Effect of hole filling by Co and hole doping by Ca on the superconductivity of GdBa2Cu3O7−δ

Effect of hole filling by Co and hole doping by Ca on the superconductivity of GdBa2Cu3O7−δ

International Journal of Inorganic Materials 3 (2001) 59–66 Effect of hole filling by Co and hole doping by Ca on the superconductivity of GdBa 2 Cu ...

421KB Sizes 0 Downloads 1 Views

International Journal of Inorganic Materials 3 (2001) 59–66

Effect of hole filling by Co and hole doping by Ca on the superconductivity of GdBa 2 Cu 3 O 72d a, a a a b D.G. Kuberkar *, Nikesh A. Shah , R.S. Thampi , S. Rayaprol , M.R. Gonal , Ram Prasad b , R.G. Kulkarni a a

Department of Physics, Saurashtra University, Rajkot 360 005, India b Metallurgy Division, BARC, Mumbai 400 085, India Accepted 8 June 2000

Abstract The structural and superconducting properties of Cu–Co and coupled Gd–Ca / Cu–Co substitution in GdBa 2 Cu 3 O 72d have been studied by X-ray diffraction for crystal structure determination, resistivity, a.c. susceptibility for T c determination and iodometry for oxygen content measurements. The oxides (Gd 12y Ca y )Ba 2 (Cu 12x Co x ) 3 O 72d have been synthesised for 0.0#x#0.15 and 0.0#y#0.45. The effect of increasing Co-concentration in GdBa 2 (Cu 12x Co x ) 3 O 72d for 0.0#x#0.15 is to lower the carrier concentration and decrease T c which is attributed to the hole filling by Co. This suppression in T c has been compensated for by appropriate hole doping with Ca, in (Gd 12y Ca y )Ba 2 (Cu 12x Co x ) 3 O 72d which shows that for a fixed Co-content T c increases dramatically as the Ca-content increases leading for instance to the oxides x 5 0.1, y 5 0.3; x 5 0.12, y 5 0.36 and x 5 0.15, y 5 0.45 with respective T c ’s of 83 K, 68 K and 63 K whereas the nonsubstituted phases x 5 0.1, 0.12, 0.15, y 5 0.0 don’t superconduct.  2001 Elsevier Science Ltd. All rights reserved. Keywords: A. superconductors; C. X-ray diffraction; D. crystal structure

1. Introduction The valency of the dopant plays an important role in monitoring the effective copper valence (the mobile carrier concentration) or the oxygen content of a YBa 2 Cu 3 O 72d (Y-123) superconductor, as superconductivity behaviour depends on it. A decreased oxygen content in the Y-123 has the same effect as M 31 (M5Al, Co, Fe) doping at the Cu(1) chain site. This results in a depression in T c and based on the observed structural changes it can be rationalised in terms of a decreased carrier concentration in the CuO 2 planes due to a transfer of the positive charge from the CuO 2 planes to the CuO chains [1–5]. To date, a systematic study is lacking that deals with Co-doped RBa 2 Cu 3 O 72d (R5rare earth ions having large intrinsic magnetic moments such as Gd 31 , Dy 31 , Ho 31 *Corresponding author. Tel.: 191-281-571-811; fax: 191-281-778-85. E-mail addresses: [email protected], [email protected] (D.G. Kuberkar).

and Er 31 ). In ‘under doped’ Y-123 (oxygen depleted or Co doped) charge carriers can be reintroduced into the CuO 2 planes by partial replacement of Y 31 by Ca 21 [6] leading to an increase of T c as expected from an increased carrier concentration. Therefore it is of interest to investigate whether substitution of Co for Cu in GdBa 2 (Cu 12x Co x ) 3 O 72d would lead to very similar changes in structure and T c to those of YBa 2 (Cu 12x Co x ) 3 O 72d [6] or the strong magnetic nature of Gd 31 compared to non-magnetic Y 31 plays different role. In order to understand the behaviour of Co with respect to the Gd-123 structure, we have undertaken the simultaneous substitution of Ca–Co for Gd–Cu trying to control the oxygen stoichiometry. In this paper we report X-ray diffraction, oxygen stoichiometry, a.c. susceptibility and resistivity measurements on the series of compounds having the stoichiometric compositions GdBa 2 (Cu 12x Co x ) 3 O 72d (0.0#x# 0.15), (Gd 12y Ca y )Ba 2 (Cu 12x Co x ) 3 O 72d for 0.0#x#0.15 and 0.0#y#0.45. The interrelationship between the super-

1466-6049 / 01 / $ – see front matter  2001 Elsevier Science Ltd. All rights reserved. PII: S1466-6049( 00 )00053-2

D.G. Kuberkar et al. / International Journal of Inorganic Materials 3 (2001) 59 – 66

60

conducting transition temperature, dopant valency and the variation of the oxygen content and orthorhombicity is discussed in the context of the carrier concentration.

