Dip effect and its disappearance

J.W. Lin a,*, X. Leng b, G. Liu b, H. Luo b, Y. Liu b, S.Y. Ding b

a College of Science, Hohai University, Nanjing 210098, PR Chinab National Laboratory of Solid State Microstructures and Department of Physics, Nanjing University, Nanjing 210093, PR China

Abstract

Measurement was made of ac susceptibility (acs) as a function of temperature T for a YBa2Cu3O6:993 single crystal

under dc fields ðHdcÞ, and dips in the acs curves were observed in both v0ðT Þ and v0ðHdcÞ curves in the same time for the

sample. We explain the dips in the experimental v0ðT ;HdcÞ curves as bulk pinning effect in higher fields and lower

temperatures. However, the dips were weakened and finally replaced by surface barriers. Careful analysis of the data

shows that there exists evidence indicating the transition of flux pinning mechanism.

� 2002 Published by Elsevier Science B.V.

PACS: 74.60.Ge; 74.72.Bk

Keywords: Dip effect; Dip depth; Bulk pinning; Surface barriers

1. Introduction

In the investigation of flux dynamics for high

temperature superconductors (HTSC), ac suscep-

tibility (acs) is one of the most important tech-

niques. In such an experiment, a small ac field hac

is used to probe the behavior of vortex matter gov-

erned by a large dc field Hdc and temperature T[1–4]. An explanation of peak effect (PE) in mag-

netization or critical current density Jc has been an

interesting object [5,6]. The corresponding dip in v0

(real part of acs) is proved as a typical bulk pin-

ning characteristics [7].

In ideal systems, one expects to see a first-ordermelting transition. However, It is suggested that

the existence surface barriers (SB) could possibly

form new vortex matter phases in HTSC, which

is expected to see over a wide range of applied

magnetic fields. Hence, different opinions about

whether a dip effect (DE) in a HTSC sample is

governed by SB or bulk pinning appears [8,9].In this paper, we carefully measured the acs for

a YBa2Cu3O6:993 single crystal at different dc fields,

and observed a quite different DE. The measure-

ment of acs in lower dc fields shows signal indi-

cating the transition of flux pinning mechanism.

2. Experimental

The sample was a ultra-pure Ba2Cu3O6:993 crys-

tal grown in a bulk BaZrO3 crucible. The chemi-

cal and structural characterization of this crystal

confirmed that it had very low level of impu-

rity elements and high degree of crystalline order.

The sample is a perfect rectangle with dimensions

* Corresponding author. Tel.: +86-25-359-3661/378-6636;

fax: +86-25-359-5535.

E-mail addresses: lin.j.w@163.com (J.W. Lin), syding@

netra.nju.edu.cn (S.Y. Ding).

0921-4534/02/$ - see front matter � 2002 Published by Elsevier Science B.V.

doi:10.1016/S0921-4534(02)02225-6

Physica C 386 (2003) 89–92

www.elsevier.com/locate/physc

1:53 � 1:28 � 0:065 mm3, and the c-axis along the

shortest dimension. The bulk of the crystal is twin-

free except a single twin boundary cutting the very

tip of one of the four corners, forming a triangle,

which has an area less than 0.12% of the total

sample area.Our experiment technique is the acs. Both dc

and ac magnetic fields are along the c-axis of the

crystal. The superimposed ac field has amplitude

hac ¼ 25 G, and frequency f ¼ 1 MHz. The acs

data were extracted from the impedance data of

the coil, measured by a two-phase lock-in ampli-

fier.

3. Results and discussion

Fig. 1 shows the experimental v0 as function of

temperature in various applied fields, where one

can see clearly that in regime from 1 to 70 kG, a

dip (a sharp minimum of v0) effect. Here we define

a ‘‘dip depth’’ of acs in superconducting state asv0

dd ¼ v0max � v0

min at a characteristic field Hdc, v0max

and v0min are maximum and minimum of acs re-

spectively, and Td is the dip bottom temperature atwhich v0 ¼ v0

min. To show the field difference at

which maximum of v0dd and v0

min occur, we re-plot

the v0dd as function of field Hdc in Fig. 2. From

the figures, we can see that the v0 dip and thus Td

moves toward high temperatures with decreasing

dc field. Interestingly, that the behavior of v0dd in

the high field regime (10–70 kG) is quite different

from that in the low field regime (1–10 kG). In theformal regime, with the decreasing Hdc, the dip

depth v0dd gradually increases from point a (Hdc ¼

70 kG) to point d (Hdc ¼ 10 kG). However, when

dc field further decreases the dip depth v0dd turns to

decreasing in spite of the magnitude of the whole

v0ðT Þ is still increasing. This decreasing tendency of

depth of v0dd is continuous as long as the applied dc

field decreases. In fact the dip v0dd at Hdc ¼ 1 kG

has already been too small to probe, becoming a

small kink at point c and one can hardly find the

v0ðT Þ dip below about 1 kG. When Hdc ¼ 0 a usual

sharp transition acs curve as a function of tem-

perature without dip is clearly seen, showing our

high quality single crystal sample. It is worthy

to point out that the v0dd manifests its maximum

at point d (Hdc ¼ 10 kG) while v0min minimums at a

lower field point b (Hdc ¼ 5 kG). Therefore, one

Fig. 1. Experimental v0ðT Þ curves under different dc fields

(Hdckhackc), showing the giant DE. Point a: the highest dc field

in which v0 has DE at v0 ¼ v0min; point b: the dc field in which

v0min turn up; point c: the dc field in which DE is too small to

probe.

Fig. 2. Re-plotting Fig. 1, we construct the v0ddðHdcÞ curves,

v0dd ¼ v0

max � v0min, showing the transition of dip depth. Point d:

the dc field Hdip in which v0dd maximums.

