Journal of Alloys and Compounds 658 (2016) 265e271

Contents lists available at ScienceDirect

Journal of Alloys and Compounds

journal homepage: http://www.elsevier.com/locate/jalcom

Anomalous Hall effect and magnetoresistance in Ge1
x
yPbxMnyTe

cluster-glass system
A. Podgo rni a, *, L. Kilanski a, M. Go
 rska a, R. Szymczak a, A. Reszka a, V. Domukhovski a,
B.J. Kowalski a, B. Brodowska a, W. Dobrowolski a, V.E. Slynko b, E.I. Slynko b
a
Institute of Physics, Polish Academy of Sciences, al. Lotnikow 32/46, 02-668 Warsaw, Poland
b
Institute of Materials Science Problems, Ukrainian Academy of Sciences, 5 Wilde Street, 274001 Chernovtsy, Ukraine

a r t i c l e i n f o a b s t r a c t

Article history: We present studies of magnetotransport properties of Ge1
x
yPbxMnyTe bulk crystals with chemical
Received 14 September 2015 composition changing from 0.180 < x < 0.311 to 0.019 < y < 0.136. Our samples show the presence of the
Received in revised form cluster-glass-like transition accompanied with magnetic irreversibility at temperatures TSG around 90 K.
15 October 2015
The presence of strong anomalous Hall effect (AHE) with hysteresis is observed at T < TSG. The data
Accepted 16 October 2015
Available online 20 October 2015
analysis allowed calculation of a correction to the Hall constant RH for T < TSG and the AHE constant, RS. RS
is around 10
6 m3/C. RS and RH do not change strongly with temperature. The scaling law analysis has
proved that the AHE in this system is due to the extrinsic mechanisms dominated by the skew scattering
Keywords:
Semimagnetic-semiconductors
with relatively small contribution of the side jump process. Strong localization occurs in our system at
Ferromagnetic-materials T < TSG and causes a presence of a minimum in the resistivity vs. temperature dependence. The
Cluster-defects magnetoresistance of our samples shows complex behavior at T < TSG. The negative magnetoresistance is
observed at T < 20 K and interpreted as spin-dependent scattering of conducting carriers on localized
magnetic moments. At higher T nearly linear positive magnetoresistance is observed at temperatures
from 20 K up to T z TSG, where it reaches maximum. Positive linear magnetoresistance is related to the
granular nature of our samples while the maximum near TSG is related to the giant spin-splitting of the
valence band.
© 2015 Elsevier B.V. All rights reserved.

1. Introduction practical applications.

Transition metal (TM) doped semiconductors are being inten-
sively studied in view of their potential applications in spintronics
for the last two decades. Carrier induced magnetic interactions are
responsible for magnetic order in many conventional IIIeV, IVeVI,
and IIeVI compound semiconductors such as TM doped PbSnTe,
GaAs, and other systems [1e3]. Room temperature magnetic
ordering due to the presence of carrier mediated magnetic in-
teractions is necessary for making use of diluted magnetic semi-
conductors in semiconductor spintronics. However, the Curie
temperature of the most intensively studied and technologically
mastered semimagnetic semiconductor Ga1
xMnxAs does not
exceed 185 K [4], which excludes the practical application of this
material. It is therefore necessary to look for alternative compounds
that can operate at room temperature opening the path for
* Corresponding author.
rni).
E-mail address: podgorni@ifpan.edu.pl (A. Podgo
IVeVI compound semiconductors alloyed with transition metal
ions, in particular the Ge1
xTMxTe alloys, show carrier mediated
ferromagnetism with high Curie temperatures reaching 200 K in
Ge1
xMnxTe with x ¼ 0.46 [5]. High Curie temperatures in
Ge1
xMnxTe alloys are due to the large values of the Mn-ion con-
ducting hole magnetic exchange constant Jpd z 0.8 and large carrier
concentration p > 1020 cm
3 and caused large interest in these
materials [6e8]. Moreover, IVeVI-based materials are recently
perceived as important representatives of 3D topological insulators.
In addition, Pb-based IVeVI crystals have the highest mobilities in
this group of semiconductors. Alloying topological insulators with
TM ions has been recently intensively studied in Bi2Te3 layers with
Mn ions [9] and similar interest is observed for IVeVI materials.
Ge1
xTMxTe alloys possess another important feature; are called
multiferroic materials due to the simultaneous presence of ferro-
electric and ferromagnetic order at T < 200 K.
The present paper extends our previous investigation of mag-
netic and transport properties of Ge1
x
yPbxMnyTe crystals into
detailed analysis and interpretation of the magnetotransport

http://dx.doi.org/10.1016/j.jallcom.2015.10.141
0925-8388/© 2015 Elsevier B.V. All rights reserved.
266 rni et al. / Journal of Alloys and Compounds 658 (2016) 265e271
A. Podgo

properties with respect to the samples having a broad range of both

Pb and Mn contents [10,11]. The detailed analysis and interpreta-
tion of the anomalous Hall effect and both negative and positive
magnetoresistance effects will be presented allowing the deter-
mination of the main physical mechanism responsible for the
observed effects.

