Anomalous temperature dependence of the peak effect in iodine-intercalated Bi2Sr2CaCu2O8+y single crystals

PHYSICA ELSEVIER

Physica C 259 (1996)280-286

Anomalous temperature dependence of the peak effect in iodine-intercalated Bi2Sr2CaCu208+ysingle crystals S. Ooi *, T. Tamegai, T. Shibauchi Department of Applied Physics, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113, Japan

Received 23 October 1995; revised manuscriptreceived9 January 1996

Abstract

Peak effects in the magnetic-field dependence of the magnetization have been studied for iodine-intercalated Bi2Sr2CaCu2Os+y (I-BSCCO) single crystals. We found that the peak field Hp in I-BSCCO is almost independent of temperature at low temperatures, whereas it shows an anomalously strong temperature dependence at high temperatures. This contrasts with the behavior of Hp in pure BSCCO where Hp is almost temperature independent. The constant value of Hp at low temperatures increases with decreasing anisotropy, which is controlled by the oxygen and iodine contents. This behavior suggests that the dimensional crossover of vortices is the origin of the peak effect at low temperatures. On the other hand, the temperature dependence of Hp at high temperatures has a correlation with the inhomogeneity of the samples. We propose a mechanism for the temperature-dependent peak effect similar to that for the fish-tail effect in YBa2Cu307_y.

1. Introduction The " p e a k effect" of the magnetization in Bi2Sr2CaCu2Os+y (BSCCO) has received much attention [1-4]. The peak effect is a phenomenon in which the absolute value of the magnetization increases anomalously at some intermediate field. Peak effects have been observed in most of the single crystals of high-T~ superconductors when the field is applied parallel to the c-axis [5-9]. Particularly the peak effect in BSCCO has the characteristic feature that the peak is very sharp and the peak field, Hp, is low (from 100 Oe to 1 kOe) and almost temperature independent [3]. On the contrary, Hp in YBa2Cu 3O7_y [5,10,11] and La2_xSrxCuO 4 [12] (in the so-

* Corresponding author. Fax: + 81-3-3816-7805.

called "fish-tall" effect) shows a strong temperature dependence and the peak is broad. For the origin of the peak effect in BSCCO, explanations such as matching effect [2,13], difference in relaxation rate of the magnetization [14], and 2 D - 3 D dimensional crossover in the vortex structure [3,15] have been proposed. Observations of the fluxline lattice by neutron diffraction [15] and muon spin rotation [16] support the 2 D - 3 D dimensional crossover. The 2 D - 3 D dimensional crossover in vortex structure is expected to occur at

~0 B20 = _Fs2,

(1)

where th0 is the flux quantum, F the effective mass ratio ( m c / m a b ) , and s the layer spacing [17]. In order to support the idea of dimensional crossover as the origin of the peak effect, it is important to study how the anisotropy of the samples affects the peak

0921-4534/96/$15.00 © 1996 Elsevier Science B.V. All rights reserved PH S0921-453 4(96)00096-2

S. Ooi et aL /Physica C 259 (1996) 280-286

effect. In BSCCO, the anisotropy tends to increase as the oxygen content decreases [18]. The relation between the oxygen content and the peak effect was studied systematically by Kishio et al. [19]. Their results are qualitatively consistent with the prediction of Eq. (1). Besides changing the oxygen content, the intercalation of other elements into BSCCO is another way to control the anisotropy in this system. It has been reported that intercalants such as iodine [20], bromine [21], and silver iodide [22], can be inserted between the two BiO layers in BSCCO. Though iodine intercalation expands the c-axis by about 20%, a torque [23] and a resistivity [24,25] measurement revealed that the anisotropy decreases by the iodine intercalation into BSCCO. Thus, it is interesting how the peak effect is influenced by the iodine intercalation. There have been conflicting reports on the peak effect in iodine-intercalated BiESr2CaCuEOs+y (I-BSCCO) single crystals. Kawamata et al. observed the peak effect in I-BSCCO similar to BSCCO by the magnetic torque method [26]. On the contrary, Ha et al. recently claimed that the peak effect in BSCCO disappears by iodine intercalation [27]. Hence, the relation between the iodine-intercalation and the peak effect is not yet clearly understood. In this paper, in order to investigate the peak effect of

281

I-BSCCO, we measured the temperature dependence of the magnetic properties in I-BSCCO single crystals having different oxygen contents using a Hall probe.

