Upload
nd-zhigadlo
View
220
Download
0
Embed Size (px)
Citation preview
High-temperature phase changes in RuSr2GdCu2O8
and physical properties
N.D. Zhigadlo a,1, P. Odier a,*, J.Ch. Marty b, P. Bordet a, A. Sulpice c
a Laboratoire de Cristallographie, CNRS, 25 Avenue des Martyrs, BP 166, F-38042 Grenoble Cedex 09, Franceb LAIMAN, Universit�ee de Savoie, 9 Rue de l�Arc-en-Ciel, BP 240, F-74942 Annecy-le-Vieux Cedex, France
c Centre de Recherche sur les Tr�ees Basses Temp�eeratures, CNRS, 25 Avenue des Martyrs, BP 166, F-38042 Grenoble Cedex 09, France
Received 30 May 2002; received in revised form 30 October 2002; accepted 5 November 2002
Abstract
A sol–gel method was successfully applied to synthesize RuSr2GdCu2O8 (Ru-1212) as a single phase. The crystal-
lization of Ru-1212 arises at 950 �C under O2 flow with a small amount of secondary phase. The pure Ru-1212 phase is
achieved after heating at 1020 �C in O2 flow. The X-ray diffraction (XRD) pattern is refined (Rietveld) in the tetragonal
space group P4/mmm with lattice parameters a ¼ 3:83904ð9Þ �AA and c ¼ 11:5678ð4Þ �AA. In situ high-temperature XRD
and differential thermal analysis coupled with thermal-weight measurement show a structural decomposition of the
phase at Td ¼ 1050 �C followed by a partial melting at Tm ¼ 1118 �C. The decomposition produces crystallized
Sr2GdRuO6, SrRuO3 phases and a mixture rich in copper containing ‘‘Sr, Gd, Cu, O’’ that does not diffract X-ray. This
phase reduces to Cuþ at Tm with an important weight loss and a significant amount of liquid. The grain size and/or
inappropriate grain boundaries of the pure phase treated below Tm does not permit to detect in sol–gel samples su-
perconductivity otherwise observed in compounds prepared by solid-state reaction.
� 2002 Elsevier Science B.V. All rights reserved.
Keywords: Rutheno-cuprates; RuSr2GdCu2O8; Sol–gel; In situ XRD; DTA, magnetic measurements
1. Introduction
A new class of rutheno-cuprates, RuSr2Ln-
Cu2O8 (Ru-1212) and RuSr2 (Ln1þx-Ce1�x)Cu2O10
(Ru-1222) (Ln ¼ Sm, Eu and Gd), belonging to the
layered cuprate family was synthesized in 1995 [1].
Coexisting magnetism and superconductivity (SC)have been claimed in the R-1222-type (R ¼ Eu and
Gd) RuSr2R1:5Ce0:5Cu2O10�y [2] and more recently
in the 1212 type (RuSr2GdCu2O8) [3,4]. However
these properties are strongly debated since SC is
apparently sample dependent. The properties of
Ru-1212 reported in the literature range from
non-SC materials with a magnetic temperature
ordering ðTAFÞ below 150 K [5,6] to ‘‘bulk SC’’samples with Tc ¼ 15–46 K and TAF ¼ 132 K de-
pending on sample preparation [3,4,7]. A recent
report [8] suggests a correlation between the or-
dered magnetic state and SC. The ‘‘bulk’’ nature of
the SC is also strongly controversial [8–10]. The
*Corresponding author. Tel.: +33-4-76-88-10-45; fax: +33-4-
76-88-10-38.
E-mail address: [email protected] (P. Odier).1 On leave from Institute of Solid State and Semiconductor
Physics, P. Brovki 17, Minsk 220072, Belarus; associate
researcher at D�eepartment des Sciences Chimiques, CNRS,
France.
0921-4534/02/$ - see front matter � 2002 Elsevier Science B.V. All rights reserved.
doi:10.1016/S0921-4534(02)02306-7
Physica C 387 (2003) 347–358
www.elsevier.com/locate/physc
most common opinion is that CuO2 planes are the
SC planes and are homogeneous but this is not
obvious to Blackstead et al. [11] who find evidences
for magnetically ordered Cu siting either in the
CuO2 plane or in the RuO2 magnetic plane. In the
former case the SC is not homogeneous, whichcontradicts Ref. [3] who claim for homogeneous SC
(CuO2) and magnetism (RuO2). The latter case also
contradicts Ref. [3] but is supported by Ref. [8] who
proves Ru substitutions by Cu to be possible. To
reconcile, Blackstead et al. [11] suggests that SrO
can be the superconducting plane. The magnetic
nature of the RuO6 plane is presently not clear
(weak ferromagnetism or AF [12]) and Ru cationsare reported to be in an intermediate valency (60%
5þ, 40% 4þ) state according to recent measure-
ments by 99;101Ru zero-field NMR and EXAFS [12–
14].
