PRZEGLĄD ELEKTROTECHNICZNY (Electrical Review), ISSN 0033-2097, R. 85 NR 5/2009 151

Bartłomiej A. GŁOWACKI1, Milan MAJOROS2, Archie M. CAMPBELL1 University of Cambridge (1), The Ohio State University (2)

Influence of magnetic materials on transport properties

of applied superconductors

Abstract. Magnetic materials can help to improve the performance of practical superconductors on the macro/micro scale as magnetic diverters and also on the nanoscale as effective pinning centres. It was established by numerical modelling that magnetic shielding of the superconducting filaments reduces AC losses in self-field conditions due to decoupling of the filaments and, at the same time, it increases the critical current of the superconducting composite. Streszczenie. Magnetyczne materiały umożlowiają polepszene wlasności naprzewodzących nadprzewodników technicznych w skali makro jako magnetyczne osłony, jak również w skali nano jako efektywne centra zaczepiania strumienia magnetycznego. Udowodniono przy użyciu modelowania numerycznego, że magnetyczne ekranowanie włókien nadprzewodnikowych redukuje straty przemiennoprądowe w wyniku magnetycznego rozłączenia włókien, jak rownież powoduje wzrost prądu krytycznego przewodu. (Wpływ materiałów magnetycznych na właściwości transportowe technicznych nadprzewodników). Keywords: superconducting conductors, magnetism, ac losses, flux pinning. Słowa kluczowe: przewody nadprzewodnikowe, magnetyzm, straty przemiennopradowe, kotwiczenie strumienia magnetycznego. Introduction In this article we will discuss the aspects of the influence of the ferromagnetic materials on the properties of the practical YBa2Cu3O7, (Pb,Bi)2Sr2Ca2Cu3O9 and MgB2 multi-filamentary conductors.

stabilisation Cu, Ag

superconductor (RE)Ba 2Cu3O7epitaxial buffer layer:NiFeO4, NiO,YSZ,CeO2, NdCuO4

metallic textured substrates: Ni, NiCr, NiV, NiW, NiFe

Fig.1. Schematic representation of RABiT architectures of the coated conductors: The key conductor components are the metallic or ceramic substrate material, the buffer layer and the superconducting layer; the latter has to be biaxially textured throughout so that, although granular, the misorientation from grain to grain should be less than a few degrees [14]

a) b) Fig.2. NiFe flexible substrates for coated conductor: a) view of the NiFe tape, 760 m long, 10 mm wide and 25 µm thick after continuous annealing; b) stress - strain curves of the cold rolled and dynamically annealed in protective atmosphere NiFe tape, 25 µm thick tape [15, 16]. It is well documented that ferromagnetic materials acting as flux diverters and magnetic flux screens can effectively reduce expose of the superconducting material to external magnetic field and therefore improve in-field performance of practical superconductors [1]. The external magnetic field may originate from the neighbouring superconducting elements or external electromagnets such as it is in the case of current leads or neighbouring strands and phases in superconducting cables [2-4], There is also calculated strong influence of the presence of the ferromagnetic substrates, coatings and buffer layers on the magnetic flux distribution from the neighbouring filaments as it is the case of practical multifilamentary YBa2Cu3O7,

(Pb,Bi)2Sr2Ca2Cu3O9 and MgB2 conductors [5-13]. Perspective of use of low cost of the highly textured Ni-based ferromagnetic substrates presented in Fig.1 and Fig.2, and magnetic buffer layers presented in Fig.3 and Fig.4, stimulates research concerning the architecture of the potential superconducting cables where ferromagnetic substrate based coated conductors can be used [2, 4], Epitaxial surface oxidation (native oxide) for YBa2Cu3O7 coated conductors Oxidation of the Ni based substrates is usually to be avoided as the oxide, which grows, is often of an unsuitable orientation or lattice match for deposition of the buffer layers. However by controlling the conditions in which the substrate is oxidised it may be possible to produce an adherent epitaxial oxide with the cube texture. Any further buffer layers can then be deposited using an oxidising environment and also substrate oxidation will not be a problem in the superconductor deposition stage. Indeed it may even be the case that if the oxide acts as a diffusion barrier it may remove the requirement for further buffer layers, in which case the surface oxidation epitaxy technique may serve as a fast and inexpensive buffer layer route. There are successful attempts to use self oxidation process of the Ni tapes [17-20], as presented in Fig.3.