90 J.W. Lin et al. / Physica C 386 (2003) 89–92

can deduce that something happens and the dip is

governed by physics different from that of ac sus-

ceptibility itself.

Meanwhile the v0ðT Þ curve clearly demonstrates

the DE in different Hdc, it seems to us an open

question that whether there is a DE in v0ðHdcÞcurves in different temperature for the same sam-

ple? Up to our best knowledge, there is no refer-

ence reporting the simultaneous observation of

the two DEs in both v0ðT Þ and v0ðHdcÞ curves, al-

though it is well known that the second PE ap-

pears in both jcðT Þ and jcðHdcÞ relationships where

jc is critical current density. Fortunately, it is pos-

sible to testify the issue based on the data points inFig. 1. Re-plotting the field dependent v0 under

different temperatures and re-displaying them in

Fig. 3, we see clearly the DE in v0ðHdcÞ curves in

temperature regime accessed by our experiment.

This is a strong conformation that both the DEs in

v0ðHdcÞ and v0ðT Þ curves take place simultaneously.

This acs result can be considered as an indication

that the PE in JcðHdcÞ is just the one observedin JcðT Þ according to the relationship v0 þ 1 /hac=Jc.

From the behavior of DE in v0 depicted above,

v0dd decreases from Hdc ¼ 10 kG and v0

min turns up

as Hdc 6 5 kG, and finally disappear in smaller

fields. This behavior is consistent with the PE ob-

served in hysteric magnetization loop that the low

field peak (Hdc 0) pronounced and the high field

peak (second peak) weakens and finally disappears

when temperature gradually ascends, see, forexample, [10]. This characteristic is explained in

terms of the competition between the geometric

barriers and bulk pinning. In the elevated tem-

peratures the SB are dominated which causes a

sharp low field magnetic hysteric peak. Whereas at

low temperatures the bulk pinning whose maxi-

mum takes place at high fields governs the hysteric

magnetization loop and results in the second PE[5,11]. According to this scenario, the sharp steps

in low field (Hdc 0) in v0ðT Þ curve corresponds to

the low field peak and the dips in elevated applied

fields manifest the second peaks in JcðT Þ curves as

DE in v0ðT Þ curve. Because it has been numerically

shown that the DE is behavior caused by bulk

pinning, the above explanation is a reasonable

one. This explanation naturally results in that thev0

dd in our experiments is caused by transformation

from the case governed by bulk pinning which is

more important in low temperatures into the state

dominated by SB which is stronger in high tem-

peratures.

It has been pointed out that the PE occurs just

below the vortex liquid to solid phase transition

in a narrow region [3], which is of course a bulkphenomenon. Based on this ideal the DE is used to

construct phase diagram of vortex matter. If the

SB are more and more important as temperature is

elevated the dip is affected more and more by the

SB, which implies that the phase transition line

HdcðT Þ at Hdc > Hdip is no longer the same as that

at Hdc < Hdip because of SB, here Hdip is the

strengths of applied field at which maximum of v0dd

occurs. We therefore construct a phase transition

line based on the v0ðT Þ dip temperature Td. Fig. 4

shows the Hdc–Td curve. In the figure two differ-

ent slopes are visible as the fitting lines. The dot-

ted line in the higher field regime with Hdc ¼H0ð1 � T=TcÞn kG, where H0 ¼ 133 kG, Tc ¼ 88:90

K and n ¼ 1:33, which is consistent with those

determined by using torque magnetometer andcalorimetric measurement for YBCO crystal, in-

dicting a first-order melting transition in the dip in

Fig. 3. Re-plotting Fig. 1, we construct the experimental

v0ðHdcÞ curves under different temperatures (80.25–88.80 K),

showing the same giant DEs in v0ðHdcÞ as in v0ðT Þ.

J.W. Lin et al. / Physica C 386 (2003) 89–92 91

v0ðT Þ on the higher fields. The crossover at

Hdc ¼ 10 kG therefore implies the effect of the SB.

In summary, we have prepared YBa2Cu3O7�d

sample. By means of measurements of ac suscep-

tibility, the DE was observed. It is shown that the

DE is observed in both v0ðT Þ and v0ðHdcÞ curves in

the same time for the sample. We explain the dips

in the experimental v0ðT ;HdcÞ curves as bulk pin-

ning effect in higher fields and lower temperatures.

However, the dips were weakened and finally re-

placed by SB. Our experimental data support theexplanation.

Acknowledgements

One of the authors S.Y. Ding acknowledges the

supports of the Ministry of Science and Technol-

ogy of China (G1999064602) and NNSFC undercontract no. 19994016.

References

[1] X.S. Ling et al., Physica C 282/287 (1997) 2191.

[2] J. Giapintzakis et al., Phys. Rev. B 50 (1994) 16001.

[3] S.S. Banerjee et al., Physica C 308 (1998) 25.

[4] J. Shi et al., Phys. Rev. B 60 (1999) R12593.

[5] E. Zeldov et al., Phys. Rev. Lett. 73 (1994) 1428.

[6] Fuchs et al., Phys. Rev. Lett. 22 (1998) 4971.

[7] P. Zhang et al., Phys. Rev. B 62 (2000) 5374.

[8] M. Benkrouda et al., Phys. Rev. B 58 (1998) 15103.

[9] E. Zeldov et al., Nature (London) 375 (1995) 373.

[10] X.F. Wu et al., Supercond. Sci. Technol. 15 (2002) 385.

[11] C.J. Van der Beek et al., Phys. Rev. Lett. 84 (1994) 4196.

Fig. 4. Dip bottom temperature Td in v0ðT Þ vs. dc magnetic

field. Two different slopes of the fitting lines are clear, showing

the effect of the SB.

92 J.W. Lin et al. / Physica C 386 (2003) 89–92