Oxygen annealing, normally associatedwith hole
doping in the high-temperature SC, has no clear
effects in the case of Ru-1212 compounds. For non-
SC samples it has no influence [5] while it progres-sively enhances the SC on superconducting samples
[8,15] but without any detectable modifications of
the oxygen stoichiometry [8]. Then ordering of Ru/
Cu atoms, chemical non-stoichiometry or changes
in structural distortions can be considered as pos-
sible reasons for varying the physical properties
[8,16]. Because both Ru and Cu can support mixed
valency in this system, modification of the magneticand SC properties can originate from modified
charge distribution between the RuO2 and CuO2
layers in response to local distortions. Recent neu-
tron diffraction experiments showed highly ordered
Cu and Ru layers and presence of cooperative ro-
tations of RuO6 octahedra [17].
In summary, the structure itself cannot distin-
guish between non-SC and SC samples. SC in Ru-1212 is often found not in the bulk and may be
associated with a very subtle cationic distribution,
perhaps in interfaces. Clearly, more work has to
be done on the material preparation and on its
characterization.
Despite several years of intensive work by the
scientific community concerning this phase, almost
no studies have been published on the formationmechanism leading to Ru-1212 neither on the re-
action path involved. This is an important feature
which has to be taken into account for under-
standing the SC in Ru-1212. In all papers, the Ru-
1212 phase has been synthesized by the solid-state
route which is always associated with SrRuO3
and/or Sr2GdRuO6. These phases are difficult to
eliminate [1,6] even after prolonged annealing athigh temperatures but do not seem to suppress the
SC. It is not clear then if the superconducting
phase concerns the stoichiometric phase or a sub-
stituted or deficient one. Conversely, sol–gel pro-
vides advantages such as chemical homogeneity
and chemical reactivity, both of which are impor-
tant requirements in obtaining ceramics of high
quality. This technique has been used successfullyin numerous cases relevant to electroceramics and
superconductors [18]. The basic idea of sol–gel is to
start from a solution of cations and to jellify it by
polymerization. The gel can then be easily manip-
ulated. Its drying gives a xerogel that transforms
to highly reactive nanosize oxides upon heating.
In this paper we report for the first time results
on the synthesis of pure Ru-1212 phase by a sol–gel method. In addition, in situ high-temperature
X-ray diffraction (XRD) and differential thermal
analysis (DTA) studies coupled with thermal-
weight analysis (TGA) have been undertaken in
order to identify the phases involved in the Ru-
1212 synthesis. To our knowledge, none of these
properties have been reported so far. Finally, we
discuss the implications of these factors on theoccurrence of SC in our compounds.
2. Experimental
2.1. Gel formation
The formation of RuSr2GdCu2O8 gels is carriedout by dissolving separately SrCO3 (Aldrich,
99.9%), CuO (Aldrich, 99.9%) and Gd2O3 (Pechi-
ney-Saint-Gobain, 99.9%) in nitric acid. Ruthen-
ium(III) nitrosyl nitrate solution, Ru(NO)(NO3) �H2O (1.5% Ru), was the source of Ru. Its pH is
adjusted to 6 with ammoniac addition. Finally, all
solutions are mixed in stoichiometric ratio, with
addition of water to reach a total volume of ap-proximately 150 ml for obtaining 5 g of Ru-1212
phase. Added to this solution was 10 wt.% of
348 N.D. Zhigadlo et al. / Physica C 387 (2003) 347–358
acrylamide monomers (Aldrich, 99%), i.e., H2C@CHCONH2; 1 wt.% of N ,N 0-methylenebisacryla-
mide (Aldrich, 99%), i.e., (H2C@CHCONH)2CH2
and a few milligrams of a reticulating agent a,a0-
azoisobutyronitrile (Fluka, 98%), i.e., C8H12N4 to
perform the polymerization which is easilyachieved by heating to 80 �C. This forms rather
hard black–brown gels that are stable with time.
Note that the acrylamide monomers react with the
copper to form a stable complex with the amino
ligand. This impedes the polymerization process
unless the equivalent quantity of monomers is
added to the system. To prevent this problem, we
used EDTA, [CH2N(CH2COOH)2]2[CH2N(CH2-CO2H)2] as an efficient chelating agent, allowing to
isolate the copper cation from acrylamide mono-
mers activity.
For the purpose, of comparison samples were
also formed by solid-state reaction. Dry RuO2
powder, SrCO3, Gd2O3 and CuO powders were
mixed in stoichiometric ratio (Ru-1212) in a an
automatic agate mortar under acetone. The re-sulting powder (2.7 g), after drying was calcined a
first time at 920 �C for 24 h in static air and then at
930 �C (24 h) under Ar flow after intermediate
grinding under acetone. After a subsequent dry
grinding, the fine powder was then reacted at 1030
�C for 24 h under O2 flow, from what resulted a
porous compact, called ‘‘ssr’’. This batch serves to
fabricate samples ssrA, ssrB, ssrC and ssrD, ac-cording to the procedure recapped in Table 1.
The phase composition of the powder samples
was investigated by XRD (Siemens D-5000 dif-
fractometer in transmission mode and Bruker D8
with a high resolution reflection set up), using
CuKa1 radiation. In situ high-temperature XRD
were carried out in a high-temperature diffrac-tometer (equipped with a CAS120 Inel detector).
