Cluster superconductivity in the magnetoelectric Pb(Fe1/2Sb1/2)O3 ceramics

V.V. Laguta1,2, M. Marysko1, V. A. Stephanovich2, I. P. Raevsky3, N.M. Olekhnovich4, A.V. Pushkarev4,

Yu.V. Radyush4, S.I. Raevskaya3, R.O. Kuzian5, V. Chlan6, H. Štěpánková6
1
Institute of Physics AS CR, Cukrovarnicka 10, 162 53 Prague, Czech Republic
2
Institute of Physics, Opole University, Oleska 48, 45-052 Opole, Poland
3
Research Institute of Physics and Faculty of Physics, Southern Federal University, Stachki Ave. 194,
344090, Rostov-on-Don, Russia
4
Scientific-Practical Materials Research Centre, NASB, Minsk, Belarus
5
Institute for Problems of Materials Science, National Academy of Sciences of Ukraine,
Krjijanovskogo 3, 03142 Kyiv, Ukraine
6
Charles University in Prague, Faculty of Mathematics and Physics, Prague 1,
Czech Republic

Abstract We report the observation of cluster (local) superconductivity in the magnetoelectric

Pb(Fe1/2Sb1/2)O3 ceramics prepared at a hydrostatic pressure of 6 GPa and temperatures 1200-1800 K to
stabilize the perovskite phase. The superconductivity is manifested by an abrupt drop of the magnetic
susceptibility at the critical temperature TC ≈ 7 K. Both the magnitude of this drop and TC decrease with
magnetic field increase. Similarly, the low-field paramagnetic absorption measured by EPR spectrometer
drops significantly below TC as well. The observed effects and their critical magnetic field dependence
are interpreted as manifestation of the superconductivity and Meissner effect in metallic Pb nanoclusters
existing in the ceramics. Their volume fraction and average size were estimated as 0.1-0.2% and 140-150
nm, respectively. The superconductivity related effects disappear after oxidizing annealing of the
ceramics.

I. Introduction

Modern technologies permit to fabricate the materials and interfaces with ordinary crystal structure
but unusual physical properties combining, for instance, long range magnetic and ferroelectric orders with
local superconductivity. Well-known examples are the interfaces between perovskite oxides [1,2], which
may have all the above properties [1,3,4] along with metallic interfacial conductivity [5-8]. Here we
report the first time observation of local superconductivity in the magnetoelectric Pb(Fe1/2Sb1/2)O3 (PFS).
This material possesses both magnetism and ferroelectricity like other much more known chemically
disordered double perovskites PbFe1/2Nb1/2O3 (PFN) and PbFe1/2Ta1/2O3 (PFT) [9-17]. PFS shows quite
intriguing magnetic properties such as existence of dynamic magnetic nanoregions with large frustrated
magnetic superspins, which on cooling freeze in superspin glass state coexisting with the long-range
ordered antiferromagnetic (AFM) phase at T < 32 K [9]. Such behavior of PFS is drastically different
from that known for its disordered counterparts PFN and PFT where the long-range AFM ordered phase
exists below TN = 150-155 K [12,18-20], while below Tg = 11-12 K it coexists with the spin glass phase
[11,19,20]. This difference was attributed to the fact that PFS is highly ordered [9]. Recently, we were
able to fabricate, for the first time, the PFS ceramics with different degree of long-range chemical

1
ordering between the magnetic Fe3+ and non-magnetic Sb3+ ions seen well by X-ray diffraction. This
fabrication flexibility permits to synthesize the samples with different chemical compositions and thus to
clarify the role of chemical order/disorder in magnetic properties of double perovskites.
At room temperature, chemically ordered PFS has a simple cubic perovskite structure with space

group Fm3m . On cooling to 200 K, the cubic structure transforms to polar one as evidenced from the
dielectric permittivity and hysteresis loops data [9]. Its X-ray diffraction data have been reported in Ref.
[21] where the crystal structure of the low-temperature polar phase was not refined yet. Other magnetic
and structure data obtained from ab initio Density functional theory (DFT) calculations are presented in
Refs. [9,22,23].
In this paper we pay attention mainly to unusual behavior of PFS ceramics at temperatures below 10
K, namely abrupt decrease of the magnetic susceptibility along with appearance of low-field EPR
absorption and its critical dependence on applied magnetic field, which is interpreted as a manifestation
of cluster (local) superconductivity due to presence of metallic Pb clusters in ceramics. Besides, as the
measurements were performed on samples ranging from nearly ordered (s=0.93) to almost disordered
(s=0.21), we observed the transformation of magnetic structure with the change of chemical ordering. In
particular, the ground magnetic state of the disordered sample contains only spin glass phase with the
freezing temperature Tg = 25 K in contrast to disordered PFN and PFT, where the long-range AFM
ordered phase exists below TN = 150-155 K [12,18-20].

