Study of the formation of the high-temperature superconducting phase

The formation of the 120 K superconducting phase in the BiSrCaCu20y system has been .... T;s of 80 and 120 K. The structure responsible for the transi...
0 downloads 0 Views 1MB Size
3864

J . Phys. Chem. 1990, 94, 3864-3868

Study of the Formation of the High-Temperature Superconducting Phase in BiSrCaCu,O, by Microwave Absorption Stephane Cuvier, Micky Puri, John Bear, and Larry Kevan* Department of Chemistry and the Texas Center for Superconductioity, University of Houston, Houston, Texas 77204-5641 (Received: August I , 1989; In Final Form: October 16, 1989)

The formation of the 120 K superconducting phase in the BiSrCaCu20ysystem has been investigated by microwave absorption. This material is characterized by an intense low-field microwave absorption (LFMA) below the superconducting transition temperature Tc and a g = 2 electron spin resonance (ESR) signal above T,. LFMA is shown to be particularly effective and discriminatory for following the formation of the high-T, phases in this system. The purity of these superconducting phases is also reflected by the intensity of the g = 2 signal. Various stages of sample preparation are studied to follow the formation of the 120 K phase. LFMA and ESR results are also compared and contrasted with resistance measurements.

Introduction Several studies'-24 of the newly discovered high-temperature superconductors by microwave absorption with an electron spin resonance (ESR)spectrometer have been reported since the work of Blazey et al.' For ceramic oxide superconductors, an intense low-field microwave absorption (LFMA) has been observed. It is now well-established that this low-field microwave absorption is diagnostic of the superconducting phase.I6 This signal is as-

( I ) Blazey, K. W.; Muller, K. A.; Bednorz, J. G.; Berlinger, W.; Amoretti, G.; Buluggiu, E.; Vera, A.; Matacotta, F. C. Phys. Reu. B 1987, 36, 7241-43. (2) Stankowski, J; Kahol, P. K.; Dalal, N . S.; Modera, J. S. Phys. Reo. B 1987, 36, 7 126-28. (3) Korczak, W.; Korczak, S. 2.; Mazurek, P.; WysokiRski, K. 1. Solid State Commun. 