2. Sample characterization

For the purpose of this work a series of bulk Ge1
x
yPbxMnyTe

crystals was grown by the modified Bridgman method. The modi-
fications applied to the growth technique are similar to those used
by Aust and Chalmers in order to improve quality of alumina
crystals [12]. A more detailed description of the growth procedure
was presented in Refs. [10] and [11].
The chemical composition of our crystals was determined by the
energy dispersive X-ray fluorescence technique (EDXRF). We used
the Tracor Spectrace 5000 EDXRF spectrometer. The maximum
relative error of the EDXRF technique does not exceed 10% of the
calculated value of x or y. The as-grown ingots were cut into thin
slices (about 1 mm thick) perpendicular to the growth direction,
mechanically polished, chemically cleaned, and etched prior to
further characterization. Each slice was characterized with the Fig. 1. Results of the SEM/EDS measurements for Ge0.570Pb0.294Mn0.136Te (the upper
EDXRF method. The obtained results indicate that the chemical left part of the figure) and Ge0.743Pb0.196Mn0.061Te (the lower left part of the figure)
composition of the slices changes continuously along the growth samples showing typical shapes of two types of magnetic clusters present in
Ge1
x
yPbxMnyTe crystals forming either stripes or dots. The EDX spectra were
direction. From all our samples we selected the ones with the
measured at selected spots on the sample surface (marked with squares). The char-
chemical composition showing the smallest changes in each indi- acteristic x-ray emission lines related to transitions to La, Ma, and Ka shells of Ge, Pb,
vidual slice and covering the widest range of x and y values. The Mn and Te elements are marked with labels.
chemical content of the samples selected for the current studies
vary in the range of 0.180 < x < 0.311 and 0.019 < y < 0.136. glass-like transition (detailed magnetic studies can be found in
The structural quality of our samples was determined with the Refs. [10] and [11]) with the transition temperature, TSG, changing
use of the standard X-ray powder diffraction (XRD) method. The between samples from 44 K to 93 K (see legend to Fig. 2), and a
main conclusions are that all our samples are two-phased con- second transition at temperatures lower than 20 K.
sisting of the two NaCl structures, one cubic phase and the other
phase showing rhombohedral distortion. A more detailed analysis
of the XRD data can be found in Refs. [10] and [11]. 3. Magnetotransport properties
Morphology of our Ge1
x
yPbxMnyTe samples was studied with
the use of high resolution Hitachi SU-70 Analytical UHR FE-SEM Electrical properties of the Ge1
x
yPbxMnyTe crystals were
scanning electron microscope (SEM) coupled with Thermo Fisher
NSS 312 energy dispersive X-ray spectrometer system (EDS)
equipped with SDD-type detector. Sample surface was polished
prior to the SEM/EDS measurements. A series of high resolution
images of the crystal surface was done at different sample spots and
magnifications. The measurements reveal the presence of
micrometer-sized inhomogeneities in all our samples (see Fig. 1).
There exist two types of clusters forming either spherical- or stripe-
shaped geometrical forms. It might therefore have significant in-
fluence on the magnetotransport properties of the material. The
detailed chemical content analysis done with the use of the EDS
microprobe revealed a large difference of the chemical content
between the matrix and the clusters. One can clearly see that the
amount of Pb in the clusters is much higher than in the matrix.
Hence it was proven that the chemical content has an influence on
the magnetic exchange constant and is responsible for the two
observed magnetic transitions [10,11].
Magnetic properties of our Ge1
x
yPbxMnyTe were studied
extensively before and the details can be found in Refs. [10] and
[11]. The temperature dependence of the magnetization M with the
sample being cooled down in the presence (FC) or the absence
(ZFC) of the applied magnetic field were measured with the use of
Quantum Design MPMS-XL5 magnetometer system. The tempera-
ture dependence of the ac susceptibility, c, was measured using the Fig. 2. Results of the temperature dependent: (i) zero-field-cooled (ZFC) and field-
cooled (FC) magnetization M(T) and (ii) the inverse of the real part of the ac mag-
LakeShore 7229 susceptometer system. The exemplary results of netic susceptibility Re(c)
1 (T) (experiment e markers, lines e fits of the full markers
our measurements are presented in Fig. 2. The magnetization and with the CurieeWeiss law) measurements for selected Ge1
x
yPbxMnyTe samples with
susceptibility data indicate that all our samples exhibit a cluster- different chemical compositions.
 rni et al. / Journal of Alloys and Compounds 658 (2016) 265e271
A. Podgo
studied by means of temperature dependent and field dependent
magnetotransport measurements. The magnetoresistance and the
Hall effect were measured simultaneously at selected temperatures
with the use of the six contact dc current technique. We have used
the superconducting magnet with the maximum magnetic field