II. Experimental

PFS ceramic samples were prepared in two stages [21]. First, we synthesized the stoichiometric
composition Pb(Fe1/2Sb1/2)O3 from the initial PbO, Fe2O3 and Sb2O5 oxides at T= 1020–1030K for 4
hours. The resulting compound had a pyrochlore crystal structure. The second stage of this synthesis has
been performed under hydrostatic pressure 6 GPa and temperature 1200 -1800 K for 2-10 min. After the
synthesis, the system was rapidly cooled to the room temperature under pressure and only then the high-
pressure apparatus was unloaded. The product of the second high-pressure synthesis was dense coarse-
grained (the grain size varied from 1 to 5 µm with a mean size of approximately 2 µm) PFS ceramics with
single phase perovskite structure. Room temperature x-ray diffraction patterns of PFS correspond to a
cubic symmetry (space group Fm3m [21]) and exhibit superstructure lines attributed to a double
perovskite unit cell caused by partial chemical 1:1 ordering of Fe3+ and Sb5+ ions. The mean value of the
chemical ordering degree varies from s=0.17 up to s=0.93 depending on synthesis conditions. It is
determined either from the ratio of the intensities of the superstructure XRD reflections to the
fundamental ones, or the fraction of a singlet component in Mossbauer spectrum [24]. The samples with
s=0.67 and 0.21 were annealed in air at T≈770 K in order to oxidize residual Pb metallic inclusions in
them.
The magnetic measurements were carried out using the SQUID magnetometer MPMS-5S (Quantum
Design) under several important protocols including field cooling (FC) and zero-field cooling (ZFC) in
the dc regime. Electron paramagnetic resonance (EPR) measurements were performed at 9.407 GHz in a
2
temperature range from 3.5 to 300 K, by employing the Bruker E580 spectrometer and Oxford Instrument
cryostat.
III. Results
Fig. 1 presents magnetic susceptibility data measured at ZFC and FC regimes in the field 500 Oe for
few PFS samples with different degree of chemical ordering between Fe and Sb ions from s = 0.21 to s =
0.93. The temperature behavior of the susceptibility in the sample with high degree of chemical ordering
is identical to that described in our previous paper [9]. On cooling, susceptibility exhibits an abrupt
increase at approximately 250 K due to the formation of superparamagnetic phase or superspins. Then, in
the temperature range between 100 and 150 K, the ZFC magnetic susceptibility (dashed lines) shows a
broad maximum related to collective freezing into a superspin glass phase. And finally, the second, sharp
maximum in both FC and ZFC data characterizes AFM phase transition with the Neel temperature TN =
32 K. One can see that with the chemical order decrease, the anomalies related to the superspin glass
phase gradually disappear. Likewise, the ZFC peak at ~ 30 K transforms into a cusp characterizing
transition to classical spin glass state where ZFC and FC curves essentially differ below the peak
temperature. However, in this paper we shall concentrate on the abrupt decrease of susceptibility on
further cooling down to T < 7 K visible in all three samples (Fig. 1a).
6
Pb(Fe1/2Sb1/2)O3 (a)

5 H=500 Oe
χ [10 emu/(g Oe)]

4
s=0.93

s=0.67
3
-5

s=0.21

1
6
0 50 100 150 200 250
5 (b)
χ [10 emu/(g Oe)]

4
s=0.67 H=500 Oe
3
-5

1
0 50 100 150 200 250
T (K)

Fig. 1. (a) Temperature dependence of the FC (dashed lines) and ZFC (solid lines) magnetic
susceptibilities for three PFS samples having different degrees of the chemical ordering: s = 0.93; 0.67;
and 0.21. (b) Magnetic susceptibility versus temperature for the sample with s=0.67 after its annealing in
air at T≈770 K.