The measurements were performed using CoKa1
radiation in a range 10–120� 2h. The temperature
was measured with a thermocouple inserted in the
sample holder close to the sample. Calibration
using melting of Ag and Au permit a rather high
degree of confidence on the T measurement (a few
degrees) up to 1050 �C. At higher temperature (inthe range of 1100 �C), the uncertainty increases
with temperature and the sample temperature may
be smaller than measured by the thermocouple
(10–20 �C). DTA and TGA measurements were
carried out in a microbalance (Setaram TAG1500-
Lyon, France) under O2 flow up to 1200 �C at a
rate of 10 �C/min. SEM (Jeol-840) and EDX were
used to observe the microstructure and the com-position of the compounds.
The zero-field-cooled (ZFC) and field-cooled
(FC) dc magnetic measurements in the range
of 2–300 K were performed in a commercial
SQUID (Metronic Ingenierie-France). The resis-
tive measurements were made by standard four-
probe technique on bar-shaped pieces cut out
from the pellets with contacts made with silverpaint.
Table 1
Samples synthesized, irreversibility between ZFC and FC in the magnetic transition regime, annealing conditions, occurrence of SC
TAF (K) DTirr (K) Annealing
Solid-state reaction
ssra 134 0 1030 �C–50 h (50 �C/h) NS
ssrAa 136 2 1030 �C–(50 h)þ 160 h (50 �C/h) S
ssrBa 136 10 1050 �C–160 h (50 �C/h) S
ssrC 1030 �C cumulated 480 h (50 �C/h) S
ssrD ssrA then sintering at 1030 �C–160 h (50 �C/h) S
Sol–gel
sga 134 0 1020 �C–(476 h) then 48 h (50 �C/h) NS
sgA (pur) 134 21 1030 �C–160 h (50 �C/h) NS
sgB (pur) 142 17 1040 �C ð�Þ (50 �C/h) NS
sgCb 136 34 1050 �C–160 h (50 �C/h) NS
SgD (pur) 1030 �C 480 h (50 �C/h)a Impurity: SrRuO3 � 5 mass%.b Impurity: SrRuO3 � 10 mass%.
N.D. Zhigadlo et al. / Physica C 387 (2003) 347–358 349
3. Results
To study the Ru-1212 phase formation, the gel
samples have been heat treated at different tem-
peratures and then characterized by XRD. At first,the gel was dried in a commercial microwave ap-
paratus for a few minutes. During this process, the
dehydration occurs first and is followed by a par-
tial ignition of the organic components. It results
an extremely porous brown xerogel that is then
easily crushed into an homogeneous powder. It was
then heat treated in a ventilated oven (100 �C/h up
to 700 �C with a dwell time of 24 h) under lowoxygen partial pressure (N2 flow). This thermal
treatment removes the remaining organic phases
and provides a fine nanoscale powder. Fig. 1(a)
shows the XRD pattern for the product after the
calcination process at 700 �C, it is composed of
SrCO3, CuO,Gd2Ru2O7, Sr2GdRuO6 and SrRuO3.
In these synthesis conditions, we do not see any
Bragg peaks belonging to the Ru-1212 phase. Af-ter treating at 950 �C in air (static) for 18 h the X-
ray diffractogram is completely changed (Fig. 1(b))
to the typical pattern of Ru-1212 phase with a
small amount of SrRuO3. The contribution of this
impurity phase decreases significantly after an-
nealing at 950 �C in O2 flow for 15 h (Fig. 1(c)) and
then for 44 h (Fig. 1(d)). The pure Ru-1212 phase
is achieved after a final treatment at 1020 �C in O2
flow for 48 h (Fig. 1(e)).
The Ru-1212 product was then analyzed by
performing Rietveld refinements using FullProf.2k
on high resolution X-ray powder diffracted data.No impurity could be detected in the spectrum.
The recorded data are in excellent agreement with
the calculated profile, Fig. 2, based on a tetragonal
space group P4/mmm with calculated lattice pa-
rameters a ¼ 3:83904ð9Þ �AA and c ¼ 11:5678ð4Þ �AAand atomic positions not significantly different
from those reported previously [17,19]. A distor-
tion typical of this phase arises due to rotations ofthe RuO6 octahedra amounting 13.2� around the
c-axis. We could not exclude possible substitutions
on Ru site (by Cu) or on Gd site (by Sr).
In situ high-temperature XRD and DTA/TGA
studies allow to identify the behavior of the phase
at high temperature, in the temperature range
where syntheses are generally conducted to provide
the SC phase. Both experiments have been done inO2 flow. The XRD spectra were taken for in-
creasing temperatures, each 15 min that is the ac-
quisition time needed for a spectrum. Each 30 min,
the temperature was raised by steps of 15–20 �C.Fig. 3 shows an overview of the measurements on
20 40 60 800.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0 SrRuO 3
a
b
c
d
e
Fig. 1. XRD diffractograms of RuSr2GdCu2O8 after treatments at 700 �C in N2 flow for 10 h (a); 950 �C in air for 18 h (b); 950 �C in
O2 flow for 15 h (c); 950 �C in O2 flow for 44 h (d) and 1020 �C in O2 flow for 48 h (e).
