In: Studies of superconductors (Advances in Research and Applications) MgB2 ed. A.Narlikar, Nova
Science Publishers, Inc., Huntington, New York, 2001, v38. p.361
THE MgB2 SUPERCONDUCTING CONDUCTORS FOR DC AND AC
APPLICATIONS
B.A.Glowacki 1,2 and M.Majoros 1,3
1) IRC in Superconductivity, University of Cambridge, Madingley Road, Cambridge CB3
0HE, UK
2) Department of Materials Science and Metallurgy, University of Cambridge, Pembroke
Street, Cambridge CB2 3QZ, UK
3) On leave from Institute of Electrical Engineering, SAS, Bratislava, Slovak Republic
1. INTRODUCTION
Of the 96 binary boride systems known, only 60 have compounds of boron with metals
of the Ia-VIIIa subgroups crystallising in more than 30 structural types. The stability of
borides formed by metals decreases with increasing atomic number, in extreme cases
borides of copper do not exist and Cu-B, Ag-B and Au-B systems are simple eutectic
alloys [1]. An essential role in the boride structures, unlike the carbides of the same metals,
is that of the electron interaction between the boron atoms. This interaction depends both
on the electron state of the metal, Me, and on the value of the atomic ratio of boron to
metal. In particular, with an increase in this ratio the structure formation evolves from one
of the isolated boron atoms to an unbroken three-dimensional framework, and this can be
shown by the following series: Me4B, Me3B, Me2B, MeB, Me3B4, MeB6,
MeB 66,....,MeB 100. The electron interaction between the boron atoms accompanied by
the formation of strong covalent bonds is one of the main reasons that the diborides
crystallise in the structure type MeB2, Figure 1. There are four phases in binary phase
Figure 1 Parallel layers of magnesium metal and boron atoms in MgB2 . [2]
361
diagram Mg-B: diboride MgB2, hexaboride MgB6, tetraboride MgB4 and dodecaboride
MgB12. The influence of the magnetic states on the properties of borides decreases in the
direction of increasing number of B atoms.
The discovery of superconductivity in the well known MgB2 material at the beginning of
the new 21st century [3] makes all of us aware of possibilities of existence of new
superconducting materials which may have critical parameters in some cases even better
than existing superconductors. As it was described in the New Scientist ‘In just a few
months, a new superconductor has upstaged its rivals’[4]. This important discovery
stimulated two types of research: first towards a combinatorial search for new
superconducting materials, the second type of research is focused on optimum thermomechanical procedures for the manufacture of the powder-in-tube and coated conductors
using doped and substituted Mg-B compounds. In this chapter we are discussing the
second approach where one is going to design and manufacture conductors for the dc and ac
energy devices applications such as MRI, NMR magnets, transformers. There is also a
strong hope that the new reliable superconducting joints and superconducting switches will
be made not only based on the MgB2 new conductors but also between MgB2 and low
temperature superconducting conductors such as NbTi, Nb3Sn and Nb3Al opening a new
range of the electromagnetic device applications.
There are two different methods: an in situ technique where an Mg+2B mixture can be used
as a central core of the powder-in-tube, PIT, conductor or as an external coating and reacted
in situ to form MgB2, and an ex situ technique where fully reacted MgB 2 powder which
may be doped or chemically modified can be used to fill the metal tube. In the case of the ex
situ coated conductors sintering will not be required if the additional densification process
in the form of multilayer sandwich structure will produced fully interconnected final
material.
Commercial exploitation of the recently discovered MgB2 superconductors will be severely
limited unless mechanically robust, high critical current density, composite conductors can
be fabricated with uniform properties over long lengths. Our experience with LTS such as
A15 conductors Nb3Sn , Nb3(Al,Ge) and also the whole range of HTS conductors suggests
that to be able to unfold the full superconducting potential of the metal-borides, one has to
address fundamental questions concerning the practical capability of the newly discovered
material to conduct intergranular supercurrents in high magnetic fields considering the short
coherence length ο =5.2nm [5], which it appears is anisotropic and may vary from 2-7nm
dependeing on the sample preparation, see Table1.
362
2. LAYERED STRUCTURE, ANISOTROPY AND CRITICAL PARAMETERS OF
MgB2
The published critical field parameters data for the first MgB2 flux grown single crystal
show that there is a noticeable anisotropy, , manifested in H c2(0) and Hc1(0) and (0)
[6]. Extensive XRD measurements on pressed pellets indicated that mechanical pressing
does not provide noticeable preferential alignment of the MgB2 powder, therefore one may
expect that making tape-form conductors will be not a necessity as it is in case of HTS.
Table 1: Properties of the different forms of MgB2.
_________________________________________________________________________
Ref.
Sample
Tc (K) Hc2(0)⊥a-b(T) Hc2(0)⊥c(T) (0)ab(nm) (0)c(nm)
_________________________________________________________________________
Single crystal
39
9.2
25.5
6.5
2.5
2.6 [6]
Alloyed film
31
19.5
39
4
2
2
[11]
(c-oriented)
Textured bulk
39
6.5
11
7
4.1
1.7 [9]
_________________________________________________________________________
Additionally from a practical point of view, it is much better to manufacture round wires
than tapes. However it may look like a final confirmation of the isotropic character of the
MgB2 but there is data in the literature which brings firm indication about anisotropy of
the MgB 2 crystal which has an AlB2 structure. In this layered structure, boron atoms form
honeycomb layers alternately with hexagonal layers of Mg atoms, Figure 1.
However as it appears the rolling procedure densified the internal MgB2 core improving
intergrain connectivity increasing that way the critical current capabilities of the conductor.
It was established [7] that in a randomly oriented grains, no preferential grain orientation
was induced and considering that different crystal directions were tested no signal of gap
anisotropy was found yet. In the framework of the BCS model, a critical temperature at the
surface (T c = 13.2 K) was found for the measured gap value (∆ = 2.0 meV). The data
concerning the gap value vary substantially and it is not currently obvious which value is
correct. Despite an overall isotropic behaviour of MgB2 in conductors observed by many
authors, there are small anisotropies in the mechanical properties. The electronic structure
under pressure reveals the 2-dimensional character of the B pxy states whose occupancy is
altered by pressure in contrast to the B p z states [8]. Anisotropic superconducting
properties of aligned MgB2 crystallites were measured recently by de Lima et al. [9]. They
observed an upper critical field anisotropy H c2(||ab)/Hc2(||c) ~1.7, which implies an
anisotropy of the critical current density Jc(H||c)/Jc(H||ab) ~ ( ab/ c)2 ~ H c2(||ab)/Hc2(||c)
363
~ 1.7. There are also other data available, see Table 1, which suggest that MgB2 is
anisotropic which may have a serious consequences for the transport current capabilities of
high current conductors.
Generally one may argue that the width of a typical high angle grain boundary is of order
ten times the lattice constant which in the case of MgB2 (a=0.314nm, c=0.352 nm) is
almost comparable to the coherence length, Table 1. Therefore one may assume that well
connected high-angle grain boundaries of MgB2 may serve as potential pinning centres as in
A15 intermetallics which have similar values of the coherence length and lattice parameter.
Some data in the literature provide the value of upper critical field, H c2(0) =18T [10]
which are lower than Nb 3Sn. This however should not exclude this material from the list of
potential superconducting wires to be used in the 20K region for cryogen-free magnets and
transformers. It was shown that the H c2(0)⊥c(T) in thin films may reach 39T which is a
very encouraging result [11].
A range of critical temperatures have been reported for ‘MgB2’ depending on the purity
and boron isotope content: MgB2 T c=37.5K (∆T~10K)(Alfa-Aesar-99.8% purity); MgB 2
Tc=38.5K (∆T~1K) (Alfa-Aesar-99.8% purity sintered at 1000oC under pressure (p~3
GPa) [12] ; Mg11B2 T c=39.2K (∆T=0.4K); Mg10B2 T c=40.2K (∆T=0.5K) [13]. (The
natural mixture of isotopes is 10B ~19%, 11B ~81%). The measured value for the gap, ∆ =
2.0 meV, and the corresponding critical temperature equal to 13.2K obtained at the surface
of small crystals of commercially available powder [7] is significantly lower than would be
expected from the conventional BCS expression ∆= 1.76k BTc (5.9 meV) taking the critical
temperature of bulk material, Tc =37.5 K. This discrepancy could be due to a deviation of
the density of states at the surface with respect to the bulk. It should be emphasised that in
tunnelling spectroscopy measurements it is the surface density of states that is probed,
which can be significantly influenced by the presence of impurities, strong defects or
stoichiometry variations close to the surface, at distances in the range of the
superconducting coherence length ο =5.2nm. At some locations on the crystal surface,
curves indicative of superconductor-insulator-superconductor, SIS, intergrain tunnelling
behaviour are observed, consistent with the gap value measured on the free surface. From
the point of view of the PIT technique, at this stage using commercial powder without any
special cleaning process such data on the superconductivity at a grain surface is of
paramount importance. Analysing the possible structure deformation during mechanical
deformation it appears to us that during mechanical deformation the superconducting
properties of the grain boundary surface can be easily modified to such extent that there is
continuous ‘boundaryless’ current transfer.
It is reported that MgB2 strands can be successfully formed by directly filling Nb-lined,
monel tubes with commercially available MgB 2 powder and then drawing and rolling into
tapes followed by sintering at 900oC [14]. A self field transport current of 105Acm-2 at
364
4.2 K was measured.
There are two additional processes relevant to conductor development reported in current
literature: one involves Mg vapour diffusion to boron fibres presented by Canfield et al.