1988, 66, 911-973. (4) Schwartz, R. N.; Pastor, A. C.; Pastor, R. C.; Kirby, K. W.; Rytz, D. Phys. Reo. B 1987, 36, 8858-59. ( 5 ) Bartucci, R.; Colavita, E.; Sportelli, L.; Balestrino, G.; Barbanera, S. Phys. Rev E 1988, 37, 2313-16. (6) Kohma, K.; Ohbayashi, K.; Udagawa, M.; Hihara, T. Jpn. J . Appl. Phys. 1987, 26>L766-L767. (7) Yu, J.; Lii, K. H. Solid State Commun. 1988, 65, 1379-83. (8) Mehran, F.; Barnes, S. E.; McGuire, T . R.; Gallagher, W. J.; Sandstrom, R . L.; Dinger, T. R.; Chance, D. A. Phys. Reu B 1987,36, 740-742. (9) Vier, D. C.; Oseroff, S. B.; Salling, C. T.; Smyth, J. F.; Schultz, S.; Dalichaouch, Y.; Lee, B. W.; Maple, M. B.; Fisk, 2.;Thompson. J. D. Phys. Rec B 1987, 36, 8888-91. (IO) Shrivastava, K. N . J . Phys. C 1987, 20, L789-L796. ( 1 I ) Khachaturyan, K.; Weber, E. R.; Tejedor, P.; Stacy, A. M.; Portis, A. M. Phys. Rev B 1987, 36, 8309-14. (12) Bist, H . D.: Khalbe, P. K.; Shahabuddin, Md.; Chand, P.; Narlikar, A. V.; Jayaraman, B.; Agrawal, S. K. Solid State Commun. 1988, 66, 899-902. ( I 3) McKinnon, W. R.; Morton, J. R.: Pleizier, G . Solid State Commun. 1988,66, 1093-95. (14) Bohandy, J . ; Adrian, F. J.; Kim, B. F.; Moorjani, K.; Shull, R. D.; Swartzendruber, J.; Bennett, L. H.; Wallace, J. S. J . Supercond. 1988, I , 191. ( 1 5) Sastry, M. D.; Kadam, R . M.; Babu, Y.; Dalvi, A. G . 1.; Gopalakrishnan, I . K.; Sastry, P. V. P. S. S.; Phatak, G. M.; Iyer, R. M . J . Phys. C 1988, 21, L607-L614. (16) Kevan. L.; Bear, J.; Puri, M.; Pan, Z.: Yao, C. L. ACS Symp. Ser. 1988, No. 58, 1143. ( 1 7) Puri, M.; Pan, 2.; Marrelli, S.; Bear, J.; Kevan, L. Unpublished work. (18) Owens, F. J.; Iqbal, 2.Solid Stare Commun. 1988, 68, 523-526. (19) Shaltiel, D.; Bill, H.; Decroux. M.; Hagemann, H.; Junod, A. Peter, M.: Ravi Sekhar, Y.; Triscone, G.; Walker, E.; Yan, Y. F.; Zhao. Z. X. Physica C 1989, 157, 240-246. (20) Adrian, F. J.; Bohandy, J.; Kim, B. F.; Moorjani, K.; Wallace, J. S.; Shull, R. D.; Swartzendruber. L. J.: Bennett, L. H. Physica C 1988, 156, 184-1 88. (21) Tyagi. S.; Barsoum, M.; Rao, K. V . ; Skumryev, V.; Yu. Z.; Costa, J. L. Physica C 1988, 156, 73-78. (22) Bombik, A.; Korczak, S. 2.; Korczak, W.; Mazurek, P.; Pacyna, A. W.: Subotowicz, M.; Wysokinski, K. I. Physica C 1989, 157, 251-256. (23) Pakulis, E. J.; Chandrashekhar, G . V. Phys. Rer B 1988, 38, 11974-1 1976. (24) Pakulis, E. J.; Chandrashekhar. G . V. Phys. Reo. B 1988, 39, 808-8 IO.

