Effect of additions of Ca compounds to the filling powder on the reduction of MgO and the critical current density properties of ex situ processed MgB2 tapes

Article abstract of DOI:10.1016/j.jallcom.2016.01.027

We have studied the effect of additions of Ca compounds, CaC₂ and CaH₂, to the filling powder on the transport critical current density (J_c) properties of ex situ processed MgB₂ tapes. These additives are expected to reduce MgO present in as-received commercial MgB₂ powder and improve the grain coupling of MgB₂ in the tapes. Heat treatment of the hand mixed powders, MgB₂ and the Ca compounds, brings about both disappearance of MgB₄ and formation of CaB₆ and Mg, compared with heat treated MgB₂ powder. They are observed much more clearly for the mixtures with CaH₂ addition. The MgO content decreases by the additions of those compounds. Neither of the a-axis and the c-axis parameters of MgB₂ are changed by the additions, suggesting no reactions of substitutions of Ca for Mg and C for B in MgB₂. The J_c properties of the tape samples using the hand mixed filling powder with CaC₂ addition degrade, whereas those with CaH₂ addition are slightly improved. When using the ball milled mixtures, slight contraction of the a-axis parameter occurs only by CaC₂ addition. Although the additions cause the reactions similar to those observed for the hand mixed powders, the MgO content reduces more clearly, compared with the MgB₂ content. The J_c properties of the tapes are enhanced in the high-field and low-field regions by the additions of CaC₂ and CaH₂, respectively. These enhancements are attributed to the carbon substitution, the improved grain coupling and the increased content of MgB₂.

Full text of DOI:10.1016/j.jallcom.2016.01.027

Journal of Alloys and Compounds 664 (2016) 650e656
Contents lists available at ScienceDirect
Journal of Alloys and Compounds
journal homepage: http://www.elsevier.com/locate/jalcom
Effect of additions of Ca compounds to the filling powder on the
reduction of MgO and the critical current density properties of ex situ
processed MgB2 tapes
,
Hiroki Fujii a *, Akira Ishitoya b, Shinji Itoh b, Kiyoshi Ozawa c, Hitoshi Kitaguchi a
a Superconducting Wire Unit, National Institute for Materials Science, 1-2-1 Sengen, Tsukuba, Ibaraki 305-0047, Japan
b Materials Analysis Station, National Institute for Materials Science, 1-2-1 Sengen, Tsukuba, Ibaraki 305-0047, Japan
c Battery Materials Unit, National Institute for Materials Science, 1-2-1 Sengen, Tsukuba, Ibaraki 305-0047, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
We have studied the effect of additions of Ca compounds, CaC2 and CaH2, to the filling powder on the
transport critical current density (Jc) properties of ex situ processed MgB2 tapes. These additives are
expected to reduce MgO present in as-received commercial MgB2 powder and improve the grain
coupling of MgB2 in the tapes. Heat treatment of the hand mixed powders, MgB2 and the Ca compounds,
brings about both disappearance of MgB4 and formation of CaB6 and Mg, compared with heat treated
MgB2 powder. They are observed much more clearly for the mixtures with CaH2 addition. The MgO
content decreases by the additions of those compounds. Neither of the a-axis and the c-axis parameters
of MgB2 are changed by the additions, suggesting no reactions of substitutions of Ca for Mg and C for B in
MgB2. The Jc properties of the tape samples using the hand mixed filling powder with CaC2 addition
degrade, whereas those with CaH2 addition are slightly improved. When using the ball milled mixtures,
slight contraction of the a-axis parameter occurs only by CaC2 addition. Although the additions cause the
reactions similar to those observed for the hand mixed powders, the MgO content reduces more clearly,
compared with the MgB2 content. The Jc properties of the tapes are enhanced in the high-field and low-
field regions by the additions of CaC2 and CaH2, respectively. These enhancements are attributed to the
carbon substitution, the improved grain coupling and the increased content of MgB2.
Received 17 October 2015
Received in revised form
21 December 2015
Accepted 4 January 2016
Available online 6 January 2016
Keywords:
MgB2
MgO reduction
Ca compounds
Ex situ processed tapes
Grain coupling
critical current density
© 2016 Elsevier B.V. All rights reserved.
1. Introduction
and result in weak coupling of grains and hence poor Jc [9e11].