2. Experiments Single crystals of BSCCO were grown by the traveling solvent floating zone (TSFZ) method. All crystals have platelet geometries with dimensions of typically 1 X 1 × 0.05 mm 3. Iodine was intercalated by sealing the samples with elemental iodine in evacuated or air-filled Pyrex glass ampoules and by heating at 150 ~ 180°C for periods up to 5 days. In order to investigate the combined effects of iodine intercalation and oxygen doping, we prepared I-BSCCO having different oxygen contents. Although in many studies the intercalation was performed for 2 weeks following Xiang et al. [20], the time required for the complete stage-1 intercalation is reported to be less than 24 h at 180°C for polycrystalline samples [28,29] and single crystals [28]. In our case, the complete stage-1 intercalation of iodine into single crystal BSCCO judged from the XRD pattern was also achieved within 24 h at 180°C. The oxygen annealing was performed in

Table 1 The transition temperature (Te), the transition width (ATe), the full width at half maximum (FWHM) of the XRD pattem, and oxygen-doping and iodine-intercalation conditions of BSCCO and I-BSCCO used in this paper. Te and ATe are determined by AC susceptibility measurements. FWHM is determined from the width of the (0 0 5) peak in the 0 - 2 0 scan. "oxygen-doping" represents the annealing process for controlling the oxygen content before the iodine intercalation. "iodine-interealation" shows the iodine-interealation conditions and "air pressure" shows the air pressure sealed in the Pyrex glass ampoules Sample name BSSCO As-grown

Te(K) ATe(K) FWHM (deg) Oxygen-doping Atmosphere Temperature (°C) Time (h) Iodine-intercalation Air pressure (Torr) Temperature (°C) Time (h)

83 3

I-BSCCO Oxygen-annealed 82 3

flowing 02 500 24

# 1

# 2

# 3

79 9.5 0.066

79 9 0.125

81 10.5 0.182

air 500 24

-

flowing 0 2

760 180 72

10 -2 180 120

760 150 36

500 24

282

S. Ooi et al./Physica C 259 (1996) 280-286 t

i

i

~

o

#1. These three samples were annealed in different oxygen atmosphere before the intercalation processes. Hysteresis measurements were carded out using a commercial GaAs Hall probe (active area: 400 × 400 ~zm2). The sample was put on the Hall probe in such a way that the magnetic field was applied parallel to the c-axis. The average sweep rate of the applied field was about 50 Oe/s. Since the active area of the Hall probe is comparable with the sample size, and it is 400 p~m away from the sample, B (local field) measured by the Hall probe is the sum of M (magnetization) and H a (applied field). In order to estimate M we subtracted a linear term for H a from B so that t h e M - H a hysteresis has a reasonable shape.

(a)

o

i I

I

I

1 I

I

1

(b) ~ oo

~o

I

J I

, I

t

I

I

I

I

(c)

{I} c-

0

I I

I

~j I

i t

i

i I

I

I

,.

(d)

"I

0

1'o

20

3'0 40 20 (deg)

s'o do

Fig. 1. XRD patterns of (a) as-grown BSCCO, (b) I-BSCCO #1, (c) I-BSCCO #2, and (d) I-BSCCO #3. These were measured in the 0 - 2 0 scan mode using Cu Kt~ radiation.