0022-3654/90/2094-3864$02.50/0

TABLE I: Synthesis Conditions for Samples 1-8 and Sample A" heat treatment at sampletype IO73 K 1093K 1143 K 1153K 1158 K 1

2 3 4 5 6 7 8 A

12 12 12 12 12 12 12 12 12

6 6 6 6 6 6 6 6

12 12 12 12 12 12

8 8 8 8 8

7 20 27 42 23

'All values are given in hours

sociated with magnetic flux trapping and may allow an estimation of the amount of superconducting phase present. A characteristic ESR signal is observed at higher field near g 2 for several copper-based superconductor^.^-^^ This g = 2 signal is associated with dipolarly broadened Cu2+species present in impurity nonsuperconducting phases. The intensity of this line is considered to inversely reflect the purity of the sample and can be used to monitor the transition temperature of the Thus, microwave absorption is a useful way of characterizing these high superconducting transition temperature (T,) materials. Recently, several groups have published interesting studies on the Bi-Sr-Ca-Cu material.3~'S~274Some controversies regarding

-

(25) Hazen, R. M.; Prewitt, C. T.; Angel, F. J.; Ross, N. L.; Finger, L. W.; Hadidiacos, C. G.; Veblen, D. R.; heaney, P. J.; Hor, P. H.; Meng, R. L.; Sun, Y. Y.; Wang, Y. Q.; Xue, Y. Y.; Huang, 2.J.; Gao, L.; Bechtold, J.; Chu. C. W. Phys. Reo. Lett. 1988, 60, 1174-77. (26) Torrance, J. B.; Tokura, Y.;La Placa, S. J.; Huang, T . C.; Savoy, R. J.; Nazzal, A. I . Solid State Commun. 1988, 66, 703-706. (27) Chu, C. W.; Rechtold, J.; Gao, L.; Hor, P. H.; Huang, Z. J.; Meng, R. L.; Sun, Y. Y.; Wang, Y. Q.; Xue, Y. Y. Phys. Reo. Lett. 1988, 60, 941-943. (28) Meng, R. L.; Hor, P. H.; Sun, Y. Y.; Huang, Z. J.; Gao, L.; Xue, Y. Y.; Wang, Y. Q.:Bechtold, J.; Chu, C. W.; Hazen, R. M.; Prewitt, C. T.; Angel, R. J.; Ross, N . L.; Finger, L. W.; Hadidiacos, C. G. Phys. Rev. Leu. 1988, 60. 1 174. (29) Xiao, G . ;Cieplak, M. 2.;Chien, C. L. Phys. Reo. B 1988, 38, 11824. (30) Onoda. M.;Yamamoto, A.; Takayama-Muromachi, E.; Takekawa. S. Jpn. J . Appl. Phys. 1988, 27, L833-L836. (31) Uehara, M.; Asada, Y.; Maeda, H.; Ogawa. K. Jpn. J . Appl. Phys. 1988, 27, L66S-L667. (32) Yang, B. C.; Li, H . C.; Xi, X. X.; Dietrich, M.; Linker, G.; Geerk, J. Z . Phys. E : Condens. Matter 1988, 70, 275-277. (33) Matsumoto, T.; Aoki, H.; Matsushita, A,; Uehara, M.; Mori, N.; Takahashi, H.; Murayama, C.; Maeda, H. Jpn. J . Appl. Phys. 1988, 27, L600-602. (34) Tarascon, J. M.; Le Page, Y.; Barboux, P.; Bagley, B. G.; Greene, L. H.; McKinnon, W. R.; Hull, G. W.; Giroud, M.; Hwang, D.M. Phys. Reo. B 1988, 37, 9382-89. (35) Kajitani, T.; Kusaba, K.; Kikuchi, M.; Kobayashi, N.; Syono, Y.; Williams, T. B.; Hirabayashi, M. Jpn. J . Appl. Phys. 1988, 27, L587-L590. (36) Ohta, M.: Takahashi, K.; Kosuge, M. Jpn. J . Appl. Phys. 1988, 27, L567-L568.

0 1990 American Chemical Society

Formation of the Superconducting Phase in BiSrCaCu20y the best method of synthesis and the corresponding properties of the resulting superconducting oxide persist. Stoichiometric variations, different sintering times, and temperatures influence the ratio of the two superconducting phases that are found with T;s of 80 and 120 K. The structure responsible for the transition close to 80 K has been described as an orthorhombic unit cell with a = 5.4 A. b = 5.4 A, and c = 30.6 A25*30,41 or as a tetragonal unit cell with a = 3.8 A and c = 30.6 A.26,34*45 The composition of this phase has been reported to be close to Bi2(Sr,Ca)3Cuz08.34 It has been established that the composition of the 120 K phase is probably close to Bi2SrzCazCu3010.47~48 But up to now this phase has never been isolated and has been only observed embedded in the 80 K phase. Therefore, no chemical analysis of the 120 K phase has been performed so far. The BizSr2CazCu3010 composition has been derived from the difference between the c lattice spacing of the two phases (30.6 A for the 80 K phase and 37 for the higher T, phase), suggesting a greater stacking number in the c direction.40 In addition, a pure isomorphic T12Ba2CazCu30,high- T, phase has been recently i ~ o l a t e d . ~ ' Here, our focus is to investigate and characterize the BiSrCaCu20ysystem at different stages of sample preparation, primarily by microwave absorption. The growth of the 80 and 120 K phases is monitored with LFMA, ESR, and resistivity methods. I t is found that LFMA is particularly discriminatory for the formation of the two high-T, phases.

