Significant enhancement of flux pinning in MgB2 superconductor through nano-Si addition

Article abstract of DOI:10.1016/S0921-4534(02)02813-7

Polycrystalline MgB₂ samples with 10 wt.% silicon powder addition were prepared by an in situ reaction process. Two different Si powders, one with coarse (44 μm) and the other with nano-size (<100 nm) particles were used for making samples. The phases, microstructures, and flux pinning were characterized by XRD, TEM, and magnetic measurements. It was observed that the samples doped with nano-sized Si powder showed a significantly improved field dependence of the critical current over a wide temperature range compared with both undoped samples and samples with coarse-Si added. J_c is as high as 3000 A/cm² in 8 T at 5 K, one order of magnitude higher than for the undoped MgB₂. X-ray diffraction results indicated that Si had reacted with Mg to form Mg₂Si. Nano-particle inclusions and substitution, both observed by transmission electron microscopy, are proposed to be responsible for the enhancement of flux pinning in high fields. However, the samples made with the coarse-Si powders had a poorer pinning than the undoped MgB₂.

Full text of DOI:10.1016/S0921-4534(02)02813-7

Physica C 385 (2003) 461–465
www.elsevier.com/locate/physc
Significant enhancement of flux pinning in MgB2
superconductor through nano-Si addition
a,
X.L. Wang , S.H. Zhou , M.J. Qin , P.R. Munroe ,
a
a
b
*
a
S. Soltanian , H.K. Liu , S.X. Dou
a
a
a
Institute for Superconducting and Electronic Materials, University of Wollongong, Northfields Avenue,
Wollongong, NSW 2552, Australia
b
Electron Microscope Unit, University of New South Wales, NSW 2052, Australia
Received 29 October 2002; received in revised form 25 November 2002; accepted 16 December 2002
Abstract
Polycrystalline MgB
2
samples with 10 wt.% silicon powder addition were prepared by an in situ reaction process.
Two different Si powders, one with coarse (44 lm) and the other with nano-size (<100 nm) particles were used for
making samples. The phases, microstructures, and flux pinning were characterized by XRD, TEM, and magnetic
measurements. It was observed that the samples doped with nano-sized Si powder showed a significantly improved field
dependence of the critical current over a wide temperature range compared with both undoped samples and samples
2
with coarse-Si added. J
MgB
c
is as high as 3000 A/cm in 8 T at 5 K, one order of magnitude higher than for the undoped
2
2
. X-ray diffraction results indicated that Si had reacted with Mg to form Mg Si. Nano-particle inclusions and
substitution, both observed by transmission electron microscopy, are proposed to be responsible for the enhancement
of flux pinning in high fields. However, the samples made with the coarse-Si powders had a poorer pinning than the
2
undoped MgB .
Ó 2002 Elsevier Science B.V. All rights reserved.
PACS: 74.70.Ad; 74.70.Ge; 74.62.BF; 74.70.Jg
2
Keywords: MgB ; Flux pinning; Si addition
To push the newly discovered MgB supercon-
2
approaches aimed at enhancement of flux pinning
in MgB such as high energy ion irradiation [1,2]
ductor towards practical applications, it is essen-
tial to enhance the flux pinning and increase the
critical current density in high fields to the level
attained by the conventional low temperature su-
2
2
for powders of MgB , inclusion of precipitates of
MgO in thin films [3], and chemical doping using
different elements, nano-particle addition/doping
seems to be the best and most practical route [3,4]
reported so far. Very recently, Dou et al. [5] have
demonstrated that nano-size SiC-doped MgB2
samples show a very significant enhancement of
critical current density in high magnetic fields over
a wide temperature range from 30 K down to 5 K.
3
perconductors NbTi or Nb Sn. Among all the
*
Corresponding author. Tel.: +61-2-422-15766; fax: +61-2-
22-15731.
E-mail address: xiaolin@uow.edu.au (X.L. Wang).
4
0
921-4534/02/$ - see front matter Ó 2002 Elsevier Science B.V. All rights reserved.
doi:10.1016/S0921-4534(02)02813-7