2. Experimental A series of compounds having the compositions GdBa 2 (Cu 12x Co x ) 3 O 72d (x50.00, 0.02, 0.04, 0.06, 0.08, 0.10, 0.12, 0.15) and (Gd 12y Ca y )Ba 2 (Cu 12x Co x ) 3 O 72d ( y 5 x, 2x, 3x, 4x; 0.0 # x # 0.15) were synthesised by the standard ceramic technique [7] under identical conditions. Stoichiometric quantities of fine powders of Gd 2 O 3 , BaCO 3 , CuO, Co 3 O 4 and CaCO 3 (all 99.98% pure) were thoroughly mixed and heated in air at 9308C for 24 h in a platinum crucible. This reacted Cu–Co powder was reground and reheated at 9308C for 24 h to obtain a homogeneous single phase sample. The black product was then pulverised and cold pressed into pellets which were sintered in air at 9308C for 24 h. To obtain oxygenated samples of Cu–Co, these pellets were annealed under flowing oxygen at 5008C for 24 h, followed by a slow cooling at a rate of 18C min 21 until room temperature was reached. The sintering time and oxygen annealing time for all Cu–Co and Gd–Ca / Cu–Co compounds were the same. The oxygen annealing for the Gd–Ca / Cu–Co samples was carried out at 5508C, which is 508C more than that of the Cu–Co samples to accommodate coupled substitutions.

All the samples were characterised at room temperature by X-ray diffraction using Cu Ka radiation. The X-ray analysis revealed that all the samples were single phase. The stoichiometric compositions of the constituents in the sample were confirmed by EDAX analysis using a JEOL scanning electron microscope. The oxygen content was determined using iodometric method. Resistance versus temperature behaviour was studied as a function of temperature on regularly shaped samples using the standard four probe technique. The a.c. susceptibility measurements were accomplished using a system with a lock-in-amplifier and an APD cryocooler.

3. Results and discussions The observed X-ray diffraction patterns were modelled by modified Gaussian functions and the refined unit cell parameters, calculated using the standard least squares program, are listed in Table 1. Five Co-doped samples (x50.00 to 0.08) and eleven Ca–Co doped samples (x5 0.10, y50.10, 0.15, 0.20, 0.25, 0.30, 0.35; x50.12, y5 0.12, 0.24, 0.36; x50.15, y50.15, 0.30, 0.45) remain orthorhombic with distortion (b 2 a) /(b 1 a) varying between 0.76% to 0.18%, while three Co-doped samples (x50.10, 0.12, 0.15) and three Ca–Co doped samples (x50.10, y50.05; x50.10, y50.40; x50.15, y50.15) show nearly tetragonal structure. A striking difference between the Co-doped Y-123 and Gd-123 is that the

Table 1 ˚ orthorhombicity, values of oxygen content (z) and effective copper valence (21p) for Gd 12y Ca y Ba 2 (Cu 12x Co x )O 72d Values of unit cell parameters (A), system Sample (x, y)

˚ Unit cell parameters (A)

Orthorhombicity (b 2 a) /(b 1 a)*100

Oxygen content (z)

Effective Cu-valence (21p)

a

b

c

(0.00,0.00) (0.02,0.00) (0.04,0.00) (0.06,0.00) (0.08,0.00) (0.10,0.00) (0.10,0.05) (0.10,0.10) (0.10,0.15) (0.10,0.20) (0.10,0.25) (0.10,0.30) (0.10,0.35) (0.10,0.40)

3.838(3) 3.834(3) 3.837(3) 3.843(3) 3.878(3) 3.899(3) 3.887(3) 3.847(3) 3.846(3) 3.845(3) 3.841(3) 3.866(3) 3.866(3) 3.867(3)

3.896(3) 3.881(3) 3.882(3) 3.880(3) 3.894(3) 3.899(3) 3.889(3) 3.887(3) 3.886(3) 3.884(3) 3.880(3) 3.887(3) 3.881(3) 3.870(3)