Compared with the in situ process, the ex situ processed wires
Since the discovery of superconductivity in MgB2 [1], this ma-
terial has been developed as a substitute for NbeTi in such appli-
cations as cryocooled systems [2,3]. As a fabrication process for the
conductor, the powder-in-tube (PIT) technique has been widely
adopted [4,5]. Alternatively, the internal Mg diffusion (IMD) pro-
cess has been developed [6e8], which uses a Mg rod and B powder.
In the PIT technique two processes, in situ and ex situ, are used.
For both processes, the quality of the filling powder is important to
obtain superior critical current density (Jc). The surfaces of the
filling powders for both processes are gradually contaminated in
air, by oxygen and moisture. These damaged surface layers act as
obstacles during sintering or reaction to the superconducting phase
and tapes show relatively poor Jc properties because the above-
mentioned weak coupling is more serious [12]. Furthermore,
high-temperature heat-treatment required for the improvement of
grain coupling results in the decomposition of MgB2 and the for-
mation of impurity phases such as MgB4 acting as obstacles to the
supercurrent path in the ex situ process.
Nevertheless the ex situ process still has the advantage of a high
density of the MgB2 layer over the in situ process [12]. The low
density of the core layer in the in situ process is due to the large
volume reduction at the formation of MgB2 from Mg and B powder
mixtures. This low density is a disadvantage for the fabrication of
wires and tapes with a fine multi-filament structure, which is
essential for stabilization of the conductors.
We previously reported that the Jc properties of ex situ processed
tapes in the high-field region are improved when using commer-
cially available MgB2 powder treated with organic acid solutions
* Corresponding author.
E-mail address: fujii.hiroki@nims.go.jp (H. Fujii).
http://dx.doi.org/10.1016/j.jallcom.2016.01.027
0925-8388/© 2016 Elsevier B.V. All rights reserved.

H. Fujii et al. / Journal of Alloys and Compounds 664 (2016) 650e656
651
[13]. This treatment brings about the dissolution of MgO on the
surface layers and helps promote subsequent carbon substitution
for boron in heat-treated tapes. The treatment with just organic
solvent such as pure benzene does not cause the carbon substitu-
tion, suggesting that the dissolution of the MgO layers is required
for the substitution reaction. Carbon substitution is one of the most
effective techniques to improve the Jc properties in the high-field
The packing factor of the core layer in the tape samples using
MgB2 powder was calculated by dividing the density of the core by
that of MgB2, 2.63 g/cm3. The density of the core was estimated by
dividing the mass of a piece of the core by the volume of the piece.
The volume was roughly calculated by measuring the size of the
piece with a vernier calliper.
Phase identification was carried out by an X-ray diffraction
(XRD) method with a Rigaku RINT-TTR III X-ray diffractometer with
region due to the enhancement of the upper critical field (Bc2
)
[14,15]. Alternatively, the MgO surface layers are mechanically
removed by a ball-milling process, which is also effective in pro-
moting the carbon substitution reaction [16e19].
MgO is reduced with CaC2, in accordance with the following
reaction [20]:
Cu Ka radiation. XRD patterns of the tape samples were recorded
after peeling off the Fe sheath. Therefore, the interface region be-
tween the core layer and the sheath was investigated. The micro-
structural observation was performed with
a Hitachi SU-70
scanning electron microscope (SEM) with an energy dispersive X-
ray (EDX) spectrometer.
MgO þ CaC2/Mg þ CaO þ 2C
(1)
The temperature dependence of the direct-current (dc)
magnetization was measured for the heat-treated tape samples by
means of a Quantum Design MPMS XL superconducting quantum
interference device (SQUID) magnetometer. The dc magnetization
measurements were carried out at temperatures above 5 K in a field
of 10 Oe in a zero-field cool (ZFC) mode. The fields were applied
perpendicularly to the tape surface after peeling off the Fe sheath.
The critical temperature (Tc) was defined as the onset temperature
at which a diamagnetic signal was observed. The dc critical current
(Ic) measurement of the tape samples was carried out at 4.2 K and
in fields up to 12 T applied parallel to the tape surface by a con-
This reaction starts to occur around 900 ꢀC. In the reaction of
degraded MgB2 and CaC2, MgO surface layers are expected to be
removed. Such reaction can also occur for other Ca compounds. In
case of CaC2, resultant product C can act as a source for the carbon
substitution, in accordance with the following reaction:
MgB2 þ pC/MgB2ꢁpCp þ pB
(2)
The released B from MgB2 in (2) may react with Mg in (1) to form
MgB2 or the excess CaC2 to form CaB6. Here the reduction of MgO
on the surface layers is expected to improve the coupling of MgB2
grains and the possible formation of MgB2 and CaB6 means the
collection of the B released from MgB2 in (2). The local formation of
MgB2 and CaB6 may suppress the deterioration of the grain
coupling compared with the released B scattered all over the
samples, although CaB6 is non-superconducting phase.
ventional four-probe resistive method. The Ic criterion was 1 mV/cm.
