See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/268325651

The Van Der Pauw method of measurements in high-tc superconductors

Article  in  MATERIALS SCIENCE-POLAND · January 2009

CITATIONS READS

5 1,293

5 authors, including:

Mario da Luz C. A. M. dos Santos

Universidade Federal do Triangulo Mineiro (UFTM) University of São Paulo
54 PUBLICATIONS   385 CITATIONS    119 PUBLICATIONS   1,116 CITATIONS   

SEE PROFILE SEE PROFILE

Carlos Shigue Antonio Jefferson Machado

University of São Paulo University of São Paulo
80 PUBLICATIONS   424 CITATIONS    112 PUBLICATIONS   631 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Abrasive Materials View project

Laser-Heated Pedestal Growth View project

All content following this page was uploaded by Mario da Luz on 24 November 2014.

The user has requested enhancement of the downloaded file.

Materials Science-Poland, Vol. 27, No. 2, 2009

The van der Pauw method of measurements

in high-Tc superconductors

M. S. DA LUZ*, C. A. M. DOS SANTOS, C. Y. SHIGUE,

F. J. H. DE CARVALHO JR., A. J. S. MACHADO

Escola de Engenharia de Lorena - USP, P. O. Box 116, Lorena, SP 12602-810, Brazil

Details on the transport measurements of high-Tc superconductors have been presented using the van
der Pauw technique. Basic procedures to obtain good and reliable results in polycrystalline samples were
discussed. The influence of heating rates and the direction of the applied magnetic field on the results of
measurements has been examined. An unexpected nonzero transverse voltage at zero magnetic field was
observed in the vicinity of the superconducting temperature transition (Tc). Measurements in two different
magnetic field directions allowed one to calculate the symmetric and the asymmetric components of the
Hall resistance. Those components were calculated for two different superconductor systems and ana-
lyzed in the framework of the recent theory about the longitudinal and transverse voltages in high-Tc
superconductors.

1. Introduction
The measurement of electrical resistivity and the Hall effect in high-Tc supercon-
ductors (HTSC) remains a subject of great interest [1–5]. In particular, the unusual
behaviour of the Hall effect in many high-temperature superconductors and in some
conventional superconductors has become a persistent problem in understanding the
flux motion in superconductors. To measure resistivity and the Hall effect employing
standard methods, the samples need to be processed into a bar structure with two end
contacts, for passing electrical current, and two pairs of thin protruding leads on each
side of the bar, where the longitudinal and Hall voltages can be measured. In an ideal
case, the Hall contacts should be correctly placed to avoid longitudinal contributions
to the Hall voltage. Since the superconductor samples are usually very small and have
no regular geometry, the fabrication of such contacts is more difficult.
In a normal state, the Hall effect in superconductors is antisymmetric in an exter-
nal magnetic field, and should be zero at zero magnetic field. However, for real sam-
__________
*
Corresponding author, e-mail: [email protected]
570 M. S. DA LUZ et al.

ples we usually find a finite voltage at zero field which can superimpose on the Hall
signal when a magnetic field is applied. This undesired contribution can result from
either a geometrical mismatch in voltage contacts, a misalignment of the magnetic
field and current directions, or the inhomogeneity of the sample. Several methods are
used to eliminate or reduce that undesired contribution. One of them is the technique
elaborated by van der Pauw (vdP) [6, 7]. The van der Pauw technique enables the
measurement of resistivities on thin samples with arbitrary shapes. In contrast to Hall-
bar measurements, the vdP technique requires only four sufficiently small contacts
placed at the circumference of an arbitrarily shaped (but simply connected) sample
[6, 7]. The methods of reversing magnetic field and without reversing its direction
may be used.
The method of reversing magnetic field is based on the fact that the Hall potential
changes its sign (but not the absolute value) after reversing the magnetic field while
the longitudinal potential remains the same. The non-Hall potentials can be thus
eliminated after two measurements in both field directions (H, –H). Here, we have to
keep in mind that the polarity of the injected current should be kept the same. So, the
longitudinal and Hall voltages can be calculated by:

⎛V
VXY = ⎜ AC , BD
( H ) − VAC , BD ( − H ) ⎞ (1)
⎟
⎝ 2 ⎠

⎛V
VXX = ⎜ AC , BD
( H ) + VAC , BD ( − H ) ⎞ (2)
⎟
⎝ 2 ⎠
where the VAC,BD is defined as the voltage between points A and C when an applied
current flows through contacts B and D (see Fig. 1).