11.700(11) 11.703(11) 11.704(11) 11.723(11) 11.747(11) 11.756(11) 11.694(11) 11.701(11) 11.724(11) 11.702(11) 11.697(11) 11.729(11) 11.714(11) 11.709(11)

0.7589 0.6157 0.5842 0.4856 0.2058 0.0064 0.0257 0.5120 0.5134 0.5213 0.5116 0.2696 0.1936 0.0270

6.96(2) 6.92(2) 6.81(2) 6.74(2) 6.70(2) 6.58(2) 6.64(2) 6.61(2) 6.69(2) 6.70(2) 6.78(2) 6.96(2) 7.01(2) 7.03(2)

2.306(2) 2.280(2) 2.206(2) 2.160(2) 2.133(2) 2.053(2) 2.093(2) 2.073(2) 2.123(2) 2.133(2) 2.186(2) 2.306(2) 2.340(2) 2.353(2)

(0.12,0.00) (0.12,0.12) (0.12,0.24) (0.12,0.36)

3.902(3) 3.884(3) 3.860(3) 3.815(3)

3.905(3) 3.898(3) 3.887(3) 3.881(3)

11.751(11) 11.689(11) 11.701(11) 11.667(11)

0.0300 0.1800 0.3500 0.8576

6.54(2) 6.78(2) 6.80(2) 6.81(2)

2.027(2) 2.187(2) 2.200(2) 2.207(2)

(0.15,0.00) (0.15,0.15) (0.15,0.30) (0.15,0.45)

3.910(3) 3.903(3) 3.859(3) 3.831(3)

3.911(3) 3.907(3) 3.899(3) 3.883(3)

11.763(11) 11.721(11) 11.730(11) 11.701(11)

0.0128 0.0512 0.5156 0.6741

6.51(2) 6.69(2) 6.75(2) 6.79(2)

2.007(2) 2.127(2) 2.160(2) 2.193(2)

D.G. Kuberkar et al. / International Journal of Inorganic Materials 3 (2001) 59 – 66

Co-doped Y-123 samples undergo an orthorhombic to tetragonal structural phase transition at x50.03 [3], whereas the Co-doped Gd-123 samples remain orthorhombic up to x50.08 (Table 1) i.e. the degree of orhtorhombicity has been extended 2.5 times, due to the redox conditions used during the synthesis. This also shows that the magnetic nature of Gd is responsible for extending orthorhombicity of Co-doped Y-123 [3] from x50.06 to x50.08 for Codoped Gd-123. It is interesting to note that the orthorhombicity of Ca-doped (Gd 12y Ca y )Ba 2 (Cu 12x Co x ) 3 O 72d samples with x50.10, 0.12 and 0.15 increases with increasing y till y 5 3x inspite of the mother sample ( y50.0) being tetragonal. The lattice parameters of pure Gd-123 is in good agreement with the reported values [7,8]. The values of the oxygen content of the single phase Co and Ca–Co doped Gd-123 samples determined by an iodometric titration technique, are listed in Table 1. The effective Cu-valence (2 1 p) or the hole concentration ( p) per Cu–O unit was calculated from these data and is included in Table 1. Moreover, the oxygen content of Co and Ca–Co doped samples varies between 6.51 and 7.03, while the effective Cu valence of the Co-doped Gd-123 system decreased with increasing Co content for x50.0 to 0.15, indicating that some of the higher valent Co 31 ions from substituted Co 3 O 4 replaces Cu 21 and the effective Cu-valence of the (Gd 12y Ca y )Ba 2 (Cu 12x Co x ) 3 O 72d system for fixed x increases with increasing Ca-concentration ( y), demonstrating that Ca 21 replaces Gd 31 . Temperature dependence of resistance for the Co and Ca–Co doped samples are displayed in Figs. 1–3. Table 2 shows the resistive superconducting transition temperature (T c ); zero resistance (T R50 ). There is a very good agreec ment between our T R50 values and those reported in the c literature [7,8] for Gd-123 (x 5 y 5 0.00). T R50 was found c to decrease from 89 K to 24 K by 8% Co substitution for Cu in Gd-123 (Table 2) with an average rate of 8.1 K per R50 atm% of Co. The suppression in T c by Co substitution 21 for Cu is compensated for by the addition of holes by Ca 21 doping for Gd 31 at constant Co-concentration. This R50 is shown by observing the increase in T c with increasing y for (Gd 12y Ca y )Ba 2 (Cu 12x Co x ) 3 O 72d (Figs. 2 and 3, Table 2). Resistance (R) values at 300 K for the samples studied are shown in Fig. 4(a) as a function of concentration (x) at constant y values. Fig. 4(b) displays T R50 versus the c Co-content x at constant Ca-concentration y in (Gd 12y Ca y )Ba 2 (Cu 12x Co x ) 3 O 72d . We have also shown in Fig. 5(a) and (b), R versus y and T R50 versus y at constant c x. It is evident from Fig. 4(a) and (b) that Co-doping in Gd-123 ( y50.0) results in an increasing resistance with decreasing T R50 of a semiconducting nature. In order to c understand the influence of Ca-content upon the T c and normal state resistance at 300 K of (Gd 12y Ca y )Ba 2 (Cu 12x Co x ) 3 O 72d oxides, we have studied a series of substituted cuprates characterised by a fixed Ca