Experimental Section The Bi-Sr-Ca-Cu samples, with a nominal composition BiSrCaCu20,,, were prepared by mixing appropriate amounts of Biz03, CaC03, SrC03, and CuO. These powders were ground with an agate mortar and pestle for 35 min. The precursor powder was heated at 1073 K for 1 2 h (sample 1) followed by heat treatment at 1093 K for 6 h in a porcelain crucible (sample 2). A compact black material was obtained which was reground and pressed into pellets prior to a reheat treatment at 1143 K for 12 h (sample 3) and then at 1153 K for 8 h (sample 4). The final step in the synthesis involved the heating of sample 4 at 1158 K for 7 h (sample 5), for 20 h (sample 6), for 27 h (sample 7), and for 42 h (sample 8). All samples were heated and cooled at 80 K/h in air. The synthesis conditions for samples 1-8 are summarized in Table I . Also, sample A is listed which is prepared with a similar composition but with a direct heat treatment at 1158 K after two steps at 1073 and 1093 K. Each sample was characterized by X-ray diffraction on a Philips Electronics Model 2500 powder diffractometer using Cu KCY irradiation and Ni filters. Microwave absorption measurements were performed on a Bruker ER300 ESR spectrometer using a modulation amplitude of 10 G and a microwave power of 2 mW. The derivative of the imaginary part of the magnetic susceptibility with respect to (37) Shaheen, S. A.; Liang, G.; Jisrawi, N.; Croft, M. Solid Stare Commun. 1988, 66, 947-95 I , (38) Inoue. 0.;Adachi, S.; Kawashima, S. JRn. J . ADPI. .. Phys. . 1988, 27, L347-L349. (39) Togano, K.; Kumakura, H.; Maeda, H.; Takahashi, K.; Nakao, M. Jpn. J . Appl. Phys. 1988, 27. L323-L324. (40) Grader, G. S.; Gyorgy, E. M.; Gallagher, P. K.; OBryan, H. M.; Johnson, D. W., Jr.; Sunshine, S.; Zahurak, S. M.; Jin, S.; Sherwood, R. C. Phys. Reo. B 1988, 38, 757-760. (41) Syono, Y.; Hiraga, K.; Kobayashi, N.; Kikushi, M.; Kusaba, K.; Kajitani, T.; Shindo, D.; Hosoya, S.; Tokiwa, A,; Terada, S.; Muto, Y . Jpn. J . ADDI.Phvs. 1988. 27. L 5 6 9 4 5 7 2 . ( i i )K i j h a , N.;'Endo, H.; Tsuchiya, J.; Sumiyama, A,; Mizuno, M.; Oguri, Y. J . Appl. Phys. 1988, 27, L821. (43) Tanaka, Y.; Fukutomi, M.; Asano, T.; Maeda, H. Jpn. J . Appl. Phys. 1988, 27, L548-L549. (44) Nobumasa, H.; Shimizu, K.; Kitano, Y.; Kawai, T. Jpn. J . Appl. Phys. 1988, 27, L846-L848. (45) Mazaki, H.; Ishida, T.; Sakuma, T . J . Appl. Phys. 1988, 27, L811.. . L813. (46) Kataria, N. D.; Tomar, V. S.; Gupta, A. K.; Kumar, M. J . Phys. C 1988, 21, L523-L27. (47) Tallon, J. L.; Buckey, R. G.; Gilberd, P. W.; Presland, M. R.; Brown, 1. W. M.; Bowden, M. E.; Christian, L. A.; Goguel, R. Nature 1988, 333, 153. (48) Torardi, C. C.; Subramanian, M. A,; Calabrese, J. C.; Morissey, K. J.; Gopalakrishnan, J.; Askew, T . R.; Flippen, R. B.; Chowdhry, U.; Sleight, A. W. Science 1988, 240, 63 I .

The Journal of Physical Chemistry, Vol. 94, No. 9, 1990 3865

I

0

250

H (GAUSS) Typical low-field microwave absorption for powdered BiSrCaCu20,, cooled in a 3-kC field a t 2-mW microwave power and 10-G modulation amplitude a t 95 K for sample 5.