[15], and the second is related to coated conductors where epitaxial MgB2 films were
deposited on Al2O3 substrates and were characterised by a magnetic critical current
density of J cmag =8x106Acm-2 at 5K [16]. The successful process of MgB2 formation
via diffusion of Mg vapour into 100µm boron fibres encapsulated in Ta tubes has some
difficulties concerning the uniformity and granularity of the final fibres. Whereas the boron
fibres were straight and moderately flexible before the reaction, the MgB2 fibres after the
reaction were brittle and deformed. There has been significant warping and bending of the
fibre as a result of the reaction with the Mg vapour. The observed increase in volume of the
B fibres during conversion to high density MgB2 fibres was estimated to be ~150%. Such
increase in volume may strongly influence the integrity of the conductors manufactured by
internal in situ reactive diffusion between B and Mg where Kirkendal voids will be present.
A similar effect was previously reported in the case of Nb3Sn where Nb which has a bcc
structure was converted to Nb3Sn which has A-15 structure causing a 37% increase in
volume [17,18].
Part of the current effort has been also focused on successful conductor formation by the ex
situ PIT technique. It should be mentioned that corresponding process whereby
superconducting Nb3Sn conductor is fabricated by starting with the actual A-15 powder is
not a practical option. For brittle intermetallic high current conductors a conventional
diffusion process should be more advantageous. Current data for ex situ PIT MgB2
conductor including thermo-mechanical and post deformation heat treatment substantially
contribut to the knowledge concerning the properties and preparation paths of the possible
future high current conductors which are better than Nb3Sn at temperatures close to 15K
and also have a significantly high irreversibility line comparable to Bi-based HTS
conductors at this temperature [2], [19-24]. The difference in the melting temperature
between Mg and B of ∆T~1500oC and the dramatic difference in microhardness of B and
Mg as well as the limited workability of B and also MgB2 (H v=1700-2800) [25] at room
temperature makes a uniform and controllable deformation process difficult. The fact that
B has hardness Hv=49000 second only to diamond is also significant. It has to be stressed
that the compaction of the powder followed by swaging, drawing and rolling of the MgB2
ex situ powder in metallic tubes does not cause any problems and the deformation is similar
as for the graphite deformation [26]. The alternative approach, an in situ PIT wire made
from fine constitute powders such as B and Mg, is expected to be a better solution,
however, deformation and solid-liquid diffusion processes are going to be complex and
difficult. This is firstly because wires manufactured by the in situ technique, diffusing Mg
to B particles would experience ~25.5% decrease in density from the initial density value
after cold deformation, due to the phase transformation from Mg + 2(β−B)=>MgB 2 all
365
with hexagonal structure and secondly because of the stability of the higher borides at
lower temperatures.
3. CONDUCTOR PREPARATION AND ARCHITECTURE
Considering the interest in development of a low cost MgB 2 conductor for use in the
medium temperature range our chapter is focused on the in situ and ex situ preparation of
Mg-B PIT conductors. The schematic outline of the processing routes for in situ and ex situ
Mg-B conductor processing is presented in Figure 2. The four major considerations which
determine the design of a suitable PIT MgB2 wire are as follows.
i) Probably the most critical problem is to prevent substantial oxidation of the final product
in order to obtain an optimum critical temperature and intergranular connectivity. As it
appeared in the case of the thin films moderate oxygen doping increases substantially
Hc2 and Jc with simultaneous but not detrimental reduction of the Tc [11]. Therefore
some trapped oxygen during PIT conductor preparation may be advantageous by
introducing the additional defects. This can be done by embedding MgB2 powder in the
metallic tube and annealing under vacuum or protective atmosphere to avoid oxidation.
Any oxygen trapped in the MgB2 core during PIT preparation will be released to the
structure at the elevated temperatures. Obviously other gasses will be trapped, so the
best way to prepare PIT conductors is to use controlled and optimised atmosphere
during compaction and also if necessary to evacuate the remaining gases from the
composite before mechanical deformation. Also considering the volatility of Mg the PIT
technique compositional uniformities of long wires may not be consistent with short
wires used often for optimisation, since loss of magnesium will occur through the ends of
the conductor.
ii) Mechanical support is of importance for a brittle material. In commercial composites
based on brittle superconductors it is common to place the superconductor under
compression, this can be done by incorporating within the composite a metal with a high
thermal expansion. One may use silver or much cheaper copper supported by stainless
steel, see Figure 3 and Figure 4. Silver and copper annealed at elevated temperatures
become too soft and do not provide a sufficient mechanical support. Silver alloys may
provide more support, however price is a major concern. Therefore copper may be a
lower cost solution characterised by a higher melting point under vacuum (1083 oC) than
Ag (960oC), and providing also a much narrower range of α-solid solution which may be
formed at the interface between Cu and Mg preventing substantial loss of Mg to the
external sheath. Because Ag and Cu sheathed MgB2 wires will be mechanically
extremely weak an attempt was made to reinforce the composite by casing with an outer
366
PIT ex situ process
MgB2 powder compaction
Jc~10 5 Acm -2
Jc~5 x10 5 Acm -2
Jc~10 6 Acm -2
swaging
rolling
annealing
sintering
PIT in situ process
Mg + 2B powder compaction
Jc ~ 5x10 5 Acm -2
swaging
rolling
Coated conductor
doctor blading
dip coating
reactive
diffusion
Jc >10 5 Acm -2
rolling
sintering
ex situ
MgB2
reactive diffusion
Mg + 2B
in situ
Figure 2 MgB2 conductors manufacture. The schematic outline of the ex situ and in situ
processes used for powder-in tube and coated conductor development.
367
layer of stainless steel. Reported use of Ta or Nb tubes for synthesis of the MgB2 [20]
appears to be more expensive solution and justification of such design has to be proven
even if Nb or Ta may be a good deoxidant (getter). The most promising cladding for high
temperature annealing or sintering (>900oC) is Fe widely used by many researchers
[21], [23-24]. For the medium temperature in situ and ex situ sintering the copper
cladding is a very good solution.
iii) In order to transfer current to the superconductor and achieve adequate cryogenic
stabilisation it is essential to clad the superconductor with a metallic coating that does
not develop an insulating oxide barrier at the superconductor interface, in our case Ag
and Cu serve this purpose, however multicomponent composite materials such as Cu
tubes lined with thin Fe inner layer as a diffusion barrier for the high temperature
sintering may shortly be introduced. As reported in the literature, quenching of the wires
with higher resistivity cladding such as Fe, Ni etc. is a problem. Therefore testing of the
high critical current during optimisation process of conductors may use the pulse current
technique widely developed for the rapid characterisation of the low temperature
conductors where the latent heat of the helium coolant at the current contacts limit the
validity of testing. Therefore development of the multifilamentary MgB2 conductors in
Cu-based matrix may appear to be a very important issue.
iv) Finally the composition, state of order, microstructure and defect structure of the
superconductor must be optimised to support a high critical current density. The
presence of oxygen may reduce intergranular current transfer in the case of the in situ
MgB2 powder route, and in the case of the ex situ route chemical stoichiometry during
diffusion formation of MgB2 between Mg and B may be also difficult to control. The
control of powder quality for the ex situ process is of a great importance for the
community since the preparation process may be largely simplified.
For the manufacture of the MgB2 conductors [2] we have adopted the advanced stainlesssteel cladding powder-in-tube technique that was introduced by our laboratory to make the
first YBa2Cu3O7-x oxide superconducting conductors [27, 28].
Similar powder-in-tube approach has recently been used to fabricate metal-clad MgB2
wires using various metals, such as Cu, Ag, Ni, Cu-Ni, or Nb [14], [19], [20], [23], [24],
[29]. In our study, the use of the Cu metal cladding has been found to be critical in
producing well sintered, high Jc MgB2 wires because magnesium in Cu-Mg-B system tends
to react and combine with Cu to form solid solutions or intermetallics with lowered melting
points which requires medium range temperature reactive diffusion processes to deliver the
high critical current densities. Usage of the Cu at temperatures above 800 oC causes
degradation of the superconducting parameters.
There are only a small number of essentially inert metals which exhibit little or no mutual
368
solubility with Mg and do not form intermetallic compounds with Mg at around 900 1000oC. These include Fe, Mo, Nb, V, Ta, Hf and W. Of these, the refractory metals (Mo,
Nb, V, Ta, Hf, W), because of their greatly inferior ductility compared to Fe, are more
difficult if not impossible to process with the substantial plastic deformation involved in
fine-wire or ribbon fabrication [21]. Thus Fe and Ni appear to be the best candidate
material as a practical cladding metal or diffusion barrier for MgB2 wire fabrication,
providing that adequate cryogenic stabilisation is maintained to avoid quenching and
destruction of the conductor. Iron and its alloys are also inexpensive, naturally abundant
and commercially widely available, therefore research towards Cu-Fe-Ni alloys based
composites as optimum cladding material is expected in a very near future.
Usage of the Fe and other inert magnetic metals in a multifilamentary MgB2 conductors
has its additional advantages because interfilamentary magnetic decoupling of the filaments
substantially reduces the transport AC losses [30], [31] and increases the critical current
density [32].
4 CONDUCTORS
4.1 Round conductors
Our conductors were prepared by the standard PIT technique where pure silver or copper
and alternatively bimetallic silver/stainless steel and copper/stainless steel tubes were used.
