Anda di halaman 1dari 6

Journal of Alloys and Compounds 288 (1999) 249254

Magnetic responses of a high-T c semi-reversible YBCO superconductor


1 , U. C
S. Celebi*, A. Ozturk
evik
Department of Physics, Faculty of Science and Arts, Karadeniz Technical University 61080 Trabzon, Turkey
Received 8 January 1999

Abstract
We report the magnetic responses of a YBCO superconducting sample prepared by solid-state reaction method, in terms of d.c.
magnetization, and a.c. susceptibility. Hysteresis loops have been analyzed in the framework of critical state model including the Meissner
current. The critical current as a function of temperature and the volume fraction of the grains have been estimated from the a.c.
susceptibility data. The zero resistance temperature and diamagnetic onset temperature were found to be 90 and 91.5 K respectively.
1999 Elsevier Science S.A. All rights reserved.
Keywords: a.c. susceptibility; Critical current density; Critical state model; High-T c superconductor; YBCO

1. Introduction
Magnetization and a.c. susceptibility measurements have
played an important role in understanding of the critical
current characteristics and flux dynamics of superconductors. It is well established that the critical current density
depends on the flux pinning properties of a superconductor. The larger the pinning strength, the larger the critical
current, hence the wider the hysteresis envelope. In a broad
qualitative sense, nonideal type II superconductors can be
classified as semi-reversible and hysteretic depending the
equilibrium Meissner contribution to their magnetic behaviour is significant or negligible. The reasons for semireversible magnetic behaviour are that at low fields intergrain currents can circulate through the specimen, and also
large intragrain current densities can exist because of
strong pinning inside the grains. These irreversible induced
currents compete with the thermodynamically reversible
Meissner currents circulating at the periphery of the
individual grains [1,2].
It is standard practice in the analysis of magnetic
behaviour of type II superconductors to exploit the simple
BeanLondon approximation [3,4] which assumes j c to be
independent of B, the local magnetic flux density. The
*Corresponding author. Fax: 190-462-325-3195.
E-mail address: celebi@risc01.ktu.edu.tr (S. Celebi)
1
Permanent address: Department of Physics, Faculty of Science and
Arts, Gaziosman Pasa University, Tokat, Turkey.

model of Kim et al. [5,6] included a dependence of critical


current on B. Irie and Yamafuji [7] and Green and
Hlawiczka [8] proposed a power law model, j c 5k /B n ,
where k and n are positive constants, and where n50.5
was proposed by Yasukochi et al. [9,10]. All these models
neglected the effect of flux exclusion below the lower
critical field Hc1 . Fitz et al. [11] proposed an exponential
model which treated the critical state model with equilibrium magnetization. Clem [12] proposed a critical state
model including the potential barrier [13] at the surface
which requires extra field increments DHen (DHex ) for the
flux entry into (exit from) the sample. LeBlanc and Lorrain
[14] calculated some simple loops where the critical
current density j c , independent of B, and DH5DHex 5
DHen 5constant. Chen and Sanchez [15] developed a
phenomenological model for interpreting the magnetic
properties of hard type II superconductors which includes
the effects of the bulk pinning, the thermal equilibrium
magnetization and the surface barrier. Recently, Tochihara
et al. [16] published the calculations of the initial magnetization and full hysteresis loops within the framework
of the modified KimAnderson critical state model .
In this work, we characterize our semi-reversible YBCO
sample by a similar model to that in [15,16]. In the
calculation of kMl vs. Ha , we have used j c 5j 0 /(Ha 1H0 2
IM )n , where Ha , is the applied field, H0 is a constant, and
IM is the Meissner current which depends on the applied
field. We recall that, in SI units, when Ha 5Hc1 , the
Meissner current IM flowing along 1 m of surface and

0925-8388 / 99 / $ see front matter 1999 Elsevier Science S.A. All rights reserved.
PII: S0925-8388( 99 )00115-2

250

S. C
elebi et al. / Journal of Alloys and Compounds 288 (1999) 249 254

transverse to Ha is also equal to Hc1 . Since Hc1 is the


applied field needed to initiate first entry of flux lines into
the specimen, this quantity can be estimated by measuring
the applied field where the slope of kMl for virgin
specimen versus Ha first deviates from linear.

