discussion_week3_FeSe - Magnetism, Bad Metals and

Report
Review of results on FeSe
P Hirschfeld, 9/19
(Data only up to 6/2014)
Thanks to:
Taka Shibauchi
Tetsuo Hanaguri
Frederic Hardy (+Anna Boehmer, Christoph Meingast)
• Basic properties N and S states
• New physics from new crystals
Relatively correlated material
Z. P. Yin, K. Haule, & G. Kotliar, Nat. Mat. 10, 932–935 (2011)
LDA+DMFT exercise:
Fix interactions U,J, vary
material
FeSe: nonmagnetic 8K superconductor, but:
ARPES gap
Wang et al. Chin. Phys. Lett. 2012
Medvedev et al 2010: Tc37K under pressure
1 layer Tc35K under tensile strain
Burrard‐Lucas et al 2012
Tc43K molecular intercalation
S. He et al aXv::1207.6823
Pressure dependence of bulk FeSe
Margadona et al 2010
Bendele et al 2012:
magnetic state at low pressure
Medvedev et al 2010
Pressure enhances spin fluctuations
Imai, Cava PRL 2009
But note difference from other systems
FeSe Spin fluctuations seem to wait until orthorhombic transition happens
Are the chalcogenides generally more correlated? “Bad metals”?
Fang et al 2009
Morosan et al (Rice group) 2013
Mizuguchi et al 2011
A tale of two Fe-chalcogenides
Mizuguchi et al 2011
Kasahara et al, unpublished (2014)
crystals from
A. Böhmer et al., PRB 87, 180505(R) (2013)
r(Tc)~0.1Wcm
Bad metal physics not evident in FeSe
High-quality stoichiometric FeSe single crystal grown @KIT
A. Böhmer et al.,
PRB 87, 180505(R) (2013).
S. Kasahara et al.,
unpublished?
• Tc ~ 10 K (cf. ~8 K for typical samples)
• Large RRR and MR indicate that samples are very clean.
How good are new KIT crystals really?
F.-C. Hsu et al., PNAS 105, 14262 (2008).
r0= 250 mWcm at 8K
RRR~6.5
S. Kasahara et al., unpublished?
r0= 10 mWcm at 10K
RRR~40
Consistent with (r(T0) =0)
Electronic specific heat
old
new
JY Lin et al, PRB 84, 220507(R) (2011)
Hardy et al, unpublished
Old and new very similar – small influence of disorder on SC
SdH (Terashima arXiv:1405.7749)
SdH
Large orbital ordering in ARPES Nakayama et al. arXiv:1404..0857
Signatures of electronic nematicity in FeSC generally
ARPES: orbital ordering
(0,p)
(p,0)
Yi et al PNAS 2011
(0,p)
(p,0)
Signatures of electronic nematicity in FeSC
STM in SC state
FeSe: CL Song et al, Science 2011, PRL 2012
topography
spectrum
defect
vortex
a and b are only ~0.1% different! But strong C4 symmetry breaking in SC state.
Tunneling spectra
 High energy spectrum (±95 mV)
 Low energy spectrum (±6 mV)
Multigap SC
Unidirectional quasi-particle interference
Hanaguri group using KIT crystals
FT-dI/dV/(I/V) T ~ 1.5 K
Topograph
dI/dV/(I/V)
Bragg
alias
bqFe
b
aqFe
a
Small orthorhombicity
bFe
yet large
anisotropy in
the band
aFe structure!
45 nm×45 nm, +50 mV/100 pA
Unidirectional dispersing features
in qa and qb directions.
cf. NaFeAs: E. P. Rosenthal et al.,
Nat. Phys. 10, 225 (2014).
QPI Bandstructure (note: over small 1-domain window!)
along qa
B = 12 T
B = 12 T
imp.
+D
-D
along qb
FT-dI/dV/(I/V)
EF
Electron-like
EF
imp.
+D
-D
Hole-like
• Orthogonal electron- and hole-like dispersions
• Extremely small EF ~ D
BCS-BEC crossover regime?
Possible intra-orbital scattering
• Orbital character changes when
we go around the FS pockets.
• If only intra-orbital scatterings
are allowed, QPI patterns may be
unidirectional.
• Why one of the orbitals is active?
Orbital order?
S. Graser et al.,
New J. Phys. 11, 025016 (2009).
Can we reproduce
orthogonal electron and
hole dispersions using the
orbital-order model?
Lifting the orbital degeneracy
Orbital character
Band calc. (by Dr. H. Ikeda)
xz
Orthorhombic
distortion only
= 0.05 eV
Eyz-Exz
= 0.1 eV
xy
More detailed
calculations are
indispensable…
xz
Eyz-Exz
yz
yz
Orthorhombic distortion alone
cannot explain the unidrectional dispersions.
xy
Orthorhomicity is
yz
xz
xy
not a player but a
spectator.
Orbital order?
