Oxygen non-stoichiometry in Ru-1212 and Ru-1222 magnetosuperconductors

Article abstract of DOI:10.1016/S0921-4534(03)01067-0

We report here the results of thermogravimetric (TG) analysis on the oxygen non-stoichiometry of RuSr₂GdCu₂O _8-δ (Ru-1212) and RuSr₂(Gd_0.75Ce _0.25)₂Cu₂O_10-δ (Ru-1222) samples. With TG measurements carried out in O₂ and Ar atmospheres it is found that the oxygen content in Ru-1212 remains less affected upon various annealings, while for Ru-1222 the wider-range oxygen-content tuning is possible. When heated in H₂/Ar atmosphere the both phases release oxygen upon breaking down to mixtures of metals (Ru and Cu) and simple oxides (CeO₂, Gd₂O₃, and SrO) in two distinct steps around 300 and 450 °C. This reductive decomposition reaction carried out in a thermobalance is utilized in precise oxygen content determination for these phases. It is found that the 100-atm O₂-annealed Ru-1212 sample is nearly stoichiometric, while the similarly treated Ru-1222 sample is clearly oxygen deficient. X-ray absorption near-edge structure (XANES) spectroscopy is applied to estimate the valence of Ru in the samples. In spite of the fact that the Ru-1212 phase was shown to possess less oxygen-deficient RuO _2-δ layer, the valence of Ru as probed with XANES is found to be lower in Ru-1212 than in Ru-1222.

Full text of DOI:10.1016/S0921-4534(03)01067-0

Physica C 392–396 (2003) 87–92
www.elsevier.com/locate/physc
Oxygen non-stoichiometry in Ru-1212 and Ru-1222
magnetosuperconductors
a
c
d
M. Matvejeff a,b, V.P.S. Awana , L.-Y. Jang , R.S. Liu ,
a
H. Yamauchi , M. Karppinen
a,
*
a
Materials and Structures Laboratory, Tokyo Institute of Technology, 4259 Nagatsuta, Midori-ku, Yokohama 226-8503, Japan
b
Laboratory of Inorganic and Analytical Chemistry, Helsinki University of Technology, FIN-02015 Espoo, Finland
c
Synchrotron Radiation Research Center, Hsinchu, Taiwan, ROC
d
Department of Chemistry, National Taiwan University, Taipei 106, Taiwan, ROC
Received 13 November 2002; accepted 27 January 2003
Abstract
We report here the results of thermogravimetric (TG) analysis on the oxygen non-stoichiometry of RuSr2GdCu2-
8ꢀd (Ru-1212) and RuSr2(Gd0:75Ce0:25)2Cu2O10ꢀd (Ru-1222) samples. With TG measurements carried out in O2 and Ar
O
atmospheres it is found that the oxygen content in Ru-1212 remains less affected upon various annealings, while for Ru-
1222 the wider-range oxygen-content tuning is possible. When heated in H2/Ar atmosphere the both phases release
oxygen upon breaking down to mixtures of metals (Ru and Cu) and simple oxides (CeO2, Gd2O3, and SrO) in two
distinct steps around 300 and 450 °C. This reductive decomposition reaction carried out in a thermobalance is utilized
in precise oxygen content determination for these phases. It is found that the 100-atm O2-annealed Ru-1212 sample is
nearly stoichiometric, while the similarly treated Ru-1222 sample is clearly oxygen deficient. X-ray absorption near-
edge structure (XANES) spectroscopy is applied to estimate the valence of Ru in the samples. In spite of the fact that
the Ru-1212 phase was shown to possess less oxygen-deficient RuO2ꢀd layer, the valence of Ru as probed with XANES
is found to be lower in Ru-1212 than in Ru-1222.
Ó 2003 Elsevier B.V. All rights reserved.
PACS: 74.25.Ha; 74.72.Jt; 75.25.+z; 75.30.Cr
Keywords: Rutheno cuprates; Oxygen stoichiometry; Thermogravimetric analysis
1. Introduction
samples of RuSr2[(Eu,Gd)0:7Ce0:3]2Cu2O10ꢀd (Ru-
1222) [1] and RuSr2GdCu2O8ꢀd (Ru-1212) [2],
these ‘‘magnetosuperconductor’’ phases have re-
ceived continuous attention. Despite the intensive
research made on the both phases, their basic
characteristics in terms of oxygen non-stoichio-
metry, carrier concentration and valence state of Ru
have not been clarified yet. In the present contri-
bution we present the results of a systematic study
Since the first reports on the coexistence of
high-Tc superconductivity and ferromagnetism in
* Corresponding author. Tel.: +81-45-924-5333; fax: +81-45-
924-5365.
E-mail address: karppinen@msl.titech.ac.jp (M. Karppi-
nen).
0921-4534/$ - see front matter Ó 2003 Elsevier B.V. All rights reserved.
doi:10.1016/S0921-4534(03)01067-0

