12
High-temperature phase changes in RuSr 2 GdCu 2 O 8 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, France b LAIMAN, Universit e de Savoie, 9 Rue de lÕArc-en-Ciel, BP 240, F-74942 Annecy-le-Vieux Cedex, France c Centre de Recherche sur les Tr es Basses Temp eratures, 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 RuSr 2 GdCu 2 O 8 (Ru-1212) as a single phase. The crystal- lization of Ru-1212 arises at 950 °C under O 2 flow with a small amount of secondary phase. The pure Ru-1212 phase is achieved after heating at 1020 °C in O 2 flow. The X-ray diffraction (XRD) pattern is refined (Rietveld) in the tetragonal space group P4/mmm with lattice parameters a ¼ 3:83904ð9Þ A and c ¼ 11:5678ð4Þ A. In situ high-temperature XRD and differential thermal analysis coupled with thermal-weight measurement show a structural decomposition of the phase at T d ¼ 1050 °C followed by a partial melting at T m ¼ 1118 °C. The decomposition produces crystallized Sr 2 GdRuO 6 , SrRuO 3 phases and a mixture rich in copper containing ‘‘Sr, Gd, Cu, O’’ that does not diffract X-ray. This phase reduces to Cu þ at T m 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 T m does not permit to detect in sol–gel samples su- perconductivity otherwise observed in compounds prepared by solid-state reaction. Ó 2002 Elsevier Science B.V. All rights reserved. Keywords: Rutheno-cuprates; RuSr 2 GdCu 2 O 8 ; Sol–gel; In situ XRD; DTA, magnetic measurements 1. Introduction A new class of rutheno-cuprates, RuSr 2 Ln- Cu 2 O 8 (Ru-1212) and RuSr 2 (Ln 1þx -Ce 1x )Cu 2 O 10 (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) RuSr 2 R 1:5 Ce 0:5 Cu 2 O 10y [2] and more recently in the 1212 type (RuSr 2 GdCu 2 O 8 ) [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 ðT AF Þ below 150 K [5,6] to ‘‘bulk SC’’ samples with T c ¼ 15–46 K and T AF 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 epartment des Sciences Chimiques, CNRS, France. 0921-4534/02/$ - see front matter Ó 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

High-temperature phase changes in RuSr2GdCu2O8 and physical properties

Embed Size (px)

Citation preview

Page 1: High-temperature phase changes in RuSr2GdCu2O8 and physical properties

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

Page 2: High-temperature phase changes in RuSr2GdCu2O8 and physical properties

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

Page 3: High-temperature phase changes in RuSr2GdCu2O8 and physical properties

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

Page 4: High-temperature phase changes in RuSr2GdCu2O8 and physical properties

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

Page 5: High-temperature phase changes in RuSr2GdCu2O8 and physical properties

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

Page 6: High-temperature phase changes in RuSr2GdCu2O8 and physical properties

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

Page 7: High-temperature phase changes in RuSr2GdCu2O8 and physical properties

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

Page 8: High-temperature phase changes in RuSr2GdCu2O8 and physical properties

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

Page 9: High-temperature phase changes in RuSr2GdCu2O8 and physical properties

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

Page 10: High-temperature phase changes in RuSr2GdCu2O8 and physical properties

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

Page 11: High-temperature phase changes in RuSr2GdCu2O8 and physical properties

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

Page 12: High-temperature phase changes in RuSr2GdCu2O8 and physical properties

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