The cross sections of the final wires manufactured are presented in Figure 3, Figure 4 and
Figure 5. The idea is very similar to that originally developed for stainless steel clad
Figure 3 Cross section of the ex situ MgB2
conductor (a) silver , (b) SS/Ag sheath.
Figure 4 Cross section of the in situ Mg + 2B
conductor (a) copper , (b) silver sheath.
369
Ni/Ag/YBCO [27], [28]. The powder used was a standard MgB2 powder purity 98% from
Johnson Matthey with average grain size of 40µm. The impurities of the powder in ppm
units are Fe>1000, Si~600, Cr~400, Mg~400, Ni~400, Al~200, Ca~100, Cu~60 and Ga~13.
The Mg (99.8%) and B (99.7%) powders were also supplied by Johnson Matthey. The
silver and copper tubes used were OD=6mm ID=5mm and purity 99.99%. The stainless
steel tube was ASTM-A269.
After completion of the cold deformation the 3 cm long pieces of conductor had OD
diameters : 2mm, 1.5mm, 1 mm and 0.5mm. They were placed in quartz containers and
sealed under vacuum and argon. The annealing process was carried out at 350oC, 600oC,
750oC, 850oC and 900 oC for between 2 and 10 hours. The producer data sheet of the
MgB2 material used recommend not to heat the powder above 800oC which was described
as a decomposition temperature. It is difficult to make conclusive judgment at this stage
concerning the maximum sintering temperature of the reacted powder if one takes into
account that the invariant temperature (probably peritectic reaction temperature) is
approximately <1050 oC for MgB2 [33]. Therefore higher temperatures can be used,
however silver should not be used at temperatures too close to its melting temperature.
The AC susceptibility measurements were conducted in the range 33-333Hz in parallel
magnetic field of 1Oe (~80Am -1). Small 3mm in diameter cylinders of MgB2 powder were
prepared for comparative study of the intergranular current and grain connectivity in a
powdered sample.
Using a SQUID magnetometer, magnetisation versus magnetic field (M-H) curves of the
round wires before and after sintering were measured at temperature of 5K and magnetic
field up to 5T to define the Jcmag. A comparison between zero-fied cooled, ZFC, and
field-cooled, FC, susceptibility measurements for different wires was conducted to define
differences in the flux pinning in the wires. Direct dc four point current measurements up
to 900A were conducted at 4.2K in self field.
stainless steel core
Figure 5 kA-class multifilamentary MgB2 conductor for the ac and dc applications; 6 x Cu/Mg2B in situ filaments twisted around central stainless steel core, insert represents individual
single core conductor diameter 0.5mm. [29].
370
The role of copper as a stabilising layer is a very important for the stability of the
conductor. As it emerges, in a simplistic design where only Ni or Fe matrix is used the
instability of the current appears to be a very important issue [19], [21], [24]. Therefore
usage of Cu as a stabilising layer is important approach to protect wire from damage in real
power applications, Figure 6. Usage of Cu is recommended in the case of the in situ wires
where the high temperature reaction is used to form MgB2 as a product of Mg+2B
reaction.
stainless steel
Cu
Ta
MgB2
a)
b)
Figure 6 MgB2 monofilamentary wire; a) Cross section of 1.18 mm diameter. The
composition of the sheath structure is given by: stainless steel/Cu/Ta or Nb, from the outside
to the filament, b) long length sample (1m length, 33 mm diam.) prepared by the wind & react
technique. (Courtesy of Dr W.Goldaker - Forschungszentrum Karlsruhe - Institut für Technische
Physik). The architecture is based on the design of Chevrel powder-in-tube wires. [20]
4.2 Tape form conductor
a)
b)
Figure 7 Transverse cross section of the MgB2 tapes: a) 180 µm thick, 4 mm wide Cu-sheathed
ex situ MgB2 tape. There are some non uniformities in the MgB2 core but in overall the selffield transport critical current density was in the range of 105 Acm -2 at 4.2K [19]; b) After
only one month the same authors improved the quality of the cross section of the MgB 2 core
and also substantially increased the critical current density which is 1.5 105 Acm -2 at 26K
[34].
371
The rapid development of the tape form conductor is clearly presented in the Figure 7
where in a period of one month the critical current value of the single core conductor was
increased by more than one order of magnitude and the uniformity of the cross section was
also dramatically improved.
4.3 Coated conductor
The important area of conductor development by non-vacuum techniques used
successfully for the manufacture of Bi-2212 and Bi-2223 coated tapes by doctor-blading
[35], screen printing [36], spraying [37] and ink-jet printing [38] can be adopted for the
preparation of the MgB 2 coated conductors by non-vacuum deposition technique. The
example of such work was presented where the fabrication, characterisation, and
superconductivity of in situ MgB2 thick films grown on stainless steel substrate were
studied [39]. For this film preparation, a suspension of mixed magnesium and amorphous
boron powder was made by stirring the powders together with acetone. The powders of
Mg and B were deposited from suspension, after evaporation of acetone, on a piece of
polished stainless steel which was placed at the bottom. This procedure was repeated
several times until the desired thickness was reached. The resulting samples were pressed
under high pressure after the deposition process in order to increase the density of the
deposited films. Another piece of stainless steel of the same size as the substrate was
placed onto the top of the deposited films so as to avoid the loss and oxidation of Mg
during the sintering at high temperatures.
It was found that MgB2 thick films can be fast formed by heating samples to 660 oC then
immediately cooling down to room temperature. XRD shows above 90% MgB2 phase and
less than 10 % MgO. However, the samples sintered at 800 oC for 4 h contain both MgB4
and MgO impurities in addition to MgB 2. The fast formed MgB2 films appear to have a
good grain connectivity that gives a Jc of 8 x 10 4 A/cm 2 at 5 K and 1 T and maintained
this value at 20 K in zero field. It was found that the MgB2 can strongly adhere to iron
sheath materials [39]. A thin layer of iron might be a suitable buffer layer between MgB2
and other metal sheath materials which could react with MgB2. Composites of this type
are under fabrication in the University of Wollongong, Australia.
5 CRITICAL TEMPERATURE
5.1 Doping, substitution and pressure
In this paragraph we are going to discuss influence of the doping and substitution of the
MgB2 and its influence on the critical temperature.
The transition temperature Tc=39 K in hexagonal MgB2 is close to the limit suggested
372
theoretically for phonon mediated superconductivity. The boron isotope effect [40]
indicates that MgB2 is a phonon-mediated superconductor. According to several authors
the high transition temperature is probably due to the low mass of boron, which yields high
phonon frequencies. A different theoretical model, the 'hole superconductivity', where
superconductivity originates from the undressing of hole carriers in almost full bands, was
proposed in [41]. It predicts hole carrier conductivity, opposite to other metal diborides
with the same crystal structure, which shows a positive Hall coefficient. In fact
measurements of the Hall coefficient on polycrystalline MgB2 bulk samples [42] and thin
films [43] showed a positive normal state Hall coefficient, which indicates hole carriers.
Hall coefficient changes its sign as the sample is cooled below T c. Some tunnelling
experiments performed on polycrystalline samples [44-46] show a good fit to BCS model,
some [47], [48] are in contradiction. While the shapes of the reported spectra support an
isotropic scenario for the pairing symmetry, the gap values are quite different, all being too
small to be realistic for the bulk MgB2. Moreover, it has been shown theoretically [49] that
MgB2 has a rather complex band structure, which in clean limit could give rise to multiple
gap superconductivity. Scanning tunnelling spectroscopy [50] showed that the
superconducting gap varied from region to region, thus indicating that the differences
between tunnelling data originate from the sample inhomogeneity. The model of hole
superconductivity [41] predicts a strong increase of T c upon shrinking the a lattice
constant, which should lead to a positive pressure effect on Tc. However, high pressure
studies [51-53] show negative pressure coefficient of Tc. Similar effect was observed also
by pre-compression of powders in pellets and wires [2], see Figure 8. Under pressure
induced by metallic sheath Tc was shifted toward a lower temperature.
0
-0.2
-0.3
-0.4
-0.5
Internal susceptibility
χ
int
-0.1
-0.6
-0.7
34
Pressed pellet
Ag/MgB2-PIT unsintered
SS/AG/MgB2-PIT unsintered
35
36
37
38
Temperature (K)
Figure 8 Internal ac susceptibility of the ex situ PIT unsintered wires. Under pressure induced
by the metallic sheath, the superconducting transition of the powder inside conductor (onset
temperature at ~ 37.5 K) is shifted toward a lower temperature 36.8K. [2]
373
To test various models, substitutions leading to hole and electron doping are crucial. Several
such experiments concerning substitutions on the Mg site have been published. Both hole
doping by a substitution of Li1+ [54] and electron doping by a substitution of Al3+ [55]
for Mg 2+ led to a decrease of Tc. Investigation of variations of the structure as a function
of doping and pressure was reported in [56]. Hydrostatic pressure has an anisotropic
influence on a and c lattice constants. According to [57] the compression along the c axis is
64% larger than along the a axis, which is in line with the weaker Mg-B bond. As a result of
the compression Tc decreases. Thermal compression along the c axis is about twice the one
along the a axis. Substitution of Al leads, apart from doping with electrons, to the
compression of the structure, due to the smaller ionic radius of Al (0.5 Å) than Mg (0.65Å)
[56]. The compression is anisotropic and for an Al content of x=0.1 leads to a structural
instability. On the other hand, shortening the B-B distance should increase coupling in the
plane. The substitution of the smaller Li ion (0.6 Å) for Mg (0.65 Å) leads to the decrease
of the a lattice constant without changing the c lattice constant [54]. So there are two
overlapping effects: doping with holes and an increase of coupling due to a shortening of
the B-B distance. As a net result Tc decreases. [58].