88
M. Matvejeff et al. / Physica C 392–396 (2003) 87–92
on the oxygen non-stoichiometry in Ru-1212 and
Ru-1222 samples based on thermogravimetric
(TG) analyses carried out in O2, Ar and H2/Ar
atmospheres. The TG experiments in O2 and Ar
atmospheres reveal us the degree of oxygen-con-
tent tunability in these phases. The experiments in
H2/Ar, on the other hand, enable us to determine
the absolute oxygen content of the sample. Fur-
thermore, we apply Ru LIII-edge X-ray absorption
near-edge structure (XANES) spectroscopy to
address the question on the valence state of Ru in
the Ru-1212 and Ru-1222 phases. Prior to the
present study, the very same samples have under-
gone thorough characterizations for structural,
magnetic and superconductivity properties. The
results of these studies have been reported else-
where [3–6].
The structure of Ru-1212 is derived from that
of RBa2Cu3O7ꢀd or CuBa2RCu2O7ꢀd (Cu-1212;
R ¼ rare earth element) with Cu in the charge
reservoir replaced by Ru, such that the CuO1ꢀd
chain is replaced by a RuO2ꢀd sheet. The Ru-1222
structure, on the other hand, results from that of
Ru-1212 when inserting a (R,Ce)–O2–(R,Ce) flu-
orite-type block, instead of the single oxygen-free
R layer, between the two adjacent CuO2 planes.
Now the question is whether the oxygen stoichio-
metry of the Ru-1212 and Ru-1222 phases can be
controlled in the same continuous manner as that
well established for Cu-1212 [7,8].
atmosphere. The phase purity and lattice para-
meters of the synthesized samples were checked
by X-ray diffraction (XRD; MAC Science:
MXP18VAHF22; CuKa radiation). The Tc values
reported for the superconductivity transition seen
in the samples are from magnetization measure-
ments carried out with a SQUID magnetometer
(Quantum Design: MPMS-XL) [3–6].
The TG analyses (Perkin Elmer: TAS-7) were
carried out in Ar, O2 and 5% H2/95% Ar atmo-
spheres. The mass of the sample was ca. 15 mg and
the heating rate was 1 °C/min. For the experiments
carried out in O2 atmosphere not only the heating
curve but also the cooling curve were recorded to
investigate the reversibility of the oxygen deple-
tion/uptake process. The Ru LIII-edge XANES
spectroscopic measurements were performed at the
BL15B beamline of the Synchrotron Radiation
Research Center in Hsinchu, Taiwan. Details of
the experimental procedure employed were as de-
scribed earlier [9–11].
3. Results and discussion
As commonly seen in the reported XRD pat-
terns of Ru-1212 samples, a small amount of
SrRuO3 was detected as an impurity phase also for
the present Ru-1212 samples. The Ru-1222 phase
was more easily obtained in single-phase form,
although a tiny trace of SrRuO3 was distinguished
from the XRD patterns even for the Ru-1222
samples. For the both phases, the relative amount
of SrRuO3 in the samples was considered to be
small enough (<5 mass%) not so as to affect the
results of TG and XANES analyses.
2. Experimental
Samples of RuSr2GdCu2O8ꢀd (Ru-1212) and
RuSr2(Gd0:75Ce0:25)2Cu2O10ꢀd (Ru-1222) were syn-
thesized through a solid-state reaction route from
stoichiometric amounts of RuO2, SrO2, Gd2O3,
CeO2 and CuO. Calcinations were carried out at
1000 °C, 1020 °C and 1040 °C for 24 h at each
temperature with intermediate grindings. The pow-
der samples thus obtained were pelletized and an-
nealed in a flow of oxygen at 1075 °C for 40 h with
a slow cooling over a span of 20 h down to room
temperature. A portion of these as-synthesized
samples was further annealed in 100 atm O2 at-
mosphere for 100 h at 420 °C and subsequently
cooled slowly to room temperature in the same
The lattice parameters for the as-synthe-
sized Ru-1212 sample were determined in space
ꢀ
group P4/mmm at: a ¼ b ¼ 3:821ð1Þ A and c ¼
ꢀ
11:476ð1Þ A. Essentially no difference was found
in the lattice parameters for the Ru-1212 sample
after annealed in 100 atm O2, suggesting that
the oxygen content was hardly changed upon the
O2 annealing. Also the Tc remained unchanged
being at 20 K. For the Ru-1222 samples the
lattice parameters were determined in space group
ꢀ
I4/mmm. They are: a ¼ b ¼ 3:834ð1Þ A and c ¼
ꢀ
28:493ð1Þ A for the as-synthesized sample and