3
 Note that this abrupt decrease of the susceptibility disappears after annealing of samples in air at
approximately 770 K as it is shown in Fig. 1b for the sample with s=0.67. Since the anomaly at T<7 K is
observed for all three samples, below we report the detailed studies of only the sample with s = 0.21 for
which the susceptibility does not expose to the influence of superspin glass phase.
Fig. 2 reports magnetic susceptibility for this sample measured at different magnetic fields from 50
to 2000 Oe. The abrupt decrease of susceptibility at TC ≈ 7 K, which is much lower than Tg ≈ 23 K, is
well seen in both FC and ZFC curves. However, the TC starts to decrease with field increase and this
anomaly completely disappears for the fields above of only 1000 Oe.

6
Pb(Fe1/2Sb1/2)O3
5 s=0.21
50 Oe
χ [10 emu/(g Oe)]

4
100 Oe
500 Oe
3
-5

2 kOe
1 kOe
2

1
0 10 20 30 40 50 60 70
T (K)

Fig. 2. Temperature dependence of the FC (dashed lines) and ZFC (solid lines) magnetic susceptibilities
of PFS sample with s = 0.21 at different magnetic fields.

We have measured also EPR spectra for these PFS samples. As an example, Fig. 3 shows EPR
spectra for the sample with s = 0.21 measured in the temperature range from 7.3 down to 3.7 K. One can
see that at T < 7 K, a new line appears at low magnetic fields. This line shifts towards higher magnetic
fields with further sample cooling and has nontypical line shape, which does not resemble usual first
derivative of the Lorentzian or Gaussian curves. Note that EPR spectrometer measures first derivative of
EPR absorption signal due to the use of a lock-in detection system. This line disappears after sample
annealing in air (Fig. 3b) similar to the anomaly in magnetic susceptibility at T<7 K.
The unusual line shape indicates that the EPR absorption decreases abruptly at the resonance peak
field. This is well seen in the absorption spectra obtained by the integration of the virgin one (Fig. 3c).
One can also observe the shift of the critical field related to the observed anomaly towards higher
magnetic fields with temperature lowering. Such low-field EPR signal is often observed in
superconductors (see, e.g. Refs. 25-27) due to change in the diamagnetic susceptibility upon transition
from the Meissner (i.e. superconductive) state [28] to the mixed or normal (i.e. non-superconductive) one.
Therefore, both magnetic susceptibility and EPR data suggest that the observed phenomenon may be
related to superconductivity in some regions of our samples. Defining the critical temperature Tc as the
onset of the negative diamagnetic contribution to susceptibility and the critical field Hc as the value of the
magnetic field at which the diamagnetic contribution disappears, we observe that the corresponding
4
values Tc ≈7 K and Hc ≈700 Oe (at 3.7 K) coincide approximately with those of superconducting Pb [29-
30]. This is not much surprising as our samples were fabricated on the base of PbO oxide. Due to specific
synthesis condition (high pressure and rapid cooling) Pb can be partially reduced, for instance, at grain
boundaries. It can be again oxidized by annealing in air. PFS ceramics after the oxidizing annealing does
not show any anomalies at T<7 K in both magnetic susceptibility and EPR spectra (see, Figs. 1b and 3b).

40 Pb(Fe1/2Sb1/2)O3 F=9458 MHz a)

30 3.7 K

20
4.5 K
10
5.2 K
0
6.2 K
-10
χ'' (arb. units)

6.9 K
-20
7.3 K
-30
2 b)
6K
0
-2
(c)
20
3.7 K

10 4.5 K
χ'' (arb. units)

5.2 K
0
6.2 K
-10 6.9 K

7.3 K
-20
0 500 1000 1500 2000
Magnetic field (Oe)
Fig. 3. EPR spectra measured in PFS (s = 0.21) at temperatures from 7.3 down to 3.7 K. Panels (a) and
(b) report the first derivative of EPR absorption signal in the virgin (a) and annealed in air (b) samples.
Integrated absorption spectra are shown in panel (c).

To verify origin of the observed EPR signal, we measured, as a test, EPR spectra of small Pb
metallic particles (∼0.1 mm) mixed with epoxy resin. One can see (Fig. 4) that below the critical
temperature Tc≈7.2 K, low-field spectral line appears similar as in PFS ceramics. The magnetic field
value at which the EPR absorption peak occurs is designated as the critical field Hc [25-27]. The
temperature dependence of this critical field for Pb particles and PFS ceramics is reported in Fig. 5. One
can see that our PFS ceramics data are in good coincidence with those in bulk testing sample and
literature data [29-30] except the lowest temperatures where PFS exhibits higher critical fields due to size
effects, i.e. the increase of the critical field in small particles [29-30].