462
X.L. Wang et al. / Physica C 385 (2003) 461–465
2
A newly reported of J
c
as high as 20,000 A/cm in
9
.5 T at 5 K was achieved for both transport and
magnetic measurements [6,7]. Mg Si phase was
found to be the major impurity phase in the SiC-
doped MgB samples [6]. Mg Si inclusions inside
MgB grains are suggested to be one of the pos-
2
2
2
2
sible sources responsible for the enhanced pinning
in addition to SiC substitutions. The intention of
the present work is to study the effect of pure
nano-size Si addition on the phase formation and
2
microstructures of MgB and to investigate its role
in the flux pinning behaviors. Our present results
indicate that, quite similar to nano-SiC-doped
Fig. 1. XRD pattern for undoped, coarse-Si and nano-Si doped
MgB , and pure nano-Si powder samples.
2
2 2
MgB [5–7], pure nano-Si-doped MgB exhibits
significantly enhanced pinning in high fields over a
wide temperature range compared with undoped
3
dimensions of a ꢀ b ꢀ c ¼ 0:56 ꢀ 2:7 ꢀ 3:73 mm
MgB
2
samples. Both Si substitution and nano-
were cut from as-sintered pellets. Transition tem-
peratures were determined by measuring the real
part of the ac susceptibility at a frequency of 117
particle inclusions were observed by transmission
electron microscopy (TEM), leading to a signifi-
cant enhancement in flux pinning.
Hz and an ac field of 0.1 Oe. The magnetic J
calculated from the height of the magnetization
c
was
2
It has been reported that MgB with high per-
formance can be formed very quickly (in as little
as a few minutes) [8] in a reaction between indi-
vidual magnesium and amorphous boron pow-
ders, the so-called ‘‘in situ’’ reaction. It has also
loops ðM–HÞ using a Bean model where J ¼
c
20DM=½a=ð1 ꢁ a=3bÞꢂ with a < b, where a and b
are the dimensions of the sample perpendicular to
the direction of applied field.
been demonstrated that the J
field dependence in low density MgB
fabricated using the ‘‘in situ’’ is better than that
those of fully dense MgB /Fe tapes [9] made using
pre-reacted MgB powders (ex situ reaction). The
c
magnitude and its
Fig. 1 shows the XRD patterns for four sam-
ples. These are reference sample (sample #1),
coarse-Si (sample #2) and nano-Si (sample #3)
doped samples, and the pure nano-Si powder, re-
2
/Fe wire [7]
2
spectively. It can be clearly seen that Mg Si is the
2
2
samples used in the present study were thus made
using the ‘‘in situ’’ solid state reaction. Magnesium
powders ()325 mesh) and amorphous boron ()325
mesh) powders were well mixed with silicon pow-
ders with different particle sizes, 44 lm (coarse)
and <100 nm (fine) at a level of 10 wt.%. Pellets 10
mm in diameter and 1 mm in thickness were made
under pressure, and sealed in Fe tubes, then heat
treated at 700–900 °C for 1 h in flowing high purity
Ar. This was followed by a furnace cooling to
room temperature. An undoped sample was also
made under the same conditions for use as a ref-
erence sample. The phases and microstructures
were examined by X-ray diffraction (XRD) and
transmission electron microscopy (TEM). The
magnetization was measured over a wide temper-
ature range between 5 and 30 K in 0–8.5 T using a
Quantum Design PPMS. Bar shaped samples with
only major impurity phase in both the coarse and
nano-Si doped samples. (A very small amount of
MgO can also be observed from the peak located
at 2H ¼ 62:4°). Some Si peaks can be found in the
2
coarse-Si-doped sample in addition to the Mg Si,
indicating that there is residual unreacted pow-
der. However, the nano-Si powder seems to react
completely with Mg and form Mg Si or substitute
2
into the MgB lattice due to the higher reactivity of
2
nano-Si compared to coarse-Si powders. This is in
agreement with the fact that the relative XRD
peak intensity from the Mg
MgB in the nano-Si-doped sample is higher than
2
Si phase to that from
2
that in the coarse-Si-doped one.
A typical TEM image of MgB
grains from a
2
nano-Si-doped MgB sample is shown in Fig. 2. A
2
number of fine grains, around 100–200 nm in dia-
meter can be seen. Most of these were shown by