M. Matvejeff et al. / Physica C 392–396 (2003) 87–92
89
ꢀ
ꢀ
a ¼ b ¼ 3:833ð1Þ A and c ¼ 28:393ð1Þ A for the
amount of oxygen released from the 100-atm O2-
annealed sample by 900 °C upon heating in O2
(Dd ¼ 0:27ð3Þ) was almost the same as that from
the same sample heat-treated in Ar. During the
slow (1 °C/min) cooling in O2 from 900 °C down to
room temperature the sample re-accommodated
part of the oxygen released from it during heating.
However, not all of the oxygen lost in heating was
taken back to the sample during the cooling. This
is due to the fact that the initial annealing of the
sample had been performed in 100 atm O2, while
the TG measurement was performed in 1 atm O2.
The difference in oxygen content between the 100-
atm and 1-atm O2-annealed samples was estimated
at 0.16(3) from the obtained TG curve for the
heating and cooling cycles.
Experiments performed for the Ru-1212-phase
samples in a thermobalance in O2 and Ar atmo-
spheres indicated that, as compared to the Ru-
1222 phase, the oxygen in the Ru-1212 structure is
relatively immovable and only small changes in the
oxygen stoichiometry were seen. In Fig. 1(b), a
representative TG curve for an Ar annealing of as-
synthesized Ru-1212 sample is presented. The
onset for a weight loss seen somewhat above
750 °C is due to the decomposition of the phase.
The absolute oxygen contents of the samples of
both Ru-1212 and Ru-1222 phases were deter-
mined based on a TG heat treatment in a strongly
reductive 5% H2/95% Ar atmosphere. We present
examples of such TG curves for the 100-atm O2-
annealed Ru-1222 sample and the as-synthesized
Ru-1212 sample in Fig. 2(a) and (b), respectively.
For both the phases the decomposition of sample
occurs in two distinct steps about 200–350 and
400–500 °C. To clarify the process of the sample
decomposition during the different steps of re-
duction, additional TG measurements were per-
formed in the same atmosphere for the simple
oxides of the constituent metals, i.e. RuO2, CuO,
Cu2O and CeO2. The TG curves obtained are
shown in Fig. 3. It is seen that RuO2, CuO and
Cu2O decompose into corresponding metals at low
temperatures of about 100, 150 and 330 °C, re-
spectively, while CeO2 remains stable at least up to
750 °C. We therefore calculated the exact oxygen
contents of the Ru-1212 and Ru-1222 samples
from the weight losses seen in the H2/Ar reduction
100-atm O2-annealed sample. The slightly de-
creased c-axis parameter of the O2-annealed sam-
ple as compared to that of the as-synthesized
Ru-1222 sample indicates successful introduction
of extra oxygen into the Ru-1222 structure upon
the high-pressure O2-annealing. Accordingly the
value of Tc increased from 23 to 43 K.
When heating the Ru-1222 samples in a ther-
mobalance, they exhibited definite changes in
weight, apparently caused by changes in the oxy-
gen content. A representative TG curve obtained
for the 100-atm O2-annealed Ru-1222 sample in an
Ar atmosphere is shown in Fig. 1(a). The phase
decomposition was found to start above 900 °C.
The total amount of oxygen loss per formula unit
by 900 °C, i.e. before the phase decomposition,
was calculated at Dd ¼ 0:28ð3Þ. As heated in O2
the Ru-1222 samples release oxygen as well. The
Fig. 1. TG curves recorded for (a) the 100-atm O2-annealed
Ru-1222 sample, and (b) the as-synthesized Ru-1212 sample in
Ar atmosphere. The heating rate was 1 °C/min and the mass of
the sample ca. 15 mg.