(a) (b) Fig.3. NiO layer on the Ni substrates: (a) pole figure of the NiO, (b) Optical image of the surface of a transparent SOE NiO/Ni grown at 1250°C in air [21, 22]. The process of oxidation of the NiFe tapes is more complicated due to the fact that mixed oxides may be formed, however the two stage of spalling and re-oxidation process produces oxide which is (100) oriented [21]. The θ-2θ scans shown in Fig.4 clearly show that the oxide which was removed has a (111) texture whilst the remaining tape is covered with an oxide which is (100) oriented. The fact that a thin (100) layer remains on the surface may be

152 PRZEGLĄD ELEKTROTECHNICZNY (Electrical Review), ISSN 0033-2097, R. 85 NR 5/2009

utilised to consistently produce cube textured oxides on the surface of the NiFe tape. If the tapes are further oxidised they maintain the cube texture, and the rocking curve in Fig.5(a) demonstrates the excellent alignment. It is likely that the oxide is neither entirely NiFe2O4 nor Fe3O4, but has some intermediate composition, best described as (NixFe1-

x)2+Fe3+2O4. This oxide has the structure known as spinel,

each unit cell consisting of 32 oxygen atoms, 16 Fe3+ in octahedral interstices and 8 Ni2+ or Fe2+ atoms in tetrahedral interstices as shown in Fig.5b).

Fig.4. The (111) oxide c) may be completely removed from the oxidised tape a) to leave behind a NiFe tape which has a cube textured oxide layer b) [21, 22]. This type of magnetic oxide buffer layer on NiFe ferromagnetic substrate will require further development concerning scaling up the native oxide buffer layers to reduce the cost of the buffered coated conductor and prevent the extensive oxidation of the NiFe substrates and contamination of YBa2Cu3O7 coating. The magnetic properties of the native oxide layer may be used also in a novel deposition approach where individual coatings could be covered by magnetic material [23]. The overall electron back-scattered patterns, EBSP, map of grain orientation as well as data about misorientation from grain to grain provides information about potential percolative current paths in the subsequent superconducting layer.

a) b) Fig.5. a) {400} rocking curve showing excellent alignment after re-oxidation; FWHM=6.5o [56, 57]; b) The spinel structure, AB2O4 after [58].

a) b)10-4

10-3

10-2

10-1

100

101

102

103

0.1 1

40Hz-c40Hz-tw40Hz-tw cor90Hz-tw cor

Qt

I/Ic

Fig.6. Transport ac losses of a helix: (a) potential taps position and potential wires arrangement on the outer surface of the central turn. White arrows - axial component of Jc flowing on outer surface, dotted white line - tangential Jc component on inner surface; (b) Measured transport ac losses. twcor represents the tw voltage signal corrected by the c voltage to provide truly transversal voltage losses [3].

NiFe/YBa2Cu3O7 coated conductors for cables However calculations conducted for model cable configuration with ferromagnetic substrates are very valuable, see Fig.7a)-c), but because they are just 2D models they do not take to account the real helical configuration of the tapes in the cable [2, 4], the influence the transversal distribution of the current in the influence the transversal distribution of the current in the conductor is not taken to account that increases the participation of the ferromagnetic substrate material in the overall losses of the cable, Fig.6. the progress in understanding of the superconducting/ ferromagnetic interfaces.

c)10-11

10-10

10-9

10-8

10-7

0.1 1

Norris stripNorris elipseFeNi outsideFeNi inside

Nor

mal

ised

loss

(J/m

/cyc

le/A

2 )

Ip/I

c

d)

e) Fig.7. Possible architectures of the cable constructed from the YBa2Cu3O7 superconductor deposited on the ferromagnetic substrate: a) superconductor on the inner surface of the cable; b) superconductor on the outer surface of the cable, (in the picture only segment of the cable with two parallel tapes has been schematically presented); c) AC transport losses for two orientations of the ferromagnetic substrates of the coated conductor in respect to the cable surface as presented in a) and b) after [2]; d) proposed new architecture of the two layer cable where the bottom layer is low cost; e) proposed architecture of the two layer cable where the both low cost layers of the superconducting coating are facing each other, see [24]. Superconductor/magnetic multifilamentary composite Jc and stability This effect is specially beneficial for coated conductors where anisotropic properties of the superconductor are amplified by the conductor architecture, Fig.8. However, ferromagnetic coatings are often chemically incompatible with YBa2Cu3O7 and (Pb,Bi)2Sr2Ca2Cu3O9 conductors, and buffer layers have to be used. From the theoretical analysis it follows that while for a superconductor with circular cross-section the presence of a magnetic coating does not influence the magnetic field distribution within the superconductor nor the transport ac losses, for a superconductor with an elliptical cross-section, the magnetic coating affects both, the magnetic field distribution and the transport ac losses. The same is expected to be valid also for Ic. When estimating the overall transport ac loss decrease by magnetic decoupling of the filaments the equation (1)