X.L. Wang et al. / Physica C 385 (2003) 461–465
463
particles, 10–30 nm in size were also observed lo-
cated on the MgB or Mg Si grains, but EDS and
2
2
diffraction data from these particles were not
conclusive as the signal from these particles is di-
luted by the surrounding grains.
There are two possibilities for the form of the Si
detected in the MgB grains: first, that it is in the
2
form of either Mg Si or Si inclusions with very fine
2
particle sizes, or second, that it is Si which has
crystal lattice.
Si was the only
substituted directly into the MgB
The former may be possible as Mg
2
2
impurity detected by XRD. EDS and microdif-
fraction studies are potentially ambiguous since
the typical probe size used (ꢃ10 nm) may be larger
than the size of the Mg Si or Si particles, and such
2
particles may be embedded in thicker MgB crys-
2
tals, where the signal from the Si-rich regions is
2
diluted by the surrounding MgB . However, no Si
was observed via electron diffraction, hence we
suggest that the Si peak in EDS spectra may partly
come from Si which occupies substitutional posi-
tions in the MgB and partly from Mg Si inclu-
2
Fig. 2. TEM micrograph of nano-Si-doped MgB .
electron diffraction to be MgB , but a number of
2
2
2
grains, especially those somewhat larger than 100
nm in diameter, were determined by electron dif-
sions inside the MgB grains.
2
The real part of the ac susceptibility of all three
samples is shown in Fig. 4. We can see that the Tc
of the coarse-Si-doped sample is slightly higher
than that of the undoped sample despite the large
2
fraction to be Mg
on several spots in different regions of the same
MgB grain. A small Si peak was always detected,
2
Si. EDS analysis was carried out
2
but varied slightly in intensity at different posi-
tions. A typical EDS spectrum is shown in Fig. 3.
EDS spectra of the Mg Si grains yielded signifi-
amount of Mg Si and the presence of residual Si.
However, the T of the nano-Si-doped sample has
c
2
dropped to 36.7 K which is higher than that of the
10 wt.% nano-SiC-doped MgB . The drop in the T
cantly more intense Si peaks. A number of smaller
2
c
Fig. 3. Typical EDS spectrum taken from one of the grains shown in Fig. 2.

464
X.L. Wang et al. / Physica C 385 (2003) 461–465
6
1
1
1
1
1
0
0
0
0
0
nano Si 10 wt %
undoped
5
4
3
2
coarse Si, 10 wt %
3
2
0 K
10K
0
4
6
8
6
Fig. 4. Real part of the ac susceptibility for coarse-Si-doped,
nano-Si-doped and undoped MgB samples.
1
0
0
2
nano Si 10 wt%
undoped
5
4
3
2
coarse Si 10 wt %
1
for the nano-Si-doped sample is another piece of
evidence for substitution of Si into the lattice. For
the coarse-Si doping, the level of substitution into
the lattice is very small due to the limited contact
area between the Si and the Mg and B. whereas the
nano-Si has much high surface energy and can
react with Mg and B more fully. Thus the substi-
tution level is expected to be much higher than is
the case for coarse-Si.
10
1
0
5 K
2
0K
10
0
2
4
6
8
Values for J
plotted in Fig. 5. The coarse-Si-doped sample
shows a poor J field dependence at T < 30 K,
c
vs field at 30, 20, 10, and 5 K are
c
Fig. 5. J vs field for undoped, coarse-Si and nano-Si doped
samples at 30 and 10 K (upper) and at 20 and 5 K (lower).
c
compared to the undoped sample. However, the
nano-Si-doped sample revealed a significant im-
values for coarse-Si-doped sample are much
smaller than for the undoped sample and decrease
in the same way, indicating that there is no im-
proved pinning as a result of the coarse-Si doping.
No nano-inclusion features were found in samples
doped with coarse-Si powders, rather large grains
provement in J field performance at all tempera-
c
tures. We can see that the J of the nano-Si-doped
c
samples is lower than that of undoped samples in
the low field regions. This is apparently due to the
decrease of the superconducting volumes inside
the nano-Si-doped samples that was caused by the
of Si and Mg
particles in SEM images, and this would create
harmful weak links between the MgB grains.
of 3000 A/cm is achieved in 8 T at 5 K for
the nano-Si-doped MgB , one order of magnitude
higher than for the control sample. This J field
behavior in high fields is similar to that observed in
nano-SiC-doped MgB [6], but it is better than for
nano-Y doped MgB [4], the best pure MgB
samples [10] which were sintered in Mg vapor, or
fully dense MgB made using the hot pressure
Si were observed around MgB
2
2
formation of Mg
5
2
c
Si impurities. Furthermore, J at
K is lower than that at 20 K in low field region is
2
2
due to the flux jumping occurring at 5 K in field
less than 3 T. The irreversibility field Hirr and Jc
values of the nano-Si-doped sample are higher
than for the coarse-Si-doped and undoped samples
in high field regions at all the measured tempera-
tures. The lower the temperature, the larger the
A J
c
2
c
2
O
2
3
2
2
difference in Hirr and J
samples. It can be seen that the J
c
at high fields for these three
for the nano-Si-
doped sample drops more slowly than for the
c
2
method [11]. The enhancement of flux pinning in
high fields may be ascribed to both substitution
reference sample as the field increases. However, Jc

