New challenges and opportunities for high- - PowerPoint Presentation

Slide1

New challenges and opportunities for high-Tc superconducting materials

Alex GurevichDepartment of PhysicsOld Dominion University, Norfolk VA

78-th Annual Meeting of SESAP, Roanoke, VA, Oct. 19-22, 2011Slide2

Superconductivity

Superconductors – frictionless conductors of electricityLow temperature superconductors such as Nb compounds (LTS)High temperature cuprate superconductors (HTS) – much more complex materials

New family of Fe-based superconductors

What makes them useful?

To save energy and reduce the dependence on oilElectric utility applicationsPower cables, fault current limiters, transformers, motors, generatorsSuperconducting magnetsAvoid 20 MW per user magnetParticle accelerators (Large Hadron collider, Free electron lasers, etc)To produce energy in fusion power reactorsITER – International Tokomak Experimental ReactorHow to make superconducting materials useful?Main parameters of merit for applications: high critical current density Jc(T,T) and the irreversibility field H*(T) require strong pinning of vortices and low anisotropy: Reduce strong thermal fluctuation of vorticesReduce current-blocking effect of grain boundaries in polycrystal conductorsSlide3

How did it start?

Heike

Kamerlingh-Onnes

Gilles Holst

Liquifying

helium by

Onnes

in 1908

led to the discovery of superconductivity in 1911

a

nd superfluidity of He in 1935 Ideas at the beginning of the 20-th century: electron liquid may crystallize at low temperaturesMetals would become insulators at low T

R

T

T

c

Hg

Electron liquid

In metalsSlide4

First proposal of superconducting magnets….

Onnes in Chicago 1913

“only” a 10 T

magnetSlide5

Phase diagram of a superconductor

Temperature

Magnetic

Field

NormalMetalPhase transition from the normal to superconducting state below the critical temperature Tc Superconductivity is destroyed by magnetic fields exceeding the upper critical magnetic field Hc2 Zero resistance disappears if the current density exceeds the critical current density Jc

Search for higher-

T

c

superconductors

Tuning materials properties to

increase Hc2 > 10 Tesla and

Jc

> 0.1 MA/cm2 at 5T Slide6

Lots of superconducting materials have been discovered

pnictid

es

Highest T

c

= 164K (under 30 GPa)Slide7

Main players in applications

NbTi

MgB

2

Nb3Sn Qubic, or hexagonal low-Tc superconductors Highly anisotropic layered high-Tc superconductors Tc seems to scale with crystalline/chemical complexities YBa3Cu3

O

7

T

c

=

92KBi2Sr2Ca2Cu

3OxTc

= 108K

T

c

= 9.2K

T

c = 18KTc = 40KSlide8

Vortices in type-II superconductors

Continuous vortex filaments

Meissner

state for H below the lower critical field H

c1

Vortex state at H

c1

< H < H

c2

For magnets we need superconductors with high upper critical field H

c2

:

Many materials have H

c2

> 50-100 T, more than ten times

the Onnes

1913 dream

Pancake vortices

in layered SCSlide9

Main parameter in applications: the critical current density Jc(B)

J

J

c

E

Current produces the Lorentz force which moves vortices

Defects pin vortices

No dissipation below the critical current density:

J

c

(B)

 10

5

-10

6

A/cm

2

fieldSlide10

Designer nanoparticle structures

S. Kang et al, Science 311, 19111 (2006)

J.L. McManus-Driscoll, Nature Materials 3, 439 (2004) (BZO); S.A. Harrington et al, SUST 22, 022001 (2009)

T.

Haugan et at, Nature 430, 867 (2004) (Y2BaCuOx nanoparticles in PLD YBCO films)Y. Yamada et al, APL 87, 132502 (2005); K. Matsumoto et al, JJAP, 44, L246 (2005).J. Gutierrez et al, Nature Materials, 6 367 (2007); X. Obradors et al, SUST 19, S1 (2006)S. Solovyev et al, SUST, 20, L20 (2007). M.W. Rupich et al, MRS Bull., 29, 572 (2004)Self assembles BZO nanpparticles

Combination of

nanoparticles

and columnar pins

B.