2. Experimental
We have fabricated the samples with the nominal
compositions Y 1 Ba 2 Cu 3 O y using the conventional solidstate reaction in air. The powders of Y 2 O 3 , BaCO 3 , CuO,
were thoroughly mixed in the appropriate amounts by a
grinding machine for 7 h. After milling, the mixed
powders were heated from room temperature to 9428C, at a
heating rate of 108C / min . The temperature was maintained for 24 h and then the material was cooled down to
room temperature for about 12 h. The reacted materials
were reground and resintered at the previous heating
process. After regrinding finely, the material was pressed
into pellets of 13 mm diameter at 375 MPa. The pellets
were heated from room temperature to 9428C, at a heating
rate of 58C / min and annealed at 9458C for 16 h and then
cooled to room temperature at a cooling rate of 18C / min.
The cooling process between 700 and 3508C was performed in oxygen. The heat treatment processes of the
sample were carried out in alumina crucibles. Dimensions
of the sample are 2.132.7310.5 mm.
The d.c. magnetization measurements were carried out
at 77 K with a commercial vibrating sample magnetometer
(VSM) upon zero field cooling (ZFC). The a.c. susceptibility measurements were done with a 7130 a.c. Susceptometer from Lake Shore with a closed cycle refrigerator at
low temperatures down to 12 K. The details of the
measuring system of the a.c. susceptibility were given in
the previous communication [17].

3. Results and discussions


Fig. 1a displays the three basic isothermal magnetization
curves measured at 77 K with the field increasing from
zero to a maximum value of 1.0 T, then decreasing to zero,
reversing in sign to a maximum negative value of 21.0 T,
and finally increasing again to 1.0 T. We call this curve a
hysteresis envelope. We show the low field part of the half
cycle in Fig. 1a, where a few traversals which just bridge
the magnetization envelope are also shown. From inspection of these magnetization curves and the values of the
applied field where the virgin curve and the curve in
hybrid region join we can estimate mo H* 36 mT, where
H* is the external field necessary for first penetration of
flux lines to the central axis of the specimen. It is
important to note that H* depends on both the size of the
specimen and on j c .
We note that in this specimen the lower envelope of the

Fig. 1. (a) Magnetization hysteresis curves for our YBCO sample


measured at 77 K. (b) Theoretical magnetization curves calculated from
the critical state model to fit the experimental data.

major hysteresis curve lies in the diamagnetic quadrant


when Ha descending in magnitude is larger than H* . This
behaviour occurs because in magnetic fields Ha $H* , the
diamagnetic moment generated by IM circulating along the
surface of the individual grains is larger in magnitude than
the paramagnetic moment produced by the flux retaining
induced currents circulating inside the volume of each
grain. It is useful at this juncture to recall that the Meissner
effect in the ideal type I and type II superconductors
manifests itself in two ways: (a) expulsion of flux upon
cooling into the superconducting state in static magnetic
fields Ha #Hc2 (T ); and, (b) expulsion of magnetic flux
when Ha is decreased below Hc2 (T ) at a fixed temperature