90
M. Matvejeff et al. / Physica C 392–396 (2003) 87–92
CeO2, and Ru and Cu metals. The results are
presented in Table 1, implying that the 100-atm
O2-annealed Ru-1212 sample is stoichiometric
within the error bars of the analysis, while the 100-
atm O2-annealed Ru-1222 sample is clearly oxy-
gen deficient. Furthermore, as already assumed
based on the lattice parameters and the Tc values,
the difference between the 100-atm and 1-atm
O2-annealed samples is larger for the Ru-1222
phase than for the Ru-1212 phase, see also the
data for relevant samples in Refs. [12,13]. In Table
1, we also show the absolute oxygen content values
for the samples annealed in Ar up to the highest
temperatures before the break-down of the struc-
ture, i.e. 750 °C for Ru-1212 and 900 °C for Ru-
1222. These values, i.e. 7.80(5) for Ru-1212 and
9.35(5) for Ru-1222, represent the minimum oxy-
gen contents tolerated by these phases.
The Ru LIII-edge XANES spectra were recorded
for the as-synthesized Ru-1212 sample and the
as-synthesized and 100-atm O2-annealed Ru-1222
samples. The obtained spectra were analyzed quan-
titatively by fitting them to certain linear combi-
nations of spectra for reference materials for
RuIV(Sr2RuO4) and RuV(Sr2GdRuO6). The same
fitting approach was for the first time applied in
Refs. [9,10]. For the two Ru-1222 samples the
spectra and fitting were shown in Ref. [11]. Here we
show only the spectrum of the as-synthesized Ru-
1212 sample in Fig. 4. All the three samples were
found to lie between the two reference materials in
terms of the ruthenium valence. Fitting the spectra
revealed a valence value of +4.74(5) for Ru in the
as-synthesized Ru-1222 sample and +4.81(5) in the
100-atm O2-annealed Ru-1222 sample [11]. It seems
that the valence of Ru in Ru-1222 depends on the
oxygen content, thus indirectly suggesting that
the changes in oxygen stoichiometry occur in the
RuO2ꢀd layer. For the as-synthesized Ru-1212
sample a valence value of +4.62(5) was obtained.
This value is very close to that reported in Ref. [9]
for another RuSr2GdCu2O8ꢀd sample, i.e. +4.60,
with the final annealing performed in O2 at 1060 °C.
Though it would be tempting to do so, we do
not calculate any precise valence values for Cu
from the obtained oxygen content and Ru valence
values. This is due to the fact that considering the
error bars for our TG and XANES analyses, we
Fig. 2. TG curves recorded for (a) the 100-atm O2-annealed
Ru-1222 sample, and (b) the as-synthesized Ru-1212 sample in
95% Ar/5% H2 atmosphere. The heating rate was 1 °C/min and
the mass of the sample ca. 15 mg.
Fig. 3. TG curves recorded for (a) RuO2, (b) CuO, (c) Cu2O,
and (d) CeO2 in 95% Ar/5% H2 atmosphere. The heating rate
was 1 °C/min and the mass of the sample ca. 15 mg.
curves by 550 °C assuming the final decomposition
product to be a mixture of oxides, SrO, Gd2O3 and

