boulder school for condensed matter and materials physics 2008...
TRANSCRIPT
Talk 1. Quasi-particle charge and heat currentsin d-wave superconductor
Boulder School for Condensed Matter and Materials Physics 2008
N. P. Ong, Princeton University
Talk 3. Magnetization of vortex liquid state
Talk 2. The Nernst effect in vortex liquid state of cuprates
Boulder, July 7-11, 2008Supported by NSF-MRSEC, ONR
http://www.princeton.edu/~npo/
Talk 1
Quasi-particle charge and heat currentsin d-wave superconductor
1. Introduction : Charge and heat currentsHall effect, Nernst effect, Thermal Hall effect
2. The Hall effect in cuprates
3. Quasiparticles and Thermal Hall conductivity
N. P. Ong, Princeton University
Collaborators: Lu Li, Joe Checkelsky, Yayu Wang, Kapeel Krishana, Wei Li Lee, Yuexing Zhang, Jeff M. Harris, Y.F. Yan, P.W. Anderson
Boulder, July 7-11, 2008Supported by NSF-MRSEC, ONR
Boulder School for Condensed Matter and Materials Physics 2008
The Hall effect
J JE .ρt
=
yxyxxxx EEJ σσ +=
yyyxyxy EEJ σσ +=
Thermopower
T∇−
V )( TEJ x−∂+= ασ
σα
=⎟⎠⎞
⎜⎝⎛
∇=
=0JTES Seebeck
coef.
heater
bath
kkk gff =− 0
Boltzmann-equation expressions for currents
∑=k
ge kk vJ 2 kkk v)(J gk
h ∑ −= με2
Charge current
Thermoelectric current )(J T−∇= α
kk
k .E lr
efgε∂
∂−=
0
).()(k
kkk T
Tfg −∇
−∂∂
−= lrμε
ε
0
EJ σ=
charge heat
conductivity
Thermoelectric cond.
kkkk ϑε
σ 20
2 cosv2 l∑ ⎟⎟⎠
⎞⎜⎜⎝
⎛∂∂
−=k
fe
kkkkk )( ϑμε
εα 2
0
cosv2 l∑ −⎟⎟⎠
⎞⎜⎜⎝
⎛∂∂
−=k T
fe
Presence of temp. gradient
kk v
xh
ε∂∂
=∂∂ 0ffk
τk
kv.x
k.k
gff kk −=∂∂
+∂∂ &
Quasi-particle excitations in normal state of Fermi liquid
particlehole
vacancy
εk
kεF
holeparticle
hk+ = c-k
A spin-down vacancy at –k translates to a spin-up hole excitation at k
Charge and heat currents in the “excitation” representation
o
Large left hole current
Large right electroncurrent
Large vacancy population
E T∇−
Large vacancy population
Both mass (entropy)currents flow to right
Charge currents nearly cancel.Difference is the Peltier charge current
Mass currents nearly cancel.Difference is the Peltier heat current
E~J α=h )(J T−∇= α
EJ σ= )(J Teh −∇= κ
k k
E(k) E(k)
Charge currents add Mass currents add
The Nernst effect (quasiparticles)
).(E.J T−∇+= ασtt
).(.E T−∇−= αρtt
( ) )( TE xyxyxy −∂+−= αρρα
Open boundaries, so set J = 0.
)(k
.Bvk
kkxyxy T
fe ll∂∂
×−
⎟⎟⎠
⎞⎜⎜⎝
⎛∂∂
−= ∑ μεε
α0
22
εθπ
∂∂
=∇
≡eTk
TE
e ByN
22
3||
Off-diag. Peltier cond.
Measured Nernst signal
Generally, very small because of cancellation between αxy and σxy
Wang et al. PRB ‘01
The 2D Hall conductivity σxy
τkk k.
kgf
−=∂∂ &
kkk gff =− 0
lr
lr
×= ∫ dBexy 213σ
Boltzmann Eq.
Eq. of motion
Hall current in 2nd order
Area swept out in ell-space!