350 N.D. Zhigadlo et al. / Physica C 387 (2003) 347–358
Ru-1212 starting on the sol–gel powder annealed at1020 �C. The diffractogram displayed at the bottom
was recorded at room temperature and the next
were recorded at increasing temperatures from
1015 �C. A number of details become obvious.
There is no phase transformation of Ru-1212 up to
1050 �C in oxygen flow. However a small change of
the main peak can be seen at this temperature, re-
flecting the onset of the transformation. Above1050 �C, the Ru-1212 phase decomposes, produc-
ing two main crystalline phases: Sr2GdRuO6 and
SrRuO3, together with a subsequent diffusion
bump in the 20–35� 2h range, suggesting a contri-
bution from a liquid phase. From both in situ X-
ray and magnetic measurements performed on
quenched samples we exclude the competitive for-
mation of Gd2CuO4. Although a large amount ofcopper oxides (2 mol of CuO/mol of Ru-1212, i.e.
23 wt.% of Ru-1212) is expected to be rejected in
this decomposition, no crystallized CuO, neither
any new clearly identified peaks, can be seen by
XRD between 1050 and 1150 �C. At 1065 �C three
crystalline phases coexist while at 1115 �C only
Sr2GdRuO6 and SrRuO3 are visible. By studies on
products formed by solid-state reaction at 1030 �C(similar to sample ssr), we have confirmed that this
Fig. 2. Rietveld refinement profile showing observed (o) and calculated (line) intensities. The markers below the profile correspond to
the Bragg peak positions for RuSr2GdCu2O8 sample annealed at 1020 �C in O2 flow for 48 h. The difference between observed and
calculated intensities is shown at the bottom.
0.00.51.01.52.02.53.03.54.04.55.05.56.06.57.0
<1150°C
1115°C
1065°C
1050°C
1030°C
1015°C
25°C
Fig. 3. An overview of a high-temperature XRD of the
RuSr2GdCu2O8. The intensity has been normalized for an
easier comparison. The Ru-1212 phase start to decompose be-
tween 1040 and 1050 �C.
N.D. Zhigadlo et al. / Physica C 387 (2003) 347–358 351
decomposition is intrinsic to the Ru-1212 phaseand not a characteristic of the sol–gel phase. The
high-temperature phases recombine with a rather
high kinetics because the phase is recovered on
cooling. Note that long annealing (160 h) at 1050
�C induces SrRuO3 impurity to come up, Fig. 4,
due to a chemical shift of the sample composition.
This figure compares two pieces of a sample made
by sol–gel and reacted under oxygen flow severaltimes (cumulated 476 h) at increasing temperatures
up to 1020 �C. This sample has a small content
of Sr2GdRuO6 impurity that was removed after
annealing 160 h at 1030 �C and slowly cooled
(50 �C/h). One part of this sample was subsequently
heated at 1050 �C for 160 h and slowly cooled
(50 �C/h). XRD, Fig. 4 shows that this sample has
now some SrRuO3 impurities (around 10 wt.%)and a broader Bragg (1 1 0) reflection suggesting
either the presence of some other phase or some
disordering.
Fig. 5 shows TGA and DTA data performed up
to 1100 �C under oxygen flow. There is a small
oxygen loss (0.1 wt.%) between 400 and 500 �C.We assume that this loss has the same nature than
in Cu-1212 (Y) phase (YBa2Cu3O7) but with amuch smaller extension [20]. It would correspond
in the case of Ru-1212 to a reduction of the O8
stoichiometry to O7:95 while the change is from O7
to O6:5 for Cu-1212. Obviously the Ru ion, due to
its high oxidation state, contributes to stabilize the
oxygen stoichiometry in Ru-1212, making the oxy-
gen annealing to have a negligible effect on the
oxygen stoichiometry [8], except eventually at the
grain boundaries [21]. At 1050 �C a sharp but
small oxygen loss (0.1 wt.%) is detected occurring
at the temperature where an endotherm indicates
the onset of the phase transformation noticed in
XRD.
In a further experiment, DTA have been per-
formed up to 1200 �C under O2 flow (Fig. 6). Itshows two endothermic peaks at 1050 �C (maxi-
mum) and 1118 �C (onset). None of these thermal
events can be assigned to Sr2GdRuO6 alone be-
cause we did not detect any transformation in this
compound up to 1450 �C nor to SrRuO3 which
melting point is above 2000 �C. The strong endo-
thermic signal recorded at 1118 �C is due to a
liquidus. The incongruent melting of this com-pound has been already mentioned to occur above
1100 �C by Lin et al. [22]. It happens with 1.2 wt.%
32.0 32.5 33.0 33.5 34.0
0.0
0.2
0.4
0.6
0.8
1.0
above decomposition (FWHM = 0.216°)
below decomposition (FWHM = 0.192°)
Ru-1212 by sol gel
Fig. 4. Comparison of the main Bragg reflection for a sol–gel
sample heated 160 h at 1030 �C and 160 h at 1050 �C.