Recently published data on doping and substitution of MgB2 by Si, Li and Al showed
progressive reduction of the critical temperature in comparison to pure MgB2 sample:
[MgB2(Si5at%) T c onset = 38.4K, MgB2(Al5at%) T c onset = 38.2K , Mg0.9Li0.1B2
Tconset = 37.2K with simultaneous improvement of Jc at 5-20K and magnetic field.
However, while the reported lattice parameters of the Li substituted, Mg0.9Li0.1B2, and
doped samples MgB2(Al5at%) and MgB 2(Si5at%) have shown no easily detectable
differences neither in the peak position nor in extra peaks, in the case of Mg0.9Li0.1B2
sample a very slight variation in the intensity of some reflections was observed. The
calculated X-ray pattern, by means of the Lazy-Pulverix program [59], agreed with the
experimental one, confirming the effective substitution on Mg site by Li. In [60] an increase
of Tc by about 0.2 K in Zn substitution was observed. Substitution of magnetic ions
Mn 2+, Fe2+, Co2+ and Ni2+ for the Mg 2+ ion suppressed Tc [60].
5.2. The 50 K anomalies with copper substitution
In [56] a study of the influence of an isovalent substitution leading only to a modification
of the structure without changing the total charge carrier concentration was performed.
They made two kinds of such substitutions, namely Zn and Cu. For Zn substitution a and
c lattice constants increased by about 0.17% and 0.2%, respectively. Tc decreased by about
0.5 K for 0.05 Zn and about 0.2 K for 0.1 Zn substitution. Cu substitution
(Mg1-xCuxB2) lead to multiphase samples without any change of lattice constants,
indicating that Cu does not enter the structure of MgB2. The Cu substitution broadened
the superconducting transition considerably, while the onset of Tc remained unchanged.
Chemical substitution on boron plane itself, which carries the superconducting property,
374
was investigated in [61-63]. Because boron shares similarity in structure with B 4C, C
(graphite) and BN in its rhombohedral phase, it may be substituted with those materials.
Carbon substitution on boron site of MgB2 showed the following effects. During the
superconducting transition, two distinct onset temperatures were observed in MgB1.8C0.2
[61]: 41 K and 38 K. The downset Tc was 33 K, i.e. the transition width was broad (~8 K).
On the other hand in [62] and [63] for MgB2-xCx with x varying from 0 to 1 it was found
that the presence of carbon was detrimental to the superconducting properties of MgB2.
The superconducting transition temperature decreased quasi linearly as a function of the
carbon concentration [63].
The results of an extensive research was presented in recently published literature
concerning doping and substitution of the MgB2 material. As it was shown almost all of
the elements used reduced T c except Cu. According to Sun et al. [64] Cu substituted MgB2
sample, Mg1-xCuxB2 with x=0.2 had an onset Tc=49 K. However, the paper was
unpublished and withdrawn by the authors. The x-ray investigation of their sample shows
as a main phase Cu2Mg resulting in a linear resistivity dependence up to Tc=50K.
According to the authors the a and c lattice constants decreased, although the ionic radius
of Cu is larger than the one of Mg. On the other hand density functional calculations [65]
showed that replacement of Mg by Cu in MgB2 structure leads to an increase in the
stiffness and doping level at the same time, while the carrier concentration can be controlled
by partial replacement of B by C. It was estimated that with full replacement of Mg by Cu
(i.e. obtaining CuB2) and fractional substitution of B by C in CuB 2, Tc values of 50 K may
be attainable (CuB2 a=5.58, c=6.28, T c=48, MgB2 a=5.74, c=6.52, T c=38) [65].
However, these are not known equilibrium phases.
5.3. Superconducting MgB2 wires in copper tubes: 50 K anomalies
We have focused our recent effort on the copper claded in situ wires to find out if the Cuinteraction with the Mg and B during in situ formation of the MgB 2 in the powder-in tube
type conductor will increase Tc of the superconductor. Also having in mind the necessity
to electrically and cryogenically stabilise the very high critical current of the single core and
future multifilamentary conductors, therefore an attempt to the pre-optimisation of the low
cost high conductivity cladding conductors was researched in our laboratories.
Reinforcement of the softened Cu wires is not a problem since twisting of the conductors
to the form of reinforced conductors, see Figure 5, is a very well known technique
incorporated from the multifilamnetary V3Ga production by bronze technique [66].
Our experience with MgB 2 wires prepared by powder-in-tube (PIT) technique using
copper tubes shows that the presence of multiphases depends on sintering conditions.
Changing the sintering temperature one can obtain the samples of different quality.
However, careful study of ac susceptibility in dependence on temperature revealed an
375
anomalous decrease in the vicinity of 50 K, regardless of the quality of the samples. Here
we present results on three typical samples.
5.3.1. Experimental
All samples were prepared by a mixture of Mg and B powders in stoichiometric amounts
and made by PIT method in Cu tubes. Outside diameters of the wires were 1.5 mm and the
diameters of MgB2 were about 1 mm. Wire 1 was sintered at 620oC during 48 hours in
vacuum (ramp rate of the furnace 180oC/h), wire 2 was sintered at 700 oC during 1 hour in
argon atmosphere (ramp rate of the furnace 300 oC/h) and wire 3 was sintered at 800oC for
1 hour in argon atmosphere (ramp rate of the furnace 300oC/h). While wire 1 and wire 2
were of good quality with sharp transition, wire 3 was poor, with a broad transition. For ac
susceptibility measurements pieces of the samples about 3 mm in length were used.
Magnetic field was applied parallel to the axis of the wires.
5.3.2. Results and discussion
All the measurements were performed in temperature range 5K-100K. The susceptibilities
were calculated using the volume of MgB 2 and corrected for magnetometric demagnetizing
factors of ideally diamagnetic cylinders in parallel external magnetic field [67]. The so called
internal susceptibilities int obtained in this way are shown in Figure 9, Figure 10 and
Figure 11 for wire 1, wire 2 and wire 3, respectively. Their values in low temperature
region carry the information about the relative amount of a superconducting phase. The
susceptibilities were normalised so as to have int =0 at 100 K. As seen, wire 1 and wire 2
have a sharp transition with onset Tc=39K , as shown in insets. However, some
anomalous decrease of the susceptibility at T~50 K can be seen in all cases. Wire 3 has a
broad transition with onset T c~50 K and downset T c~25 K. There is some frequency
dependence of the susceptibility due to skin effect in copper. At 50 K the resistivity of Cu
is =4x10-10 Ωm [68] and the skin depth =( /µoπf)0.5 at frequencies f=333.3 Hz starts
to be comparable with wall thickness of the copper tube of our samples, which is about
0.25 mm.To avoid the screening effects we grounded the inner core of the wire 2 and wire 3
into a powder and measured ac susceptibilities again. The results (Figure 12) still have an
anomalous decrease of the susceptibility at T~50K. DC SQUID measurements performed
on wire 3 in form of a powder confirmed the anomalous magnetic moment decrease at
T~50K (Figure 13). It should be mentioned however, that a presence of different
paramagnetic phases was seen in SQUID measurements and the magnetic moments were
small, so a special care had to be taken when preparing the sample. The powder was glued
directly onto a straw using a non magnetic glue, so there was no capsule present nor any
other discontinuity giving rise to a signal detected by SQUID. The only discontinuity was
the powder itself.
376
-0.024
-0.032
-0.04
χ
int
-0.048
0
333.3 Hz
H
-0.056
||=1 Oe
ext
rms
-0.2
-0.064
-0.4
χ
int
-0.6
-0.072
-0.8
-0.08
-1
0
20
40
60
80
100
T(K)
-0.088
40
45
50
55
60
65
T(K)
Figure 9 The AC susceptibility of in situ Cu-Mg-B wire sintered at 620o C for 48 hours under
protective argon atmosphere, measured at frequency 333.3 Hz in applied magnetic field 1 Oe
(rms). The inset shows the full curve in the temperature range from 5K to 100K.
-0.04
-0.06
0
333.3 Hz
H
χ
int
-0.08
||=1 Oe
ext
rms
-0.2
-0.4
-0.1
int
χ
-0.6
-0.12
-0.8
-0.14
-1
0
20
40
60
80
100
T(K)
40
45
50
55
60
65
T(K)
Figure 10 The AC susceptibility of in situ Cu-Mg-B wire sintered at 700 o C for 1hour under
protective argon atmosphere, measured at frequency 333.3 Hz in applied magnetic field 1 Oe
(rms). The inset shows the full curve in the temperature range from 5K to 100K.
377
-0.025
-0.03
-0.035
χ
int
-0.04
-0.045
0
333.3 Hz
H
-0.02
-0.05
||=10 Oe
ext
rms
-0.04
int
χ
-0.055
-0.06
-0.06
-0.08
-0.1
0
-0.065
20
40
60
80
100
T(K)
40
45
50
55
60
65
T(K)
Figure 11 The AC susceptibility of in situ Cu-Mg-B wire sintered at 800 o C for 1hour under
protective argon atmosphere, measured at frequency 333.3 Hz in applied magnetic field 10 Oe
(rms). The inset shows the full curve in the temperature range from 5K to 100K.