Penetration depth and
thermal conductivity
results
Introduction: FeSex
Superconducting gap symmetry ---- A key for the mechanism
Thermal Conductivity
MBE-STM
J.K. Dong, et al., PRB (2009).
Specific heat
 The simplest structure
 Strong correlation
F.C. Hsu, et al.,
PNAS (2008).
J.-Y.Lin, et al., PRB (2011).
Can-Li Song, et al.,
Science 332. 1410 (2010).
 Single crystals (off-stoichiometry)
 Defect-free stoichiometric films
 Nodeless multiple gaps
 Nodal superconductivity
Magnetic field penetration depth
Large temperature dependence
Quasi T-linear at T/Tc < 0.2
Dl ~T1.4
No Curie term
(No excess irons)
Finite qusiparticle
excitation
at low temperatures
Presence of line nodes
Thermal conductivity in a stoichiometric FeSe single
crystal
Increase of the
quasiparticle
life time below Tc
Large residual value
k0/T~ 0.4 (W/K2m)
kn/T
Wiedemann-Franz law
kn/T=L0/r0 ~ 1.43
(W/K2m)
L0: Lorentz number
Tc
r0 ~ 1.70 mWcm
~ 30-40% of the normal state
value
Strong evidence
for the line nodes
Discussion: Origin of the different behavior
D
Present study (Clean single crystals)
Quasi T-linear l(T)
Finite residual k0/T
f
0
Nodal Superconductivity
Accidental nodes
Earlier study (Dirty crystals)
Negligibly small k0/T at 0 T
Gap anisotropy is smeared
by strong scattering
G
D
0
f
J.K. Dong, et al.,
PRB (2009).
Nodeless
(Anisotropic s-wave)
Nodes can be removed
Nodal s-wave state in FeSe
Discussion: Origin of the different behavior
Magnitude of the residual term
D
Present results
0
x: coherence length ~ 5 nm
~ 0.3-0.4
l: mean free path
~ 200 nm
1/m ~ 6 - 8
node
Inconsistent with d-wave
Slope parameter of gap at nodes
Nodes are nearly vanishing
G
D
f
Accidental nodes
Gap anisotropy is smeared
by strong scattering
0
D
f
0
f
2-band model
V. Mishra et al.,
PRB, 80, 224525 (2009).
Nodes can be removed
Nodal s-wave state in FeSe
Anomalous field dependence of thermal conductivity
Strong reduction of k/T at low fields
FeSe
Plateau at high fields
kel/T ~ N(EF)vFl
Different from ordinal
behaviors
Doppler shift N(E)~ H1/2
m0
Anomalous field dependence of thermal conductivity
Strong reduction of k/T at low fields
FeSe
Plateau at high fields
kel/T ~ N(EF)vFl
① Vortex scattering due to long mean free
path
CeCoIn5
Y. Kasahara et al.,
PRB, 72, 214515 (2005
Doppler shift N(E)~ H1/2
m0
Long m.f.p.
l ~ H-1/2
& vortex scattering(av ~ H-1/2)
Cancelation  Plateau
Anomalous field dependence of thermal conductivity
Strong reduction of k/T at low fields
FeSe
Plateau at high fields
kel/T ~ N(EF)vFl
① Vortex scattering due to long mean free
path
FeSe
Magnetoresistance
Dr/r0 ~ (wct )2
l =vFt ~ 0.2 mm
Long mean free path
m0
Hard to explain
a sharp kink at low fields
and a plateau in a nearly whole vortex state
Anomalous field dependence of thermal conductivity
Strong reduction of k/T at low fields
FeSe
Plateau at high fields
② Possible phase transition in the SC
state
BSCCO
Field induced change of gap
symmetry
dx2-y2  dx2-y2 + idxy
or
dx2-y2 + is
K. Krishana, et al.,
Science (1997).
m0
FeSe
s  s + id
(???)
Anomalous field dependence of thermal conductivity
Strong reduction of k/T at low fields
FeSe
Plateau at high fields
③ Lifting nodes under magnetic field
m0
V. Mishra et al.,
Phys. Rev. B, 80, 224525 (2009)
Plateau with finite k/T
 Small SC gap already suppressed at low
fields
High-field anomaly in thermal conductivity
H*
1.0
0.70
0.9
1.5 K
k/T (W/K m)
0.8
2
1.12 K
2
k/T (W/K m)
0.65
0.7
0.76 K
0.6
0.76 K
0.60
0.58 K
0.55
0.58 K
0.50
0.39 K
0.45
0.39 K
0.5
0.4
0
4
8
m0H
(T)
12
16
0
4
8
m0H
(T)
12
16
Proposed new high-fied phase
Summary
• FeSe Tc very sensitive to pressure
• Apparent strong orbital ordering in ARPES, STM,
no magnetism strong nematic ordering
(resistivity anisotropy???)
Big challenge to electronic structure theory!
• SC state consistent with weak nodes
(easily removed by perturbation)

similar documents