-100 0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300-0.7
-0.6
-0.5
-0.4
-0.3
-0.2
-0.1
0.0
0.1 Ru-1212, sol gel (1020°C)O2 flowTGA: blanc substracted
-50
-40
-30
-20
-10
0
10
Fig. 5. TGA and DTA of RuSr2GdCu2O8 powder in oxygen
flow. The powder was processed from sol–gel and annealed at
1020 �C in O2 prior to its analysis.
700 750 800 850 900 950 1000 1050 1100 1150 1200 1250 1300 1350 1400-40
-20
0
20
II I
III
I (onset): 1040 °C, max 1053°CII(onset): 1118 °C, max 1129°C
DTA SG Ru-1212O2 flow
Fig. 6. DTA of the same powder as in Fig. 5 up to its melting.
352 N.D. Zhigadlo et al. / Physica C 387 (2003) 347–358
loss of oxygen which can be attributed to the
transformation of almost all Cu2þ into Cuþ. Note
that a significant drop of the X-ray diffracted in-
tensity (not shown because the spectra were nor-
malized for comparison in Fig. 3) and an increased
diffusion in the range 20–30� ð2hÞ is observed
above this temperature in agreement with a melt-
ing phenomenon. On cooling, a first peak appears
at the same temperature as for heating, it is at-
tributed to the same transformation which is then
non-hysteretic. This reflects a fast recombination
process as mentioned earlier. The shape of this
0
0.02
0.04
0.06
0.08
0.1
0 5 0 100 150 200
M(u
em
/g)
T
0
0.01
0.02
0.03
0.04
0.05
110 120 130 140 150 160
M(u
em
/g)
T
0
0.02
0.04
0.06
0.08
0.1
0 5 0 100 150 200
M(u
em
/g)
T
ZFCF C
0
0.005
0.01
0.015
0.02
0.025
0.03
0.035
0.04
110 120 130 140 150 160
M(u
em
/g)
T
ZFC
F C
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0 5 0 100 150 200
M(u
em
/g)
T
ZFC F C
0
0.005
0.01
0.015
0.02
0.025
0.03
0.035
0.04
110 120 130 140 150 160
M(u
em
/g)
T
ZFC
F C
sg(a)
ssr(b)
ssrB(c)
Fig. 7. ZFC and FC versus T , measured at 10 Oe, for selected samples to show that DTirr changes with synthesis conditions. (a) sg
obtained at 1020 �C; (b) ssr obtained at 1030 �C; (c) ssrB obtained at 1050 �C. The right hand part is a zoom in the range 110–160 K.
N.D. Zhigadlo et al. / Physica C 387 (2003) 347–358 353
peak suggests an invariant transformation. Its
larger intensity is due to a better thermal contact
with the crucible after the melting. The second
peak is then due to the crystallization of the
remaining liquid occurring with a rather large
undercooling (�80 �C) inherent with the reoxy-dation of Cuþ into Cu2þ. In most cases the oxygen
is not completely recovered on cooling and the
properties of the resultant specimens may be
drastically changed.
To confirm the Ru-1212 phase changes at high
temperatures a quench study was performed. Two
samples made from a pellet of a sol–gel powder
were fired in oxygen at 1060 and 1185 �C, held atthese temperatures for 1 h and then quenched on a
metal plate in air. The cooling rate was estimated
to be �250 �C/min. Even though the XRD pat-
terns of the sample quenched from 1060 �C do not
indicate the formation of impurities, secondary Sr-
rich and Cu-rich phases were detected in fractured
pieces by EDX and significant amount of liquid
phase by SEM. However their weight fraction was
small (<10%) because they were rather insensitive
to X-ray. The XRD pattern and EDX analysis
of the samples quenched from 1185 �C shows
Sr2GdRuO6, SrRuO3 as main phases, plus minor
phases: Sr3Ru2O7, and Cu2O (and CuO due to the
oxidation of the latter during cooling) and per-haps Gd2CuO4 that is difficult to distinguish from
SrRuO3. This also point towards a possible in-
variant nature of the transformation at 1050 �C.ZFC and FC magnetization curves at low fields
(10, 4, 2, 0.4 Oe), MðHÞ at 1.6 K have been
undertaken to detected any possible signature of
SC in the samples prepared. Considering that most
samples studied in the literature have been madeby solid-state reaction, we have also prepared one
batch of powder by solid-state reaction according
to the procedure described in part 2 to compare it
with our sol–gel samples. The magnetization ver-
sus T curves are described first, see Fig. 7. In all
samples we observe the typical Gd moments or-
dering (AF) at �3 K. In some samples, a positive
bump is detected at �20 K [5,21]. It is due to a
-0.001
0
0.001
0.002
0.003
0.004
0.005
0.006
0 10 20 30 40 50 60T
ZFC
FC
-0.007
-0.006
-0.005
-0.004
-0.003
-0.002
-0.001
0
0.001
0 10 20 30 40 50 60
M -
(C
uri
e-W
eiss
law
)
T
ZFC
FC
0.0005
0.001
0.0015
0.002
0.0025
0.003
0.0035
0.004
0 10 20 30 40 50 60 70
M(u
em/g
)
T
ZFC
FC
-0.0035
-0.003
-0.0025
-0.002
-0.0015
-0.001
-0.0005
0
0.0005
0 10 20 30 40 50 60 70
M -
(C
uri
e-W
eiss
law
)
T
FC
ZFC
M(u
em/g
)
(a)
(b)
ssrA
ssrB
Fig. 8. ZFC and FC magnetization measured at 0.4 Oe for (a) ssrA and (b) ssrB. In both cases the curve at left is after deducing the
Curie–Weiss contribution. In both samples the MðHÞ had a negative slope below 0.4 Oe.