3. Results and discussion
3.1. Powder sample
In this paper first we study the reactions of MgB2 with Ca
compounds, CaC2 and CaH2, and then the superconducting prop-
erties of ex situ processed tapes using MgB2 powder with the ad-
ditions of those compounds.
The as-received CaC2 powder exhibited light red brown colour
and XRD data recorded under Ar atmosphere indicated that it
contains CaO as main impurity. By a reference intensity ratio (RIR)
method, the CaO content was estimated about 7wt%. On the other
hand, the as-received CaH2 powder showed white colour and some
negligible small peaks unassignable to MgB2 in the XRD pattern.
2. Experimental
MgB2 (Alfa Aesar; 325 mesh) was mixed with CaC2 (Aldrich; 80%
in purity; pieces) or CaH2 (Strem Chemicals; 95%, granules) with
molar ratios of MgB2: R ¼ 100: x (R ¼ CaC2 and CaH2; x ¼ 0, 1, 3, 6
and 10). The CaC2 pieces and the CaH2 granules were preliminary
crushed in an agate mortar in an Ar gas-filled glove box. The above-
mentioned mixtures were obtained by hand and through a ball
milling process. In the milling process, 2 g of those mixtures were
put in a WC jar and ball milled at a speed of 500 rpm with 5-mm
diameter WC balls for 1 h. The mixtures obtained by those two
techniques were pelletized and then sealed in a stainless steel tube
in an Ar atmosphere. Subsequently they were heat treated at
800e920 ꢀC for 1 h.
Some of those mixtures were packed into a Fe tube with an
outer diameter of 6.35 mm and an inner diameter of 3.5 mm and
subsequently sealed in the glove box. Then, they were deformed
into a rectangular-shaped tape, with a width of 4 mm and a
thickness of 0.6 mm, by means of groove rolling and flat rolling
machines. These tapes were heat treated at 800e950 ꢀC for 1 h in a
flow of Ar gas. Multiple powder and tape samples were prepared to
check reproducibility.
Chemical analyses on the ball milled powders were carried out
by inductively coupled plasma (ICP) - optical emission spectroscopy
(OES) with an Agilent 720-ES. The analyses were performed twice
for the same powder to check reproducibility. Thermogravimetric
Analysis (TGA) was performed with a MAC Science TG-DTA2000
differential thermogravimetric analyser at a heating rate of 5 ꢀC/
min in a flow of Ar gas.
Fig. 1. A TGA curve of as-received CaH2 powder. The sample weight starts to decrease
around 700 ꢀC, as indicated by an arrow.

652
H. Fujii et al. / Journal of Alloys and Compounds 664 (2016) 650e656
Fig. 1 shows a TGA curve of as-received CaH2 powder. The profile
shows a decrease of the sample weight starting around 700 ꢀC. This
suggests that CaH2 starts to release H2 gas around this temperature.
Fig. 2(a) and (b) shows XRD patterns of sintered hand mixed
powders with the additions of CaC2 and CaH2 with x ¼ 10. Here the
XRD pattern of sintered pure MgB2 powder is also shown in
Fig. 2(c). The sintering temperature for those samples was 920 ꢀC,
which is optimized for the Jc properties of the tape samples using
those hand mixed powders. The sintered MgB2 contains a trace of
MgO and MgB4. The MgB4 was not observed in the as-received
MgB2 powder. MgB4 is formed by the decomposition of MgB2 in
the sample without addition. Some XRD peaks which cannot be
assigned to MgB2 are observed for Fig. 2(a) and (b). These impu-
rities were identified as MgO, CaB6 and Mg. Although they are
observed for both samples, those other than MgO are less clearly
observed for the sample with CaC2 addition. XRD peaks assigned to
CaB6 were more appreciable with increasing x, and those assigned
to other Ca compounds and MgB4 are not detected for both
samples.