B ¼ 13 T and the sweep speed of about 0.5 T/min, equipped with the
cryostat allowing the control of the temperature of the sample in
the range of 1.4  T  300 K. The samples, cut to size of about
1  1  10 mm, were etched and cleaned before making electrical
contacts. The contacts were made with the use of gold wire and
indium solder. The ohmic behavior of each contact pair was
checked prior to the measurements.
3.1. Anomalous Hall effect
The magnetic field dependence of the resistivity component
perpendicular to the current and magnetic field direction, ryx(B),
was measured at several stabilized temperatures below, near, and
above the cluster-glass-like transition temperatures TSG (values
gathered in the legend to Fig. 2) in all our samples. Our results
indicate clearly that, for all Ge1
x
yPbxMnyTe crystals below TSG, the
ryx(B) dependencies show large anomalous Hall effect (Fig. 3b) and
the presence of hysteresis in ryx(B) curves (Fig. 3c). The selected
ryx(B) curves visible in Fig. 3b show that the AHE makes a signifi-
cant contribution to the total Hall effect in our Ge1
x
yPbxMnyTe
samples at T < TSG. The analysis of the magnetometric and the Hall
effect hysteresis loops shows that the coercive field, HC, values
obtained by using the two types of measurements coincide with a
good accuracy and point into similar values of TSG. It is a signature
that the asymmetric carrier scattering occurs in our cluster-glass-
like system and can be directly linked to the magnetic properties
of the Ge1
x
yPbxMnyTe alloy.
The typical analysis of the Hall effect in a nonmagnetic
conductor assumes that the magnetic field dependence of the off-
diagonal resistivity component, ryx(B) ¼ RH$B, i.e. the Hall effect is
Fig. 3. Results of magnetometric and magnetotransport measurements obtained for
the selected Ge1
x
yPbxMnyTe samples with different chemical compositions including
(a) the magnetic field dependence of the magnetization necessary for the proper AHE
analysis, (bec) magnetic field dependence of the off-diagonal resistivity component ryx
shown at high fields (b) and low-fields (c), (d) the temperature dependence of the
coercive field, HC, determined with the use of magnetization (open symbols) and Hall
effect (full symbols) hysteresis loop data.
267
a linear function of B. The temperature dependence of the Hall
constant, RH, calculated from the dependence of ryx on B is plotted
in Fig. 4. As we can see the RH does not show significant changes
with the temperature at T > TSG and increases with decreasing
temperature at T  TSG. The carrier transport in our samples is not
thermally activated and thus we do not expect any significant
temperature dependence of the Hall constant. It is then evident
that the increase of the RH(T) dependence at T  TSG is an artifact of
the data analysis. In order to determine the strength of the AHE and
to properly estimate the Hall carrier concentration and mobility for
T  TSG an appropriate fitting procedure must be employed.
The Hall effect in a material doped with magnetic ions shows
the usual Lorentz term RHB and a second contribution, namely AHE
term RSM, caused by the asymmetric carrier scattering. The AHE is
due to the spineorbit coupling in the presence of spin-polarization
(for details see Ref. [13] and references therein). The AHE term in
some cases dominates the total Hall effect below the temperature
of the magnetic transition, thus making the precise estimation of
the carrier concentration and mobility very difficult. The magnetic
field dependence of the Hall resistivity tensor component, ryx,
measured in the standard six contact Hall geometry can be
expressed using the following relation:
ryx ¼ RH B þ m0 RS M; (1)
where m0 is the magnetic permeability constant. Both the ordinary,
RH, and anomalous, RS, Hall coefficients can be extracted from the
total Hall effect with the knowledge about the magnetic field
dependence of the magnetization, namely M(B) curves at given
temperatures. The use of the M(B) curve (example in Fig. 3a) is
crucial, especially for a system in which the magnetization does not
show saturation even at relatively high fields B ¼ 9 T, which is the
case for our samples. In such a system the AHE term gives a
contribution that is not constant as a function of the magnetic field
over the entire measurement range. Therefore, both the ordinary
and the anomalous terms of the Hall effect affect the ryx(B)
dependence at high fields and cause it to be an increasing function
Fig. 4. The ordinary and anomalous Hall constants, RH and RS, respectively, as a
function of temperature for the selected Ge1
x
yPbxMnyTe samples with different
chemical compositions. The Hall constant RH is plotted without (open symbols) and
with (closed symbols) the AHE contribution included in the analysis of the Hall effect
data.
268 rni et al. / Journal of Alloys and Compounds 658 (2016) 265e271
A. Podgo
of the applied magnetic field. Thus, a complex fitting procedure
needs to be employed in order to properly determine the Hall effect
parameters i.e. to calculate precisely the RH and RS at low temper-
atures T < TSG.
The proper determination of the strength of the observed AHE