PRZEGLĄD ELEKTROTECHNICZNY (Electrical Review), ISSN 0033-2097, R. 85 NR 5/2009 153

(1) K=QmonoQmulti

=aIc

2F( i)

aNIc12 F( i)

=Ic2

NIc12=

(NIc1)2

NIc12

=N

must be modified by taking into account the ac loss increase in the superconductor itself by the presence of the magnetic coating, its Ic degradation and the hysteretic losses in the magnetic coating as well. Because the theoretical results are unknown, we introduced an additional parameter k which contains the effect of ac loss increase in both materials, superconductor and magnetic coating. The critical current of a coated single filament is Ic1coated = αIc1 where the parameter α characterizes the Ic degradation by the coating. Then ac losses of a single coated filament is described by eq.(2)

(2) Q1coated =kaIc1coated2 F( i)=kaα 2Ic1

2 F( i)

a)

buffer layer(RE) BaCuO

stabiliser

metal tapesubstrate

b)

mag. buffer layermetal tapesubstrate

magnetic stabiliser

Fig.8. Schematic representation of the design of straight multifilamentary structures of the coated conductor, where: a) highly conductive stabilizer on the top of the filaments and in between filaments with the conductive buffer layer providing additional stability from the metallic substrates (filaments are fully coupled in the conductive matrix; b) magnetic buffer and magnetic cap layer to provide magnetic decoupling and even electric decoupling of the filaments.

a) b)

Fig.9. Magnetic field lines (scale 6 x 10-10 Wb) of a model YBCO coated conductor with 7 strips embedded in a ferromagnetic cover of rectangular cross-section, when the current with current density j = 109 A/m2 passes through the central filament only: (a) μr =1 of all the materials (substrate, buffer layer and cover); (b) ferromagnetic cover mr =1000, substrate μr=1, ferromagnetic buffer layer μr =1000. (dimensions of the filaments: 8 μm x 2μm, thickness of the ferromagnetic layer: 0.5mm, spacing between the covered filaments = 1μm, thickness of the buffer layer= 0.5μm, thickness of the substrate equal 9.5μm).

Then ac losses in the multifilamentary superconductor with magnetically coated filaments presented schematically in Fig.9 can be described by eq.(3)

(3) Qmulticoated=NQ1coated=Nkaα2Ic1

2 F( i)

In comparison with a monocore superconductor at the same reduced current i one obtains

(4) K ''=Qmono

Qmulticoated=

aIc2F( i)

Nkaα 2Ic12 F(i)

=N 2Ic1

2

Nkα 2Ic12 = N

kα 2

where the parameters k and α are to be determined from experiment. The example results of the ac losses calculation for 11 filamentary coated conductor covered by ferromagnetic material are presented in Fig.10.

In contrast, in MgB2 conductors, an iron matrix may remain in direct contact with the superconducting core. Application of superconducting-magnetic heterostructures requires consideration of the thermal and electromagnetic stability of the superconducting materials used. On one hand, magnetic materials reduce the critical current gradient across the individual filaments, Fig.11 but, on the other hand, they reduce the thermal conductivity between superconducting core and the cryogen, which may cause destruction of the conductor in the event of thermal instability, Fig.12 [1].

1

2

3

4

5

6

S 1S 2

S 3

0 .0 0 E + 0 0

2 .0 0 E - 0 4

4 .0 0 E - 0 4

6 .0 0 E - 0 4

8 .0 0 E - 0 4

1 .0 0 E - 0 3

1 .2 0 E - 0 3

1 .4 0 E - 0 3

1 .6 0 E - 0 3

1 .8 0 E - 0 3

2 .0 0 E - 0 3

AC

Los

s (J

/m)

Colu

mn

num

ber

S/FeS

No cover

1 2 3 4 5 Fig.10. AC loss distribution in individual filaments when they are covered with Fe layers and without the covering. (S+Fe - total losses, S - part of the losses only in the superconductor). Only the filaments from the cross-section marked in black are shown.

a) b)