Fig. 1. Schematic view of the electrical

contacts in the sample

In the method without reversing the applied magnetic field, the voltage due to
the transversal electric field can be detected by permutation of voltage and current
contacts (see Fig. 2). Any pair of contacts can be used as current contacts and the
remaining pair as voltage contacts. From the reciprocity theorem it follows that
VAC,BD(–H) = VBD,AC(H) [8, 9]. Thus, the Hall and longitudinal voltages can now be
reformulated as:
 The van der Pauw method of measurements in high-Tc superconductors 571

⎛V
VXY = ⎜ AC , BD
( H ) − VBD, AC ( − H ) ⎞ (3)
⎟
⎝ 2 ⎠

⎛V
VXX = ⎜ AC , BD
( H ) + VCD, AB ( − H ) ⎞ (4)
⎟
⎝ 2 ⎠

Fig. 2. Scheme of a four probe configuration used for the longitudinal (left) and Hall
voltages (right) measurements. The solid and dashed arrows represent current
and potential pairs of the contacts, respectively. In the top, the two configurations denote
VAB,CD and VAC,BD and the other two below VAB,CD and VBD,AC voltages, respectively

Based on the methods described above, this work reports an important dc tech-
nique which uses the van der Pauw method for measurements HTSC polycrystalline
sample. Firstly, a modified method has been presented. Basic procedures to obtain
good and reliable results have been discussed. The influence of heating rates and of
the direction of the applied magnetic field is described. The results are analyzed in the
framework of the recent theory of the longitudinal and transverse voltages in high-Tc
superconductors.

2. Experimental

Polycrystalline samples of YBa2Cu3O7–δ were prepared by a solid state reaction.

Samples of the Bi2Sr2Ca1–XPrXCu2O8+δ superconducting system were also prepared
with x = 0.2 and 0.3. X-ray powder diffractometry showed that all the samples are
572 M. S. DA LUZ et al.

single phase. For more details about sample preparation and characterization see the
references [10, 11].
In order to obtain a good voltage signal, the sample was mechanically polished un-
til very thin square prism had been obtained. The contacts were placed in the corners
as shown schematically in the Fig. 1. The electrical terminals were prepared using low
resistance sputtered Au contacts (~ 0.1 Ω). Magneto-transport properties were studied
using a Maglab Oxford system of 9T. The magnetic field was applied perpendicular to
the sample surface. A constant current was applied using a programmable current
source Keithley model 220, and the corresponding voltage was measured using
a nanovoltmeter Keithley model 181. VXX and VXY were measured by means of the van
der Pauw technique, with permutation of the voltage and current contacts. The con-
tacts were appropriately switched using a computer-controlled relay circuit (Fig. 3). In
this method, the Hall and longitudinal voltages were measured simultaneously at each
temperature and each applied magnetic field. Data acquisition and calculation of VXX
and VXY were done in real time using the LabVIEW software [12].

Fig. 3. Measurement setup for the modified switching contacts exploiting

the van der Pauw technique. The contacts are switched using four different relays

Noise (thermopower effects, etc.) was eliminated by reversing the transport cur-
rent polarity (I) in the samples under measurement. Thus, to obtain longitudinal (VXX)
and transverse (VXY) voltages, eight voltage signals were measured. The VXX and VXY
components were calculated based on the combination of eight records, which can be
expressed as

VXX =
(V AB , CD ( I ) − VAB,CD ( − I ) ) (VBC , AD ( I ) − VBC , AD ( − I ) )
(5)
4
and
 The van der Pauw method of measurements in high-Tc superconductors 573

VXY =
(V AC , BD ( I ) − VAC , BD ( − I ) ) (VBD , AC ( I ) − VBD , AC ( − I ) )
(6)
4
By the method without field reversal, all voltages may be detected due to the
transversal electric field. However, using Eq. (6), we can only partially compensate for
those voltages added to the Hall effect. In this case we should also revert the magnetic
field to separate the asymmetric and symmetric part of the Hall effect.