and the calculation of the Hall constant and carrier mobility re-
quires employing a fitting procedure of the experimental data to Eq.
(1). We performed
coefficient was estimated for each sample and measurement tem-
perature. The nH values obtained for our samples change from 1.06
to 1.18. The values of nH indicate that the AHE in our
Ge1
x
yPbxMnyTe samples is mainly due to the extrinsic skew
scattering process. Since nH > 1 the other scattering mechanisms
giving a small contribution to AHE are also present in our system.
The scaling analysis of our data indicates that extrinsic skew scat-
 a series of least square fits of the experimental tering is dominant AHE mechanism in our cluster-glass-like
ryx(B,M(B))T¼const dependence to Eq. (1) with the use of the Minuit
functional minimization package [14]. Fitting procedure was
divided into two steps. As a first step both RH and RS were taken as
the fitting parameters. For the second series of fits we assumed a
constant value of RH (averaged value calculated using the values
obtained during the first series of fits). This is a reasonable result,
since our samples do not show thermally activated carrier trans-
port. During the second series of fits only the anomalous Hall
constant RS was taken as a fitting parameter. The obtained values of
RS as a function of temperature for the selected Ge1
x
yPbxMnyTe
samples with different chemical compositions (see legend) are
plotted in Fig. 4.
The values of RS obtained in this work are of the same order of
magnitude as the values reported for other IVeVI based diluted
magnetic semiconductors such as Sn1
x
yMnxEryTe,
Ge1
x
yMnxEuyTe and Ge1
x
ySnxMnyTe [15,17,16]. We found no
trace of any trend between the Mn content and the value of the RS.
A small minimum in RS(T) is observed at T z 10 K, the effect most
probably associated with the low temperature magnetic transition
in our samples. Besides that effect there seem to be no significant
changes in RS(T) functions for all our samples. The absence of RS(T)
dependence is typical for IVeVI semiconductors [18,16]. The
physical mechanisms responsible for the observed AHE in our
samples are yet unknown and the proper scaling analysis of the
experimental data must be employed in order to shed light on the
physical origin of the AHE in this material.
Physical mechanisms responsible for the AHE can be divided
into intrinsic and extrinsic mechanisms. It is a fact known in the
literature, that there are two major extrinsic mechanisms leading to
the formation of AHE, namely skew scattering and side jump, which
have been described theoretically and distinguished by appropriate
linear [19] and square [20] dependencies between the diagonal and
off-diagonal resistivity components, using the relation RS f rnxxH ,
where nH ¼ 1 or 2, respectively. There also exists an intrinsic
mechanism that can be responsible for the AHE. This mechanism is
related to the topological effect occurring in the semiconductor.
This intrinsic AHE mechanism was described theoretically based on
the Berry phase theory and was used to describe the AHE in
Ga1
xMnxAs with a metallic type of conductivity [21]. The topo-
logical explanation of the AHE was also employed theoretically for
IVeVI semiconductors [22,23]. The Berry phase theory predicts the
square scaling dependence RS f r2xx .
The scaling analysis of the AHE in our samples was based on the
fitting of the experimental data using the following equation
Ge1
x
yPbxMnyTe samples. On the other hand, for ferromagnetic
Ge1
xMnxTe and Ge1
xCrxTe layers the side jump mechanism of
AHE becomes more pronounced [24,25]. The extrinsic scattering
mechanism becomes more pronounced in a system with strong
magnetic disorder and low electrical resistivity [26]. The strong sp-
d exchange interaction and low carrier mean free path can intro-
duce a contribution from side-jump mechanism to the system. The
nH values obtained for our Ge1
x
yPbxMnyTe samples are generally
smaller than the ones obtained for Ge1
x
ySnxMnyTe [17] and
Ge1
x
yMnxEuyTe [16] crystals. The nH values from Refs. [17] and
[16] change from 1.1 to 1.3. Therefore, it seems that for the Pb-
alloyed GeTe crystals the extrinsic skew scattering process is a
more pronounced physical mechanism responsible for the AHE
than in the tho other representatives of IVeVI semimagnetic
semiconductors.
3.2. Resistance and magnetoresistance studies
The transport studies of our Ge1
x
yPbxMnyTe samples included
the measurements of the temperature dependence of the resistivity
parallel to the current direction, rxx, in the absence of an external
magnetic field. The results of our measurements are presented in
Fig. 5 and gathered in Table 1. The resistivity values measured at
T ¼ 300 K (see Fig. 5b) show a slight increase with the amount of Pb
in the alloy. It is a consequence of an increasing concentration of
scattering centers in the alloy. Carrier concentration and mobility
do not exhibit such clear trend as a function of the average lead
content. This means that the increase in resistance as a function of x
is most probably related to the increasing (with x) concentration of