354 N.D. Zhigadlo et al. / Physica C 387 (2003) 347–358
contribution of Sr2GdRuO6, a frequent impurity
of the system, that should not be confused with
SC.
At TAF ¼ 132–142 K a cusp in the ZFC curve
signs an AF ordering attributed to Ru atoms of
the Ru-1212 structure [23,24]. In FC experimenta spontaneous magnetization anticipates the AF
ordering and shows an high-temperature irrevers-
ibility that depends of the sample preparation
(Table 1, Fig. 7). We estimate this irreversibility by
the difference DTirr between the merging tempera-
ture of FC with ZFC curves (of course this de-
pends of the sensitivity of our apparatus) and TAF.
Remarkably DTirr depends of sample preparation,and varies from 0 to 34 K. This is not correlated
with the presence or not of SrRuO3 which ferro-
magnetically orders at 165 K, except eventually in
the samples called sgC (not shown) that contains a
rather large amount of SrRuO3 linked with a de-
gradation of the sample.
Systematic exploration of MðHÞ above 1.6 K
have been performed in order to detect possible SC.Sample ssr was not superconducting, while ssrA,
ssrB, ssrC and ssrD were found to be supercon-
ducting. For both former samples the SC is unam-
biguously a volume effect as evidenced by a negative
slope of the first magnetization MðHÞ at low field
and low temperature. The first critical field Hc1 is
very low, Hc1 < 0:4 Oe at 1.65 K. Below this field,
the magnetic shielding represents more than 50% ofthe sample volume. Furthermore, the ZFC and
FC measurements present an irreversible behavior
0 50 100 150 200 250 300
0.0
0.2
0.4
0.6
0.8
1.0
powder 1030°C 160 h ssrApowder 1050°C 160 h ssrB
0 50 100 150 200 250 300
0.0
0.2
0.4
0.6
0.8
1.0
powder 1030°C 160 h ssrApowder 1030°C 480 h ssrC
0 50 100 150 200 250 300
0.0
0.2
0.4
0.6
0.8
1.0
1.2
powder 1030°C 160 h ssrAsintered 1030°C 160 h ssrD
0 50 100 150 200 250 3000.5
1.0
1.5
2,0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
0 50 100 150 200 250 300
powder 1030°C 160 h ssrApowder 1050°C 160 h ssrB
0 50 100 150 200 250 300
8
0
powder 1030°C 160 h ssrApowder 1030°C 480 h ssrC
0 50 100 150 200 250 300
powder 1030°C 160 h ssrAsintered 1030°C 160 h ssrD
0 50 100 150 200 250 300
2.0
powder 1030°C 160h sgApowder 1050°C 160h sgC
(a) (c)
(b) (d)
Fig. 9. RðT Þ for samples ssrA, ssrB, ssrC and ssrD, sgA and sgC. This figure shows clear trends of annealing and sintering.
N.D. Zhigadlo et al. / Physica C 387 (2003) 347–358 355
below 12K that we attribute to the SC (Fig. 8). This
Fig. 8(a) and (b) shows on its right part, the mag-
netization after subtraction of the Curie–Weiss
contribution due to the magnetism of the sample. It
shows the usual behavior of superconducting ma-
terials. We observe a weak deviation to the Curie–Weiss law around 40 K which can be due to the
appearance of SC but could be also the sign of a
magnetic anomaly.However, the resistivity starts to
decrease very fast below 44 K (Fig. 9). It reinforces
the assumption that SC should be responsible for
the magnetization anomaly. At lower temperature,
for example below 27 K for ssrA, the resistance
become vanishingly small ðRlow T=RTc onset < 10�6Þ.It reproduces the results reported by Lorentz et al.
[25]. The high-temperature part is attributed to
Fig. 10. SEM microstructure of RuSr2GdCu2O8 after various annealing under pure O2 flow. (a) ssrA was a portion of ssr after an-
nealing 160 h at 1030 �C (cooled at 50 �C/h); ssrC was a part of ssrA grounded, compressed and sintered 160 h at 1030 �C (cooled at 50
�C/h). (b) sg was the as received powder from synthesis; sgD, the same after annealing under oxygen flow at 1030 �C for 480 h (cu-
mulated time at this temperature, cooling 50 �C/h).
356 N.D. Zhigadlo et al. / Physica C 387 (2003) 347–358
intra-grain while the low-temperature part is due to
the inter-grain counterpart. Samples ssrA, ssrB and
ssrC were powders annealed at different time and
temperatures with no intermediate grinding to
preserve grains integrity as suggested [25], they
originate from the same batch ssr. The annealingprovides some sintering of the grains of the powder
with a possible percolating path. Sample D was
sintered after a gentle grinding before (dry) com-
paction and sintering. Fig. 10(a) shows the micro-
structure of both with an obvious effect of sintering
process.