Gauss mapping to …
NPO, PRB (1991)
B)vE(.kk
k ×+∂∂
−= efg τ
τε yy eEefeJ v
0
).(k
.Bvk
k lr
∂∂
×⎟⎟⎠
⎞⎜⎜⎝
⎛∂∂
−= ∑
( ) yxxyfBe llk
k
k .t̂ ∇⎟⎟⎠
⎞⎜⎜⎝
⎛∂∂
−= ∑ εσ
03
The 2D Hall conductivity σxy
∫ ×= lr
lrr
d21A
0
22φ
σ A.B)/( hexy =
swept area
σxy is the area swept out by mfp (for arb. anisotropy)
Harris, Yan, NPO, PRB ‘92
Temp. dependences of Hall coef. and Hall angle in YBCO
Hwang Batlogg et al., PRL ‘94 Ono, Komiya, Ando, PRB ‘07
Similar T dependence of RH seen in LaSrCuO
Kanigel, Campuzano et al. Nature Phys. 2007
Antinode node
ky
kx
Δ(φ)
φ
Ferm
i arc
leng
th
T/T*
NPO and Wang unpubl.
Model: Only states on Fermi archave long mean free path
NPO and Wang unpubl.
y = 6.95
6.63
6.7
6.8
Fits to T dependence of RH in YBCO
Fits to cotθ and resistivity ρ
Vortex in cuprates
CuO2 layers
2D vortex pancake
ξ
Gap Δ(r) vanishes in core
|Ψ| = Δ
Vortex in Niobium
Js
superfluidelectrons
ξ
= 500 A
Js
b(r)Normal core
H
Vortex motion in type II superconductor
Applied supercurrent Js exerts magnus force on vortex core
Velocity gives induced E-field in core (Faraday effect)Current enters core and dissipates (damping viscosity)
Motion of vortices generates observedE-field
Consequence of Josephson equation
Tilt angle of velocity gives negative vortex Hall effect
In clean limit, vortex v is || - Js
E = B x v
(Bardeen Stephen, Nozieres Vinen)
ηρρ /02
Φ== BHH
cNxx
0Φ×=r
sM JF
Vortex Hall current
Vortex Hall σxy is negative.Appearance is abrupt
Sxy
Nxyxy σσσ +=
Quasiparticle and vortex Hallconductivities are additive
22yxxx
yxxy ρρ
ρσ
+=
Invert matrix
~H ~ -1/H
Harris et al. PRB ‘95
Thermal Hall conductivity of quasi-particles in cuprates
Problem: Separate the QP current from vortex currents?
K. Krishana, Yuexing Zhang, J. M. Harris, NPO
κxx vs. T in 90-K YBCO (twinned and untwinned)
Is peak from QP or phonons?
Monitor thermal currents.
Excitations of an s-wave superconductor
remove
Energy cost Ek = [ ξk2 + Δ2 ]1/2
Ek
ξk
quasiparticles
Δ
QP’s cost energy, but increase entropy S (lower free energy F)
γk = uk c k + vk c-k
Heat transport in low-Tc superconductors
JQ = κtot (- T)
Condensate does not carry heat (zero entropy)
κtot
T
Tc
0
Pb
Heat-current density
κtot = κel + κph
electrons phonons
T > Tc , κtot mostly electronic
Below Tc , QPs carry heat
T << Tc , QP population 0
Phonons are long-lived
κph
Jericho, Odoni, Ott, PRB ‘85
Dirac-like spectrum of QP at nodes
Quasiparticle dispersion E vs. k is linear.
Δ(φ) = Δ0 cos 2φ
Quasiparticles in a d-wave superconductor
gap max. (antinode)
node
Δ(φ) = Δ0 cos 2φ
Ek
ξk
Δ0 antinode
Ek
ξk
node
fk
εk - μ
0 kF-kF
ptcl mass currenthole mass current
Q-ptcl mass current
Q-hole mass current
- T
Heat current in normal and superconducting states
Fermi Surface
Εk
0 ξkξ−k
nodalQP’s
holeexcitations
electrons
Separating electronic and phonon κ
in 93-K YBa2 Cu3 O7
κa = κe + κph
vortex Dirac quasiparticle
H
Quasiparticles in the CuO2 plane
Hall thermal current from asymmetric scattering of QP by vortex
heater
asymmetric scattering of QP by vortex
scattering of phonons:no asymmetry
warm
cool
heaterbath
Doppler shift
vs
counter-moving qp
co-moving
vs
counter-moving qp
co-moving
J = Js + Jqp ( Js || -Jqp )
In a supercurrent vs , energy of counter-moving QP’slowered.
