A. Kikuchi et al. / Physica C 412–414 (2004) 1174–1178
1175
(
diffusion reaction at low temperatures.
650 °C), thus Mg
2
Cu is expected to promote a
SQUID magnetometer (Quantum Design: MPMS-
5). X-ray diffraction (XRD) analysis for the phase
identification was made for both specimens, and
microstructures were observed using a field-emis-
sion scanning electron microscope (FE-SEM).
Elemental mapping and quantitative composition
analysis were carried out with an energy dispersive
X-ray (EDX) spectrometer.
2
. Experimental
Mg Cu ingot was made in an Ar atmosphere
2
with a Tammann furnace. The ingot was then
crushed and mechanically ground into a coarse
powder of about 50 lm in average diameter. Phase
identification was carried out with X-ray diffrac-
tion (XRD) analysis using Ni-filtered CuKa radi-
3. Results and discussion
ation for the coarse Mg
2
Cu powder. Thermo
Fig. 1 shows the TGA–DTA results for Mg
and pure Mg. The Mg
is lower than that of pure Mg (650 °C). Both
2
Cu
Cu melting point (568 °C)
gravimetric analysis (TGA) and differential ther-
mal analysis (DTA) were also carried out for the
coarse Mg Cu powder. Both TGA and DTA were
2
weight ratios, Mg Cu and pure Mg, increase with
increasing temperature, which indicates that both
powders oxidize at high temperatures. However,
2
2
carried out with an auto simultaneous instrument
(
2
Shimadzu: TGA-60H) in open air. The Mg Cu
powder was mixed with amorphous B powder
using an atomic ratio of Mg:B ¼ 1:2. The average
size of the amorphous B powder was a few tens of
the weight ratio for Mg
pure Mg, therefore, Mg
tive to oxidation compared to pure Mg.
2
Cu is smaller than that of
Cu should be more resis-
2
microns in diameter. The Mg
2
Cu and B mixture
c
Fig. 2 shows the relationship between T versus
was tightly packed into Fe tube (99.99% purity)
with an outer diameter and inner diameter of 10
and 6 mm, respectively. Wire drawing was carried
out using a grooved-roller and cassette-roller-dies.
Precursor wires had a diameter of 0.72 mm. Nor-
mal precursor wires using pure Mg and amor-
phous B powders were also made for comparison.
temperature for Mg Cu/B/Fe and Mg/B/Fe wires.
When the temperature is elevated to around the
2
melting point of Mg
Mg Cu/B Fe-sheathed wire specimens rapidly in-
creased and reached about 37.5 K due to the for-
mation of the MgB phase through the liquid
(Mg Cu)–solid (B) diffusion reaction. The T val-
ues are comparable to the reported values for
2
Cu (568 °C), the T
c
for
2
2
2
c
2
For the other study, a thick MgB layer on a metal
substrate was fabricated. Amorphous B layer (200
lm thick) was formed on the pure Nb substrate-
tape by using a plasma spraying technique. Then,
Mg
coated on the surface of the plasma-sprayed
amorphous B layer creating Mg Cu/B/metal-sub-
2
Cu slurry using ethyl alcohol as a solvent was
2
strate layered composites. Both the PIT Fe-
sheathed wires and layered composites were heated
to between 550 and 650 °C in a dynamic vacuum
ꢀ
3
under a pressure of less than 10 Pa.
After heat treatment, T and transport critical
current I (4.2 K) were measured by a DC four-
probe method. I (4.2 K) measurements were made
under magnetic fields up to 10 T using a super-
conducting magnet. J for the PIT Fe-sheathed
wires is defined as the I divided by a total cross-
c
c
c
c
Fig. 1. Comparison of TGA–DTA results for Mg
2
Cu and pure
c
Mg. Solid circles and triangles represent increasing weight ra-
sectional area of the powder filled core. Magnetic
transition curves were also measured using a
tios calculated from TGA data for the Mg
respectively.
2
Cu and pure Mg,