S. C
elebi et al. / Journal of Alloys and Compounds 288 (1999) 249 254

T ,T c (Ha ) where T c (Ha ) is the critical temperature for the


normal-superconducting transition in the magnetic field Ha .
The steep descent from diamagnetism to paramagnetism
seen in Fig. 1a as Ha is lowered below H* can be attributed
to three factors:
1. IM (Ha , T ) hence the diamagnetic moment of the grains,
is expected to decrease linearly with Ha in the range
0,Ha ,Hc1 , thereby allowing the paramagnetic moment
generated by the flux retaining induced currents to
become manifest;
2. the intragrain j c (H ) increases steeply with decreasing H
in the low field range; and,
3. the intergrain critical current (the current carrying
capacity of the Josephson junctions) is partly restored in
the low field range.
We stress, however, that the intergrain current density is
not completely restored even when Ha is returned to zero
after excursions of the magnetization to the envelopes of
the major hysteresis curve. The reason for this is that now
the links between the grains (the Josephson junctions)
bathe in the return fields of the fully magnetized grains
[18].
Fig. 1b shows the theoretical magnetization curves we
calculated. In these calculations, for simplicity and convenience, we let
IM 5 Ha

251

H 2 (r) 5 [(Ha 2 IM 1 H0 )n11 1 H* (1 2 r /R)] 1 / n11 2 H0


(6)
In our calculations we neglected the surface barrier
effects since we did not observe the straight line in the
traversal magnetization curves between the upper and the
lower branches [23]. Good agreement with observation can
be found staking Hc1 50.4H* and n50.4; p50.2 and
H0 50 (see Fig. 1b). For comparison of the effects of Hc1 ,
n, and p we presented three calculated curves in Fig. 2. In
these calculations the adjustable parameters are H* , n, p,
and Hc1 . The former is determined from the measured
magnetic hysteresis curve and the ratio Hc1 /H* , is selected
to generate semi-reversible behavior corresponding semiquantitatively with the observations, i.e. a lower envelope
for the major hysteresis curve which lies in the diamagnetic quadrant when Ha descending is $H* . The exponent
n for bulk current and p for Meissner current are determined by the best fit of the calculated and experimental
curves. The ratio of the remnant magnetization to the
maximum diamagnetic response during the initial application of Ha is smaller than the expected value obtained from
the critical state calculations. This is dominantly related to
the effective volume fractions of the grains. Furthermore,
we recall that magnetization and susceptibility data for

(1)

for Hc1 $Ha


p11
IM 5 H c1
/H pa

(2)

for Ha $ Hc1 and introduce the simple analytic function,


namely,
j c (H, T ) 5 j 0 (T ) / [(Ha 1 H0 2 IM ) /Href ] n

(3)

where the parameter Href is introduced so that the important quantity j 0 (T ) has the units of current density. We
then calculated the magnetization curves for various zones
or stages (see, for example, Refs. [16,19]) numerically. For
example, for a cylinder of radius R, with its axis parallel to
an applied field Ha , within the critical state model and
Maxwells equation, the magnetization kMl is given by
R

2
kMl 5 ]
H(r)r dr 2 Ha
R2

(4)

for Ha $H* in the field increasing case. When the slope of


the flux density profile is positive or negative we have used
the following two equations H 1 (r), H 2 (r) for H(r),
respectively:
H 1 (r) 5 [(Ha 2 IM 1 H0 )n11 2 H* (1 2 r /R)] 1 / n11 2 H0
(5)

Fig. 2. The theoretical magnetization curves calculated from the critical


state model using three sets of parameters given in the legend.

252

S. C
elebi et al. / Journal of Alloys and Compounds 288 (1999) 249 254

granular samples are complicated to analyze. The fraction


of intergranular material (matrix), size, shape and demagnetization factors of the sample and grains, anisotropy
of the current flow inside and between the grains, flux
pinning properties of the grain and matrix and volumetric
distributions of superconducting parameters and coupling
properties between grains are primary factors that influence
magnetic data.
Fig. 3 shows the a.c. susceptibility data x 9(T ) and x 0(T )
as a function of temperature measured on the ceramic
YBCO sample where f 5125 Hz and Hac 580 A / m . Since
the demagnetizing correction would cause x 95 21 for low
enough temperature, we normalized xac (T ) to the u x 9u at
the lowest temperature. We estimate the diamagnetic onset
temperature as T c 591.5 K. The zero resistance temperature was obtained from resistive measurements as 90 K.
Fig. 4 depicts the temperature dependence of j cm . The
inset displays the low critical current density range in
enlarged scale. For these values of j cm , we have used a.c.
susceptibility data taken at Hac 580 A / m(rms), f 5125 Hz.
In hysteretic type II superconductors, when the flux lines
fully penetrate the material, i.e. when Ha (Ha 5
2Hac (rms)) is equal to the first full penetration field H*,
the losses reach a maximum [20] for infinite cylinder case.
Residual trapped flux calculations [18] reveal that the
results for the square sample are identical to that of the
infinite cylinder except that the factor of 4 should be
replaced by p. According to Bean model [3], the critical