XRD patterns of the sintered ball milled powders with the ad-
ditions of CaC2 and CaH2 with x ¼ 6, and without addition are
shown in Fig. 3(a) e (c), respectively. The sintering temperature for
those samples was 850 ꢀC, which is optimized for the Jc properties
of the tape samples using those ball milled powders. Compared
with the sintered pure MgB2 powder shown in Fig. 2(c), the cor-
responding sintered ball milled sample shows additional XRD
peaks, which are assigned to WC. The content of WC in the ball
milled MgB2 powder was around 1mass% by ICP analysis. Further-
more, XRD peaks assigned to MgB4 are much more appreciable. The
additions of the Ca compounds result in the formation of CaB6 and
the reduction of MgB4, which was also observed for the sintered
hand mixed samples, as shown in Fig. 2(a)e(c). In the samples with
the additions, the Ca compounds react with MgB4 or decomposed
MgB2 to form CaB6 and reduce the formation of MgB4.
Fig. 3. XRD patterns of sintered ball milled powders with the starting compositions of
MgB2: R ¼ 100: 6 (R ¼ (a) CaC2 and (b) CaH2), and (c) sintered ball milled pure MgB2
powder. The sintering temperature was 850 ꢀC. XRD peaks assigned to MgB2 are
indexed, whereas the peaks assigned to MgO, CaB6, MgB4 and WC are indicated by
triangles, rectangles, rhombuses and circles, respectively.
unknown phases which are possibly formed by the reaction be-
tween MgB2 and the impurities in the as-received CaC2 are super-
imposed on those of CaB6.
Fig. 4(a) and (b) shows the region of 102 peak of MgB2 and 220
peak of MgO in the XRD patterns of the sintered hand mixed and
ball milled powders shown in Fig. 2(a)e(c) and Fig. 3(a)e(c),
respectively. Tables 1 and 2 list the content of each phase observed
in those sintered samples. Regarding the relative intensity between
the 102 peak and the 220 peak, the relative intensity of the 220
peak does not change remarkably by the additions of the Ca com-
pounds for the sintered hand mixed samples. The MgO content was
estimated to be 12.0(19)wt% for sintered pure MgB2, whereas those
were 10.4(9) and 8.5(3)wt% for the additions of CaC2 and CaH2,
respectively, by an RIR method. The content ratios of MgB2 to MgO
were 6.8, 7.7 and 8.1 for those samples.
XRD peaks assigned to Ca compounds other than CaB6 were not
detected for both sintered hand mixed and ball milled samples with
the additions of the Ca compounds. However, some of the observed
XRD peaks assigned to CaB6 were broad for those samples with
CaC2 addition. This suggests that the XRD peaks assigned to
For the sintered ball milled samples, in contrast, the relative
intensity clearly decreases by the additions. The MgO content was
estimated to be 8.4(5)wt% for sintered pure MgB2, whereas those
were 4.9(6) and 8.1(3)wt% for the additions of CaC2 and CaH2,
respectively. The corresponding MgB2 contents were 65.0(17),
81(3) and 78.6(13)wt%, and the content ratios of MgB2 to MgO were
7.7,16.5 and 9.8 for those samples. The decrease of MgB4 and MgO is
linked with the increase of MgB2 and CaB6 for the samples with the
additions of the Ca compounds.
Although 220 peak of CaB6 and 303 peak of MgB4 are super-
imposed on 102 peak of MgB2 [21,22], the contents of CaB6 and
MgB4 are low in those samples, compared to that of MgB2.
Furthermore, the relative intensities of the 220 and the 303 peaks
to the strongest peaks of CaB6 and MgB4 are 47.3/1000 and 1.3/
1000, respectively. Therefore, the contribution to the intensity of
the 102 peak of MgB2 by those impurities is negligible. Thus, the
reduction of MgO and the increase of MgB2 occur by the additions
of the Ca compounds. This is more effective through the ball milling
process, probably because the as-received materials are damaged
by moisture and oxygen and the ball milling process successfully
improves the reactivity of those materials.
Fig. 2. XRD patterns of sintered hand mixed powders with the starting compositions
of MgB2: R ¼ 100: 10 (R ¼ (a) CaC2 and (b) CaH2), together with the pattern of (c)
sintered pure MgB2 powder. The sintering temperature was 920 ꢀC. XRD peaks
assigned to MgB2 are indexed, whereas the peaks assigned to MgO, CaB6, MgB4 and Mg
are indicated by triangles, rectangles, rhombuses and circles, respectively.
Fig. 5(a) and (b) shows 002 and 110 XRD peak profiles of MgB2 in
various sintered ball milled powders, respectively. The profiles are
shown for the samples with the additions of the Ca compounds and

H. Fujii et al. / Journal of Alloys and Compounds 664 (2016) 650e656
653
Fig. 4. The profiles of 102 XRD peak of MgB2 and 220 XRD peak of MgO for (a) sintered hand mixed powders with the additions of (i) CaC2 and (ii) CaH2 with x ¼ 10 and (iii) without
addition. (b) The corresponding profiles for sintered ball milled powders with x ¼ 6 and without addition.
axis parameters of Mg1ꢁqCaqB2 increase with increasing q [24].