3. Results and discussion

Separating the Hall effect signal from its resistive offset signal is hard enough un-
der normal circumstances, but it becomes even more challenging when the resistivity
is drifting rapidly, as happens in materials transforming from metallic to superconduc-
tor behaviour. An important prerequisite for the validity of Eqs. (5) and (6) is that the
measurements of the eight components are made under the same conditions, especially
at the same temperature. Thus, the van der Pauw measurement of high-Tc supercon-
ductors requires consideration of some additional factors such as the influence of heat-
ing conditions of the sample during the measurement.

Fig. 4. Longitudinal resistance (upper curves) and transverse resistance of YBa2Cu3O7–δ

(lower curves) in function of the temperature measured for the various heating rates
at B = 0, and I = 30 mA

We have performed a number of tests focused on the temperature stabilization of

the sample. Figure 4 presents longitudinal (RXX) and transverse (RXY) resistances in
function of temperature for the sample of YBa2Cu3O7–δ measured at various heating
rates under zero applied magnetic field. Measurements were taken during heating at
574 M. S. DA LUZ et al.

the rates of 0.2 K/min and 2 K/min. At 2.0 K/min, the onset of the superconducting
transition is far from the well-reported critical temperature for YBa2Cu3O7–δ
(Tci ≈ 90 K). However, at 0.2 K/min, Tci is ca. 89 K which suggests that the measure-
ment was performed under thermal equilibrium. Note that RXX is essentially independ-
ent of the heating rate but the longitudinal resistance moves towards lower tempera-
tures for higher heating rates. However, in the RXY signal, both the intensity and the
temperature dependence are altered with the change in the heating rate. The peak in-
tensity (positive) at Tc increases while the second peak intensity (negative), at lower
temperatures, decreases as the heating rate increases. Also both peaks move towards
lower temperatures at 2 K/min. Based on these results, we conclude that lower heating
rates are crucial for reliable transport measurements in superconductor materials, es-
pecially for Hall effect measurements.
The fact of occurrence of the Hall resistance at zero magnetic field is worth not-
ing. In the mixed state, the Hall resistance in superconductors results from hydrody-
namics vortices produced by the current flowing through the sample in an external
magnetic field [1–5]. Then, in zero magnetic field we should not observe the appear-
ance of a transverse voltage. However, transverse voltages in zero magnetic field have
been observed in various HTSC materials [1, 5, 13–17]. Near Tc, a free vortex in the
mixed state was observed which can be generated without the application of an exter-
nal magnetic field. In this case, vortex–antivortex pairs may be excited by thermal
fluctuations in the superconducting state, or induced by transport current passing
through the sample on the opposite sides of the sample [13–17]. From Fig. 4, we can
clearly see nonzero longitudinal transverse voltage close to Tc at zero magnetic field.
The VXY curves are non-zero close to the critical temperature with the maximum value
approximately in the middle of the longitudinal superconducting transition, which has
been reported in the literature [14, 15]. Moreover, a sign change from positive to nega-
tive is observed as the temperature decreases. In a true superconducting state (R = 0 in
the longitudinal resistance) and in the normal state (T > Tc), no finite transversal vol-
tage was detected. These results can be explained based on induced vortex–antivortex
pairs applying the guiding vortex model [15]. The theory of the guiding vortex was
proposed to explain the even effect in superconductors which assumes the existence of
a new force acting on the vortex. This force (the guiding force) impels the vortex to
move only in a given direction that is determined by the direction of the pinning po-
tential valley. However, the nature of this pinning potential has not been completely
solved. One of the promising models is the intrinsic pinning model, which assumes
that the origin of the guiding forces is due to the layered structure of the HTSC [18].
Other mechanisms such as grain boundary guiding in polycrystalline materials for
example, should be kept in mind. This model seems to offer a consistent explanation
of our results.
In general, transverse voltage consists of three components: a component coming
from the geometrical misalignment of the contacts, a Hall voltage component, and
a component originating from the guided motion of vortex and antivortex. The geo-
 The van der Pauw method of measurements in high-Tc superconductors 575

metrical misalignment can be detected by normal state measurements in zero magnetic

field. In our case, in the normal state, no transverse voltage was detected at zero ap-
plied magnetic field. Therefore, the transverse voltage in zero magnetic field cannot
originate from a geometrical misalignment of the contacts. The other two components
can be distinguished by comparing the voltages measured in two opposite magnetic
field directions: while in the Hall voltage, the sign of the measured voltage depends on
the applied magnetic field direction because the vortex motion is only governed by the
Lorentz force: in the case of the guided vortex–antivortex motion, it is independent of
the field direction. This difference gives us the chance to distinguish between genera-
tion mechanisms of these two voltages. Figure 5a shows the transverse voltage in
function of the temperature measured in the same sample that was presented in Fig. 4
under magnetic field in positive (H = 9 T) and negative (H = –9 T) directions.