flowing oxygen at 500°C for 24 h before the intercalation. The oxygen-annealing and the iodine-intercalation conditions for each samples are listed in Table 1. Samples were characterized by X-ray diffraction (XRD) and AC susceptibility measurements. The superconducting transition temperature (T~) determined from the AC susceptibility for the as-grown and the oxygen-annealed BSCCO are 83 and 82 K, respectively. The transition width (ATe) for both samples is 3 K. The Tc and AT~ for I-BSCCO samples used in this study are listed in Table 1. Fig. 1 shows XRD patterns for pure BSCCO and three I-BSCCO samples. These patterns indicate that I-BSCCO # 1 and # 2 are completely single-phased. We evaluated the homogeneity of the sample from the FWHM (full width at half maximum) of the (0 0 5) peak in the XRD patterns (see Table 1). The FWHM of the (0 0 l) peak indicates the distribution of the c-axis lengths as modulated by the iodine intercalation. From this value, it is concluded that I-BSCCO # 3 is more inhomogeneous than I-BSCCO

3. Results and discussion Fig. 2 shows M - H a hysteresis curves for the as-grown BSCCO single crystal at various temperatures. As reported earlier [3], sharp peaks in the M - H a curves are observed. Here we estimate the value of the peak field, Hp, in the first quadrant of t h e M - H a plane. Arrows indicate the peak positions•

60

I

I

I

Bi~SrzCaCu~Os.y 50 40 30

20

0

i

20.OK

200

400 Ha (Oe)

600

800

Fig. 2. Hysteresis curves in the first quadrant of the M - H a plane for an as-grown BSCCO single crystal at various temperatures. The field was applied parallel to the c-axis. Arrows indicate the peak positions. The peak field, Hp, is estimated by the intersection of two lines drawn on both sides of the peak.

S. Ooi et al./Physica C 259 (1996) 280-286

8~t

~27.5K

6

I III \\

0

I

400

800

1200

H a (Oe)

,,o[_

'

'

~.~6t" 0 ~ 40 I "-'-.=J

/ 20.oK

1200 121104 I-Bi2Sr2CaCu2~ 0

N

400

800 Ha (Oe)

6

Woo.o / 0

400

800 H a (Oe)

1200

1600

283

Hp for this crystal is almost temperature independent and is about 200 Oe. Figs. 3(a), (b), and (c) show M - H a hysteresis curves for three I-BSCCO # 1, #2, and #3, respectively. Peak effects similar to the as-grown BSCCO are observed in all our I-BSCCO crystals. Even in the very inhomogeneous I-BSCCO #3, clear single peaks can be seen. Comparing Fig. 2 with Fig. 3, we find an interesting difference between the peak effects in BSCCO and in I-BSCCO. While Hp is almost independent of temperature in BSCCO, it shows a clear temperature dependence at high temperatures in I-BSCCO. However, Hp in I-BSCCO at low temperatures is almost temperature independent similar to BSCCO. Namely, I-BSCCO has two different temperature regimes. As introduced above, there are two types of peak effects in high-T~ superconductors. One shows a strongly temperature dependent Hp as in the case of YBa2Cu307_y and La2_xSrxCuO4, and the other shows an almost temperature-independent Hp as in the pristine BSCCO. The temperature dependent Hp observed here in I-BSCCO is different from these two and therefore it could be a new type. Fig. 4 shows the temperature dependence of Hp for the as-grown and the oxygen-annealed BSCCO, and intercalated samples (I-BSCCO # 2 and #3). As far as the low-temperature regime is concerned, Hp is almost constant for I-BSCCO as well as for BSCCO [30]. The constant Hp at low temperatures increases with the iodine intercalation as well as the oxygen doping. Thus, it is clear that the iodine intercalation has an effect similar to the oxygen doping on the peak effect in BSCCO. From these observations, the temperature-independent peak effect at low temperatures in I-BSCCO is considered to have the same origin as that in BSCCO. Actually based on the 2D-3D dimensional crossover, this behavior is consistently explained as follows. Eq. (1) shows that the peak field increases as the anisotropy decreases, if s is constant. Since it is reported that s is almost unchanged and the anisotropy decreases by oxygen doping in BSCCO [18], the increase in Hp by the oxygen doping is

Fig. 3. Hysteresis curves in the first quadrant of the M - H a plane for (a) I-BSCCO # 1 , (b) I-BSCCO # 2 , and (c) I-BSCCO # 3 at various temperatures. The field was applied parallel to the c-axis. Arrows indicate the peak positions.