Table 1
The content of each phase observed in the sintered hand mixed powders with the
additions of the Ca compounds with x ¼ 10 and without addition. The content was
estimated with the XRD patterns of Fig. 2 by a RIR method in wt %.
Therefore, taking into account the results that no appreciable
variation of the lattice parameters of MgB2 is observed by the ad-
ditions of the Ca compounds and that the CaB6 content increases
with increasing x for those samples, Ca is probably not incorporated
into the MgB2 lattice. This is presumably due to the insufficient
pressure for the reaction.
Sample
MgB2
MgO
MgB4
CaB6
Mg
CaC2 (x ¼ 10)
CaH2 (x ¼ 10)
MgB2 (x ¼ 0)
80(3)
69(3)
82(5)
10.4(9)
8.5(3)
12.0(19)
e
5.2(3)
18(5)
e
4.53(9)
3.2(1)
e
e
5.8(4)
Therefore, slight contraction of the a-axis parameter suggests
that carbon substitution occurs in the sintered ball milled samples
with CaC2 addition. CaC2 acts as carbon source for the substitution
reaction. The carbon substitution decreases the a-axis parameter of
MgB2, while the c-axis parameter is not influenced by the substi-
tution [15,25,26]. The substitution amount was estimated to be
p ¼ 0e0.01 in MgB2ꢁpCp from the a-axis parameter.
Table 2
The content of each phase observed in the sintered ball milled powders with the
additions of the Ca compounds with x ¼ 6 and without addition. The content was
estimated with the XRD patterns of Fig. 3 by a RIR method in wt %.
Sample
MgB2
81(3)
78.6(13)
65.0(17)
MgO
MgB4
CaB6
WC
1.7(1)
0.98(3)
0.36(7)
CaC2 (x ¼ 6)
CaH2 (x ¼ 6)
MgB2 (x ¼ 0)
4.9(6)
8.1(3)
8.4(5)
8.3(12)
3.45(3)
26.3(11)
3.9(1)
8.9(3)
e
3.2. Tape sample
No binary compounds are reported in the CaeFe system [27]
and XRD analyses on the tape samples indicated that only FerB
(r ¼ 1 and 2) are present at the interface region.
without addition. The positions of those XRD peaks do not shift
appreciably by CaH2 addition. The corresponding sintered hand
mixed sample did not show an appreciable peak shift, either. These
indicate no variation of the lattice parameters of MgB2 by CaH2
addition for both sintered hand mixed and ball milled samples.
Regarding the CaC2 addition, the positions of those XRD peaks
did not shift appreciably for the sintered hand mixed sample,
either. In contrast, for the sintered ball milled sample, a slight shift
of the 110 peak is observed, whereas a shift of the 002 peak is not
appreciable. This indicates the contraction of the a-axis parameter
of MgB2.
Although attempts of preparation of Ca-substituted MgB2,
Mg1ꢁzCazB2, have been carried out so far, the substitution reaction
does not take place even under high-pressure (HP) heat-treatment
under 3.5 GPa and 900 ꢀC when using CaH2 as a starting material,
judging from that the CaH2 or Ca doping does not change the lattice
parameters of MgB2 [23]. On the other hand, when using Ca3B2N4 as
a Ca source, the HP heat-treatment under 5 GPa and 1100 ꢀC brings
about the formation of Mg1ꢁqCaqB2 and both the a-axis and the c-
Fig. 6(a) and (b) shows dc magnetization curves as a function of
temperature for various tapes using filling powders with the ad-
ditions of CaC2 and CaH2, respectively, together with the curves of
the tapes using MgB2 powders without addition. The magnetiza-
tion M is normalized to that measured at 5 K, M(5K), for each tape.
The transition widths (DTcs), defined as 10e90% transition, of the
magnetization curves increased with increasing x for the tape
samples using the hand mixed filling powders with the additions of
the Ca compounds. The increase of the
DTc was much larger for the
samples with CaC2 addition. The Tc values were 7.6 and 4.3 K for
D
the samples with the additions of CaC2 and CaH2 with x ¼ 10,
respectively, whereas the corresponding value was 0.9 K for the
sample without addition.
