good catalysts for CO reactions.

Acknowledgment. We thank Mr. S.K. Tung for his valuableassistance in the S EM study and Mrs. S.W. Phua and T. J. Pangfor their assistance in the XRD measurements.(21 ) Didziulis, S. V.;Butcher, K. .; ohen, S.L.; olomon, E. I. J . Am .(22) Lin, L.; ones,P.; uckert, J.; Solomon,E. I . J . Am . Chem.Soc.1991,Chem.Soc. 1989, 111,7110.113,8312.