QP excitations
Origin of QP Hall current
counter-moving
co-moving
vortex core
vs
Incident QP’s
Scattered QP’s flow against vs circulation
Doppler shift lowers energy of counter-moving QPs
σH
T/H1/2
Skew scattg cross-section (Durst,Vishwanath, Lee, PRL ‘03)
σH ~ T / H1/2
Thermal Hall Conductivity κxy in high-purity YBCO7(normal state)
Thermal Hall Conductivity κxyIn high-purity YBCO7
1) Hall signal much larger below Tc
2) Giant increase in initial slope 85 to 40 K
3) Strongly non-linear in H
Thermal Hall Conductivity κxy in high- purity YBCO7
(12.5 to 35 K)
Plot initial slope limB-0 κxy /Bvs. T.
Initial slope increases by 1000 between 85 and 30 K
Steep increase in QP mean-free-path ~ 120
QP mean-free-path l derived from Hall angle
Increases by 120 from 85 to 30 K
Abrupt increase at Tc (coherence effect?)
κxy = C0 (TH)1/2
F(x) ~ x
Simon-Lee scaling
κxy = T2 F( H1/2/T )
Calculated fits to Kxx and Kxy(Adam Durst, Ashvin Vishwanath, P.A. Lee, 2003)
κxx = ce vF l ~ T2T-1
κxy = κxx tan θ
= κxx nV σH l
~ T2T-1 . H .TH-1/2. T-1 ~ (TH)1/2
Durst, Vishwanath, Lee
Explains observation
κxy = C0 (TH)1/2
Quasi-particles are hole like in UD YBCO y=6.60
Summary
Below Tc , we observe
. 1000-fold increase in κxy (weak field)
. 200-fold increase in QP mfp l.(80 Angstrom to 2 microns)
. Giant anomaly in κtot is entirely from QP.
. Steep increase in mfp starts just below Tc
(conflicts with ARPES)
. Intriguing scaling behavior in κxy (Simon-Lee)
. No evidence (yet) for Landau quantization
References for Lect. 1 (website http://www.princeton.edu/~npo/)
1. N. P. Ong, Phys. Rev. B 43, 193 (1991).2. J.M. Harris, Y.F. Yan, and N.P. Ong, Phys. Rev. B 46, 14293 (1992).3. J.M. Harris, Y.F. Yan, O.K.C. Tsui, Y. Matsuda and N.P. Ong, Phys. Rev. Lett. 73, 1711 (1994).4. J. M. Harris, N. P. Ong, P. Matl, R. Gagnon, L. Taillefer, T. Kimura and K. Kitazawa, Phys. Rev. B 51, 12053 (1995).5. K. Krishana, J. M. Harris, and N. P. Ong, Phys. Rev. Lett. 75, 3529 (1995).6. Y. Zhang, N.P. Ong, Z.A. Xu, K. Krishana, R. Gagnon, and L. Taillefer, Phys. Rev. Lett. 84, 2219 (2000).7. Y. Zhang, N.P. Ong, P. W. Anderson, D. A. Bonn, R. X. Liang, and W. N. Hardy, Phys. Rev. Lett. 86, 890 (2001).8. Adam C. Durst, Ashvin Vishwanath, and Patrick A. Lee, Phys. Rev. Lett.
90, 187002 (2003).
Vortices in cuprates
CuO2 planes
2D vortex pancake
ξ
Gap amplitude vanishes in core
|Ψ| = Δ
A new length scale ξ* Cheap, fast vortices
ξ0
Js (r)
Δ(r) Δ(r)
Js (r)
ξ0
00
H* = φ0
2πξ*2
Is H* determined byclose-packing of fat vortices?
ξ*