Fig.11. Spatial distribution of the critical current density in a 19-filament MgB2 wire cross-section in the self-field, for different values of relative magnetic permeability, μr, of the concentric multi-screens: (a) μr = 1, Ic = 442 A; (b) nonlinear Fe μrmax = 9000, Ic = 628 A [1]. The electromagnetic stability is a issues one may see that the critical current density gradient across the individual filament can be as high as 20% of its nominal value. The gradient is different for the internal filaments and also the lowest Jc values are achieved for the external filaments. Ferromagnetic pinning centres Improvement of the transport properties of the conductors by addition of the nanoscale/sub-micron magnetic pinning centres has been recently considered more widely driven by Prozorov [26]. A possible nanoscale method of improving the critical current density of superconducting conductors is introduction of nano-scale/sub-micron magnetic pinning centres. However, the volumetric density and chemical compatibility of magnetic nano-inclusions has to be controlled to avoid suppression of the superconducting properties over larger distance that the coherence length [3]. An important aspect of coexistance of the superconducting and ferromagnetic interface especially in the case of MgB2 is more widely considered in respect of magnetic properties of the ferromagnetic particles at low temperatures 4.2K-77K. There is an open question if the magnetic pinning centres should be made from Fe or from iron oxides. It was discussed by Grovenor [25] that the MgB2/Fe interface shows some detectable interaction causing formation of the Fe-B compounds.

154 PRZEGLĄD ELEKTROTECHNICZNY (Electrical Review), ISSN 0033-2097, R. 85 NR 5/2009

10-1

100

101

102

103

104

105

106

100 101 102 103 104

MgB2MgB2/Fe; t=0.25dMgB2/Fe; t=0.5dMgB2/Fe; t=d

Ele

ctric

fiel

d E

(μV

cm-1

)

Total current , Itotal

(A)

Film boiling onset

Tmax

=Tc (39K)

d t

Fig.12. Calculated voltage–current characteristics (points) of the MgB2 superconducting composite conductors immersed in liquid He: a) Fe/MgB2 composite wires. Bare MgB2 wire has diameter d = 0.6 mm where the Fe/MgB2 composite wires (insert) have different Fe thickness t. The straight vertical line is the isothermal power-law voltage–current characteristic of the bare MgB2 wire at 4.2 K. ( - bare MgB2, - MgB2 covered with Fe of thickness t = 0.25d, - MgB2 covered with Fe of thickness t = 0.5d and - MgB2 covered with Fe of thickness t = d).

a) -30

-20

-10

0

10

20

30

-0.6 -0.4 -0.2 0.0 0.2 0.4 0.6

MgB2MgB2+MoO6MgB2+Fe2O3

M (e

mu

cm-3

)

B (T) b) -0.2

-0.15

-0.1

-0.05

0

0.05

0.1

0.15

0.2

-0.6 -0.4 -0.2 0 0.2 0.4 0.6

Mag

netic

mom

ent

M (e

mu)

B (T) Fig.13. Magnetisation of MgB2 nanocomposites: a) comparison of the unreacted MgB2 with high-intensity ultrasound treated MgB2+Mo2O6 and MgB2+Fe2O3 after [26]; b) comparison of unreacted MgB2 (solid line) with unreacted mixture of MgB2+Fe in the (Cu/MgB2+10%Fe) conductor dotted line, magnetic field applied along the wire axis. It was proven that magnetically induced critical current density for Fe2O3 nanoparticles doped MgB2 was higher than for nonmagnetic nanoparticles consistently in the whole range of temperatures [27], Fig.13a). Data presented in Fig.13b) was achieved for iron powder doped copper clad PIT MgB2 wires and shown no improvement of the Jc , which may be due to interfacial interaction between MgB2 and Fe [25]. Further research is going to clarify if the smaller level of Fe doping may improve the pinning.

REFERENCES [1] G lowack i B .A . , Ma jo ros M. , V i cke r V . , E is te re r M. ,

Toen ies S . , Weber H .W. , Fuku tomi M. , Komor K . , Togano K . , Supercond. Sci. Technol., 16 (2003) 297

[2] M iyag i D . , Umabuch i M. , Takahash i N . , Tsukamoto O. , Physica C, 463–465 (2007) 785

[3] Ma jo ros M. , G lowack i B .A . , Campbe l l A .M. , Han Z . , Vase P ., Physica C, 314 (1999) 1

[4] Sa to S . , Amemiya N . , IEEE Trans on Appl. Supercond., 16 (2006)127

[5] G lowack i B .A . ,Ma jo ros M. , Supercon. Sci. Technol., 13 (2000) 971

[6] Ma jo ros M. , G lowack i B .A . , Campbe l l A .M. , Physica C, 334 (2000) 129