Fig. 4. Intergranular critical current density j cm , as a function of


temperature at Hac 580 A / m (rms) and f 5125 Hz.

current density at the peak temperature T p , can be written


j c (T p )5H* /aH* / ab, and where the cross-section of
the rectangular bar-shaped sample is 2a32b. Under these
approximations and using Clems model, we have estimated and displayed the intergranular critical current
density j cm for our YBCO sample in Fig. 4. Under the
above-mentioned approximation, the estimated value of the
intergranular current density is in good agreement with the
transport critical current density ( j cm 525 A / cm 2 ) obtained by means of the four-probe technique at 77 K.
In Fig. 5, we display the x 9(T ) and x 0(T ) data on our
YBCO at f 5125 Hz and Hac 5320 A / m . Different from
the case of lower Hac , for this higher field amplitude,
x 0(T ) shows two peaks, the high temperature peak is
associated with the flux (or supercurrents) penetration
inside the grains. The effective volume fraction of the
grains can be estimated by a.c. susceptibility study [21,22],
flux trapping study [18], and d.c. magnetization study [23].
For the granular superconductors, one can have the following equation for the imaginary part of the complex
susceptibility [21],

x 0 5 (1 2 fg )x 99
m

Fig. 3. Temperature dependence of a.c. susceptibility for YBCO.

(7)

where x 0 and x 99
m are the measured and extracted matrix
(intergranular) susceptibility, respectively, and fg is the
volume fraction of the grains. For crude approximation if
we use the Bean model where x 99
m (max)50.212 for infinite
cylinder, x 0(max)50.04953 then we obtain fg 50.766. If
we multiply the calculated remnant magnetization by (12
fg ) [24], then we approximately obtain the agreement
between the measured and calculated remnant d.c. magnetization value (see Fig. 1). Note that the theoretical
calculations we have presented address a specimen which

S. C
elebi et al. / Journal of Alloys and Compounds 288 (1999) 249 254

253

and extracted a.c. susceptibility data for the matrix, the


best fit for pinning strength parameter hence critical
current density was found to be j cm 5j cm0 (12T /T c )s where
s51.5, which is consistent with the irreversibility line
exponent for YBCO reported in the literature [25].

4. Summary and conclusion

Fig. 5. A.c. susceptibility versus temperature data for YBCO at Hac 5320
A / m and f 5125 Hz.

is isotropic and has idealized geometry, whereas the actual


sample is an orthorhombic containing a tightly packed
collection of highly anisotropic grains consisting of irregularly shaped platelets with nearly random orientation of
their c-axis and spanning a range of sizes around 10 mm.
It is remarkable, in our view, that we obtain such a degree
of correspondence with observations for the standard
hysteresis curves.
Fig. 6 shows the calculated (employing Bean-cylinder)

We have measured the d.c. magnetization and a.c.


susceptibility of our YBCO sample prepared by the
conventional solid-state reaction method.We have carried
out numerical calculations based on the critical state model
for magnetization hysteresis curves in order to account for
the behaviour of semi-reversible type II superconductors
where the Meissner current IM (Ha , T ) plays an important
role. We present our extension of the basic model described
in the text. A.c. susceptibility measurements have been
performed at two different a.c. field amplitudes. Employing Beans and Clems model, we determined the temperature-dependent critical current density. The volume
fraction of the grains was estimated as fg 50.766 from the
a.c. susceptibility data . The temperature dependence of the
pinning strength parameter, hence critical current density,
for the matrix fits quite well to the power law relation
j cm 5j cm0 (12T /T c )1.5 for our YBCO sample, which shows
the zero resistance temperature and diamagnetic onset
temperature as 90 and 91.5 K, respectively.