ryx ¼ RH B þ cH rnxxH M; (2)

where cH and nH are the scaling coefficients. This analysis enabled

the determination of the dominant scattering mechanisms
responsible for the AHE in the studied system. Scaling relation of
the AHE was found for each set of magnetotransport curves by
fitting the experimental results to Eq. (2). The data analysis was
performed with RH ¼ const. taken from the previous series of fits.
The cH and nH constants were taken as fitting parameters. The re- Fig. 5. Results of the resistivity, rxx, measurements for the selected Ge1
x
yPbxMnyTe
samples with different chemical compositions: (a) normalized rxx(T) curves, (b)
sults of the fitting procedure indicate a good agreement (with rxx(T ¼ 300 K) as a function of the Pb content, x, and (c) rxx(T
1/4) dependence obtained
variance smaller than 10
12) between the experimental data and experimentally (markers) and fitted (using full markers) to the scaling law (lines) given
the theoretical curves given by Eq. (2). The values of the nH scaling by Eq. (3).
 rni et al. / Journal of Alloys and Compounds 658 (2016) 265e271
A. Podgo 269

Table 1 The measurements were performed over a wide temperature range

Chemical compositions, x and y, and electronic properties of Ge1
x
yPbxMnyTe below 200 K. The rxx(B) curves were obtained by averaging the
crystals at room temperature: carrier concentration p, mobility m, and resistivity rxx.
results for positive and negative current in order to remove all the
x y p [1020 cm
3] m [cm2/(V s)] rxx [10
3 Ucm] artifacts due to the contact asymmetry. The rxx(B) curves at
0.232 0.019 8.8 ± 0.9 15.4 ± 1.5 0.5 ± 0.1 different temperatures were normalized to the zero-field resistivity
0.196 0.055 6.1 ± 0.6 14.1 ± 1.4 0.8 ± 0.1 value, r(B ¼ 0). We determined the relative magnetoresistance
0.196 0.061 5.6 ± 0.6 11.7 ± 1.2 1.0 ± 0.1 given by the formula Drxx =rxx ð0Þ ¼ ðrxx ðBÞ
rxx ðB ¼ 0ÞÞ/rxx(B ¼ 0).
0.180 0.074 4.6 ± 0.5 2.0 ± 0.2 13.6 ± 1.4
The exemplary results are presented in Fig. 6. The experimental MR
0.183 0.074 6.6 ± 0.7 12.8 ± 1.3 0.6 ± 0.1
0.254 0.074 6.0 ± 0.6 8.2 ± 0.8 1.3 ± 0.1 results let us speculate that there are three main contributions to
0.311 0.081 8.1 ± 0.8 3.5 ± 0.4 2.2 ± 0.2 the total MR observed at T < 120 K. Above 120 K only the classical
0.294 0.136 9.9 ± 1.0 1.6 ± 0.2 3.0 ± 0.3 positive MR showing fB2 dependence due to the orbital carrier
movement in the presence of the external magnetic field is
observed. In the temperature range from T z 120 K to T z TSG, for
extended defects. Those defects do not affect the microscopic pa- all our samples, there is visible at least one additional contribution
rameters but create barriers for the electrical conductivity. There- to the total MR causing an increase of the total MR with decreasing
fore the resistivity of the material increases. T to the maximum values observed at T z TSG (see Fig. 7). For
All our samples exhibit a metallic-like rxx(T) curves at T > 100 K temperatures below TSG the MR decreases with decreasing T.
while at lower temperatures for most of our samples a minimum There are several different physical mechanisms that can
followed by the insulating-like behavior is observed. Metal- introduce both positive and negative MR in nonmagnetic semi-
insulator transition observed in our samples can be explained in conductors. Moreover, there are a few mechanisms related to the
terms of localization of quantum states. The classical example of presence of magnetic impurities inducing MR in diluted magnetic
localization is described with Mott's T
1/4 law for low temperature semiconductors. In our case we have to consider also MR related to
conductivity for disordered semiconductors [27]. Phonon-assisted the presence of clusters in the samples. At low temperatures, weak
hopping through localized states (PAHTLS) is the physical mecha- localization or antilocalization phenomena are responsible for the
