Full text of DOI:10.1051/epjap:2001128

Eur. Phys. J. AP 13, 157–160 (2001)
THE EUROPEAN
PHYSICAL JOURNAL
APPLIED PHYSICS
c
ꢀ EDP Sciences 2001
Top seeding melt texture growth of NdBaCuO pellets in air?
E. Guilmeau, F. Giovannellia, I. Monot-Laffez, S. Marinel, J. Provost, and G. Desgardin
CRISMAT Laboratoryb, 6 boulevard du Mar´echal Juin, 14050 Caen, France
Received: 31 July 2000 / Revised: 30 October 2000 / Accepted: 14 November 2000
Abstract. Most of NdBaCuO synthesis were realized under oxygen controlled melt growth (OCMG) which
allows to avoid the solid solution formation. However, for industrial applications air synthesis is preferable
due to its simplicity and its cost. In this work, different seeds and thermal cycles have been investigated in
order to optimize the nucleation conditions for the obtention of a single domain. Several criteria have been
defined for the choice of the seed. The solid solution Nd1+xBa2−xCu3Oy formed in these air synthesised
samples appears in the range of x = 0.10 to 0.15, corresponding to Tc
between 60 K and 80 K. The
onset
nominal composition and the synthesis conditions (substrate, thermal treatments.) have to be accurately
defined in order to avoid neodymium-rich solid solution formation.
PACS. 74.60.Jg Critical currents – 74.72.Jt Other cuprates
1 Introduction
Table 1. Crystallographic parameters and melting tempera-
ture of the different seeds used.
The NdBa2Cu3Oy (NdBaCuO) high-Tc superconduc-
tor has shown promising properties for high field ap-
plications. Unlike YBaCuO, the NdBaCuO is a non
stoichiometric compound leading to the solid solution
Nd1+xBa2−xCu3Oy formation [1,2]. The superconducting
properties decrease dramatically with x [2,3]. Yoo et al. [4]
have proposed an oxygen controlled melt growth (OCMG)
process to avoid the solid solution formation. However
an air synthesis method is benefit for potential industrial
applications. In this paper, the top seeding melt texture
growth (TSMTG) method [5] is applied to NdBaCuO pel-
lets in air. First of all we were interested in NdBaCuO tex-
ture formation. We tested several seeds and optimised the
thermal cycle for obtaining large domains. In another part
we tried to decrease the solid solution formation to reach
optimal superconducting properties already obtained un-
der air by Hu et al. [6]. Different pellet supports and nom-
inal composition have been tested.
6HHGV
&ULVWDOORJUDSKLF
SDUDPHWHUV
&XELF
7PHOWLQJ
ꢃꢄꢅꢅꢀƒ&
ꢃꢅꢄꢅꢀƒ&
ꢆꢅꢄꢋꢀƒ&
ꢆꢅꢄꢋꢀƒ&
7PD[
ꢆꢆꢅꢅꢇꢀꢆꢆꢁꢅꢇ
ꢆꢃꢅꢅꢀƒ&
ꢆꢆꢅꢅꢇꢀꢆꢆꢁꢅꢀƒ&
0J2
Dꢀ ꢀꢁꢂꢃꢃꢀ$
&XELF
ꢀVLQJOHꢁFU\VWDOꢂ
672ꢁꢀ6U7L2ꢃꢂ
ꢀVXEVWUDWHꢂ
672ꢁꢄꢁ1Gꢅꢆꢃ
ꢀWKLQꢁILOPVꢂ
1Gꢅꢆꢃ
Dꢀ ꢀꢈꢂꢉꢅꢊꢀ$
7HWUDJRQDO
Dꢀ ꢀꢈꢂꢉꢆꢀ$
7HWUDJRQDO
Dꢀ ꢀꢈꢂꢉꢆꢀ$
ꢆꢆꢅꢅꢇꢀꢆꢆꢁꢅꢀƒ&
ꢆꢆꢁꢅꢀƒ&
ꢀDXWRꢇVHHGꢂ
controlled by XRD measurements. Nd123 and an excess of
15%mol. Nd422 were mixed in agate ball mill for 45 min-
utes. After cold isostatic pressing under 300 MPa, 16 mm
diameter pelletised samples were submitted to TSMTG
process. A seed, which acts as a nucleation centre, was
placed on the top of the pellets. It must have a melting
temperature (Tmelting) higher than Nd123’s one and the
nearest lattice parameters. Several seeds, which seemed to
fit more or less these criteria and which resist to the max-
imum temperature used during the thermal cycle (Tmax),
were tested (see Tab. 1). The “Auto seed” method was
tested. This method consists in placing a Nd123 seed at
1065 ◦C, after the Nd123’s decomposition step during the
temperature decreasing. This experiment was realised in
an isothermal furnace in which one hole has been made in
order to introduce an alumina tube (see Fig. 1a), designed
to deposit the seed (see Fig. 1b). Nd422 powder layer or
sintered Nd422 bars were used as support. In the last ex-
periments, 0.5% weight of CeO2 were added to Nd123 +
2 Experimental
NdBa2Cu3Oy (Nd123) and Nd4Ba2Cu2O10 (Nd422) were
prepared by solid state reaction mixing stoichiometric
compounds of Nd2O3 (3N), CuO (5N) and BaCO3 (5N).