Maiorov

et al, Nature Materials 8, 398 (2009) weaker flux creep at high fields

weaker field dependence (reduced  in Jc  H-

)Slide11

Enhancement of Jc by “designer” nanoparticle structures

T. Haugan, et al. Nature 430, 867 (2004)

8 nm YBa

2

CuO5 nanoparticlesSelf-assembled chains of BZO nanoparticlesAFOSR 10

P. Mele, K. Matsumoto, T. Horide, A. Ichinose,

M. Mukaida, Y. Yoshida,S. Horii, R. Kita

SUST 21, 032002 (2008)Slide12

Superconducting cables 50-60 years later: Avoid Joule losses at the expense of cryogenic refrigeration

Superconducting

cable

Cryostat to keep

T <

T

c

Cooling by liquid helium

at

4.2K

Cooling by liquid

nitrogen at

77K (much cheaper)

The

higher the

temperature

, the

more efficient the superconducting systems are: Search for high-Tc materials

Nb

3Sn filaments in Cu

Bi-2212 in silverSlide13

Power magnet applications.

Research magnets

Medical MRI

HTS motors & generators

Power transmission lines

MagLev

FusionSlide14

U.S. HTS Cable Installations

Albany, NY

Carrollton, GA

Long Island, NY

Columbus, OH

New Orleans, LA

New Project

New York, NY (DHS)

Early tests have been done with silver-sheathed BSCCO wires, now being replaced by better and cheaper YBCO wiresSlide15

Power RF applications

ILC: 20000 cavites, 500 tons of high purity Nb; 20 kW refrigeration at 2K

Spallation neutron source (ORNL)

X-ray free electron laser

Superconducting LINAC

Tunable 0.25-14

m

light source at JLabSlide16

1500 tonnes of SC cables

27 km Tunnel

3286 HTS Leads

Large Hadron Collider

15000 MJ of magnetic energy

1232 SC Dipoles

Large

Hadron

Collider-CERN – 2009 turn on

Switzerland

France

Mont Blanc

Lake GenevaSlide17

Conventional LTS approachIncrease Hc2 by alloying the material with nonmagnetic impurities

The highest impurity concentration which does not produce significant Tc suppressionDirty limit: Hc2(0)



0 /20  nProduce appropriate defect structures to pin vorticesThe more pinning defects the betterMake multifilamentary conductors to suppress thermo-magnetic instabilities and reduce ac losses in alternating electromagnetic fieldsNot easy to implement in high-Tc cuprates and Fe-based superconductorsSlide18

Figures of merit for magnet applications

bad metal

LTS

HTS

Vortex pinning and critical current density

J

c

(T,H)

Irreversibility field H*(T) below which

J

c

(T,H) = 0

Thermal fluctuations of vortices

H/H*

pinned

vortex

solid

vortex

liquid

It is neither

T

c

nor Hc2, but the high

Jc and H*(T), which make

superconductors

useful Slide19

Strong suppression of H* in anisotropic HTS

Strong anisotropy

can eliminate all benefits

of higher

Tc and Hc2 YBCO (Tc

= 92K) is

much better than

Bi-2223 (

T

c

= 110K) MgB2 (Tc = 40K) or oxypnictides (Tc < 52K) can be as good as

Bi-2223 for 20K < T < 35K, and B < 15T

H

c2

Nd(F,O)FeAsSlide20

Thermal fluctuations in superconductors

Ginzburg

parameter:

Critical fluctuation region:

T =

T

c

– T < T

c

Gi

LTS:

Gi

 10

-8, T  10

-7 K

YBCO, higher-Tc Fe-pnictides: Gi  10-2, T  1K BSCCO: Gi  0.1, T  10K

Tc reduction by phase fluctuations (Emery & Kivelson, 1995) Low anisotropy and high superfluid density reduce thermal fluctuations