284

S. Ooi et al./Physica C 259 (1996) 280-286

1000

i

I

I

I

' 30

' 40

800 .--. 600 0 o ~:= 400

200

~ I-BSC -e-- BSCCO - t - O 2 annealed I-BSCCO #3 - e - O 2 annealed BSCCO

0

'

I0

' 20

Temperature (K) Fig. 4. Temperature dependence of Hp for the as-grown BSCCO before (open diamonds) and after intercalation, i.e., I-BSCCO # 2 (closed diamonds). Open and closed circles indicate the data for the oxygen-annealed BSCCO before and after intercalation (I-BSCCO #3), respectively.

consistent with Eq. (1). On the other hand, the iodine intercalation expands s from 15.4 ,~ to 18.9 ,~. However, if the decrease in the anisotropy by the iodine intercalation is larger than the increase in s, Hp is still expected to increase. I-BSCCO is reported to show metallic c-axis resistivity [24], which supports the above speculation that the anisotropy is largely reduced by the iodine intercalation. Next, we consider the origin of the temperaturedependent peak effect in I-BSCCO at high temperatures. If we interpret the temperature-dependent Hp by Eq. (1), the anisotropy must increase with increasing temperature. However, the torque measurement [23] suggests that the anisotropy decreases with increasing temperature. Thus, we conclude that the temperature-dependent peak effect at high temperatures is not due to the 2D-3D dimensional crossover. Fig. 5 shows Hp versus temperature curves for I-BSCCO #1 and #3. If we compare I-BSCCO # 1 and #3, Hp starts to decrease at a lower temperature in I-BSCCO #3. Since these two samples have a similar almost temperature-independent value of Hp at low temperatures [31], they are considered to have similar oxygen contents. So, we believe that the strongly temperature-dependent peak effect is caused by the other parameter. The other difference between

these samples is the inhomogeneity of the iodine distributions. As described before, I-BSCCO # 3 is more inhomogeneous than #1. Therefore, we infer that the inhomogeneity of iodine distributions makes the temperature-dependent Hp at high temperatures shift to lower fields. The temperature-dependent part of the peak effect in I-BSCCO is similar to the "fish-tail" effect in YBa2Cu307_y and La2_xSrxCuO 4 single crystals which also shows a strong temperature dependence of Hp. The "fish-tail" effect is considered to originate in enhanced flux pinning by oxygen-deficient regions [5,11,12]. In other words, as the magnetic field increases, the oxygen-deficient regions which have a lower He2 [5,12] change into the normal state. When the magnetic field is further increased, the magnetization becomes smaller by the field-induced granularity [5] or by percolatively connecting the reversible regions throughout the sample [11]. According to these idea, Hp correlates Hc2 [5] or Hi~r (irreversibility field) [1 Considering the similarity between the peak effects in I-BSCCO and the "fish-tail" effect in YBa2Cu307_y and La2_xSrxCuO 4, we believe that the peak effect in I-BSCCO at high temperatures is due to the enhanced flux pinning induced by the weak superconducting region caused by microscopic irregularities of the iodine distributions. Based on this idea, the correlation between the peak effect in 1000

800 #1 0

600

0

a:~ 400 I - B S C C O #1 I-BSCCO #3

200 0

0

r

I

I

I

10

20

30

40

50

Temperature (K) Fig. 5. Temperature dependence of Hp for I-BSCCO # 1 and #3 having a different inhomogeneity of the iodine intercalation.