Figure 1.

!O TEMPERATURE (K) Figure 2. Intensity of the maxima of the low-field microwave absorption versus temperature for 5 mg of samples 1-3.

magnetic field (dx"/dH) was recorded as a function of the applied magnetic field. The temperature in the cryostat was controlled with an Oxford Instruments ESR 900 helium flow system. In addition to the internal thermocouple mounted in the Oxford cryostat, an external thermocouple was used to monitor the temperature at the sample position. This thermocouple was taped in contact with the sample tube. Approximately 5 mg of each reground pellet was placed into 2-mm4.d. by 3-mm-0.d. Suprasil quartz tubes, evacuated and sealed for ESR measurements. The resistivity measurements were performed by a standard four-probe method. Copper wires were connected to four indium contacts with indium solder. A lock-in amplifier was used to generate an ac current with a frequency of 17 Hz and to measure the resulting voltage through the sample. A Dupont 9900 thermal analyzer system was used to carry out thermogravimetric (TGA) and differential thermal analysis (DTA). Samples of 20-30 mg were analyzed in a atmosphere of air with a heating rate of 10 K/min. A scanning electron microscope (SEM) from Cambridge Instruments (Model 250 MK3) was used to characterize the samples and examine crystal formation as well as grain size distribution.

Results X-ray Diffraction Patterns. The X-ray patterns obtained for 20 = 15-60' for samples 4-8 are characteristic of the a ~ e . Nevertheless, ~ ~ ~the ~ Bi 2Sr2CaCu Z0y low- TC P h presence of an additional diffraction peak at 20 = 32' suggests the multiphase character of the samples. This diffraction peak has been reported previously and attributed to Ca2C~0333v42 or S r C a 2 C ~ 7 0 , 4formed 3 during the growth of the high-T, phase. This peak increases from samples 4 to 6, where it is maximum, and then decreases for samples 7 and 8. Thus, this peak correlates with the development of the high-T, phase as monitored by LFMA and resistivity (vide infra). Low-Field Microwave Absorption ( L F M A ) . An intense lowfield microwave absorption is observed for all samples. After cooling the sample in a 3-kG field, the magnetic field is switched

~

~

3866 The Journal of Physical Chemistry, Vol. 94, No. 9, 1990

Cuvier et al.

)O TEMPERATURE (K)

TEMPERATURE (K) Figure 3. Intensity of the maxima of the low-field microwave absorption versus temperature for 5 mg of samples 4-8.

Figure 5. Resistance versus temperature profiles for samples 4-8. The ordinate scale is shown for sample 8. The zero of the other samples is offset vertically, but the scale is the same. a

5 0051 E

\

Y I

5 -022

-0 26

W

Ir 1

a z P m

40 K

TEMPERATURE(K)

20 K

Figure 6. DTA for sample 7 carried out with a I O K/min ramp from (a) I

2000

I

I

3000 H (GAUSS)

I

4000

Figure 4. Typical ESR spectra in the g = 2 region at 9.4 GHz from 20 to 120 K for sample 5 .

373 to 1273 K and (b) enlargement of the temperature scale between 1 1 13 and 1233 K. The arrows indicate the positions of the two endothermic peaks.