The ball milling process decreased the Tc and increased the DTc
for the tapes using pure MgB2 powder. They are probably caused by
the stress in MgB2 crystal and the increased content of MgB4 and
WC particles, which were brought about by the milling process.
Tapes using ball milled powders with CaH2 addition does not show
the increase of the
DTc appreciably as those using the

654
H. Fujii et al. / Journal of Alloys and Compounds 664 (2016) 650e656
Fig. 5. (a) 002 and (b) 110 XRD peak profiles of MgB2 in sintered ball milled powders with the starting compositions of MgB2: R ¼ 100: 6 (R ¼ (i) CaC2 and (ii) CaH2), together with
(iii) the profiles in sintered ball milled pure MgB2 powder. The dashed lines are drawn at 2
q
¼ 52.0ꢀ and 60.0ꢀ for (a) and (b), respectively.
Fig. 6. Temperature dependence of the dc magnetization curves for various tapes using the hand mixed and ball milled powders with the starting compositions of x ¼ 0, 1, 6 and 10
in MgB2: R ¼ 100: x (R ¼ (a) CaC2 and (b) CaH2). The measurements were carried out in a magnetic field of 10 Oe in a zero-field cool (ZFC) mode. The magnetization M is normalized
to that measured at 5 K, M(5K), for each tape.
corresponding hand mixed powders. In contrast, CaC2 addition
causes tremendous broadening. The sample with CaC2 addition
with x ¼ 6 shows very broad transition with a kink at about 22 K.
sample without addition. The sample without addition shows
smaller Tc, compared with those with CaC2 addition.
D
Fig. 7(a) and (b) shows the transport Jc as a function of magnetic
field for the tapes using the filling powders with the additions of
CaC2 and CaH2, respectively. For the hand mixed filling powders,
the optimized heat treatment temperature at which the best per-
formance in Jc was obtained did not change remarkably by the
additions of the Ca compounds. The Jc properties of the tapes using
powders with CaC2 addition slightly degrade with increasing x. In
contrast, they are improved by the addition of a slight amount of
CaH2, x ꢃ 3. However, further addition degrades the Jc properties
and tapes using the powders with x ꢂ 6 in the Jc property are
inferior to those with x ¼ 0.
The increase of the DTc suggests weak coupling of grains. Therefore,
CaH2 addition does not cause remarkable weak coupling, whereas
severely weak coupling occurs by CaC2 addition when x ꢂ 6. This
severely weak coupling is probably brought about by the impurities
in the as-received CaC2, since the addition of CaC2 reduces MgO,
which is considered the obstacles to the supercurrent path, more
effectively than that of CaH2, as shown in Fig. 4(b) and CaB6 formed
in the sintered samples is main impurity phase for both additives.
Furthermore, the content of another main impurity phase MgB4 in
the samples with CaC2 addition is much lower than that in the

H. Fujii et al. / Journal of Alloys and Compounds 664 (2016) 650e656
655
Fig. 7. Transport Jc at 4.2 K as a function of magnetic field for various tapes using the hand mixed and ball milled powders with the starting compositions of x ¼ 0, 1, 3, 6 and 10 in
MgB2: R ¼ 100: x (R ¼ (a) CaC2 and (b) CaH2). The magnetic field was applied parallel to the tape surface.
The ball milling process decreased the optimized heat treatment
temperature by 50e80 ꢀC, which was independent of the additives.
The Jc property of the tape with x ¼ 0 is enhanced by the process in
spite of the increase of the volume fraction of MgB4 and the Jc value
at 9 T is improved by two orders of magnitude. Such Jc improve-
ment by using ball milled filling powder was reported for the ex situ
processed wires and tapes [16e19,28]. The difference between the
packing factors of the core of the tapes with x ¼ 0 using the as-
received and the ball milled MgB2 powders was not appreciable
and the factors were 0.55e0.60 for both powders. This Jc
enhancement is probably due to the increased pinning force and
enhanced Bc2 by grain refinement [28].