61

Fig. 1. Resistance versus temperature for (Gd 12y Ca y )Ba 2 (Cu 12x Co x ) 3 O 72d . (a) x50.12, y50.0, 0.12, 0.24, 0.36. (b) x50.15, y50.0, 0.15, 0.30, 0.45.

substitution rate ratio with Co, namely y /x5`, 1, 2 and 3. The introduction of Ca in GdBa 2 (Cu 12x Co x ) 3 O 72d altered the normal state and superconducting properties substantially. Particularly intriguing is that Ca-doped samples have higher T c values and lower resistivities, compared to the non Ca-doped systems [Fig. 4(a) and (b)]. The T c and resistance of these series show strong dependence on Co-content (Fig. 4), for y 5 x, 2x and 3x. For increasing y, T c increases with y, attaining a maximum value at y 5 3x and resistance decreases with y displaying a minimum value for y 5 3x. Fig. 5(a) and (b) indicate that the resistance decreases and T c increases with increasing Ca concentration y for a fixed Co-content. All these observations suggest that the higher concentration of Ca in the material transforms it into better superconductor. The temperature dependence of a.c. susceptibility, xac , for Co and Ca–Co doped Gd-123 samples are shown in Figs. 6 and 7 and values of T on c ( xac ) are listed in Table 2. There is very good agreement between the resistive T cR50 and T on ( xac ) values, showing the consistency of the c measurements and the high quality of the samples. The observed suppression of T c by Co-substitution in GdBa 2 (Cu 12x Co x ) 3 O 72d is explained on the basis of the oxygen effect, assuming it is real through the T c versus x and hole concentration ( p) versus x plots shown in Fig. 8.

62

D.G. Kuberkar et al. / International Journal of Inorganic Materials 3 (2001) 59 – 66

Fig. 2. Resistance versus temperature for GdBa 2 (Cu 12x Co x ) 3 O 72d ; x50.02, 0.04, 0.06, 0.08, 0.10.

Fig. 3. Resistance versus temperature for (Gd 12y Ca y )Ba 2 (Cu 12x Co x ) 3 O 72d ; x50.10, y50.05, 0.10, 0.15, 0.20 and x50.10, y50.25, 0.30, 0.35, 0.40.

D.G. Kuberkar et al. / International Journal of Inorganic Materials 3 (2001) 59 – 66

63

Table 2 Values of T Rc 50 (K), T on c ( xac ) for Gd 12y Ca y Ba 2 Cu 12x Co x O 72d system Sample (x,y)

Resistivity T R50 (K) c

a.c. susceptibility T con ( xac )

(0.00,0.00) (0.02,0.00) (0.04,0.00) (0.06,0.00) (0.08,0.00) (0.10,0.00) (0.10,0.05) (0.10,0.10) (0.10,0.15) (0.10,0.20) (0.10,0.25) (0.10,0.30) (0.10,0.35) (0.10,0.40)

89(1) 77(1) 63(1) 50(1) 24(1) N.S. 20(1) 36(1) 50(1) 56(1) 64(1) 83(1) 76(1) 73(1)

88.0(1) 79.0(1) 66.0(1) 52.5(1) 28.5(1) N.S. 21.5(1) 35.5(1) 53.0(1) 59.0(1) 66.0(1) 84.5(1) 78.0(1) 75.0(1)

(0.12,0.00) (0.12,0.12) (0.12,0.24) (0.12,0.36)

N.S. 27(1) 40(1) 68(1)

N.S. 26.5(1) 41.5(1) 67.0(1)

(0.15,0.00) (0.15,0.15) (0.15,0.30) (0.15,0.45)

N.S. 14(1) 37(1) 63(1)