intensity of this signal can be used to determine the transition temperature of the high-temperature superconducting phase. to 0 G and the microwave absorption is recorded as a function 120 K, this ESR When the temperature is lowered below T, of the magnetic field for temperatures from 4 to 125 K. line disappears due to the Meissner effect exclusion of the magnetic Below the second transition temperature ( T , 120 K), the field. The temperature associated with the disappearance of the microwave absorption shows a maximum (Figure 1) which has g 2 ESR signal is similar to the temperature where the LFMA been described as the critical field where flux slippage 0ccurs.I disappears (Figure 3). The position of this maximum is slightly shifted toward lower Resistance Measurements. Samples 4-8 show a two-step drop magnetic fields as the temperature is raised from 4 to 120 K. Also, in their resistance around T = 120 and 90 K (Figure 5 ) , but zero as the temperature increases the signal maximum becomes narresistance is only achieved at 70 K for the best sample (sample rower and higher. 6) or below 70 K for the other samples. Samples 1 and 2 exhibit When the intensity of (dx"/dH), is plotted versus temperature only a transition temperature close to 85 K and are not shown (Figures 2 and 3), two maxima are observed similar to those in this figure. These data are in agreement with the microwave obtained by Adrian et aLzo This plot also shows the growth of absorption results which suggest that the amount of high-T, phase the high-T, phase from sample 1 to 8. A prolonged heat treatment grows from sample 3 to sample 6, reaches a maximum, and then at 1158 K seems to increase the amount of the high-T, phase as decreases slightly from samples 6 to 8. Furthermore, T, observed well as the amount of the low-T, phase. by resistance measurements is consistent with the temperature ESR. For samples 2-8, an ESR signal near 3000 G with typical where the low-field absorption ceases and where the asymmetric g values of glI = 2.27 and g, = 2.04 is observed above the second g = 2 signal appears as reported by Owens et aI.'* transition temperature of I20 K (Figure 4). This asymmetric DTA, TGA, and SEM. Differential thermal analysis performed line is attributed to dipolarly broadened cupric ion and has been on samples 1-8 does not provide any information concerning the and observed in YBa2Cu30,, EuBaCu30xr3,5-8~10,12,i6~17 development of two superconducting phases. For all samples a Bi2Sr2CaC~20,.18J9 Below the 120 K transition temperature, this superposition of sharp ( 1 170 K) and broad (1 210 K) endothermic asymmetric ESR line disappears and prominent oscillations appear peaks is obtained without any significant temperature shift or which are associated with periodic temperature fluctuations in change of peak intensity (Figure 6). our gas flow system as demonstrated by correlation with an inThermogravimetric analysis in Figure 7 of a sample in air shows dependent temperature recording. a weight loss occurring in two steps at about 1123 and 1198 K. The behavior of the BiCaSrCu20y system is similar to that of At a I O K/min cooling rate the weight loss is largely recovered. YBa2Cu30J regarding the g 2 ESR signal above T,. The apparently by absorption of oxygen from the air.

-

-

-

-

-

Formation of the Superconducting Phase in BiSrCaCu20y

97*5* 97.0 773

073

The Journal of Physical Chemistry. Vol. 94, No. 9. 1990 3867

973 1073 1173 1273

TEMPERATURE (K) Figure 7. TGA for sample 7 carried out with a 10 K/min heating and cooling rate from 773 to 1298 K.

Figure 8. SEU pictures for samples 2 (top) and 8 (bottom).

The scanning electron microscope micrographs in Figure 8 show a slablike structure for this material. The average sizes of the slabs appear to increase from samples 1 to 8. These pictures show also that the material becomes more compact with longer sintering time.

Discussion Several samples having a starting composition of BiSrCaCu20y were found to melt partially and react with the walls of the crucible on heating directly to 1 153 K. A two-step heat treatment seems to be essential for the synthesis of a material with a significant high- T, phasei4.26*36*39.4244 which develops at tempera tures close It is interesting to note that sample 1 to the melting which was heated for only 12 h at 1073 K gave a transition temperature (from resistance measurement) around 85 K and a microwavc rcsponsc at both high and low magnetic fields. However, thc high-T, phase incrcases at higher sintering temperaturcs ncar thc mclting point of the material and is very

1

~

1

2000

3000

4000

~

1

2000

3000 4000 H (GAUSS) H (GAUSS) Figure 9. Powder ESR spectra in the g = 2 region at 9.4 GHz (a) for a small piece of sample 7 cut from the pellet and reground (sample 7’) and (b) for the reground and mixed remaining pellet (sample 7).