-0.005
-0.005
-0.01
-0.01
-0.015
0
-0.02
χ (a.u.)
χ (a.u.)
-0.015
333.3 Hz
Hext=10 Oerms
-0.02
-0.05
-0.025
-0.03
-0.1
-0.025
(a.u.)
χ
-0.15
-0.03
-0.035
-0.02
-0.03
χ(a.u.)
-0.05
-0.06
0
-0.25
0
a)
666.6 Hz
-0.04
-0.2
-0.04
0
-0.01 Hext=10 Oerms
20
40
60
80
100
20
40
60
80
100
T(K)
T(K)
-0.035
40 45 50 55 60 65
40 45 50 55 60 65
T(K)
T(K)
b)
Figure 12 The AC susceptibility of the powder extracted from the core of the in situ Cu-Mg-B
wire after sintering for 1hour under protective argon atmosphere, measured at frequency 333.3
Hz in applied magnetic field 10 Oe (rms). The inset shows the full curve in the temperature
range from 5K to 100K; a) at 700o C , b) 800o C.
378
-6
6.5 10
-6
6.4 10
-6
m(emu)
6.3 10
-6
6.2 10
8 10
-6
H =60 Oe
ext
6 10
-6
6.2 10
4 10
-6
-6
2 10-6
-6
6.1 10
0
m(emu)
-2 10-6
-6
ZFC
6.0 10
-4 10-6
-6 10
-6
-6
0
20
40
60
80
100
T(K)
5.9 10
40
45
50
55
60
65
T(K)
Figure 13 DC SQUID measurement of the powdered core of the Cu/Mg-2B conductor sintered
at 800 o C for 1 hour.
Our measurements using microanalyser with wavelength dispersive spectrometers,
MAWDS, conducted on the sintered sample showed that as a result of Cu-Mg-B
interaction multiphase conductors have been obtained and there is no interdiffusion
between Cu and B, see Figure 14 and Figure 15. X-ray powder pattern conducted on the
powdered cores of the conductors shows a mixture of MgB2 and other phases, namely
MgCu2, MgO, MgB4 and possibly also higher borides. With increasing temperature Mg
content in the copper matrix increases considerably. The lattice constants of Cu-‘doped’
SEM
a)
Cu
b)
Mg
c)
B
d)
Figure 14 Scanning Electron Micrograph and MAWDS surface analysis of Cu-MgB2 interface
region in a in situ Cu-Mg-B conductor sintered at 700o C for 1h.
379
PIT MgB2 wires are the same as for non-substituted samples, which indicates that Cu does
not enter the structure of MgB2. It is important to conduct further XRD investigations
[59] on the larger quantities of the powdered core of the superconductor to understand the
possible substitution of the Cu in the MgB2 structure.
SEM
Cu
Mg
B
a)
b)
c)
d)
Figure 15 Scanning Electron Micrograph and MAWDS surface analysis of Cu-MgB2 interface
region in a in situ Cu-Mg-B conductor sintered at 800o C for 1h.
6. CRITICAL MAGNETIC FIELD
An electronic anisotropy arising from its layered crystal structure was revealed by Patnaik
at al. [11]. They reported data on three c-axis oriented thin films, showing that the upper
critical field anisotropy ratio Hc2 || /Hc2⊥ is 1.8 to 2.0, and the ratio increasing with higher
resistivity. Measurements of the magnetic field-temperature phase diagram show that flux
pinning disappears at H*≈0.8H c2⊥ (T) in untextured samples. Hc2||(0) is strongly
enhanced by oxygen alloying to 39 T for the highest resistivity film, more than twice that
seen in bulk samples.
Figure 16 Schematic representation of the field-temperature boundaries of superconductivity
for bulk MgB 2 and for film 1, assuming dirty-limit extrapolations of data near Tc to low
temperature are valid [11]; --- H* irreversibility lines, —— Hc2 upper critical field lines.
(Courtesy of D.C.Larbalestier, The Applied Superconductivity Center, University of
Wisconsin–Madison)
380
7. CRITICAL CURRENT
The critical current density regularly achieved in in situ and ex situ type MgB 2 conductors
is in range of >10 5 A/cm 2. From the economical point of view ex situ process without
post deformation sintering is economically very attractive since it is the only one
superconducting material which will not require sintering even when starting deformation
from the lightly compacted powder initial density of 1g/cm3. Obviously one may densify
effectively soft superconducting alloys but e.g. densification of the Nb3Sn is impossible
since grains have high microhardness. Further progress of the tape transport properties is
still expected by improving both the tape cold working process and its design, that would
be definitely facilitated by the fact that the materials eventually chosen to sheath the
superconductor may not be subjected to any heat treatment. Optimised MgB2 with a
higher degree of purity and with controlled grain size would also help to improve the cold
working process and the density of pinning centres associated to the interconnected grain
boundaries. The intergranular mechanical welding of the individual MgB2 crystallites in the
metallic tubes during mechanical deformation to some extent has some similarities to the
deformation of graphite which was used in manufacture of the multifilamentary composite
Nb3Sn conductors [26].
7.1 Magnetic cladding, Fe, Ni
The magnetic metal clad wires appeared to be the best high current conductors currently
manufactured. There are two categories: round conductors and tapes. It appeared once
again that alignment of the superconducting core by rolling is probably the best way to
interconnect the superconducting grains in this layered MgB2 superconductor.
7.1.1.Round wires
It was established that in situ formation of the MgB2 powder in iron matrix PIT conductor
can be performed rapidly without need of prolonged heat treatment during the fabrication
[23]. A total time in the furnace of several minutes is more than enough to form nearly pure
MgB2 with high performance characteristics. The results from Tc, Jc and Hirr show that
the samples which were sintered for 3 minutes above 800 oC are as good as those sintered
for longer times. In fact, the Jc - magnetic field performance for the most rapidly sintered
sample is slightly better than for all other samples. We have achieved also a very similar
results in the case of the copper claded in situ wires where the best performing wires were
achieved in shortest time. The Jc of 4.5x105 A/cm2 in zero field and above 105 A/cm2 in
2T at 15K has been achieved for the best Fe-clad MgB 2 wires. As a result of such a short
sintering there is no need for using high purity argon protection and it is possible to carry
381
out the heat treatment in a much less protective atmosphere or in air. These findings may
substantially simplify the fabrication process, making it possible to have a continuous
process for fabrication and reducing the costs for large-scale production of MgB2 wires.
A high transport Jc of the values 3x105 and 7x104Acm-2 at 4.2K and 20K respectively
were recently reported for copper claded Fe/MgB2 manufactured by powder-in-tube
technique and sintered at 900oC for 1h [21].
7.1.2 Tapes
The critical current of several pieces of the earlier Ni-cladded tape manufactured by Grasso
et al. [19] has been measured at 4.2 K, and the critical current values of short pieces of this
unsintered Ni-sheathed MgB2 tape are found within the range 240A and, by considering a
superconducting cross section of about 0.23 mm 2, we can therefore estimate its average
transport critical current density in 105 A/cm 2. The second series of tapes recently
manufactured at the University of Genova, were characterised by the highest values of the
critical transport current density reported in the literature where Jc is reaching value above
1.5x105 Acm-2 at 24.5K the lowest temperature for liquid Neon, see Figure 17.
c
250
5
1.5 10
200
5
1 10
150
100
50
4
LH2
5 10
LNe
c
Critical current, I (A)
5
2 10
300
Critical current density, J (Acm )
350
0
0
25
30
Temperature ( K)
35
40
-2
20
Figure 17 Critical current and the critical current density of the single core ex situ Ni-sheathed
tape after additional heat treatment. The values are already above 1.5 105 Acm -2 at 25K [34].
(Courtesy of G.Grasso from Genova University, to be published)
It is suspected , that the strong pressure applied during the cold working procedure of the
Ni-sheathed tapes could have induced some additional shear stress in the MgB2 crystal
382
structure, improving then its intrinsic current density and also interconnecting the layered
structure of the MgB2 substantially improving intergranular connectivity. Irreversibility
line measurements would definitely help to understand whether this mechanism is
effectively taking place.
4.2 K
Figure 18 Critical current density vs magnetic field of the ex situ monofilamentary MgB2 /Ni
and MgB2/Fe superconducting tapes as drawn and annealed. (Courtesy of R.Flukiger,
Departement de Physique de la Matiere Condensee, Universite de Geneve) [24].
Another example of high quality dense ex situ monofilamentary MgB2/Ni and MgB2/Fe
superconducting tapes with high transport critical current densities fabricated using a
straightforward powder-in-tube method was recently presented by H-L.Suo et al [24], see
Figure 18. The density of the MgB 2 core appears to be a decisive factor for reaching high
critical current densities Jc. They have measured transport Jc values up to 2.3 × 105 A/cm2
at 4.2 K in a field of 1.5 T (corresponding to an Ic of 300 A) in an annealed MgB 2/Ni tape
and up to 104 A/cm2 at 4.2 K in a field of 6.5 T in a MgB2/Fe tape. Currently these are
probably reaching the one of the highest Jc values at 4.2K reported so far in tapes with this
material. For higher currents these monofilamentary tapes quenched, but extrapolating their
field dependence yields self-field Jc values around 1 MA/cm 2. With both sheath materials,
annealing causes clear sintering densification, leading to Jc values which are about ∼10 times
higher than those measured in the as-deformed tapes as it is presented in Figure 18. Ni
shows a limited reaction with MgB2, leading to the formation of Mg2Ni reaction layers
between the filament and the matrix. On the other hand, Fe stays chemically inert and turns
out to be an excellent candidate material for multifilamentary tapes with self-field Jc values
at 4.2 K in excess of 1 MA/cm2 [24], [34].