5
 It is worth also to mention that the spectra similar to those in Figs. 3 and 4 were reported for small
(0.1-1 mm) superconductive thin spheres [31].
The above experimental facts show convincingly that the observed phenomenon in PFS ceramics is
related to superconductivity of local Pb clusters in it. Such clusters can emerge at the grain boundaries.
The full screening of some grains by Pb shell cannot be excluded as well. However these inclusions of Pb
do not form the percolative cluster as the room-temperature ac conductivity of all the samples studied is
rather low (10-8 - 10-7 S/cm at 1 kHz and 10-5 - 10-4 S/cm at 1 MHz). Annealing of the samples in air at
700-800 K completely destroys the superconductivity as the metallic Pb inclusions transform into PbO
207
due to oxidation. It is worth noting that Pb NMR measurements also show presence of Pb ions which
do not belong to intrinsic magnetic composition.

F=9750.9 MHz bulk Pb

2 3.3 K
4.5 K
5.1 K
5.7 K
χ'' (arb. units)

6.2 K

1
6.6 K

7.1 K
0
7.3 K

0 500 1000 1500 2000

Magnetic field (Oe)

Fig. 4. EPR spectra measured in bulk metallic Pb sample.

To assess the volume of Pb clusters, VPb, in our PFS sample (for definiteness we choose s=0.21), we
use the results of the work [29] dealing with the superconductivity of Pb nanoparticles. In this work, the
experimental magnetic field dependence of superconducting critical temperature Tc has been
approximated by the relation [29]:

⎧⎪ ⎛ T ⎞α ⎫⎪
H C (T ) = H C (0) ⎨1 − ⎜ C ⎟ ⎬ , (1)
T (0) ⎠ ⎪
⎩⎪ ⎝ C ⎭
where TC(0) and HC(0) are, respectively, the critical temperature at H=0 and critical magnetic field at
T=0. The exponent α is related to the characteristics of magnetic field penetration into a superconductor
and increases from 2 (bulk Pb), to 3 for the nanoparticles with diameter 6 nm [29]. We consider the value
of TC =7.2 K to be equal to TC(0). As our TC(0) coincides with that of bulk Pb, it is reasonable to put α=2
in subsequent calculations. The best fit of the above expression to the data in the Fig. 5 (solid line) yields

6
HC(0)≈950 Oe. The estimate of linear dimension d of a Pb cluster can be done with the empirical relation
(see Fig. 8 of Ref. 29) HC(0)=170d -3/2 (where d is in nanometers and HC(0) is in Tesla) from which we
obtain d ≈140-150 nm.

1000
bulk Pb
magnetic suscep.
EPR data
750 fit by Eq. (1)
HC (Oe)

500

250

0
0 1 2 3 4 5 6 7
T (K)

Fig. 5. Critical field versus temperature in bulk Pb (dashed line) and PFS ceramics determined from
magnetic susceptibility (triangles) and EPR spectra (stars). The solid line is the fit to Eq. (1).

To estimate the volume portion of metallic Pb in our PFS sample, we use the simple argument
stemming from the low-temperature behavior of magnetic susceptibility in Fig. 2. Namely, ZFC
susceptibility in the low filed H=50 Oe decreases by Δχm ≈ 9.8⋅10-6 emu/(g Oe) on cooling from 7.2 K
down to 2 K. This mass susceptibility decrease corresponds to the volume value ΔχV ≈9⋅10-5 emu/(cm3

Oe). Using this value we can estimate the entire volume of Pb fraction that becomes superconductive, as
approximately ΔχV / χ 0 ≈ 0.0012 where χ0 = −1/ 4π is the volume susceptibility of a bulk
superconductor. Taking into account penetration depth of magnetic flux into Pb cluster (∼40 nm), the total
volume of metallic Pb clusters may be few times larger than the above estimated value 0.12%. Such
minute amount of Pb clusters randomly distributed in ceramics cannot be detected by transmission
electron microscopy. However, they are reliably detectable via observation of their superconductive
properties in magnetic susceptibility and magnetic resonance spectra even in a quantity as low as few
micrograms.
Discovered local superconductivity of PFS ceramics may take place in other Pb contained complex
oxide materials, including many double perovskites (PFN, PFT, Pb(Fe1/2W1/2)O3 (PFW)) as well as PZT
and PMN-PT ceramics. In particular, we detect the metallic Pb inclusions in PFT ceramics synthesized
with 6% excess of PbO oxide [32]. Such ceramics exhibits properties similar to described above.
Inclusions of metallic Pb have been already revealed in PbVO3 [33] and Ba1−xLaxPbO3 [34] ceramics
obtained by high-pressure synthesis. In the latter material, the anomaly of magnetic susceptibility at about
7 K has also been observed. It was attributed to a transition of Pb inclusions into the superconducting
state, though this anomaly was masked by the transition into the superconducting state of the
Ba1−xLaxPbO3 phase, which occurs at somewhat higher temperature. Of course, this effect will not
influence markedly the electric or magnetic properties of a material at temperatures above TC≈7 K. But it