S. Ooi et al./Physica C 259 (1996) 280-286

I-BSCCO and the inhomogeneity of the sample is explained as follows. In the region where the iodine is imperfectly intercalated, the superconductivity is expected to be destroyed at a relatively small field because of the weak-link nature of this region. The weak-link nature of this region is enhanced by the microscopic irregularity of the iodine-intercalation. Hence, I-BSCCO with a higher degree of inhomogeneity is expected to have a smaller value of Hp. It should be noted that Hp in I-BSCCO # 2 and # 3 starts to decrease at about 20 K, although they have a different inhomogeneity of the iodine distribution. This is because judging from the low-temperature value of Hp, I-BSCCO # 2 has a larger anisotropy and thus has a smaller irreversibility field [32], which could be related to Hp [11]. Hence, the effect of the inhomogeneity of the iodine distribution on the temperature dependence of Hp is most clearly seen in samples having a similar anisotropy, like I-BSCCO #1 and #3. Finally, we explain why we did not observe two separate peak effects at a given temperature. As is evident from the data for the pure BSCCO (Fig. 2), the magnitude of the peak effect with constant Hp diminishes very fast with increasing temperature. The peak structure is hardly observable above 30 K, where the temperature-dependent peak effect in I-BSCCO starts to emerge. Hence, we cannot observe the low-temperature peak effect with constant Hp at high temperatures if any. On the other hand, at the 2D-3D dimensional crossover field, the form of the vortices changes drastically and at higher fields they are pinned so that the pinning energy is maxim i z e d only in each layer. In this case, an additional pinning due to the field-induced normal region is not expected to give a large effect. Thus, we observe only the peak effect with constant Hp at low temperatures.

4. Conclusion We have measured the M - H a hysteresis for I-BSCCO single crystals having different iodine distributions and oxygen contents using a Hall probe. We found an anomalous temperature-dependent peak effect in I-BSCCO. Hp is almost independent of

285

temperature at low temperatures, whereas it shows a strong temperature dependence at high temperatures. The constant Hp at low temperatures has a correlation with the anisotropy of the samples which is controlled by the oxygen and iodine contents. This behavior supports the 2D-3D dimensional crossover as the origin of the peak effect at low temperatures. On the other hand, the temperature-dependent peak effect at high temperatures correlates with the inhomogeneity of the iodine distribution in the sample, not with the oxygen content. We propose a mechanism for the temperature-dependent peak effect similar to the fish=tail effect; an imperfectly iodine-intercalated region with microscopic irregularities of the iodine distributions plays a role as magnetic-field induced pinning center like the oxygendeficient region in YBaECU307_y.

Acknowledgement This work was supported in part by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science and Culture of Japan.

References [1] V.N. Kopylov, A.E. Koshelev and I.F. Schegolev, Physica C 170 (1990) 291. [2] G. Yang, P. Shang, S.D. Sutton, I.P. Jones, J.S. Abell and C.E. Gough, Phys. Rev B 48 (1993) 4054. [3] T. Tamegai, I. Oguro, Y. lye and K. Kishio, Physica C 213 (1993) 33. [4] A.K. Pradhan, S.B. Roy, P. Chaddah, C. Chen and B.M. Wanklyn, Phys. Rev. B 49 (1994) 12984. [5] M. Daeumling, J.M. Seuntjens and D.C. Larbalestier, Nature (London) 346 (1990) 332. [6] F. Zuo and S. Khizroev, Phys. Rev. B 49 (1994) 12326. [7] F. Zuo, S. Khizroev, G.C. Alexandrakis and V.N. Kopylov, Phys. Rev. B 52 (1995) R755. [8] M. Xu, D.K. Finnemore, G.W. Crabtree, V.M. Vinokur, B. Dabrowski, D.G. Hinks and K. Zhang, Phys. Rev B 48 (1993) 10630. [9] V. Hardy, A. Wahl, A. Ruyter, A. Maignan, C. Martin, L. Coudrier, J. Provost and Ch. Simon, Physica C 232 (1994) 347. [10] J.L. Vargas and D.C. Larbalestier, Appl. Phys. Lett. 60 (1992) 1741. [11] L. Klein, E.R. Yacoby, Y. Yeshurun, A. Erb, G. Miiller-Vogt, V. Breit and H. Wiihl, Phys. Rev. B 49 (1994) 4403. [12] T. Kobayashi, Y. Nakayama, K. Kishio, T. Kimura, K.