sensitive to the sintering condition^.^^ It has been reported that heat treatment at 1 163 K results in the deterioration of the high-T’ pha~e.~~.~’ The appearance of a byproduct during the formation of the high-T, phase is supported by sample A showing a relatively intense diffraction peak at 26 = 32’ but which contains only a very small amount of the high-T, phase according to microwave and resistance measurements. Several groups have reported the formation of Ca2Cu03 during the synthesis of the high-T, p i a ~ e . ~Furthermore, ~.~~ Tarascon et al. have observed that the slow decomposition of the superconducting oxide gives a bronze-colored product on the back of pelletized samples which has a composition rich in calcium and copper.34 This type of behavior was obtained for our samples rich in high-T, phase showing some small signs of partial melting. The LFMA observed for this type of superconductor is a consequence of its morphology. According to Blazey et aI.,I the bulk material consists of grains of superconducting material in a matrix of nonsuperconducting material. These grains couple to form superconductingclusters. When the critical field is attained (i.e., interpreted as the position of the maximum of the LFMA), these clusters begin to decouple resulting in flux slippage. The two maxima obtained in the (dX’’/dH)maxplot as a function of the temperature (Figure 3) may also be explained in terms of thermally induced grain decoupling. I t is also possible through these microwave measurements to explain the development of the two superconducting phases. Sample 1 shows the presence of a low-T, phase (-80 K) and possibly of another phase with a transition temperature between 20 and 30 K. This second phase appears clearly for sample 2 and could be the calcium deficient compound Bi2Sr2Cu207described earlier.49 This phase becomes much less pronounced for higher sintering temperatures. The appearance of the high-T, phase (- 120 K ) in sample 3 is observed with a significant shift of the lower transition temperature to 90-95 K. Therefore, the ideal sintering temperature in order to obtain the low-T, phase with a transition temperature close to 90 K seems to be between 1093 and 1 143 K. The LFMA associated with the 120 K phase appears with a decrease of the intensity of the LFMA attributed to the low-T, phase. This behavior is also encountered when the sintering temperature is increased from 1 153 to 1 158 K (samples 4 and 5 ) . Samples 5-8 show clearly that the intensity of the LFMA due to the high-T, phase grows quicker than that of the low-T, phase (samples 5 and 6) and then decreases slowly as the contribution of the low-T, phase to the LFMA increases (samples 6-8). These observations can be explained by conversion of the 90 K phase to the 120 K phase followed by an equilibrium between these two phases. The slow decrease of the intensity for the 120 K phase (sample 6-8) may be due to the partial decomposition of this phase. The origin of the asymmetric ESR signal observed around 3000 G is assigned to a byproduct formed during the synthesis of the (49) Michcl, C.; Hervieu, M.;Borel, M. M.; Grandin, A.; Deslandes. F.; Provost, J.; Ravcau. R. 2.Phys. B Condens. Mailer 1987, 68. 421.

3868

The Journal of Physical Chemistry, Vol. 94, No. 9, 1990

Cuvier et al.

0.I

TEMPERATURE (K) Figure 10. Intensity of the low-field absorption versus temperature for samples 7 and 7‘; see Figure 9 caption for sample description.