The additives further improve the Jc properties. The Jc properties
are enhanced in the high-field and low-field regions for the tapes
with the additions of CaC2 and CaH2, respectively. These Jc en-
hancements are observed just for x ꢃ 3 for both additives. The
samples with x ꢂ 6 show significant Jc degradation, that is inde-
pendent of the additives. Especially, the samples with CaC2 addition
with x ꢂ 6 did not carry supercurrent at all. This is probably because
CaB6 formed by the additions of the Ca compounds is not negligible
to the grain coupling for x ꢂ 6 and the as-received CaC2 contains a
substantial amount of impurities such as CaO. Those already
existing impurities and/or impurities produced after heat treat-
ment of the samples other than CaB6 can act as obstacles to the
supercurrent path. The formation of such newly produced impu-
rities is suggested by the broadenings of XRD peaks in the sintered
ball milled powder samples with CaC2 addition, as mentioned
above. For the tapes with CaC2 addition, those impurities cause
The Jc enhancement by CaC2 addition was observed only when
using the ball milled filling powders. This is probably because the
as-received CaC2 and MgB2 powders were contaminated by mois-
ture and oxygen. The ball milling process can remove those
damaged surface layers and activate those powders. Thus, the re-
actions of the reduction of MgO and the substitution of carbon
which is supplied from the CaC2 are promoted, although the sub-
station amount is currently close to 0.
For the samples with CaH2 addition, when the using hand mixed
filling powders, Jc enhancement in the high-field region is observed
for small x. In contrast, when using the ball milled filling powders,
the Jc enhancement shows an opposite trend. The Jc enhancement
in the low-field region is clearly observed, while that in the high-
field region is not appreciable. The field dependence of the Jc
properties of the tapes using the hand mixed powders is almost the
same as that using the ball milled powders. This enhancement is
probably due to the improved grain coupling and the increased
content of MgB2, since the
that of the tape with x ¼ 0, as shown in Fig. 6(b).
D
Tc of the tape with x ¼ 1 is smaller than
Fig. 8(a) e (c) shows SEM images of the fractured surface of the
core layer of the tapes using the ball milled powders without
addition and with the additions of CaC2 and CaH2 with x ¼ 6,
respectively. The average grain size of MgB2 decreases by the ball
milling process from several hundred to a few hundred microns for
the tape using the powder without addition. The CaC2 addition
does not change the grain size remarkably, whereas the CaH2
addition improves the grain connectivity and promotes the grain
growth up to 1 mm. CaH2 starts to decompose and release H2 at
large
DTc and the kinks in the transition curves shown in Fig. 6(a),
around 700 ꢀC and the melting point of Ca is 842 ꢀC. The Ca acts as
reducing agent for MgO and probably promotes the grain growth of
MgB2.
and hence weak coupling of MgB2, although substantial Jc
enhancement was expected from the large reduction of MgO
observed for the sintered ball milled samples with CaC2 addition.
The field dependence of the Jc properties for the tapes using ball
milled powders with CaC2 addition with x ꢃ 3 was lower than that
for the corresponding tape with x ¼ 0. This is due to enhanced Bc2
brought about by carbon substitution [14,15]. The sintered ball
milled sample with CaC2 addition shows a slight shift of the 110
XRD peak of MgB2, as shown in Fig. 5(b), suggesting a slight amount
of carbon substitution.
Elemental mapping for the fractured surface of the core layer of
the tapes using the ball milled powders with the additions of the Ca
compounds revealed that Ca-rich region is present in the core layer
in those tapes. However, CaB6 grains were not detected clearly by
SEM. In contrast, WC particles with the size of a few hundred
nanometres were observed clearly.

656
H. Fujii et al. / Journal of Alloys and Compounds 664 (2016) 650e656
Fig. 8. SEM images of the fractured surface of the core layer of the tapes using the ball milled (a) MgB2 powder and mixtures with the starting compositions of x ¼ 6 in MgB2:
R ¼ 100: x (R ¼ (b) CaC2 and (c) CaH2).
285e291.
4. Conclusion
[7] J.M. Hur, K. Togano, A. Mastumoto, H. Kumakura, H. Wada, K. Kimura,
Supercond. Sci. Technol 21 (2008) 032001.
[8] G.Z. Li, M.D. Sumption, M.A. Susner, Y. Yang, K.M. Reddy, M.A. Rindfleisch,
M.J. Tomsic, C.J. Thong, E.W. Collings, Supercond. Sci. Technol. 25 (2012)
115023.
[9] H. Fujii, H. Kumakura, K. Togano, Phys. C 363 (2001) 237e242.
[10] J.M. Rowell, Supercond. Sci. Technol. 16 (2003) R17eR27.
[11] J. Jiang, B.J. Senkowicz, D.C. Larbalestier, E.E. Hellstrom, Supercond. Sci.
Technol 19 (2006) L33eL36.
[12] A. Yamamoto, H. Tanaka, J.-I. Shimoyama, H. Ogino, K. Kishio, T. Matsushita,
Jpn. J. Appl. Phys. 51 (2012) 010105.