N.S. 38.5(1) 38.5(1) 65.5(1)

The oxygen content z of a superconductor is directly related to the hole concentration p or the effective Cuvalence 2 1 p, which controls the superconductivity. The concentration of holes can be varied by varying the Co 31 doping concentration. It is evident from Fig. 8 that T c decrease from 89 K to 0 K as p decreases from 0.306 to 0.053 with a corresponding increase in x from 0.00 to 0.10. T c approaches nearly zero around x50.10, so that the x50.12 and 0.15 samples exhibit pure semiconducting behaviour with p50.0. The trend is clear: The hole concentration p decreases as the samples are doped with a higher concentration of Co for Cu 21 . Co 3 O 4 doing at Cu-site under redox conditions of synthesis results in to the formation of Co 21 and Co 31 ions in the final phase. The comparison of T c versus oxygen content and hole concentration for Co-doped Y-123 [3] and Co-doped Gd-123 (present work) shows that, the rate of T c suppression is lower in case of Co-doped Gd-123 similar to Y-123 suggesting that majority of Co-ions are in Co 21 state and the fewer Co 31 ions provide the additional electrons which are expected to fill mobile holes in the Cu–O chain site, thereby reducing the conduction and decreasing T c at slower rate. Our results agree with the T c -hole concentration behaviour observed in orthorhombic YBa 2 Cu 3 O 72d in which T c and oxygen content or hole concentration [9] decreases slowly which is attributed to Co 21 doping at Cu-site. These results clearly establish that the effect of increasing the Co concentration in GdBa 2 (Cu 12x Co x ) 3 O 72d is equivalent to that of decreasing the hole concentration in Gd-123. This also strongly suggests that Co exits in two valence states, Co 21 and

Fig. 4. (a) Room temperature resistance versus concentration (x) for y50, x, 2x and 3x in (Gd 12y Ca y )Ba 2 (Cu 12x Co x ) 3 O 72d . (b) T R50 versus c concentration (x) for y 5 0, x, 2x and 3x in (Gd 12y Ca y )Ba 2 (Cu 12x Co x ) 3 O 72d .

64

D.G. Kuberkar et al. / International Journal of Inorganic Materials 3 (2001) 59 – 66

the T c of these co-substituted phases, we have studied three series of oxides, corresponding to a constant Cocontent x50.10, 0.12 and 0.15. It is evident from Fig. 5(b) that the T c of all three oxides increase with increasing Ca content y up to y 5 3x inspite of all exhibiting semiconducting behaviour at y50.0. These results show the remarkable influence of the introduction of Ca in enhancing the superconductivity. The most spectacular proof is the phases x50.10, 0.12 and 0.15 and y 5 3x which exhibit T R50 ’s of Ca-free cuprates x50.10, 0.12 and 0.15 c with y50.0 do not superconduct. Here, Ca increases the carrier number and counter balances the effect of Co 31 which goes into the Cu–O chains and is hole filling. In particular, the T R50 of (Gd 12y Ca y )Ba 2 (Cu 0.90 Co 0.10 ) 3 O 72d c increases with increasing y up to y 5 3x 5 0.30. This implies that the reduction in T R50 from 89 K (x50.0) to 0 c K (x50.10) is compensated for by the addition of holes by 5–30 atm% Ca ( y50.05 to 0.30). The introduction of Ca increases T R50 from 0 K to 83 K in c (Gd 12y Ca y )Ba 2 (Cu 0.90 Co 0.10 ) 3 O 72d for y50.0 to 0.30 as Ca increases the hole concentration (Table 2) and counterbalances the effect of Co 31 , which goes into Cu–O chain site and is hole filling. Such a behaviour has indeed been observed and displayed in Fig. 5(b) for three constant Co-concentrations x50.10, 0.12 and 0.15. The dependence of T c on Ca concentration in (Gd 12y Ca y )Ba 2 (Cu 0.90 Co 0.10 ) 3 O 72d (Fig. 8) is explained as follows. Both oxygen content and hole concentration increase as the samples are doped with higher concentrations of Ca up to y50.40 in comparison with the y50.0 sample as shown in Fig. 8 (Table 2). The T c of (Gd 12y Ca y )Ba 2 (Cu 0.90 Co 0.10 ) 3 O 72d increases with increasing y as p increases (y50.30, T max |83 K). The c oxide at x50.10 and y 5 3x 5 0.30 is identified as the compensated oxide displaying a T R50 of 83 K. The T R50 c c of this oxide lies closer to that of pure Gd-123 (T cR50 589 K). Moreover, in this compensated oxide, the hole filling by Co is completely balanced by hole doping from Ca.