[7] Ma jo ros M. , G lowack i B .A . , Campbe l l A .M. , Physica C, 338 (2000) 251

[8] Ma jo ros M. , G lowack i B .A . , AC losses and Flux Pinning and Formation of Stripe Phase, Nova Science Publishers, Inc., Huntington, New York, 32 (2000) 1-51

[9] Ma jo ros M. , G lowack i B .A . , Campbe l l A .M. , IEEE Trans. Applied Supercond., 11 (2001) 2780

[10] G lowack i B .A . , Ma joros M. , N .A .Ru t te r and A .M.Campbe l l , Physica C, 357-360 (2001) 1213

[11] G lowack i B .A . , Ma jo ros M. , Rut te r N.A . , Campbe l l A .M. , Cryogenics 41, (2001) 103

[12] G lowack i B .A . , Ma jo ros M. , Advances in Research and Applications MgB2, ed. A. Narlikar, Nova Science Publishers, Inc., Huntington, New York, 38 (2001) 361

[13] G lowack i B .A . , Ma jo ros M. , Physica C, 372-376 (2002) 1235

[14] G lowack i B .A . , in: Studies of HTS superconductor applications (Advances in Research and Applications): ed. A.Narlikar, Springer Verlag, Volume1-Materials, 2003 p.239

[15] G lowack i B .A . , V i cke rs M. , Ru t te r N . , Maher E . , Pasot t i F . , Ba ld in i A . , Ma jo r R . , Journal of Materials Science, 37 (2002) 157

[16] Denu l J . , Van Dr i ssche , te L in te lo H . , De Gryse R , Tse tsekou A , Hos te S , G lowack i B A , R .Gar re R , Maher E , Good J , Ranucc i E . , La r rau r i E., Inst. Phys. Conf. Ser. 167 (2000) 335

[17] Kursumov ic A . , Huhne R . , Tomov R . , G lowack i B .A . , Eve t ts J .E . , Ho lzap fe l , Acta Materiala 51 (2003), 3759

[18] Kursumowic A . , Tomov R . , G lowack i B .A . , Eve t ts J .E . , Tu iss i A . , V i l l a E . Ho lzap fe l B . , Physica C, 385 (2003), 337

[19] Lockman Z . , Q i X . , Go ldaker W. , Nas t R . , deBoer B . , Kursumov ic A . , Tomov R. , Hühne R . , Evet ts J .E . , G lowack i B .A . , MacManus-Dr isco l l J .L . , Ceramic Transactions 140 (2003), 185

[20] Huhne R . , Kursumov ic A . , Tomov R . I . , G lowack i B .A . , Ho lzap fe l B . Eve t ts J .E . , J. Phys. D: Appl. Phys. 36 (2003), 1053

[21] Ru t te r N .A . , G lowack i B .A . , Dur re l l J .H . , Eve t ts J .E . , t e L in te lo H . , De Gryse R . , Denu l l J., Inst. Phys. Conf. Ser., 167 (2000), 407

[22] Ru t te r N A , PhD Thesis, supervisor prof B.A. Glowacki Dep. Materials Science and Metallurgy, Cambridge 2001

[23] Ma jo ros M. , G lowack i B .A . , Campbe l l A .M. , Physica C, 334 (2000), 129

[24] http://www.msm.cam.ac.uk/ascg/lectures/applications/cables.php [25] Grovenor C R M, Goods i r L , Sa l te r C J , Kovac P

and Husek I , Supercond. Sci. Technol., 17 (2004), 479 [26] P rozorov R . , P rozorov T . , Snezhko A . , IEEE Trans.

Appl. Supercon., 15 (2005), 2, 3277 [27] Snezhko A . , P rozorov T . , P rozorov R . , Phys. Rev.

B, 71 (2005), 024527 Authors: prof. dr Bartłomiej A. Głowacki, Department of Materials Science and Metallurgy, University of Cambridge Pembroke Street, Cambridge CB2 3QZ, England, E-mail: bag10@cam.ac.uk; dr Milan Majoros, College of Engineering, The Ohio State University, 555 MacQuigg Laboratory 105 W Woodruff Ave. Columbus, OH 43210, USA, E-mail majoros@ecr6.ohio-state.edu; prof. dr Archie Campbell, Department of Engineering, University of Cambridge, Trumpington Street, Cambridge CB2 1PZ, England, amc1@cam.ac.uk.