7
will have impact below the critical temperature leading to unusual behavior of magnetic and dielectric
characteristics, which could be misinterpreted. For instance, the abrupt decrease of the magnetic or
dielectric susceptibility and relaxation at TC can be interpreted as a manifestation of phase transition or
even quantum effects, like the quantum spin tunneling [35]. Strong influence of the external magnetic
field on susceptibility in Pb contained magnetoelectrics at low temperatures can be wrongly interpreted as
manifestation of magnetoelectric coupling as well.
In summary, we have synthesized magnetoelectric Pb(Fe1/2Sb1/2)O3 ceramics with controlled degree
of chemical ordering ranging from s=0.17 (almost disordered) up to s=0.93 (nearly perfectly ordered). In
contrast to chemically disordered Pb(Fe1/2Nb1/2)O3 or Pb(Fe1/2Ta1/2)O3, the disordered PFS ceramics has a
unique magnetic ground state in the form of short-range ordered spin-glass phase with the freezing
temperature around 25 K. We have additionally found that above PFS ceramics shows quite unusual
behavior of magnetic susceptibility and EPR signal. Namely, both these quantities sizeably drop at the
critical temperature ≈7 K, which depends on applied magnetic field. We explained these observations as
manifestation of local superconductivity of Pb metallic clusters at grain boundaries. Estimated total
volume of these metallic clusters is of the order (1-2)⋅10-3 of the sample volume and its typical linear
dimension is around 140-150 nm.

ACKNOWLEDGMENTS
The research was supported by the GA CR under project No. 13-11473S, projects
SAFMAT LM2015088 and LO1409 and partially by the Russian Foundation for Basic Research
(grant T16R-079), Belarusian Republican Foundation for Fundamental Researches (grant 16-52-00072
Bel_a), the Ministry of education and science of the Russian Federation (research project 2132)