Fig. 5. Transverse resistance in function of the temperature measured

in the sample of YBa2Cu3O7–δ at the applied magnetic field in positive (H = 9 T)
and negative (H = –9 T) directions (a); and symmetric
and asymmetric transverse resistances calculated from the data of Fig. 4a (b)

From Figure 5a we can observe two distinct behaviours: above Tc (ca. 90 K), RXY
is dependent on the direction of the applied magnetic field; below Tc, the transverse
resistance is independent of its direction. Using RXY measurements in both directions
of the magnetic field, we can calculate the symmetric (RXY S
) and asymmetric (RXY A
)
resistance components. To obtain the symmetric transverse resistance which is due to
the applied magnetic field, we should calculate RXY S
= (RXY(H) – RXY(–H))/2, where
RXY(H) and RXY(–H) are measured under magnetic field applied in positive and nega-
tive directions, respectively. The asymmetric transverse resistance was calculated as
A
RXY = (RXY(H) + RXY(–H))/2. The results for both components are presented in Fig. 5b.
In the normal state (T > Tc), the symmetric component presents a positive signal, as
576 M. S. DA LUZ et al.

observed for several superconductor materials. In the superconducting state (T < Tc)
no transverse voltage was detected for this value of magnetic field. Thus, we can say
that this component presents the true Hall resistance effect. The inset (Fig. 5b) shows
the asymmetric component which can be due to guided motion of vortex and anti-
vortex, similar to Hall measurements in zero applied magnetic field.

Fig. 6. RXY
S
in function of temperature for the same sample YBa2Cu3O7–δ for various magnetic
fields. The inset shows the transverse resistance in function of magnetic field at ca. 100 K

In Figure 6, we show other results for RXY

S
in function of temperature for the sam-
ple YBa2Cu3O7+δ for various magnetic fields. In the normal state (see inset), the trans-
verse voltage is positive and increases linearly as the strength of the applied magnetic
field increases, as is expected of hole-doped high-Tc superconductors [19, 20]. The
resistance changes from positive to negative below Tc for low and moderate magnetic
fields. Higher fields suppress this sign reversal and restore RXY > 0 near Tc, while at
low temperatures RXY tends to zero. There are several models to explain this effect, but
common agreement has not been achieved up to now [3, 5]. In addition to the sign
reversal near Tc, the RXYS
measured for the sample Bi2Sr2Ca0.8Pr0.2Cu3O8+δ suggest the
occurrence of a second sign reversal which has been observed in most high-Tc super-
conductors (see inset of Fig. 7) [21, 22]. The error is about 5%. Figure 7 shows longi-
tudinal resistance (upper) and symmetric transverse resistance (lower), calculated from
the results of the van der Pauw measurements at 5 T. In the RXX we can see a broad
superconducting transition which can be related to granular superconductivity. The
S
RXY is positive in the normal state and changes to negative around Tc. These sign re-
versals in the transverse component are believed to be related to the quasiparticle or
vortex-core contributions which are associated with the normal state excitation, super-
 The van der Pauw method of measurements in high-Tc superconductors 577

conducting dissipation resulting from vortex hydrodynamics and superconducting

fluctuations [22].

Fig. 7. Longitudinal resistance (upper) and symmetric transverse (lower) resistances

calculated from the results of the van der Pauw measurements at 5 T, for the sample of
Bi2Sr2Ca0.8Pr0.2Cu3O8+δ. In the inset, the results suggest a double sign reversal
of the RSXY in the vicinity of Tc

Fig. 8. Longitudinal resistance (upper) and symmetric transverse (lower)

resistances calculated from results of the van der Pauw measurements at 3 T,
for the sample of Bi2Sr2Ca0.7Pr0.3Cu3O8+δ

Finally, the longitudinal resistance and symmetric transverse resistance, calculated

from results of the van der Pauw measurements at 3 T for Bi2Sr2Ca0.7Pr0.3Cu3O8+δ are
shown in Fig. 8. We have noted a very wide superconducting transition in the tem-
perature dependence of RXX. In contrast with that observed for Bi2Sr2Ca0.8Pr0.2Cu3O8+δ
578 M. S. DA LUZ et al.