Anisotropy parameter

in

a uniaxial superconductor:



T

c

 T

c

5



2

/n

3Slide21

Thermal fluctuations of vortices

Elastic energy of a distorted vortex line Brandt, Rep. Prog

. Phys. 58, 1465 (1995);

Blatter et al, RMP 66, 1125 (1994) Dispersive line tension of a single vortex

c

2

 

3

nm

λ



150 -200

nm



rigid

rods

Anisotropy strongly reduces

bending rigidity of the vortex:



ℓ

 3 K/Å (YBCO @ 0K)ℓ  0.5 K/Å (YBCO @ 77K)ℓ  103 K/Å for LTS



soft filaments

Mostly determined by

superfluid

density and mass anisotropy!Slide22

Melting of vortex lattice

Weak pinning:

J

c = 0 in the vortex liquid phase B > Bm Lindemann criterion: <u2(T,Bm)> = cL20/Bm, cL  0,1-0.3 (Nelson et al;

Blatter

et al, Brandt et al; …)

Upper branch of the melting field B

c1

<<

Bm << Bc2:

For YBCO, Bm(77K)  9T, B

c2(77K)

 20T

Similar relation between Bm

and Bc2 in Nd-1111,

but weaker reduction of the melting field in lower-

Tc FBSMain material parameter: Calorimetric measurements by Schilling et al PRL, 78, 4833 (1997);Nature, 382, 791 (1996)

YBCOG. Blatter et al, RMP 66, 1125 (1994)

Slide23

Magnetic granularity causedby grain boundaries

Only small currents can pass through GBs

despite strong pinning of vortices caged in

the grains

Fragmentation of uniform current flow into decoupled current loops in the grains

Magnetic granularity in HTS

polycrystals

16

O

[001] tilt

GB

in YBCO





J

AG and L. Cooley, PRB 50, 13563 (1994)

d = b/2sin(

/2

)

dSlide24

The grain boundary problem in cuprates and FBS

Dimos

,

Chaudhari and Mannhart, PRB 41, 4038 (1990)Hilgenkamp and Mannhart, APL 73, 265 (1998); RMP 74, 485 (2002)

d

Similar nearly exponential drop of

J

c

with the

misorientation

angle both in the cuprates and Fe-based superconductorsFirst measurement of Jc

on a Ba

2(Fe1-xCo

x)2

As2 bicrystal

S. Lee, et al. APL 95, 212505 (2009).Slide25

Similarities of cuprates and Fe-based superconductors

short coherence length,   1-2 nm charging and strains effects of dislocation cores

competing orders:

nonsuperconducting

AF phase precipitates on GB low carrier density  long Thomas Fermi screening length lTF  1-2 nm AG and Pashitskii, PRB 57, 13875 (1998);

1

nm

Slide26

15 years of R&D to overcome the current-limiting GBs: “YBCO single crystal by the mile”

Eliminate high-angle GBs by growing YBCO films of textured substrates

State

of the art: complex, expensive, only a small fraction carries current, high ac losses

Jc of YBCO layer must be pushed to its limit Industry produces km long second generation YBCO coated conductors Slide27

Iron-based superconductorsFe-based superconductors: unconventional multiband superconductivity originated from magnetic

Fe ionsSuperconductivity competing with antiferromagnetic states in low-carrier density semi-metalsHigh Tc and huge upper critical magnetic fields. Interplay of orbital and paramagnetic pairbreaking in multiband SCs and their effect on H

c2

(T)

Effective tuning of Hc2 by doping-induced small shifts of the Fermi energy, Instead of the conventional way of introducing disorder.Strong Pauli pairbreaking in FBS can lead to exotic effects at high magnetic fields, such as FFLO state.Good prospects for magnet applications if grain boundary problem is resolved Slide28

Diverse family of Fe-based superconductors (FBS)Slide29

Phenomenology of pnictides

Tetragonal

Orthorhombic

Paramagnetic

AF

La-1111

H.