nism responsible for the shape of the rxx(T) curves at low temper- presence of negative or positive MR in the semiconductors,
atures. In order to test the possibility that the phonon assisted respectively [34]. Typically the product of the Fermi wave vector kF
hopping is present in our samples the following scaling relation and the mean free path l, namely kFl is the quantity needed to
should be fitted to the experimental data: establish the regime at which the carrier conductivity occurs in the
system. For kFl z 1 the system remains in the weak localization
  regime while for kFl > 1 degenerate regime is present and locali-
rxx ¼ r0 exp bT
1=d with d ¼ 2; 3 or 4; (3) zation effects should be suppressed. For most of our samples kFl > 1
indicating the degenerate regime. Moreover, weak localization and
where r0 is the resistivity minimum value, b is phenomenological antilocalization effects are suppressed by the presence of magnetic
constant, and d is the scaling coefficient. Fits were done with three impurities in the alloy. We consider it unlikely that these effects are
fitting parameters: r0, b, and d. We obtained a good agreement responsible for MR in our samples.
between the experimental rxx(T) curves and Eq. (3) (see Fig. 5c) for The presence of magnetic impurities can induce a negative MR
d z 4. It is a proof that the PAHTLS is present in our samples and with relatively low values at low temperatures, as in our
responsible for the appearance of metal-insulator transition Ge1
x
yPbxMnyTe crystals, due to the formation of bound magnetic
observed in Fig. 5a. The presence of PAHTLS is commonly attributed polarons. The negative MR observed in our samples at T < 20 K is
to materials such as CdTe [28] with carrier concentration having caused by reduction in the scattering effective cross-section due to
much lower value range (n < 1017 cm
3) than the values observed the spin alignment of the magnetic moments. This effect induces
for our samples. According to Mott [29] the critical carrier con- MR only at low temperatures, whereas at high temperatures spin-
centration at the metal-insulator transition satisfies the criterion ordering is strongly suppressed and the positive component to the
n1=3 Z2 ε=m e2 z0.27, where ε is the dielectric constant, Z is the MR dominates. Moreover, this negative MR decreases with an in-
Planck constant divided by 2p, and m* is the effective mass. Using crease of carrier concentration [35,36], as it is in our samples (see
the dielectric constant ε ¼ 36 for GeTe [30] and the hole effective Fig. 6). The other possible mechanism, especially important at low
mass m* ¼ 1.2me for GeTe [31] we estimate the corresponding temperatures, is related to the variable range hopping present in
critical carrier concentration to be approximately 5  1018 cm
3. our samples. However, the MR at T < 20 K cannot be fitted with
This value is more than order of magnitude lower (for some sam- exp(B1/3), relation typical for hopping conduction [37]. Moreover,
ples more than two orders of magnitude) than the carrier con- the negative MR diminishes at a temperature much lower than the
centration values observed for our samples (see Table 1) indicating temperature at which hopping regime is present. Thus, we believe
that our samples are on the metallic side of the Mott's transition. that the negative MR present in our samples at T < 20 K is due to the
The presence of magnetic impurities and, as a consequence, the spin-dependent carrier scattering on the localized magnetic
exchange interaction between the localized magnetic moments moments.
and conducting carriers is known to modify drastically the carrier The MR observed at T > 20 K changes sign to positive, scales with
transport in the solid. The exchange interaction induces a giant fB2 at B < 2 T, remains nearly linear at B > 2 T, and its value
spin-splitting of electronic bands and shifts the critical carrier (measured at B ¼ 13 T) increases as a function of temperature up to