X.L. Wang et al. / Physica C 385 (2003) 461–465
465
effects and fine nano-particle inclusions inside
MgB grains.
Acknowledgements
2
It should be noted that there is a clear difference
between nano-additive induced pinning such as in
The authors thank Drs. T. Silver, M. Ionescu, J.
Horvat, E.W. Collings, M. Sumption, M. Tomsic
and R. Neale for their helpful discussion and
Australian Research Council, Hyper Tech Re-
search Inc. OH USA and Alphatech International
Ltd. for financial support.
2 3
the case of nano-Y O , and the situation for nano-
SiC and nano-Si. The former gives a strong en-
hancement of flux pinning at low temperatures
but not at higher temperatures (above 20 K). In
contrast, the nano-SiC and nano-Si doping with
substitution as well as induced nano-particle in-
clusions exhibit strong enhancement in flux pin-
ning over all the entire temperature range despite
References
[
[
1] Y. Bugoslavsky, L.F. Cohen, G.K. Perkins, et al., Nature
11 (2001) 561.
c
the fact that the T was depressed more than for
4
the nano-particle addition. Thus, it is essential to
introduce intra-grain pinning into MgB2 to
achieve the best high field performance.
2] M. Eisterer, M. Zehetmayer, S. Toenies, H.W. Weber,
et al., Supercond. Sci. Technol. 15 (2001) L9.
[3] C.B. Eom, M.K. Lee, J.H. Choi, et al., Nature 411 (2001)
58.
5
4] J. Wang, Y. Bugoslavsky, A. Berenov, et al., Appl. Phys.
Only a 10 wt.% addition of Si was added to the
[
2
MgB in our present work. It is believed that a
Lett, in press, cond-mat/0204472.
5] S.X. Dou et al., cond-mat/0207093.
further enhancement can be achieved when the
nano-Si addition level and the heat treatment
conditions are further optimized.
[
[
6] S.X. Dou, S. Soltanian, J. Horvat, X.L. Wang, P.G.
Munroe, S.H. Zhou, M. Ionescu, H.K. Liu, M. Tomsic,
cond-mat/0207223.
In summary, nano-Si particle sizes (<100 nm)
doped MgB demonstrated significantly improved
pinning in high fields over a wide temperature
[7] S.X. Dou, J. Horvat, S. Soltanian, X.L. Wang, et al.,
Presented at ASC 2002, Houston, USA, also cond-mat/
2
0
208215.
range compared to undoped MgB
coarse (44 lm) Si-doped sample had weaker pin-
ning than pure MgB . TEM results indicated that
both Si substitutional effects and fine particle in-
clusions inside MgB grains are responsible for the
2
. In contrast, a
[
[
8] X.L. Wang, S. Soltanian, J. Horvart, A.H. Li, M.J. Qin,
H.K. Liu, S.X. Dou, Physica C 361 (2001) 149–155.
9] H.L. Suo, C. Beneduce, M. Dhall ꢀe , N. Musolino, J.-Y.
Genoud, R. Fl u€ kiger, Appl. Phys. Lett. 79 (2001) 3116–
2
3
10] V. Braccini et al., cond-mat/0208054.
118.
2
[
[
improved intra-grain flux pinning in the nano-Si-
doped samples.
11] M. Dhalle, P. Toulemonde, C. Beneduce, N. Musolino, M.
Decroux, R. Fl u€ kiger, Physica C 363 (2001) 155.