M. Matvejeff et al. / Physica C 392–396 (2003) 87–92
91
Table 1
The value of oxygen content as determined from the TG analysis for the variously treated Ru-1212 and Ru-1222 samples (see the text)
Synthesis/annealing conditions
RuSr2GdCu2O8ꢀd (Ru-1212)
RuSr2(Gd0:75Ce0:25)2Cu2O10ꢀd (Ru-1222)
100-atm O2, 420 °C
1-atm O2, 1075 °C (as-synthesized)
1-atm Ar, 750/900 °C (Ru-1212/Ru-1222)
7.98(5) (Tc ¼ 20 K)
7.93(5) (Tc ¼ 20 K)
7.80(5)
9.63(5) (Tc ¼ 43 K)
9.54(5) (Tc ¼ 23 K)
9.35(5)
The value of Tc is given in parentheses when being measured.
chiometric, while the Ru-1222 phase was clearly
oxygen-deficient even after 100-atm O2-annealing.
Also shown was that the oxygen content in the Ru-
1212 phase remains less affected by various
annealings, while for the Ru-1222 sample a wider-
range oxygen-content tuning is possible. We also
derived the conditions for the TG H2/Ar reduction
to determine the absolute oxygen contents in
samples of Ru-1212 and Ru-1222.
Acknowledgements
Fig. 4. Ru LIII-edge XANES spectrum collected for the as-
synthesized Ru-1212 sample. Also shown are the spectra for
reference samples Sr2RuO4 (RuIV) and Sr2GdRuO6 (RuV).
The present work has been supported by a
Grant-in-Aid for Scientific Research (grant no.
11305002) from the Ministry of Education, Science
and Culture of Japan and by the Academy of
Finland (decision no. 46039). M. Matvejeff ac-
knowledges Finnish Cultural Foundation, Sasak-
awa Foundation and the Komppa Foundation for
grants for Ph.D. studies.
believe that such estimation would end up with
not-reliable-enough Cu valence values. However, a
rough assessment suggests that the valence of Cu is
lower in the Ru-1222-phase samples than in the
samples of the Ru-1212 phase. The considerably
larger oxygen non-stoichiometry of the Ru-1222
phase, as compared to the Ru-1212 phase, might
let us to believe that both the RuO2ꢀd sheet and the
(Gd0:75Ce0:25)–O2–(Gd0:75Ce0:25) fluorite block in
Ru-1222 are non-stoichiometric in terms of oxygen
content. However, for another M-1222 phase, i.e.
References
[1] I. Felner, U. Asaf, Y. Lavi, O. Milio, Phys. Rev. B 55
(1997) 3374.
[2] C. Bernhard, J.L. Tallon, Ch. Niedermayer, Th. Blasius, A.
€
Golnik, E. Brucher, R.K. Kremer, D.R. Noakes, C.E.
Co-1222 at the stoichiometry of CoSr2(Y0:75
-
Stronack, E.J. Asnaldo, Phys. Rev. B 59 (1999) 14099.
[3] V.P.S Awana, J. Nakamura, M. Karppinen, H. Yamauchi,
S.K. Malik, W.B. Yelon, Physica C 357–360 (2001) 121.
[4] V.P.S. Awana, S. Ichihara, M. Karppinen, H. Yamauchi,
Physica C 378–381 (2002) 249.
Ce0:25)2Cu2O9ꢁd, we recently confirmed that its
(Y0:75Ce0:25)–O2–(Y0:75Ce0:25) fluorite block is oxy-
gen-stoichiometric even after the sample had been
exposed to various reductive and oxygenative
annealings [14].
[5] V.P.S. Awana, S. Ichihara, J. Nakamura, M. Karppinen,
H. Yamauchi, J. Yang, W.B. Yelon, W.J. James, S.K.
Malik, J. Appl. Phys. 91 (2002) 8501.
[6] V.P.S. Awana, M. Karppinen, H. Yamauchi, Pramana 58
(2002) 809.
4. Conclusion
[7] M. Karppinen, H. Yamauchi, Mater. Sci. Eng. B 54 (1998)
92.
In the present study we have shown that the
Ru-1212 phase is easily obtained as oxygen stoi-
€
[8] M. Karppinen, L. Niinisto, H. Yamauchi, J. Therm. Anal.
48 (1997) 1123.

92
M. Matvejeff et al. / Physica C 392–396 (2003) 87–92
[9] R.S. Liu, L.-Y. Jang, H.-H. Hung, J.L. Tallon, Phys. Rev.
B 63 (2001) 212505.
[12] I. Felner, U. Asaf, F. Ritter, P.W. Klamut, B. Dabrowski,
Physica C 364–365 (2001) 368.
[10] G.V.M. Williams, L.-Y. Jang, R.S. Liu, Phys. Rev. B 65
(2002) 64508.
[13] R.W. Henn, H. Friedrich, V.P.S. Awana, E. Gmelin,
Physica C 341–348 (2000) 457.
[11] V.P.S. Awana, M. Karppinen, H. Yamauchi, M. Matvejeff,
R.S. Liu, L.-Y. Jang, J. Low. Temp. Phys. 131 (2003) 1211.
[14] M. Karppinen, V.P.S. Awana, Y. Morita, H. Yamauchi,
Physica C (2003); these Proceedings.