concentration of the metal-insulator transition [32]. It is most T z TSG (see Fig. 7). The temperatures at which the MR reaches a
probably the case of our samples. Similar effects are known in maximum (see Fig. 7) are for each our sample different than the
literature for Ge1
xMnTe thin layers [33] and for both our samples temperatures at which a minimum in rxx(T) is observed (see Fig. 5)
and literature data the observed minimum in the rxx(T) dependence indicating that the maximum in MR effect is not related to the
is interpreted in terms of PAHTLS. PAHTLS effect present at low temperatures. The positive linear MR
The magnetic field dependence of the resistivity component observed in our samples is most probably the superposition of a
parallel to the current direction, namely magnetoresistance (MR), very small contribution from the classical quadratic MR and linear
rxx, was measured in parallel with the Hall effect measurements. MR related to the granular nature of our samples. Since the
270 rni et al. / Journal of Alloys and Compounds 658 (2016) 265e271
A. Podgo
Fig. 6. The magnetic field dependencies of the magnetoresistance obtained at selected stabilized temperatures (see labels) for the studied Ge1
x
yPbxMnyTe samples with different
chemical compositions.
Fig. 7. Temperature dependence of magnetoresistance at B ¼ 13 T for selected
Ge1
x
yPbxMnyTe crystals with different chemical compositions.
maximum MR effect occurs in our samples in the vicinity of TSG it is
highly probable that we observe the MR effect with maximum near
the Curie temperature due to giant spin-splitting of the valence
band observed in many ferromagnetic semiconductors. Giant Zee-
man effect observed near TC leads to a maximum in positive
magnetoresistance.
4. Summary
To conclude, we have studied magnetotransport properties of
bulk Ge1
x
y PbxMnyTe crystals with chemical composition varying
from 0.180 < x < 0.311 to 0.019 < y < 0.136. Our previous studies
showed that the cluster-glass-like state appears at temperatures
lower than 100 K.
The magnetic order of the alloy has a significant influence on the
magnetotransport properties of our samples. A strong anomalous
Hall effect with hysteresis was observed at T < TSG in all our sam-
ples. The data analysis indicated the ordinary Hall constant, RH, at
T < TSG does not change with temperature. The AHE coefficient was
found to show small changes with the chemical composition,
changing from 1  10
6 to 3  10
6 m3/C. The scaling law analysis
of the AHE shows that the extrinsic skew scattering mechanism,
accompanied with the extrinsic side jump, is the main physical
mechanism responsible for the AHE in the Ge1
x
yPbxMnyTe
system.
The presence of localized states influencing carrier conduction
occurs in our system at T < TSG and is reflected in the presence of a
minimum in the resistivity vs. temperature dependence. Such
dependence of resistivity on temperature is characteristic of the
phonon-assisted hopping through localized states. The magneto-
resistance of our samples shows complex behavior at T < TSG. The
negative magnetoresistance observed at T < 20 K is interpreted as
spin-dependent scattering of conducting carriers on localized
magnetic moments. At temperatures T > 20 K a nearly linear pos-
itive magnetoresistance increasing as a function of temperature up
to T z TSG, where it reaches maximum, is observed. Positive linear
MR is related to the granular nature of our samples while the
maximum near TSG is related to the giant spin-splitting of the
valence band. At T > TSG only a positive classical MR is observed.
Acknowledgments
The research was supported by the Foundation for Polish Sci-
ence e POMOST/2011-4/2 Programme co-financed by the European
Union within European Regional Development Fund. The research
was partly financed by the National Center for Science of Poland