7.1.3 Critical currents - numerical modelling
383
As we have mentioned before the advantage of the magnetic metal surrounding individual
filaments in the multifilamentary superconducting conductors is in magnetic decoupling of
the filaments leading to a reduction of the transport AC losses [30], [31] and also an
increase of the critical current [32]. Multilayer screening is more effective that using the one
layer of the identical total thickness [31].
Let us consider the case where 19 filamentary conductor will be manufactured by
restacking the Cu/Fe claded MgB 2 monocore conductor into the 19 filamentary conductor
as is presented in Figure 19, and let us calculate the critical currents of such a conductor
with and without the magnetic screening.
a)
b)
Figure 19 Schematic cross section of the 19 filamentary MgB2 conductor. a) uncovered
filaments, b) covered filaments; each filament is covered by 2 magnetic layers, each 5µm thick
separated by 5µm thick non-magnetic material, also all interfilamentary region is covered by 2
magnetic layers separated by non-magnetic material of the same thickness.
The Jc data obtained from magnetisation measurements for the wire (660oC/0.1h-quenched,
see Figure 23) have been fitted by a general Kim relation [69] with 5th order polynomial:
J co Bo
Jc =
2
3
4
5
Bo + B + a 2 B + a3 B + a 4 B + a5 B
The fit of the experimental data is shown in Figure 20a. The parameter Jco=1.0666 10 9
A/m2. The critical currents have been calculated using this non linear Jc(B) characteristic by
Finite Element Method, where B is the magnitude of the local overall magnetic field, i.e. the
self magnetic field generated by the transport current added vectorially to the applied
magnetic field. Comparison of the back calculated data with experimental ones is shown in
Figure 20b. The calculation has been made for a monocore wire 0.8718 mm in diameter
having the same cross-section of the superconducting area as the 19 filament wire. The
experimental data of the critical current were determined from experimental values of Jc
simply dividing them by the cross-section of the wire. A good agreement between
experimental and numerical data has been achieved. Numerical results gave the self-field
384
T=5K
H ||
100
ext
I (A)
c
c
2
J (A/m )
8
10
10
7
10
a)
T=5K
0.0 1.0 2.0 3.0 4.0 5.0
µ H (T)
o
b)
ext
0.0 1.0 2.0 3.0 4.0 5.0
µ H (T)
o
ext
Figure 20 a) The fit of the experimental data b) Comparison of the back calculated data with
experimental ones.
critical current 390.53 A. The value of the critical current determined by multiplying Jco
parameter by the cross section of the wire is Ico=636.58 A. Comparison of these two
values reflects the influence of self-field.
=1
Figure
21 Comparison of the calculated critical current versus magnetic field
monofilamentary and 19 filamentary wires with different magnetic coatings.
385
for
a)
b)
c)
Figure 22 Spatial distribution of the critical current density in a 19 filamentary wire cross
section in self field, for different values of relative magnetic permeability r of the concentric
multi-screens as presented in Figure 19b: a) r =1, Ic=442A , b) r =10, Ic= 462A, c) Nonlinear Fe rmax =9000, Ic=628A.
386
The magnetic field dependence of calculated critical currents of a monocore wire and 19
filament conductor with either bare filaments ( r magnetic =1) or with filaments covered by
magnetic layers (see Figure 19b) with r magnetic=10 are shown in Figure 21. Also shown
in the figure is the self-field critical current of the 19 filament wire (Figure 19b) when the
magnetic layers are made of a non linear iron with r max =9000 [70]. Using the iron
cladding a significant increase of self-field critical current was achieved (Ic=628 A as
compared with 442 A without the iron cladding or with 462 A when cladding was made of
a magnetic Material with r magnetic=10). A spatial distribution of the critical current
density at self-field in these three cases is shown in Figure 22. Iron cladding gives a very
homogeneous distribution of the critical current density with a much higher value due to the
screening of self-field of the filaments (Figure 22c). Comparing a monocore wire with the
19 filament conductor without magnetic cladding, (Figure 21), the 19 filament wire has
higher Ic in external magnetic fields lower than 0.2 T because of smaller influence of selffield due to separation of the filaments.
7.1.4. AC Losses aspects
AC losses in the superconductor at the critical current value can be estimated simply as a
half of the product of the critical current and the electric field 1µV/cm. For the model 19
filamentary conductor with iron clading, Fig 19b, sing the self-field value of Ic=628.76 A,
Fig 22c, we obtain the loss of 31.438 mW/m. The loss in the Fe coating is 2.5 mW/m if we
take a frequency of 50 Hz. Numerical calculations in self-field showed that to obtain the
same Ic in a monocore wire, its diameter should be 1.157 mm, corresponding to the crosssection of the superconducting area of 1.05 mm2, a value 1.76 times larger than the
superconducting area of the multifilamentary wire.
7.2 Nonmagnetic cladding, Cu, Ta, Nb
Using a SQUID magnetometer, magnetisation versus magnetic field (M-H) curves of the
round in situ Cu clad MgB2 wires after sintering were measured at temperature of 5 K and
magnetic fields up to 5 T to define the Jcmag, see Figure 23. The Jc mag was determined
using the Bean critical state model [71] for an infinitely long cylinder in parallel external
magnetic field corresponding to fully penetrated state. A comparison between zero-fieldcooled (ZFC) and field-cooled (FC) susceptibility measurements for different wires was
conducted to define differences in the flux pinning in the wires. Direct four point current
measurements were conducted at 4.2 K in self field. The exemplary transport data for the
Cu claded wire is incorporated in Figure 23. Use of copper also allows to minimise the
overall cost of the tape or round wire, because expensive non magnetic barrier materials
surrounding the MgB2 phase to avoid its diffusion, as typically done by adding Nb or Ta
387
layers, can be neglected, giving further room to improve the conductor stabilisation.
Goldaker et al. [20] prepared and characterized monofilamentary MgB2 wires with a
mechanically reinforced composite sheath of Ta(Nb)/Cu/steel, which leads to dense
filaments and correspondingly high transport currents up to Jc = 10 5 Acm-2 at 4.2 K, self
field. The reproducibility of the measured transport currents was excellent and not
depending on the wire diameter. Using different precursors, commercial reacted powder or
109
Transport Jctrans data Ic=750A self
T=5K
field 0.3T (sintering at 700oC/1h)
8
2
Jc (A/m )
10
10
7
10
6
0
550deg/48h
620deg/48h
660deg/0.1h-quenched
660deg/1h
700deg/0.1h
700deg/1h
1
2
µ oHext(T)
3
4
Figure 23 The transport and magnetic critical current density, J cmag, defined from
magnetisation, (M-H) curves of the round in situ Mg-2B wires after reactive sintering
formation of MgB2 ; samples were measured at 5K, using a SQUID magnetometer. The
measured transport current density J ctrans ~105 Acm -2 is represented by red solid square which
was calculated for the whole cross section of the B rich area superconducting core diameter,
however it is an underestimated value since only 70% of the whole cross section can be
estimated to be an effective superconductor, see Figure 14.
an unreacted Mg/B powder mixture, a strong influence on the pinning behaviour and the
irreversibility field was observed. The critical transport current density showed a nearly
linear temperature dependency for all wires being still 52 kAcm-2 at 20 K and 23 kAcm -2
at 30 K. Detailed data for Jc(B,T) and Tc(B) are also presented by Goldaker et al. [20].
However it has to be stressed that the values of the Jc achieved for the Nb or Ta layered
PIT conductor are very impressive and there is a noticeable difference between in situ and
388
ex situ preparation concerning the maximum pinning force for the ex situ beginning at higher
magnetic fields, Figure 24 [20].
7.3. Critical current and flux penetration in thin films
Magnetooptic experiments conducted on the bulk samples [72], [73] and also on the high
current thin films [74] revealed very important issues concerning the flux dynamic in the
MgB2 material including PIT and coated conductors. For the dc and ac applications of the
superconducting conductors an issue of primary importance is the static and dynamic
behaviour of the magnetic vortices. This issue was addressed to high-quality epitaxial c-axis
oriented MgB2 film using magneto-optic (MO) imaging. The direct evidence for a complex
a)
b)
Magnetic Field µo H (T)
Magnetic Field µo H (T)
Figure 24 Critical current density and the volume pinning force of in situ Mg-2B wires vs field
at 4.2 K compared to ex situ MgB2 wires, a) Critical current density vs field b) volume pinning
force of the Jc vs B graphs of -Fig.a. ● -MgB2 ex situ wires ■ - Mg-2B in situ wires. [20]
flux behaviour was recently presented by T.H.Johansen et al. [74] and D.V.Shantsev, et al
[75], where below T=10K the global penetration of vortices is dominated by dendritic
structures abruptly entering the thin film. The first four frames in Figure 25, show a
sequence of MO images of the high quality MgB2 film taken as the applied field increases
from 0 to 60 mT, after first zero-field-cooling the sample to 5 K. The images reveal flux
penetration patterns where an invasion of dendritic structures gradually covers the entire
film area [77]. The dendrites nucleate at seemingly random places near the sample edge, and
develop to their final size in less than 1 ms. During subsequent field reduction the flux
redistributes in the same abrupt manners, last two frames of the Figure 25, leading to a
389
remanent state with an overlapping mixture of two types of dendrites, one containing
trapped flux of initial polarity and one with anti-flux due to penetration of the reverse
return field created by the trapped vortices.