The powders were calcined for 2 hours at 1010 ◦C and
then for 10 hours at 970 ◦C. The purity of the powder were
?
This paper has been presented in “Sixi`emes journ´ees de
cryog´enie et supraconductivit´e” at Aussois on May 16–19,
2000.
a
e-mail: fabien.giovannelli@ismra.fr
UMR 6508 CNRS/ISMRA
b

158
The European Physical Journal Applied Physics
1XFOHDWLRQꢀZLQGRZ
ꢅꢂꢆꢂƒ&
7PD[
ꢅꢂꢄꢅƒ&
ꢁꢂꢂƒ&ꢃꢄK
(a)
7LPH
Fig. 2. Optimised thermal cycle.
$ꢇ
(b)
ꢅꢄꢀPP
Fig. 1. (a) Scheme of the auto-seed furnace. (b) Detail of
Figure 1a.
15%mol. Nd422. All the experiments were realised with
the same thermal cycle previously optimised by a series
of quenches at different temperature to determine the nu-
cleation window. The samples have been characterised by
scanning electron microscopy (SEM) coupled with energy
dispersive spectroscopy analysis (EDS) and the supercon-
ducting properties have been measured by SQUID mag-
netometer (quantum Design MPMS) between 30 K and
100 K with an applied field of 10 Oe parallel to the c-axis
after an annealing of the samples in oxygen atmosphere.
The texture formation was controlled by pole figure mea-
surements with a four-circle X-ray diffractometer.
%ꢇ
Fig. 3. A) Photo of multi-domains obtained with MgO seed.
B) Photo of a single-domain obtained with the auto-seed
method.
3 Results and discussion
heterogeneous nucleation and the final sample is generally
multi-domain (see Fig. 3a). The existence of multido-
According to previous quenching experiments performed mains could be explained by the difference of the crys-
in air, the heterogeneous and homogene◦ous nucleation be- tallographics parameters between MgO and NdBaCuO.
◦
gin respectively at 1061 C and 1050 C. This range of A crystallographic study on thin films has shown that
temperature (11 ◦C) defines the solidification window. the epitaxial growth between YBaCuO and the substrate
The cooling rate value is fixed at 0.5 ◦C/h in this window. MgO can be observed with two orientations. On 75% of
The optimised thermal cycle is determined owing to these the layed films, the a, b axes of YBaCuO are aligned with
quenching experiments (see Fig. 2). SrTiO3 and SrTiO3 the MgO axes while, for the 25% remaining, the axes are
+ Nd123 seeds have reacted with the melt contrary 45◦ misoriented [7,8]. As aNdBCO is very close to aYBCO
,
to the MgO seed. However, this latter seed induces an this feature could be transposed to the NdBaCuO system.

E. Guilmeau et al.: Top seeding melt texture growth of NdBaCuO pellets in air
159
1Gꢀꢁ[
F
Fig. 4. Photo of a transversal section of a single-domain ob-
tained with the auto-seed method.
(a)
1Gꢀꢁ[
ꢁꢇꢀꢄ
ꢁꢇꢀꢆ
ꢁꢇꢀꢈ
ꢁꢇꢁ
ꢁꢇꢁꢂ
ꢁꢇꢁꢄ
ꢀ
ꢁ
ꢂ
ꢃ
ꢄ
ꢅ
ꢆ
Fig. 5. (006) pole figure of a single domain obtained with the
auto-seed method.
(b)
The multinucleation from the seed was foreseeable. Never-
theless, single domain in RE-123 system have already been
obtained [9], but the reproducibility is not controlled. The
presence of several domains in the sample creates grain
boundaries acting as weak links which decrease the super-
conducting properties in the material. Thus, it seems to be
interesting to develop the autoseed alternative method [10]
in order to obtain single domains. By this technique sin-
gle domains of about 1 × 1 × 0.5 cm3 are obtained (see
Fig. 3b). However, a pellet transverse sliding underlines
the orientation of the (a, b) plane (see Fig. 4) which shows
that the domain is not correctly oriented. Indeed the seed
seems to play the role of a heterogeneous nucleation centre
like an impurity in the melt and the growth from the seed
is not epitaxial. However, the pole figure confirms that the
domain has a single orientation of c-axis (see Fig. 5) mean-
ing that the thermal cycle provide good conditions for a
continuous growth. Further experiments are necessary to
solve the problem of the orientation of the (a, b) plane.