under Decision No. DEC-2012/05/D/ST3/03161.
References
[1] J. Kossut, W. Dobrowolski, Handbook of Magnetic Materials, 1993, p. 231305.
North-Holland, Amsterdam.
[2] F. Matsukura, H. Ohno, T. Dietl, Handbook of Magnetic Materials, Elsevier,
Amsterdam, 2002, p. 187. Chaps. III
V.
 rni et al. / Journal of Alloys and Compounds 658 (2016) 265e271
A. Podgo
[3] W. Dobrowolski, J. Kossut, T. Story, Handbook of Magnetic Materials, Elsevier,
Amsterdam, 2003, p. 289377. Chaps. II
VI.
[4] M. Wang, R.P. Campion, A.W. Rushforth, K.W. Edmonds, C.T. Foxon,
B.L. Gallagher, Appl. Phys. Lett. 93 (2008) 132103.
[5] R.T. Lechner, G. Springholz, M. Hassan, H. Groiss, R. Kirchschlager, J. Stangl,
N. Hrauda, G. Bauer, Appl. Phys. Lett. 97 (2010) 023101.
[6] M. Hassan, G. Springholz, R.T. Lechner, H. Groiss, R. Kirchschlager, G. Bauer,
J. Cryst. Growth 323 (2011) 363.
[7] W. Knoff, V. Domukhovski, K. Dybko, P. Dziawa, R. Jakieła, E. Łusakowska,
 ̧tek, B. Taliashvili, T. Story, K. Szałowski, T. Balcerzak, Acta
A. Reszka, K. Swia
Phys. Pol. A 116 (2009) 904.
[8] Y. Fukuma, H. Asada, S. Miyawaki, T. Koyanagi, S. Senba, K. Goto, H. Sato, Appl.
Phys. Lett. 93 (2008) 252502.
[9] J.S. Lee, A. Richardella, D.W. Rench, R.D. Fraleigh, T.C. Flanagan, J.A. Borchers,
J. Tao, N. Samarth, Phys. Rev. B 89 (2014) 174425.
[10] A. Podgo rni, L. Kilanski, W. Dobrowolski, M. Go rska, V. Domukhovski,
B. Brodowska, A. Reszka, B.J. Kowalski, V.E. Slynko, E.I. Slynko, Acta Phys. Pol. A
126 (2014) 1180.
[11] A. Podgo rni, L. Kilanski, K. Szałowski, M. Go rska, R. Szymczak, A. Reszka,
V. Domukhovski, B.J. Kowalski, B. Brodowska, W. Dobrowolski, V.E. Slynko,
E.I. Slynko, Magnetic properties of Ge1
x
yPbxMnyTe Cluster-Glass System,
J. Alloys Comp. 649 (2015) 142.
[12] K.T. Aust, B. Chalmers, Can. J. Phys. 36 (1958) 977.
[13] J. Sinova, T. Jungwirth, J. Cerne, Int. J. Mod. Phys. B 18 (2004) 1083.
[14] F. James, M. Roos, Comput. Phys. Commun. 10 (1975) 343.
[15] B. Brodowska, W. Dobrowolski, M. Arciszewska, E.I. Slynko, V.K. Dugaev,
J. Alloys Comp. 423 (2006) 205.
[16] L. Kilanski, M. Go rska, R. Szymczak, W. Dobrowolski, A. Podgo rni, A. Avdonin,
V. Domukhovski, V.E. Slynko, E.I. Slynko, J. Appl. Phys. 116 (2014) 083904.
[17] L. Kilanski, R. Szymczak, W. Dobrowolski, A. Podgo  rni, A. Avdonin, V.E. Slynko,
271
E.I. Slynko, J. Appl. Phys. 113 (2013) 063702.
[18] B. Brodowska, I. Kuryliszyn-Kudelska, M. Arciszewska, K. Dybko,
V. Domukhovski, W. Dobrowolski, V.E. Slynko, E.I. Slynko, V.K. Dugaev, Mater.
Pol. 26 (2008) 927.
[19] J. Smit, Phys. Amst. 12 (1955) 877.
[20] L. Berger, Phys. Rev. B 2 (1970) 4559.
[21] T. Jungwirth, Q. Niu, A.H. MacDonald, Phys. Rev. Lett. 88 (2002) 207208.
[22] A. Dyrdal, V.K. Dugaev, J. Barnas, Phys. Rev. B 78 (2008) 245208.
[23] A. Dyrdał, V.K. Dugaev, J. Barnas, B. Brodowska, W. Dobrowolski, Acta Phys.
Pol. A 115 (2009) 287.
[24] Y. Fukuma, M. Arifuku, H. Asada, T. Koyanagi, J. Appl. Phys. 91 (2002) 7502.
[25] Y. Fukuma, H. Asada, T. Taya, T. Irisa, T. Koyanagi, Appl. Phys. Lett. 89 (2006)
152506.
[26] N. Nagaosa, J. Sinova, S. Onoda, A.H. MacDonald, N.P. Ong, Rev. Mod. Phys. 82
(2010) 1539.
[27] A.L. Efros, B.I. Shklovskii, J. Phys. C. Solid State Phys. 8 (1975) L49.
[28] N.V. Agrinskaya, V.I. Kozub, D.V. Shamshur, JETP 80 (6) (1995) 1142.
[29] N.F. Mott, Metal-insulator Transitions, second ed., Taylor and Francis, London,
1990.
[30] N.E. Zein, V.I. Zinenko, A.S. Fedorov, Phys. Lett. A 164 (1992) 115.
[31] J.E. Lewis, Phys. Stat. Sol. B 59 (1973) 367.
[32] Y. Shapira, D.H. Ridgley, K. Dwight, A. Wold, K.P. Martin, J.S. Brooks, J. Appl.
Phys. 57 (1994) 3210.
[33] W.Q. Chen, K.L. Teo, S.T. Lim, M.B.A. Jalil, T. Liew, T.C. Chong, Appl. Phys. Lett.
90 (2007) 142514.
[34] P.W. Anderson, Phys. Rev. 109 (1958) 1492.
[35] Y. Toyozawa, J. Phys. Soc. Jpn. 17 (1962) 986.
[36] R.P. Khosla, J.R. Fischer, Phys. Rev. B 2 (1970) 4084.
[37] H. Bottger, V.V. Bryksin, Hopping Conduction in Solids, Akademie, Berlin,
1985.