Acknowledgements
We are very grateful to Dr M.A.R. LeBlanc for helpful
discussions Dr S.K. Hasanain and Dr A. Gencer for the
measurements of d.c. magnetization and a.c. susceptibility,
respectively. This work was supported by the Research
Fund of Karadeniz Technical University, Trabzon, Turkey,
under grant contract No: 97.111.001.1.

References

Fig. 6. Intergranular a.c. susceptibility (imaginary part), open circles


show the extracted experimental data described in the text and the solid
curve is for the theoretical calculation.

[1] Bulk Materials, S. Jin (Ed.), Processing and Properties of High-T c


Superconductors, Vol. 1, World Scientific, Singapore, 1993.
[2] E.H. Brandt, Am. J. Phys. 58 (1990) 43.
[3] C.P. Bean, Rev. Mod. Phys. 36 (1964) 31.
[4] H. London, Phys. Lett. 6 (1963) 162.
[5] Y.B. Kim, C.F. Hempstead, A.R. Strnad, Phys. Rev. Lett. 9 (1962)
306.
[6] Y.B. Kim, C.F. Hempstead, A.R. Strnad, Phys. Rev. 131 (1963)
2486.
[7] F. Irie, K. Yamafuji, J. Phys. Soc. Jpn. 23 (1967) 255.
[8] I.M. Green, P. Hlawiczka, Proc. IEE 114 (1967) 1326.
[9] K. Yasukochi, T. Ogaswawra, N. Usui, H. Kolayashi, S. Ushio,
Phys. J. Soc. Jpn. 19 (1964) 1649.
[10] K. Yasukochi, T. Ogaswawra, N. Usui, H. Kolayashi, S. Ushio, J.
Phys. Soc. Jpn. 21 (1966) 80.
[11] W.A. Fietz, M.R. Beasley, W.W. Webb, Phys. Rev. 136 (1964) A355.

254
[12]
[13]
[14]
[15]
[16]
[17]

S. C
elebi et al. / Journal of Alloys and Compounds 288 (1999) 249 254

J.R. Clem, J. Appl. Phys. 50 (1979) 3518.


C.P. Bean, J.D. Livingston, Phys. Rev. Lett. 12 (1964) 14.
M.A.R. LeBlanc, J.P. Lorrain, Cryogenics 24 (1984) 143.
D.-X. Chen, A. Sanchez, Phys. Rev. B. 45 (1992) 10793.
S. Tochihara, H. Yasuoha, H. Mazaki, Physica C 295 (1998) 101.
S. Celebi, S. Nezir, A. Gencer, E. Yanmaz, M. Altunbas, J. Alloys
Compounds 255 (1997) 5.
[18] S. Celebi, M.A.R. LeBlanc, Phys. Rev. B 49 (1994) 16009.

[19]
[20]
[21]
[22]
[23]
[24]

D.-X. Chen, R.B. Goldfarb, J. Appl. Phys. 66 (1989) 2489.


J.R. Clem, Physica C 153 (1988) 50.
D.X. Chen, J. Nogues, K.V. Rao, Cryogenics 29 (1989) 800.
S. Ravi, V. Seshu Bai, Physica C 230 (1994) 51.
M.A. R LeBlanc, Cryogenics 32 (1992) 813.
S.J. Penn, N.McN. Alford, T. Button, G.R. Court, IEEE Trans. Appl.
Supercond. 3 (1993) 1157.
[25] Y. Yeshurun, A.P. Malezemoff, Phys. Rev. Lett. 60 (1988) 2202.

Anda mungkin juga menyukai