4. Discussion
The results reported in the present study show
that Ru-1212 decomposes above 1050 �C and re-
combines with a fast kinetic on cooling suggest-
ing an invariant transformation. On the other
hand, the sample composition may be irreversibly
changed if the sample is heated for a too long timein the T range above 1050 �C. It expresses that
either some species are removed by vaporization
(Ru oxides?) or by some liquid phase expelled
from the sample. It rises the question of the exact
stoichiometry of the samples after long standing
time at 1060 �C where most of the samples re-
ported so far are treated.
The high-temperature irreversibility betweenZFC and FC is an interesting point because it re-
flects some intimate changes in the samples. Ac-
cording to previous observations, the Ru moments
order antiferromagnetically (G type) at 132 K, but
the Ru moments are canted, which results in a net
ferromagnetic moment parallel to the ðabÞ plane.
This canting is influenced by an external field and
causes the irreversibility. Contrarily to the AF or-dering, it is very sensitive to the presence of a
magnetic rare earth (Gd versus Eu) on the rare
earth site and presumably to disorder in the RuO2
layer. Cu is a good candidate to this substitution as
shown first by the existence of Ru1�xSr2GdCu2þxO8
compositions [26] (reminiscent to CuBa2RECu2O7
structure) and to the recent observation of mag-
netically ordered Cu impurities on Ru sites witha locally inhomogeneous magnetization [11]. The
results summarized in Table 1 assume that our sol–
gel procedure or high-temperature heating in oxy-
gen favor such substitutions. In this case, copper
might be deficient in the CuO2 planes, which is not
good for development of SC.
According to the literature most samples are
prepared above Td, close to 1060 �C. Althoughrecombination occurs rapidly on cooling it may
be not quantitative and superconducting samples
may be off-stoichiometry. Our study on solid-state
reaction samples shows that SC appears in samples
prepared below the decomposition temperature Td(samples ssrA, ssrC and ssrD) or at Td (sample
ssrB), but our samples do contains some impuri-
ties. Therefore, we cannot conclude whether stoi-chiometry is essential or not for the onset of SC.
After analyzing the effect of heat treatments on
SC in our solid-state reaction samples, it becomes
clear that the annealing time has a major effect on
strengthening the inter-grain superconducting part
(Fig. 9(a)). However if the annealing temperature
is too high, i.e. above the decomposition T , thereverse is observed (Fig. 9(b)). We thought thatsintering could strengthen even more the grain
boundaries. The reverse effect is observed (Fig.
9(c)), proving that oxygen diffusion, which is in-
hibited in the sintered material, plays an important
role in the inter-grain SC development.
The samples made by sol–gel technique which
are pure from a structural point of view, are not
superconducting (after magnetization and resistiv-ity studies), although a weak accident is detected by
resistivity measurements in sgA sample below 25–
30 K (Fig. 9(d)). Again a too high annealing tem-
perature degrades the resistivity. Perhaps the grain
boundaries have not the appropriate composition.
By EDX we noticed a loss of Ru at the surface of
our sg samples that could be due to RuO2 selective
loss after such long heat treating in oxygen. Sucheffect is enhanced by the small grain size that is
shown in Fig. 10(b). On the other hand, the as
synthesized powder has a mean grain size of 0.6 lm(mean value) typical of this type of synthesis [27,28]
and even after 476 h (cumulated from two con-
secutive annealing at 1030 �C), the grains do not
grow larger that 1.7 lm (mean value). According to
the large penetration depth measured in similarsamples [25,29], the SC is probably very difficult to
be detected by magnetization studies.
N.D. Zhigadlo et al. / Physica C 387 (2003) 347–358 357
5. Conclusions
Pure phase of the rutheno-cuprate RuSr2-
GdCu2O8 has been synthesized by a sol–gel
method. High-temperature phase changes havebeen investigated through in situ XRD and DTA
studies. The pure phase is stable up to 1050 �C.Above this temperature, it decomposes into crys-
tallized Sr2GdRuO6, SrRuO3 phases and other
compositions containing a liquid phase with ‘‘Sr,
Gd, Cu, O’’ that is not seen by XRD. At 1120 �Cthe compound reaches a liquidus and looses oxygen
due to reduction of Cu2þ to Cuþ. High-temperatureheat treatments, in the temperature range where the
decomposition occurs, may induce irreversible
changes in the sample structure and composition.
We do observe bulk SC in samples prepared in a
temperature range where no decomposition has
occurred but in a sample where the stoichiometry is
not perfect. It is then difficult to conclude as re-
garding the role of off-stoichiometry on the SCoccurrence in Ru-1212 phase. Inter-granular prop-
erties have been shown to be sensitive to annealing
and to microstructure. In very small grains, intra-
grain SC cannot be detected due to too large
penetration depth and inappropriate grain bound-
aries assumed to have lost RuO2.