Luetkens

et al, Nature Mat. 8, 305 (2009)

C. Lester et al, PRB 79, 144523 (2009)Slide30

Huge Hc2 in pnictides

High slopes H

c2

/

= 2-100 T/K at TcHc2(0) for 1111 and 122 FBS, extrapolate to > 100TShort GL coherence lengthsAG, Nature Mat. 10, 255 (2011)

result from high

T

c

and low

carrier density in semi-metallic FBS

Dirty limit can hardly be reachedSlide31

Does increasing Hc2 by disorder work in FBS?

Effect of the elastic mean free path ℓ on the orbitally-limited

Clean limit:

ℓ >>

0 ⟹ ξ = 0= vF/ and Dirty limit : ℓ <<



0

⟹ ξ = (ℓ0 )1/2

(

Werhamer-Helfand-Hohenberg

, 1966)

Works in conventional superconductors: 10 –fold increase of H

c2

in MgB

2Does not work in FBS because ℓ <0

 1-2 nm implies the Joffe-Regel limit and ℓkF < 1 for which the conventional dirty limit BCS theories failHc2 in semi-metallic FBS can be effectively tuned by dopingSlide32

Orbital or Pauli-limited Hc2?

Werthamer-Helfand-Hohenberg

,

1963-1965

FFLOOrbitally limitedMostly Pauli limitedSarma, Maki, 1963-1964Gruenberg and Gunther, 1966Slide33

Pauli pairbreaking

k

-k

magnetic

energy

condensation

energy

Chandrasekhar –

Klogston

limit

Using BCS

yields a useful relation

First order phase transitionSlide34

Relation between orbital and Pauli pairbreaking

Maki parameter M = 21/2H

c2

orb

/Hp : In ordinary metallic BCS superconductors with mab  m0 and  << EF , paramagnetic pairbreaking is negligible ,



M

<< 1

Pauli-limited superconductors with

M > 1 Heavy fermions with

mab/m0

 103

Highly anisotropic materials with m

c/m0

 106

: layered organic SC, high-Tc cuprates (BSCCO), etc for H||ab Semi-metalic, strongly correlated FBS with

EF < 0.01-0.1 eV, and mab/m0  10 Slide35

Cooper pairing with nonzero momentum Q = 2q: modulation of the order parameter along H

(z) = 0

cos

(

Qz) (Larkin-Ovchinnikov) (z) = 0 exp(iQz) (Fulde-Ferrel)Ekk

k

F

- q

-k

F

- q

FFLO

Orbital and Pauli coupling: FFLO state

Q

FS nesting facilitates

the FFLO stateSlide36

FFLO in heavy fermions and organics

B. Lortz et al, PRL 99, 187002 (2007)

Bianchi et al, PRL 91, 187004 (2003)

CeCoIn

5Layered organic SCHeavy fermionsSlide37

Equation for Hc2 and Q (single band)

 = 2

FFLO transition for

 > 1 Spontaneous FFLO vector Q(T) appears at low T The FFLO period (T) = 2/Q(T) diverges at the spinodal: T = TFFLO At zero T: (0)  

0

.

First order transition line between two

spinodals

.Slide38

Electron spectrum from ab-initio calculations and ARPES

multiple bands

crossing

the Fermi

level two hole pockets at  and two electron pockets at MFeSe0.42Te0.58LaFeP(O,F)Slide39

New features of FBS revealed by ARPES

Small Fermi energies: EF  0.02-0.5

eV

Large effective mass renormalization:

m*  (2-16)me Several shadow bands near the FS: Lifshitz transition upon doping Strongly correlated semimetals Good candidates for the FFLO state:  > 1 Example of a Pauli-limited SC: FeSe0.5Te0.5 :