(see Fig. 7), no sign inversion was observed in the transverse component. This differ-
ence in behaviour probably reflects the difference in the flux dynamics and the nature
of the pinning mechanism between these two superconductor materials. Similar behav-
iour is observed in some of the conventional high-Tc superconductors in the mixed
state, but is not fully understood yet [23].

4. Conclusion
A modified van der Pauw technique has been presented to measure transport prop-
erties of HTSC superconductors. It was observed that a low heating rate is crucial for
obtaining reliable transport measurements in high-Tc superconducting materials. It is
necessary to carry out measurements in two opposite magnetic field directions in order
to separate the transverse resistance components. We believe that this technique makes
real-time Hall and longitudinal measurements practical in undergoing superconducting
transitions.

Acknowledgements

This work has been supported by the FAPESP (97/11113-6, 00/03610-4) and CNPq.

References

[1] DA LUZ M.S., DE CARVALHO F.J.H. JRDOS SANTOS C.A.M., SHIGUE C.Y., MACHADO A.J.S., Physica
C, 419 (2005), 71, and references therein.
[2] JANEČEK I., VAŠEK P., Physica C, 468 (2008 ), 17.
[3] RICHTER H., PUICA I., LANG W., PERUZZI L.M., DURRELL J.H., STURM H., PEDARNIG J.D., BÄUERLE D.,
Phys. Rev. B, 73 (2006), 184506.
[4] WÖRDENWEBER R., SANKARRAJ J.S.K., DYMASHEVSKI P., HOLLMANN E., Physica C, 434 (2006) 101.
[5] PUICA I., LANG W., GÖB W., Phys. Rev. B, 69 (2004), 104513.
[6] VAN DER PAUW L.J., Philips Res. Rep., 13 (1958), 1.
[7] VAN DER PAUW L.J., Philips Tech. Rev., 20 (1959), 220.
[8] JANEČEK I., VAŠEK P., Physica C, 390 (2003), 330.
[9] JANEČEK I., VAŠEK P., Physica C, 402 (2004), 199.
[10] DOS SANTOS C.A.M., MACHADO A.J.S., Physica C, 354 (2001), 213.
[11] DOS SANTOS C.A.M., MOEHLECKE S., KOPELEVIC Y., MACHADO A.J.S., Physica C, 390 (2003), 21.
[12] National Instruments LabVIEW Graphical Programming for Instrumentation, Austin, TX, Version
5.0, 1998.
[13] VAŠEK P., Supercond. Sci. Technol., 20 (2007), 67.
[14] FRANCAVILLA T.L., HEIN R.A., IEEE Trans. Mag., 27 (1991), 1039.
[15] VAŠEK P., SHIMAKAGE H., WANG Z., Physica C, 411 (2004), 164.
[16] FRANCAVILLA T.L., CUKAUSKAS E.J., ALLEN L.H., BROUSSARD P.R., IEEE Appl. Sup., 5 (1995),
1717.
[17] VAŠEK P., Physica C, 364–365 (2001), 194.
 The van der Pauw method of measurements in high-Tc superconductors 579

[18] VAŠEK P., JANEČEK I., PLECHÁČEK V., Physica C, 247 (1995), 381.
[19] CHIEN T.R., JING T.W., ONG N.P., WANG Z.Z., Phys. Rev. Lett., 66 (1991), 3075.
[20] MALINOWSKI A., CIEPLAK M.Z., GUHA S., WU Q., KIM B., KRICKSER A., PERALI A., KARPINSKA K.,
BERKOWSKI M., SHANG C.H., LINDENFELD P., Phys. Rev. B, 66 (2002), 104512, and references therein.
[21] LANG W., GÖB W., PEDARNIG J.D., RÖSSLER R., BÄUERLE D., Physica C, 364-365 (2001), 518.
[22] GÖB W., LIEBICH W., LANG W., PUICA I., SOBOLEWSKI R., RÖSSLER R., PEDARNIG J.D., BÄUERLE D.,
Phys. Rev. B, 62 (2000), 9780.
[23] IYE Y., NAKAMURA S., TAMEGAI T., Physica C, 159 (1989), 616.

Received 2 August 2008

Revised 28 October 2008

View publication stats