The comparative self-exemplary MO images underlining such situations are demonstrated
in Figure 26 where flux penetration at increasing magnetic field up to 17mT, and after
reduction of field to zero is presented. In Figure 26b there are two pronounced bands of the
flux : the external band represents the anti-flux associated with the internal band
representing the remanent trapped flux in the sample. A striking change in the morphology
of the flux patterns was observed when the experiment was carried out at different
temperatures [74], [75]. At the lower temperature, Figure 25, each dendrite has only one or
a few branches. With increasing temperature the branching tendency grows steadily, where
mT
8.5mT
mT
60mT
21mT
0mT
Figure 25 Magneto-optic images of the magnetic flux penetrating MgB2 thin film (Tc=39K,
DT~0.7K [76]) in an external magnetic field at 5K, ( brightness represent flux density).
Increasing magnetic field from the virgin state is marked by upward arrow for 3.4mT,
8.5mT,17 mT and 60mT; where decreasing magnetic field from 60mT to zero field value is
marked by downward arrow for 21mT and 0mT. (Courtesy of D.V.Shantsev, Oslo University).
390
a large tree-like structure was formed in one single burst of flux motion, Figure 26. It was
observed that above temperature ~ 10 K dendrite formation never occurred, instead only
smooth flux distributions, as expected from the critical-state model were observed.
This remarkable complexity in flux dynamics, where the detailed dendrite pattern never
repeats itself, may suggest that a thermo-magnetic instability is taking place. The heat
released by moving vortices causes local temperature rise facilitating further motion, which
in turn can trigger an avalanche-like flux invasion [78]. The critical current density Jc(T) ,
see Figure 27, was calculated from MO images. Below 8 K, Jc exceeds 107 A/cm 2, thus
establishing that state-of-the-art MgB2 films have now reached a benchmark for practical
applications. Whether the complex flux dynamics will limit their technological potential in
coated conductors and devices for dc and ac applications remains an open question. Some
future answers may be found in incoming publications [66], [79].
a)
b)
Figure 26 Magneto-optic images of the magnetic flux penetrating one of the corner of the
MgB2 thin film (presented in Figure 25) in an external magnetic field at 9.9K, (brightness
represents flux density) a) increasing magnetic field from the virgin state, 17 mT, b) after
decreasing field from 17mT to zero (remanent state).
Very important information on the magnetic field and current percolation was presented
recently by Wisconsin group [72]. Comparison of the light and MO images in Figure 28
shows a complex multi-scale microstructure. X-ray analysis of this sample showed only
~5% MgO as a minority phase. The MO images all show direct evidence of
superconducting inhomogeneity, as exemplified in Figure 28b, where very strong
superconducting regions of sizes up to about ~150 µm form a minority fraction on several
length scales throughout the whole sample. The flux gradient dH/dx across these regions can
be observed at all temperatures from 11 to 38 K. The maximum gradients correspond to
105 A/cm 2 at 11 K, 0.05 T. Investigation of the microstructure of the strongest
superconducting regions shows that they are mulitiphase, too. In fact these strong
391
Jc x 106 (Acm-2 )
12
10
MgB2 thin film
(magnetooptic measurements)
8
6
4
2
0
5
10
15 20 25 30
Temperature (K)
35
Figure 27 Critical current density of MgB2 thin film calculated from the magneto-optic
measurements as a function of temperature.
superconducting regions are subdivided on a scale of ~100 nm, perhaps due to partial
decomposition in situ. Electron microprobe analysis verified that the darker central area
corresponds to a mixture of MgB2 and boron-rich phases. Since none of the X-ray
examinations suggest any texture, we conclude that these regions consist of an untextured
two-phase nanomixture of MgB2, MgO, pores and perhaps some amorphous B-rich phase.
The important implication is that there is a very large number of high-angle grain
a)
b)
Figure 28: (a) Light microscopy and (b) magneto-optical image of the same area of sample in
figure 28a), showing extensive inhomogeneity of the sample. In (b), the dark regions exhibit
very strong superconductivity, the lighter areas weaker superconductivity [72]. (Courtesy of
D.C. Larbalestier et al., University of Madison).
392
boundaries and blocking insulating phases within each of the strongly shielding regions.
Nonetheless, these regions support high current densities of order 105 Acm-2 , which they
could not sustain if there was any inherent strong suppression of current across the grain
boundaries. Indeed, the true local current densities within the superconducting particles
must be significantly larger. As it was presented also by Botta et al. [73] that in the
sintered bulk MgB2 samples even at 35K and perpendicular magnetic field of µ oH=51mT
the induced currents does not show granularity proving that the grain boundary does not
present problem in the newly discovered superconductors.
8. CONCLUSIONS
One has to assume that although all possible efforts have been made to update the values of
the critical parameters and also to describe the manufacturing techniques used, in a such
rapidly developing subject, development of new high current conductors and applications
will take place driving the subject to the new areas of science and technology.
There is also a strong hope that the new reliable superconducting joints and
superconducting switches will be made not only based on the MgB2 new conductors but
also between MgB 2 and low temperature superconducting conductors such as NbTi,
Nb3Sn and Nb3Al opening a new range of the electromagnetic device applications.
Because an adequate cryogenic stabilisation is required to avoid quenching and destruction
of the conductor research towards ex situ and in situ formation of the MgB2 PIT
conductors and coated conductors using Cu-Fe-Ni-alloys based composites as the optimum
cladding material is expected in a very near future.
It is still an open and maybe even speculative question how the substituted atoms or doped
atoms such as Cu enter the structure of MgB 2 and whether it may cause critical
temperature increases. Currently in a X-ray analysis of pure MgB2 powder and MgB2
powder currently extracted from the in situ Cu-Mg-B PIT conductor there are no easily
detectable differences either in the peak positions or in the appearance of extra peaks. Even
so thorough investigations of the XRD patterns [59] conducted on the larger volume of the
extracted powder may reveal a very slight variation in the intensity of some reflections
which, may possibly confirm some substitution on Mg site by Cu and may help us to
understand the 50K anomalies observed in the in situ Cu-Mg-B PIT conductors.
The currently manufactured superconductors as composite conductors already are reaching
critical current values of 1MA/cm2 at 4.2 K and more than 105 Acm-2 at 25K, but it
appears to us that those transport current values at the given temperatures and magnetic
fields will be improved in the very near future. This will make this material a very
competitive conductor for the future ac and dc applications not only for the power
393
applications, but also will make its way to low temperature electronics and devices.
Let’s take this challenge.
9. ACKNOWLEDGEMENTS
The authors would like to thanks Prof J.E.Evetts for his valuable comments, M. Vickers
for XRD measurements, P. Hunneyball for EDAX measurements, E.A.Robinson for
development of the Ic testing unit and also Prof A.M.Campbell for providing testing
facilities at the IRC in Superconductivity and Prof.D.Fray for providing processing
facilities in the Department of Materials Science and Metallurgy University of Cambridge.
M.Majoros acknowledges the AFRL/PRPS Wright-Patterson Air Force Base, Ohio for the
financial support.
10. REFERENCES
[1] T. Serebryakova, Journal of the Less-Common Metals, 67 (1979) 499.
[2] B.A. Glowacki, M. Majoros, M. Vickers, J.E. Evetts, Y. Shi and I. McDougall, Supercond. Sci. Technol., 14
(2001) 193.
[3] J. Nagamatsu, N. Nakagawa, T. Muranaka, Y. Zenitani, J. Akimitsu, Nature 410 (2001) 63.
[4] M.Cooper, ‘Current champion’ Frontiers Emerging Technologies, New Scientist, 2 June 2001 No.2293
(2001) 21.
[5] D.K. Finnemore, J.E. Ostenson, S.L. Bud'ko , Lapertot G and Canfield P C Phys. Rev. Lett. 86
(2001) 2420.
[6] M. Xu, H. Kitazawa , Y. Takano , J.Ye , K. Nishida , H. Abe , A. Matsushita and G. Kido, cond.mat./0105271.
[7] G. Rubio-Bollinger, H. Suderow, S. Vieira, cond- mat/ 0102242.
[8] T. Vogt, G. Schneider, J.A. Hriljac, G. Yang and J. Abell, cond-mat/ 0102480.
[9] O. F.de Lima, R.A. Riberio, M.A. Avila, C.A. Cardoso, and A.A. Coelho, cond-mat/ 0103287.
[10] K-H. Müller, G. Fuchs, A. Handstein, K. Nenkov, V.N. Narozhnyi, D. Eckert, cond- mat/ 0102517.
[11] S. Patnaik, L.D. Cooley, A. Gurevich, A.A. Polyanskii , J. Jiang , X.Y. Cai , A.A. Squitieri , M.T.
Naus, M.K. Lee, J.H. Choi, L. Belenky, S.D. Bu, J. Letteri, X. Song , D.G. Schlom , S. E.
Babcock, C. B. Eom, E.E. Hellstrom, and D. C. Larbalestier, Superconductor Science and
Technology, 14 (2001) 315.