T

c

= 16K, E

F

= 25

meV, mab = 10me  = 1.5 even for H||c

In-plane coherence lengths   1-2 nmSlide40

Multiband pairing gap symmetries

s



pairing:

gaps with opposite signs Mazin, Singh, Johannes, Du, PRL 101, 057003 (2008); Kuroki et al, PRL 101, 087004 (2008) Extended s-wave or d-wave gapsKuroki et al, PRL 101, 087004 (2008); Graser, Maier, Hirshfeld, Scalapino, NJP 11, 025016 (2009)

e

h

Q

e

h

Q

+

-

Strong

interband

repulsion:



12

21 > 1122

Phonons are not sufficient to explain high Tc

Pairing coupling constants

Impurity scattering ratesSlide41

Multiband superconductivity on repulsion

BCS gap equations for two bands:

where E = (



2 + 2)1/2 s pairing for repulsive interaction 12

< 0

and

opposite signs

of



1 and 2 Pairing glue due to AF spin fluctuations,

w = 11

22

-

12

21

< 0Slide42

Upward curvature of Hc2(T) in two-band models

Bilayer model of two-band SC

Interaction of two bands with

conventional H

c2(T) can produce unconventional Hc2(T) with upward curvature Model independent mechanism Slide43

Hc2 for two coupled bands (clean limit) H||c

Band coupling parameters:

a

1

= (0 + -)/2w, a2 = (0 - -)/2w, - = 11 - 22, 0 = (

-

2

+ 4



12

21)1/2, w = 1122 - 1221

Band asymmetry parameters:

AG, PRB 82, 184504 (2010)

Rep.

Prog

. Phys. 74, 124501 (2011)Slide44

Band competition: hidden FFLO

Due to the significant differences in the band parameters,

one band can be FFLO unstable (1 > c) but another one is not (2 < c). Passive band reduces manifestations of the FFLO in the WHH-like shape of Hc2(T), but FFLO is still there “Hidden” FFLO: no apparent signs in H

c2

(T) but

can be revealed as the first order PT by

magnetic torque and specific heat or NMR Slide45

Experiment-I: LiFeAs

FFLO

Undoped

composition corresponds to the maximum Tc No suppression of FFLO by doping - induced disorder Good candidate to search for FFLO, mean free path >> K. Cho, H. Kim, M. A. Tanatar, Y. J. Song, Y. S. Kwon,W. A. Coniglio, C. C. Agosta, AG, R. Prozorov, PRB, 83, 060502(R) (2011)

Small jump in magnetic torque develops

below 8K

N. Kurita et al. J. Phys. Soc.

Jpn

. 80, 013706 (2011) Slide46

Suppression of orbital pairbreaking in srained FeSe

0.5Te0.5 filmsC. Tarantini et al. cond

-mat.

arXiv

: 1108.5194 FFLOFFLOSlide47

Experiment-III:tuning Hc2 by doping in Ba

1-xKxAs2Fe2

C.

Tarantini

et al. cond-mat. arXiv: 1108.5194 Tuning the shapes of Hc2(T) due to expansion and contraction of FS pocketsHighest so far Hc2 in the optimally doped Ba-122Change from upward to downward curvature of Hc2

(T) upon doping

x

= 0.4 (1); x = 0.25 (2); x = 0.15 (3)Slide48

FFLO triggered by the Lifshitz transition

H

c2

equation in effective 2-band form:  = (1221 + 2332 )1/2 reduction of the FFLO instability thresholdSlide49

Summary

The higher Tc, the less relevant for high-temperature applications it becomes

The key parameters to be optimized irrespective to pairing mechanisms:

Carrier density (the higher the better)

Electron mass anisotropy (the smaller the better)Thomas Fermi screening length (the smaller the better)The higher Tc the less parameter space we have to satisfy the constrains on the carrier density and mass anisotropy The symmetry of the order parameter and competing orders: non-s-wave pairing and competing orders greatly complicate applications Reduce current-blocking effect of grain boundariesDesigner pinning nanostructures would be required to minimize vortex fluctuations and produce high critical currentsFe-based superconductors: huge Hc2 and good prospects for applications