References
[1] J. Heber, Nature, 459, 28 (2009).
[2] H. Y. Hwang, Y. Iwasa, M. Kawasaki, B. Keimer, N. Nagaosa and Y. Tokura, Nat. Mater., 11, 103
(2012).
[3] L. Li, C. Richter, J. Mannhart and R. C. Ashoori, Nat. Phys. 7, 762 (2011).
[4] J. A. Bert, B. Kalisky, C. Bell, M. Kim, Y. Hikita, H. Y. Hwang and K. A. Moler, Nat. Phys. 7, 767
(2011).
[5] A. Ohtomo and H. Y. Hwang, Nature 427, 423 (2004).
[6] G. Khalsa and A. H. MacDonald, Phys. Rev. B 86, 125121 (2012).
[7] V. A. Stephanovich and V. K. Dugaev, Phys. Rev. B, 93, 045302 (2016)
[8] V. A. Stephanovich, V. K. Dugaev and J. Barnaś, Phys. Chem. Chem. Phys., 18, 2104 (2016)
[9] V. V. Laguta, V. A. Stephanovich, M. Savinov, M. Marysko, R. O. Kuzian, I. V. Kondakova,
N. M. Olekhnovich, A. V. Pushkarev, Yu. V. Radyush, I. P. Raevski, S. I. Raevskaya and S. A.
Prosandeev, New J. Phys. 16, 113041(1) - 113041(19), (2014).
[10] V. A. Stephanovich and V. V. Laguta Phys. Chem. Chem. Phys., 18, 7229 (2016)
[11] W. Kleemann, V. V. Shvartsman, P. Borisov and A. Kania, Phys. Rev. Lett., 105, 257202 (2010).
8
[12] V. V. Laguta, M. D. Glinchuk, M. Marysko, R. O. Kuzian, S. A. Prosandeev, S. I. Raevskaya, V. G.
Smotrakov, V. V. Eremkin, I. P. Raevski, Phys. Rev. B 87, 064403 (2013).
[13] V. V. Laguta, A. N. Morozovska, E. A. Eliseev, I. Р. Raevski, S. I. Raevskaya, E. I. Sitalo, S. A.
Prosandeev, L. Bellaiche. J. Mater. Sci. 51, 5330 (2016).
[14] S. A. Prosandeev, I. P. Raevski, S. I. Raevskaya, H. Chen. Phys. Rev. B. 2015. 92, 220419(R)
(2015).
[15] R. Martinez, R. Palai, H. Huhtinen, J. Liu, J. F. Scott, and R. S. Katiyar, Phys. Rev. B 82, 134104
(2010).
[16] I. P. Raevski, M. S. Molokeev, S. V. Misyul, E. V. Eremin, A. V. Blazhevich, S. P. Kubrin, H. Chen,
C.-C. Chou, S. I. Raevskaya, V. V. Titov, D. A. Sarychev, M. A. Malitskaya. Ferroelectrics. 475, 52
(2015).
[17] I. P. Raevski, V. V. Titov, M. A.Malitskaya, E. V. Eremin, S. P. Kubrin, A. V. Blazhevich, H. Chen,
C.-C. Chou, S. I.Raevskaya, I. N. Zakharchenko, D. A. Sarychev, S. I. Shevtsova. J. Mater. Sci. 49,
6459 (2014).
[18] S. Ivanov, R. Tellgren, H. Rundlof, N. W. Thomas, S. Ananta, J. Phys.: Condens. Matter 12, 2393
(2000).
[19] G. M. Rotaru, et al., Phys. Rev. B 79, 184430 (2009).
[20] S. Chillal, M. Thede, J. F. Litterst, S. N. Gvasaliya, T. A. Shaplygina, S. G. Lushnikov and A.
Zheludev, Phys. Rev. B 87 220403(R) (2013).
[21] S. V. Misjul et al., J. Siberian Federal Univ. Math. Phys. 6, 227 (2013).
[22] R. O. Kuzian, I. V. Kondakova, A. M. Dar´e, and V. V. Laguta, Phys. Rev. B 89, 024402, (2014).
[23] I. P. Raevski, S. P. Kubrin, S. I. Raevskaya, D. A. Sarychev, S. A. Prosandeev, and M. A.
Malitskaya, Phys. Rev. B 85, 224412 (2012).
[24] I. P. Raevski, N. M. Olekhnovich, A. V. Pushkarev, Y. V. Radyush, S. P. Kubrin, S. I.
Raevskaya, M. A. Malitskaya, V. V. Titov, V. V. Stashenko. Ferroelectrics. 444, 47 (2013).
[25] C. Rerroti, D. Davidov, I. Belash, I. Felner, Phys. Rev. B 36, 4028 (1987).
[26] K. W. Blazey, K. A. Muller, J. G. Bednorz, W. Berlinger, Phys. Rev. B 36, 7241 (1987).
[27] S. K. Misra, L. Misiak, J. Phys.: Condens. Matter 1, 9499 (1989).
[28] W. Meissner and R. Ochsenfeld, Naturwiss 21, 787 (1933).
[29] W.-H. Li, C. C. Yang, F. C. Tsao, and K. C. Lee , Phys.Rev. B 68, 184507 (2003).
[30] P. Zolotavin, Ph. Guyot-Sionnest, ASCnano 4, 5599 (2010).
[31] A. S. Kheifets and A. I. Veinger, Physica C 165, 491 (1990).
[32] These results will be published elsewhere.
[32] R. V. Shpanchenko, V. V. Chernaya, A. A. Tsirlin, P. S. Chizhov, D. E. Sklovsky, E. V. Antipov, E.
P. Khlybov,V. Pomjakushin, A. M. Balagurov, E. E. Kaul, C. Geibel. Chem. Mater. 16, 3267 (2004).
[33] A. P. Menushenkov, A. V. Tsvyashchenko, D. V. Eremenko, K. V. Klementev, A. V. Kuznetsov, V.
N. Trofimov, L. N. Fomichev. Phys. Solid State, 43, 613 (2001)
[34] P. C. E. Stamp, E.M. Chudnovsky, B. Barbara, Inter. J. of Modern Phys. B, 6, 1355 (1992).