[13] H. Fujii, K. Togano, K. Ozawa, J. Mater. Res. 22 (2007) 1281e1285.
[14] T. Masui, S. Lee, S. Tajima, Phys. Rev. B 70 (2004) 024504.
[15] R.H.T. Wilke, S.L. Bud’ko, P.C. Canfield, D.K. Finnemore, R.J. Suplinskas,
S.T. Hannahs, Phys. Rev. Lett. 92 (2004) 217003.
[16] B.J. Senkowicz, J.E. Giencke, S. Patnaik, C.B. Eom, E.E. Hellstrom,
D.C. Larbalestier, Appl. Phys. Lett. 86 (2005) 202502.
[17] V. Braccini, A. Malagoli, A. Tumino, M. Vignolo, C. Bernini, C. Fanciulli,
G. Romano, M. Tropeano, A.S. Siri, G. Grasso, IEEE Trans. Appl. Supercond 17
(2007) 2766e2769.
Heat treatment of MgB2 powder with the additions of Ca com-
pounds, CaC2 and CaH2, results in the formation of CaB6 and Mg.
The MgO content in those samples decreases compared with that in
the sintered pure MgB2 powder. This is due to the reduction of MgO
by the Ca compounds. The lattice parameters of MgB2 do not
change appreciably by CaH2 addition, whereas the a-axis param-
eter contracts slightly by CaC2 addition when using ball milled
mixtures. Ca is probably not incorporated into the MgB2 lattice and
the variation of the lattice parameters in the samples with CaC2
addition is attributed to carbon substitution. The Jc properties of the
tapes are enhanced in the high-field and low-field regions by the
additions of CaC2 and CaH2, respectively when using the ball milled
mixtures as filling powders. These enhancements are attributed to
the carbon substitution, the improved grain coupling and the
increased content of MgB2.
€
ꢀꢁ
ꢁ
[18] A. Kario, V. Grinenko, A. Kauffmann, W. Haßler, C. Rodig, P. Kovac, T. Melisek,
B. Holzapfel, Phys. C 483 (2012) 222e224.
[19] H. Fujii, K. Ozawa, H. Kitaguchi, Supercond. Sci. Technol 25 (2012) 065008.
[20] R. Matsuura, M. Uchino, H. Tsuda, Kougyoukagakuzassi 42 (1939) 590e597.
[21] C.H. Chen, T. Aizawa, N. Iyi, A. Sato, S. Otani, J. Alloys Compd. 366 (2004)
L6eL8.
[22] A. Guette, R. Naslain, J. Galy, C.R. Seances, Acad. Sci. Ser. C 275 (1972) 41e44.
[23] P. Toulemonde, N. Musolino, R. Flükiger, Supercond. Sci. Technol 16 (2003)
231e236.
Acknowledgements
This research was supported by JSPS KAKENHI Grant Number
25420696.
[24] Y. Sun, D. Yu, Z. Liu, T. Wang, J. He, J. Xiang, D. Zheng, Y. Tian, Supercond. Sci.
Technol 20 (2007) 261e266.
References
[25] T. Takenobu, T. Ito, D.H. Chi, K. Prassides, Y. Iwasa, Phys. Rev. B 64 (2001)
134513.
[26] S. Lee, T. Masui, A. Yamamoto, H. Uchiyama, S. Tajima, Phys. C 397 (2003)
7e13.
[27] E. Schürmann, H. Jacke, Steel Res. 58 (1987) 399e405.
[28] A. Malagoli, V. Braccini, M. Tropeano, M. Vignolo, C. Bernini, C. Fanciulli,
G. Romano, M. Putti, C. Ferdeghini, E. Mossang, A. Polyanskii, D.C. Larbalestier,
J. Appl. Phys. 104 (2008) 103908.
[1] J. Nagamatsu, N. Nakagawa, T. Muranaka, Y. Zenitani, J. Akimitsu, Nature 410
(2001) 63e64.
[2] http://www.columbussuperconductors.com/.
[3] http://www.hypertechresearch.com/.
[4] B.A. Glowacki, M. Majoros, M. Vickers, J.E. Evetts, Y. Shi, I. McDougall,
Supercond. Sci. Technol 14 (2001) 193e199.
[5] S. Jin, H. Mavoori, R.B. van Dover, Nature 411 (2001) 563e565.
[6] G. Giunchi, S. Ceresara, G. Ripamonti, A. Di Zenobio, S. Rossi, S. Chiarelli,
M. Spadoni, R. Wesche, P.L. Bruzzone, Supercond. Sci. Technol 16 (2003)