Acknowledgements
One of us (NDZ) thanks the staff of the Labo-
ratoire de Cristallographie, Grenoble, for kind
hospitality and assistance during his stay and ac-
knowledge the CNRS for financial support. It is a
pleasure to acknowledge Prof. J.L. Jorda (Annecy)
and Dr. A.T. Matveev (Minsk) for fruitful dis-cussions.
References
[1] L. Bauernfeind, W. Widder, H.F. Braun, Physica C 254
(1995) 151.
[2] I. Felner, U. Asaf, Y. Levi, O. Millo, Phys. Rev. B 55
(1997) R3374.
[3] C. Bernhard, J.L. Tallon, Ch. Niedermayer, Th. Blasius, A.
Golnik, E. Br€uucher, R.K. Kremer, D.R. Noakes, C.E.
Stronach, E.J. Ansaldo, Phys. Rev. B 59 (1999) 14099.
[4] J.L. Tallon, C. Bernhard, M.E. Bowden, P.W. Gilberd,
T.M. Stoto, D.J. Pridgle, IEEE Trans. Appl. Supercond. 9
(1999) 1696.
[5] R.W. Henn, H. Friedrich, V.P.S. Awana, E. Gmelin,
Physica C 341–348 (2000) 457.
[6] I. Felner, U. Asaf, S. Reich, Y. Tsabba, Physica C 311
(1999) 163.
[7] I. Felner, U. Asaf, Y. Levi, O. Millo, Physica C 334 (2000)
141.
[8] P.W. Klamut, B. Dabrowski, S.M. Mini, M. Maxwell, S.
Kolesnik, M. Mais, A. Shengelaya, R. Khasanov, I. Savic,
H. Keller, T. Graber, J. Gebhardt, P.J. Viccaro, Y. Xiao,
Physica C 364–365 (2001) 313.
[9] C.W. Chu, Y.Y. Xue, S. Tsui, J. Cmaidalka, A.K.
Heilman, B. Lorenz, R.L. Meng, Physica C 335 (2000) 231.
[10] C. Bernhard, J.L. Tallon, E. Br€uuchner, R.K. Kreme, Phys.
Rev. B 61 (2000) R14960.
[11] H.A. Blackstead, J.D. Dow, D.R. Harshman, D.B. Pulling,
Z.F. Ren, D.Z. Wang, Physica C 364–365 (2001) 305.
[12] Y. Furukawa, S. Tanaka, Y. Yamanaka, K. Kumagai,
Physica C 341 (2000) 453.
[13] K. Kumagai, S. Takada, Y. Furukawa, Phys. Rev. B 63
(2001) 180509.
[14] R.S. Liu, L.-Y. Jang, H.-H. Hung, J.L. Tallon, Phys. Rev.
B 63 (2001) 212507.
[15] I. Felner, U. Asaf, F. Ritter, P.W. Klamut, B. Dabrowski,
Physica C 364–365 (2001) 368.
[16] G.M. Kuz�micheva, V.V. Luparev, E.P. Khlybov, I.E.
Kostyleva, A.S. Andreenko, K.N. Gavrilov, Physica C 350
(2001) 105.
[17] O. Chmaissem, J.D. Jorgensen, H. Shaked, P. Dollar, J.L.
Tallon, Phys. Rev. B 61 (2000) 6401.
[18] A. Sin, P. Odier, Adv. Mater. 12 (2000) 649.
[19] J.W. Lynn, B. Keimar, C. Ulrich, C. Bernhard, J.L. Tallon,
Phys. Rev. B 61 (2000) R14964.
[20] K. Conder, Mater. Sci. Eng., R 32 (2001) 41.
[21] T.P. Papageorgiou, T. Herrmannsd€oorfer, R. Dinnebier, T.
Mai, T. Ernst, M. Wunschel, H.F. Braun, Physica C 377
(2002) 383.
[22] C.T. Lin, B. Liang, C. Ulrich, C. Bernhard, Physica C 364–
365 (2001) 373.
[23] C. Artimi, M.M. Carnasciali, G.A. Costa, M. Ferretti,
M.R. Cimberle, M. Putti, R. Masini, Physica C 377 (2002)
431.
[24] J.D. Jorgensen, O. Chmaissem, H. Shaked, S. Short, C.W.
Klamut, B. Dabrowski, J.L. Tallon, Phys. Rev. B 63 (2001)
054440.
[25] B. Lorenz, R.L. Meng, J. Cmaidalka, Y.S. Wang, J. Lenzi,
Y.Y. Xue, C.W. Chu, Physica C 363 (2001) 251.
[26] P.W. Klamut, B. Dabrowski, S. Kolesnik, M. Maxwell,
J. Mais, Phys. Rev. B 63 (2001) 224512.
[27] F.J. Gotor, A. Fert, P. Odier, N. Pellerin, J. Am. Ceram.
Soc. 78 (1995) 2113.
[28] A. Sin, P. Odier, M. Nu~nnez-Requeiro, Physica C 330 (2000)
9.
[29] Y.Y. Xue, B. Lorentz, R.L. Meng, A. Baikalove, C.W.
Chu, Physica C 364–365 (2001) 251.
358 N.D. Zhigadlo et al. / Physica C 387 (2003) 347–358