Thermoelectricity in mesoscopic devices due to ... - Robert S Whitney

THEORY µHAND-WAVING PICTURE path1 path2. Andreev reflection. Phase−shift fluctuates from path to path φ δτ e h e h. AVERAGE: e µh h µe св. ¢. ¾¯ ф ·.
90KB taille 2 téléchargements 263 vues
Institut Laue Langevin Grenoble, France

TransalpNano 2010 Como, June 2010

Thermoelectricity in mesoscopic devices due to Aharonov-Bohm interference Robert S. Whitney [1] Thermoelectricity

| density-of-states | interferen e e.g. Aharonov-Bohm

[2] Nanostructures with SC loops — explain puzzling expts : Chandrasekhar’s group (1998,2009), Petrashov’s group (2003)

Ph. Ja quod & R.W. arXiv:0910.2943 see also R.W. & Ph. Ja quod, PRL 103, 247002 (2009)

THERMOELECTRICITY — Peltier or Seebeck effect

L lead

black box

R lead

VOLTAGE difference

TEMPERATURE difference

CHARGE current

HEAT current

APPLICATIONS at nanoscale (sub-Kelvin):

| Refrigeration =)

colder than cryostat ? Expts: Courtois-Pannetier Group

| Novel transport probe of nanostructures =) di eren e between physics at (EF + kB T )

(

and at EF

kB T

)

ORIGINS OF THERMOELECTRICITY DENSITY OF STATES L lead HOT

HOT Fermi sea

R lead

n−type semi− conductor

111 000 000 111 000 111 000 111

T hot

COLD

SUB-KELVIN: semicond

semiconductor density 000 111 of states 000 111

111 000 000 111 000 111 000 111 000 111

!SUPERCOND

Expt: Courtois-Pannetier Group

INTERFERENCE:

COLD Fermi sea

oheren e ) only sub-Kelvin

path 1

L lead HOT

R lead

Aharonov− Bohm ring

HOT

COLD

φ

COLD

path2

PHASE-DIFFERENCE:

 = p(l2

l1 with

) + p p/

ENERGY

Electrons − contructive Holes − destructive

PUZZLING EXPTS on THERMOPOWER house

(1998) Chadrasekhar’s group HOT

asym. house

(2003) Petrashov’s group

HOT

φ

parallelogram

(1998) Chadrasekhar’s group

hot−middle (2009) Chadrasekhar’s group HOT

HOT

φ

φ

SC loop

EVEN in φ

φ

ODD in φ

PREVIOUS THEORY: quasiclassical \Usadel equations" Virtanen-Heikilla (2004-07), Titov (2008)

Thermopower ALWAYS odd in  & will VANISHES in symmetric case; i.e. house ... BUT theory misses mesoscopic fluctuations

FULLY QUANTUM THEORY — random matrix theory or semiclassics house

(1998) Chadrasekhar’s group HOT

asym. house

(2003) Petrashov’s group

HOT

parallelogram

(1998) Chadrasekhar’s group

HOT

HOT

φ

φ

hot−middle (2009) Chadrasekhar’s group

φ

φ

SC loop

L

R φ

φ δτ

δτ

δτ

φ

δτ

THEORY path1

e h

path2

=) HAND-WAVING PICTURE Phase−shift

e

fluctuates from path to path

h

φ AVERAGE:

δτ e )h

Andreev reflection

h)e

os 2 Æ +  os 2 Æ

SAMPLE-TO-SAMPLE FLUCTUATIONS:

os (2 Æ + e+h) + ( + e h) 

os (2 Æ + e+h) ( + e h) For sym. house:





=)

symmetry kills ODD terms

=) ODD in  ODD & EVEN but small

OVERVIEW : NANO-THERMOELECTRICITY

PHYSICAL ORIGINS:

| Density of states | Interference (need v. low temp)

...

anything else??

APPLICATIONS:

| Refrigeration

NEEDS BIG EFFECT

| Probe of di eren e in physics at (EF + kB T )

(

and EF

kB T

)

WORKS WITH SMALL EFFECT