[12] C.U. Jung, M-S. Park, W.N. Kang , M-S. Kim, S.Y. Lee and S-I Lee, cond- mat/0102215.
[13] S.L. Budko, C. Petrovic, G. Lapertot, C.E. Cunningham, P.C. Canfeld, M-H. Jung and A.H. Lacerda,
mat/0102413.
[14] M.D. Sumption, X. Peng, E. Lee, M. Tomsic and E.W. Colling, cond- mat/ 010244.
[15] P.C. Canfield, D.K. Finnemore, S.L. Budko, J.E. Ostenson, G. Lapertot, C.E. Cunningham and C.
Petrovic, Phys. Rev. Lett., 86 (2001) 2423.
[16] W.N. Kang, H-J. Kim, E-M. Choi, C.U. Jung and S-I. Lee, cond- mat/ 0103179.
[17] J.R. Cave and C.A.F. Weir, IEEE Trans on Magnetics, 19 (1983) 1120.
394
c o n d -
[18] B.A. Glowacki and Z. Bukowski, J.Mater.Sci., 22 (1987) 775.
[19] G. Grasso, A. Malagoli, C. Ferdeghini, S. Roncallo, V. Braccini, M.R. Cimberle and A.S. Siri, condmat/0103563, also subitted to Appl.Phys Lett.
[20] W. Goldacker, S.I. Schlachter, S. Zimmer, H. Reiner, cond-mat/0106226 , also sumitted to SST
[21] S. Jin, H. Navoori, C. Bower and R.B. van Dover, Nature, 411 (2001) 563.
[22] K. Togano, Y. Tokano, H. Takeya, H. Fujii, H. Kumakura and T. Hatano, 10th International Workshop
on Critical Currents, 4-7 June 2001, Gottingen, Germany.
[23] X.L. Wang, S. Soltanian, J. Horvat, M.J. Qin, H.K. Liu and S.X. Dou, con-mat 0106148.
[24] H-L.Suo, C. Beneduce, M. Dhallé, N. Musolino, X-D. Su, E. Walker, P. Toulemonde and R.
Flükiger, submitted to Applied Physics Letter.
[25] Y. Takano, H. Takeya, H. Fujii, H. Kumakura, T. Hatano, K. Togano, H. Kito and H. Ihara, condmat/ 0102167.
[26] B.A. Glowacki and Z.M. Kosek, Cryogenics, 27 (1987) 551.
[27] B.A. Glowacki and J.E. Evetts, Mat.Res.Soc.Symp.Proc., 99 (1988) 419.
[28] B.A. Glowacki and J.E. Evetts, Proceedings First European Workshop, High Tc Superconductors and
Potential Applications, Genova, Italy, 1-3 July 1987, p.447.
[29] B.A. Glowacki et al. to be published.
[30] B.A. Glowacki and M. Majoros, Supercon.Sci.Technol., 13 (2000) 971.
[31] M. Majoros, B.A. Glowacki and A.M. Campbell, Physica C, 334 (2000) 129.
[32] M. Majoros, B.A. Glowacki and A.M. Campbell, IEEE Trans. Applied Superconductivity, 11 (2001)
2780.
[33] L. Larkovskii, G.V. Kapturovskaya and Y.D. Kondrashev, Zh. Obshch. Khim., 25 (1955) 433; (
J.Gen.Chem. USSR, 25 (1955) 409).
[34] G. Grasso to be published
[35] M. Chen, D.M. Glowacka, B.Soylu, D.Watson, J.K.S.Christiansen, Y.Yan, B.A. Glowacki
and J.E. Evetts, IEEE Trans. Appl. Superconductivity, 5 (1995) 1467.
[36] D-J. Kang, B.A. Glowacki, P.R. Winter and E.R. Wallach, IEEE Trans. on Mag., 32 (1996) 2970.
[37] A.P. Baker, B.A. Glowacki and R. Riddle, Advances in Cryogenics Engineering, 40A (1994) 201.
[38] B.A. Glowacki, Supercon.Sci.Technol., 13 (2000) 584.
[39] A.H. Li, X.L. Wang, M. Ionescu, S. Soltonian, J. Horvat, T. Silver, H.K. Liu, and S.X. Dou, condmat/0104501, submitted to Physica C.
[40] S. L. Budko et al, Phys. Rev. Lett., 86 (2001) 1877.
[41] J. E. Hirsch, cond-mat/0102479.
[42] W. N. Kang et al, cond-mat/0102313.
[43] R. Jin, M. Paranthaman, H. Y. Zhai, et al, cond-mat/0104411.
[44] A. Sharoni, I. Felner, O. Millo, cond-mat/0102325.
[45] G. Rubio-Bollinger, H. Suderov and S. Vieira, cond-mat/0102242.
[46] G. Karapetov et al, cond-mat/0102312.
[47] A. Plecenik, S. Benacka, P. Kus, cond-mat/0104038.
[48] P. Szabó, P. Samuely, J. Kacmarcík, Th. Klein, J. Marcus, D. Fruchart, S. Miraglia, C. Marcenat, A.
G. M. Jansen, cond-mat/0105598.
[49] A. Y. Lin et al, cond-mat/0103570.
395
[50] F. Giubileo et al, cond-mat/0105146.
[51] B. Lorenz, R. L. Meng, C. W. Chu, cond-mat/0102264.
[52] V. G. Tissen, M. V. Nefedova, N. N. Kolesnikov, M. P. Kulakov, cond-mat/0105475.
[53] S. Deemyad, J. S. Schilling, J. D. Jorgensen, D. G. Hinks, cond-mat/0106057.
[54] Y. G. Zhao et al, cond-mat/0103077.
[55] J. S. Slusky et al, Nature, 410 (2001) 243.
[56] S. M. Kazakov, M. Angst, J. Karpinski, cond-mat/0103350.
[57] J. D. Jorgensen, D. G. Hinks, S. Short, cond-mat/0103069.
[58] M.R. Cimberle. M. Novak, P. Manfrinetti and A. Palenzona, submitted to Superconductor Science and
Technology.
[59] K. Yvonj, W. Jeitschko and E. Parthe, J. Appl. Crystallogr., 10 (1977) 73.
[60] Y. Moritomo, Sh. Xu, cond-mat/0104568.
[61] J.S. Ahn, E.J. Choi, cond-mat/0103169.
[62] M. Paranthaman, J. R. Thompson, D. K. Christen, cond-mat/0104086.
[63] T. Takenoku, T. Ito, Dam H. Chi, K. Prassides, Y. Iwasa, cond-mat/0103241.
[64] Y.P. Sun, W.H. Song, J.M. Dai, B. Zhao, J.J. Du, H.H. Wen, Z.X. Zhao, cond-mat/0103101
(unpublished - withdrawn).
[65] M. J. Mehl, D. A. Papaconstantopoulos, D. J. Singh, cond-mat/0104548.
[66] http://www.cus.cam.ac.uk/~bag10/
[67] D.X. Chen, J. A. Brug, R. B. Goldfarb, IEEE Trans. Magn., 27 (1991) 3601.
[68] G.K. White, Experimental techniques in low-temperature physics, Oxford, Claredon Press (1959).
[69] Y.B. Kim, C.F. Hempstead, A.R. Strnad, Phys. Rev. Lett., 9 (1962) 306.
[70] R.M. Bozorth, Ferromagnetism, New York, IEEE Press 1993.
[71] C.P. Bean, Phys. Rev. Lett., 8 (1962) 250.
[72] D.C. Larbalestier, L.D. Cooley, M. Rikel, A.A. Polyanskii, J. Jiang, S. Patnaik, X.Y. Cai, D.M.
Feldmann, A. Gurevich, A.A. Squitieri, M.T. Naus, C.B. Eom, E.E. Hellstrom, R.J. Cava, K.A. Regan, N. Rogado,
A. Hayward, T. He, J. S. Slusky, P. Khalifah, I. Inumaru, and M. Haas, Nature, 410 (2001) 186.
[73] D..Botta, A..Chiodoni, R. Gerbaldo, G. Ghigo, L. Lozzelio, F. Laviano, E. Mezzetti, S. Ceresara, G.
Giunchi, M. Poyer and G. Ripamonti, 10th International Workshop on Critical Currents, 4-7th
June 2001, Goettingen, Germany.
[74] T.H. Johansen, M. Baziljevich, D.V. Shantsev, P.E. Goa, Y.M. Galperin, W.N. Kang, H.J. Kim, E.M.
Choi, M.-S.
Kim, S.I. Lee, cond-mat/0104113 , and also submitted to Nature.
[75] D.V. Shantsev, T.H. Johansen, M. Baziljevich, P.A. Goa and Y.M. Galperin, 10th International
Workshop on Critical Currents, 4-7th June 2001, Goettingen, Germany.
[76] W.N.Kang et al, Science, 292 (2001) 1521.
[77] A similar dendritic behaviour was observed in Nb-films, see C.A. Duran, et al., Phys. Rev.B, 52 (1995) 75; V.
Vlasko-Vlasov Physica C, 341-348 (2000),1281, as well as in YBa 2 Cu 3 O7 films, see P.Leiderer et al Phys
Rev.Lett., 71 (1993) 2646 and U. Bolz et al., Physica B, 284-288 (2000) 757.
[78] R.G. Mints, and A.L. Rakhmanov, Rev. Mod. Phys., 53 (1981) 551.
[79] http://www.fys.uio.no/faststoff/ltl/english/index.html
396