superconductor. This interpretation has been suggested by several g r o ~ p s ~and ~ ~is~consistent ’ ~ ~ ~ *with the following results obtained for sample 7. Before being reground and mixed prior to ESR measurements, a small part of sample 7 was cut from the pellet and ground separately (sample 7’). Figure 9 shows dramatic differences in the g = 2 signal intensity between these two samples. Sample 7 exhibits a very small g = 2 signal which disappears below 120 K (Figure 9b), but for sample 7’, an observable g = 2 signal remains even below 90 K (Figure 9a). This result suggests that the sample is very inhomogeneous. In addition, the LFMA for these two samples shows a greater microwave absorption for sample 7 than for sample 7’ (Figure lo), indicating that sample 7 has more superconducting phase. So, the g = 2 signal is attributed to a nonsuperconducting impurity phase and does not arise from the 120 K phase as suggested by Shaltiel et aI.l9 The LFMA signal for the BiSrCaCuO material is similar to that obtained for other oxide superconductors. However, the LFMA intensity is weaker than that of YBa2C~307.15 A possible explanation given by Tyagi for this intensity difference is the stronger grain coupling in the Bi-Sr-Ca-Cu-0 ceramic2’ Blazey et al.’ suggested that the field B,,, at which (dX”/dH),,, occurs is inversely related to the area of superconducting clusters S consisting of coupled Josephson junctions between grains which maintain superconducting current loops below 7,by B,,,S = 4 = 2.07 X G cm2 where q5 is the magnetic flux quantum. For S = Rr2 this simple expression gives r -0.3 pm for B,,, -60 G at lower temperatures near 50 K and r -0.7 pm for B,,, -14 G at higher temperatures above 100 K. The DTA of these samples shows a superposition of two endothermic peaks. The peak at 1170 K probably corresponds to phase melting. However, the two peaks cannot be correlated with the melting of two different phases because DTA does not distinguish between the different sample preparations. The loss of weight observed in the thermogravimetric analysis is totally reversible and is attributed to oxygen (Figure 7). This observation is in contrast with the suggestion made by Grader et a1.,40who considered that bismuth evaporation could occur in this compound. Assuming the theoretical composition to be B~S~C~CU the~ oxygen O ~ , ~stoichiometry , is reduced to 4.9 over the first larger step beginning about 1120 K and to 4.8 over the second smaller step beginning about 1 195 K. This loss of oxygen is probably due to the reduction of Cu2+species. The shape of the small plates observed by SEM arises from the nature of the crystalline lattice. This oxide has been described as an oxygen-deficient layered perovskite type structure having

20

30 40 SINTERING TIME AT 1158K IN HOURS Figure 11. Estimation of the contribution from the high-T, phase to the LFMA signal (a) and to the resistance data (0) as a function of the sintering time at 1 1 58 K for samples 5-8. 0

IO

some quite long interlayer Bi-0-Bi distances (3.0 A). Thus, it is expected to observe cleavages of these layers to give this slablike arrangement.33 The size of the grains observed by S E M (-2-3 pm) is much larger than the calculated average size of the current loops from microwave absorption data (0.3-0.7 km). Bombik et al. have also observed such a difference between the current loop size and the actual grain size determined by SEMSZ2This indicates that the grains coupled by Josephson junctions to form an intergranular superconductor are significantly smaller than the macroscopic grains observed by SEM. The intensity of the LFMA as well as the positions of the maxima of the microwave absorption (Figures 2 and 3) is dependent upon the presence and quality of the superconducting phases in the sample. The comparison of these LFMA results from Figure 3 with resistance results from Figure 5 is shown in Figure 1 1 as a function of the sintering time at 1 158 K for samples 5-8. For the resistance data, the relative contribution of the high-T, phase was estimated from the value of the normalized resistance (RT/RT= 300 K) drop between 120 and 90 K in Figure 5 divided by the value at 120 K. For the LFMA data, the relative contribution of the high-T, phase was estimated from the value of its intensity maximum divided by the sum of the intensity maxima for the low- and high-T, phases of Figure 3. These two contributions of the high-T, phase are plotted versus the sintering time at 1 158 K in hours for samples 5-8. The general trends for the LFMA and the resistance are the same. The 120 K phase contribution increases from 0 to 20 h sintering time and then decreases and levels off. This last point supports the phase equilibrium idea suggested above. Conclusions The low-field microwave absorption is a useful technique to estimate the transition temperature of a superconducting material. The LFMA is semiquantitatively consistent with resistance data but seems more sensitive to small changes in sample quality than resistance. The g = 2 ESR signal is inversely related to the purity of the superconducting phases but is nonetheless a useful diagnostic to determine the transition temperature. Acknowledgment. This material is based upon work supported by the Texas Center for Superconductivity at the University of Houston under prime grant MDA 972-88-G-0002 from the Defense Advanced Research Projects Agency and the State of Texas. Registry No. BiSrCaCu,O,, 114901-61-0