4. Conclusion

Fig. 5. (a) Room temperature resistance versus concentration ( y) for x50.10, 0.12, 0.15 in (Gd 12y Ca y )Ba 2 (Cu 12x Co x ) 3 O 72d . (b) T R50 versus c concentration ( y) for x50.10, 0.12, 0.15 in (Gd 12y Ca y )Ba 2 (Cu 12x Co x ) 3 O 72d .

Co 31 in agreement with the redox conditions used during the synthesis. In order to understand the influence of Ca-content upon

In conclusion, the comparatively slower decrease of T c with Co-doping in GdBa 2 (Cu 12x Co x ) 3 O 72d suggests that the Co exits in Co 21 and Co 31 valence states and few Co 31 ions fills the mobile holes thereby reducing the hole concentration and suppresses superconductivity around x5 0.1. This suppression can be compensated by appropriate hole doping with Ca. The introduction of Ca for Gd in (Gd 12y Ca y )Ba 2 (Cu 12x Co x ) 3 O 72d shows that Ca-doped samples have higher T c values and lower resistivities contrary to the non Ca-doped systems. In particular, Cadoping is able to revive the superconductivity (T R50 583 c K, 68 K, 63 K) for x50.10, 0.12 and 0.15, respectively for y 5 3x in GdBa 2 (Cu 12x Co x ) 3 O 72d having x50.10, 0.12 and 0.15 which are semiconductors.

D.G. Kuberkar et al. / International Journal of Inorganic Materials 3 (2001) 59 – 66

65

Fig. 6. Temperature dependence of a.c. susceptibility for GdBa 2 (Cu 12x Co x ) 3 O 72d ; x50.00, 0.02, 0.04, 0.06, 0.08, 0.10.

Fig. 7. (a) Temperature dependence of a.c. susceptibility for (Gd 12y Ca y )Ba 2 (Cu 12x Co x ) 3 O 72d ; x50.10, y50.05, 0.10, 0.15, 0.20, 0.25, 0.30, 0.35, 0.40. (b) Temperature dependence of a.c. susceptibility for (Gd 12y Ca y )Ba 2 (Cu 12x Co x ) 3 O 72d ; x50.12, y50.00, 0.12, 0.24, 0.36. (c) Temperature dependence of a.c. susceptibility for (Gd 12y Ca y )Ba 2 (Cu 12x Co x ) 3 O 72d ; x50.15, y50.0, 0.15, 0.30, 0.45.

D.G. Kuberkar et al. / International Journal of Inorganic Materials 3 (2001) 59 – 66

66

Fig. 8. Dependence of T c and hole concentration on x and y (x50.10) for (Gd 12y Ca y )Ba 2 (Cu 12x Co x ) 3 O 72d .

Acknowledgements This research was supported by Department of Atomic Research (BRNS) India. Authors are thankful to Professor Ajay Gupta and Dr. V. Ganesan, IUC-DAEF, Indore for extending X-ray diffraction facilities. Author DGK is thankful to AICTE, New Delhi for the financial assistance in the form of R&D Project.

References [1] Kakihana M, Eriksson SG, Borjesson L, Johansson LG, Strom C, Kall M. Phys Rev B 1993;47:5359.

[2] Xu Y, Sabatini RL, Moddenbauch AR, Zhu Y, Shyu SG, Suenaga M, Dennis KW, McCallum RW. Physica C 1990;169:205. [3] Tarascon JM, Barboux P, Miceli PF, Greene LH, Hull GW. Phys Rev B 1988;37:7458. [4] Smith MG, Taylor RD, Oesterreicher H. Phys Rev B 1990;42:4202. [5] Kulkarni RG, Kuberkar DG, Baldha GJ, Bichile GK. Physica C 1993;217:175. [6] Suard E, Maignan A, Caignaert V, Reveau B. Physica C 1992;200:43. [7] Kulkarni RG, Upadhaya RV, Bichile GK, Shaikh IA. Cryogenics 1992;32:770. [8] Tarascon JM, Makinnon WR, Green LH, Hull GW, Vogel EM. Phys Rev B 1987;36:226. [9] Mckinnan WR, Post ML, Selwyn LS, Pleizier G, Tarascon JM, Barboux P, Greene LH, Hull GW. Phys Rev B 1988;38:6543.