The second problem encountered in this work was
the solid solution formation. The first tests, realised
with a Nd422 powder layer support, present an average
Nd1.16Ba1.9Cu2.94Oy composition (compared to the nom-
inal composition which was stoichiometric Nd123) which
results in a Tc close to 60 K. No (Ba-Cu) compounds are
found by EDS analysis in the textured matrix. In fact, the
melt, after the peritectic decomposition, is a (Ba-Cu) rich
liquid. The Nd422 powder layer support, due to its im-
portant porosity, absorbs a lot of liquid thus leading to a
Fig. 6. (a) Evolution of solid solution with the depth from the
seed in a Nd123 +15%mol. Nd422 pellet. (b) Evolution of solid
solution with the depth from the seed in a Nd123 +15%mol.
Nd422 +0.5% weight CeO2 pellet.
decrease of Ba and Cu concentrations in the melt. There-
fore, a sintered Nd422 support, which limits the support
porosity and its contact with the pellets, should allow to
decrease the liquid losses. Effectively, in this configura-
tion, an average Nd1.1Ba1.95Cu2.95Oy solid solution was
obtained with a Tc of 80 K.
Nevertheless, in each case, the sample composition has
been found to continuously vary with the depth from the
seed (see Fig. 6a). This evolution is certainly related to
the variation of the liquid composition near the growth
front due to the liquid losses and liquid migration from
top to bottom. The addition of 0.5% weight of CeO2 to
the Y123 and 15%mol. Y211 composition was reported to
increase the viscosity of the melt [11,12]. Such an addition
was tested on Nd123 system The liquid losses was found
less important and the melt more homogeneous. This was
confirmed by EDS analysis which reveals an homogeneous
composition of Nd1.08Ba1.97Cu2.95Oy in the whole domain
(see Fig. 6b). Tc is measured at 80 K showing that the
composition of the solid solution and the annealing cycle
have to be further optimised.

160
The European Physical Journal Applied Physics
4 Conclusion
3. M.J. Kramer, S.I. Yoo, R.W. McCallum, W.B. Yelon, H.
Xie, P. Allenspach, Physica C 219, 145 (1994).
4. S.I. Yoo, N. Sakai, H. Takaichi, T. Higuchi, M. Murakami,
Appl. Phys. Lett. 65, 633 (1994).
5. C. Leblond, I. Monot, J. Provost, G. Desgardin, Physica
C 311, 211 (1999).
6. A.M. Hu, Z.X. Zhao, M.Z. Wu, C. Wende, T. Strasser, B.
Jung, G. Bruchlos, W. Gawalek, P. Goernert, Physica C
278, 43 (1997).
7. C. Prouteau, J.F. Hamet, B. Mercey, M. Hervieu, B.
Raveau, D. Robbes, L. Coudrier, G. Ben, Physica C 248,
108 (1995).
8. N. Hari Babu, W. Lo, D.A. Cardwell, Y.H. Shi, Sup. Sci.
Tech. 13, 468 (2000).
9. M.R. Koblischka, M. Muralidhar, M. Murakami, Mater.
Sci. Eng. B 65, 58 (1999).
10. H. Chauhan, M. Murakami, Proceeding 2nd PASREG
Morioka, Japan, 19 Oct., 1999.
In this work, we have demonstrated the possibility of syn-
thesising NdBaCuO bulk ceramics under air. The resulting
superconducting properties are conditioned by the accu-
rate control of the liquid phase. The choice of the nominal
composition, of the support placed bellow the pellet, and
of the appropriate doping, reduces the liquid losses and
maintains the matrix composition close to NdBa2Cu3Oy.
In those conditions, we expect to obtain good supercon-
ducting transition. The growth conditions for large sin-
gle domain have been established by optimisation of the
thermal cycle during TSMTG. But the control of the ori-
entation of the (a, b) plane remain unsuccessful, despite
several attempts with different seeds. This point has to be
now optimised in order to find out an effective seed with
reproducible results.
11. S. Pinol, F. Sandiumenge, B. Martinez, V. Gomis, J.
Fontcuberta, X. Obradors, E. Snoeck, Ch. Rouceau, Appl.
Phys. Lett. 65, 1488 (1994).
References
1. Y. Matsui, S. Takekawa, N. Ivy, Jpn J. Appl. Phys. 26,
L1693 (1987).
2. S.I. Yoo, R.W. McCallum, Physica C 219, 157 (1994).
12. S. Marinel, I. Monot, J. Provost, G. Desgardin, Supercond.
Sci. Technol. 11, 563 (1998).
To access this journal online:
www.edpsciences.org