286

[13] [14] [15]

[16]

[17] [18]

[19]

[20]

[21] [22] [23] [24]

& Ooi et al./Physica C 259 (1996) 280-286

Kitazawa and K. Yamafuji, Appl. Phys. Lett. 62 (1993) 1830. R. Yoshizaki, H. Ikeda and D.S. Jeou, Physica C 225 (1994) 299. Y. Yeshurun, N. Bontemps, L. Burlachkov and A. Kapitulnik, Phys. Rev. B 49 (1994) 1548. R. Cubbit, E.M. Forgan, G. Yang, S.L. Lee, D.M. Paul, H.A. Mook, M. Yethiraj, P.H. Kes, T.W. Li, A.A. Menovsky, Z. Tarnawski and K. Mortensen, Nature (London) 365 (1993) 407. S.L. Lee, P. Zimmermann, H. Keller, M. Qarden, I.M. Savic, R. Schauwecker, D. Zech, R. Cubitt, E.M. Forgan, P.H. Kes, T.W. Li, A.A. Menovsky and Z. Tamawski, Phys. Rev. Lett. 71 (1993) 3862. V.M. Vinokur, P.H. Kes and A.E. Koshelev, Physica C 168 099O) 29. Y. Kotaka, T. Kiraura, H. Iknta, J. Shimoyama, K. Kitazawa, K. Yamafuji, K. Kishio and D. Pookc, Physica C 235-240 (1994) 1529. K. Kishio, J. Shimoyama, Y. Kotaka and K. Yamafuji, Proc. 7th. Int. Workshop on Critical Currents in Superconductors, ed. H.W. Weber (World Scientific, Singapore, 1994) p. 339. X.-D. Xiang, S. Mckernan, W.A. Vareka, A. Zettl, J.L. Corkill, T.W. Barbee III and M.L. Cohen, Nature (London) 348 (1990) 145. Y. Koike, T. Oknbo, A. Fujiwara, T. Noji and Y. Saito, Solid State Commun. 79 (1991) 501. H. Kumakura, J. Ye, J. Shimoyama, H. Kitaguchi and K. Togano, Jpn. J. Appl. Phys. 32 (1993) L894. S. Kawamata, K. Okuda, T. Mochiku and K. Kadowaki, Physica B 194-196 (1994) 1545. X.-D. Xiang, W.A. Vareka, A. Zettl, J.L. Corkill and M.L Cohen, Phys. Rev. Lett. 27 (1992) 530.

[25] K. Kishio, D. Pooke, H.J. Trodahl, C.K. Subramaniam, Y. Kotaka, M. Seto, S. Kitao and Y. Maeda, J. Supercond. 7 (1994) 117. [26] S. Kawamata; private communication. [27] D.H. Ha, K.W. Lee, H.-C. Ri, K.H. Yoo, Y.K. Park, K. Oka, Y. Yamaguchi and Y. Nishihara, Physica C 247 (1995) 137. [28] H. Nakashima, D.M. Pooke, N. Motohira, T. Tamura, Y. Nakayama, K. Kishio, K. Kitazawa and K. Yamafuji, Physica C 185 (1991) 677. [29] D. Pooke, K. Kishio, T. Koga, Y. Fukuda, N. Sanada, M. Nagoshi, K. Kitazawa and K. Yamafuji, Physica C 198 (1992) 349. [30] Hp at low temperatures in oxygen-annealed BSCCO is not precisely temperature independent. It slightly increases with increasing temperature before it disappears at high temperatures. However, the increase in Hp is much smaller than the variation of Hp in I-BSCCO at high temperatures. The mechanism for this slight increase in Hp in the intermediate temperature range in oxygen-annealed BSCCO is unknown and needs to be clarified in the future. [31] Hp in I-BSCCO #1 below 30 K shows a similar temperature dependence to that in oxygen-annealed BSCCO (see Ref. [30]). The mechanism for this temperature dependence in I-BSCCO # t is unknown but could be common in both I-BSCCO and oxygen-annealed BSCCO. Though this effect could also exist in I-BSCCO #3, it might be hidden in I-BSCCO # 3 by the opposite temperature dependence of the peak effect induced by inhomogeueity. [32] K. Kishio, J. Shimoyama, T. Kimura, Y. Kotaka, K. Kitazawa, K. Yamafuji, Q. Li and M. Suenaga, Physica C 235-240 (1994) 2775.