Transcript PowerPoint プレゼンテーション - University of Tokyo

’06 07/25 ICPS TuA3o.3 Tue 17:30 Poster
Band-edge divergence and Fermi-edge singularity
in an n-type doped T-shaped quantum wire
T. Ihara1, M.Yoshita1, H. Akiyama1, L. N. Pfeiffer2 and K.W. West2
1Institute
for Solid State Physics, University of Tokyo and CREST, JST, Chiba 2778581, Japan
2Bell Laboratories, Lucent Technologies, Murray Hill, NJ 07974, USA
Introduction
Inverse-square-root divergence
Density of state (DOS)
D1D ( E )  1 / E
Interesting phenomena appear.
・Large exciton binding energy.
・Strong absorption of
exciton groundstate.
・High speed optical modulation device.
・High performance laser device.
Researches on 1D doped quantum wires
Experiments
’91 J. M. Calleja et al.
’93 M. Fritze et al.
’01 D. Y. Oberli et al.
(PL & PLE)
(PL)
(PL & PLE)
Experimental investigations
on the 1D FES effect.
’02 S. Sedlmaier et al.
’02 H. Akiyama et al.
(PL)
(PL)
Experimental investigations
on the 1D BGR effect.
Theories
’92 T. Ogawa et al.
’96 Y. Oreg et al.
’96 H.V. Grünberg et al.
Calculations of the critical exponents
of Fermi edge power-law singularity.
’92 Das Sarma et al.
’01 M. Stopa et al.
Calculations of
the Band Gap renormalization effect.
’92 J.F. Mueller et al.
’92 P. Hawrylak et al.
’93 F.J. Rodriguez et al.
Calculation of optical spectra
with 1D DOS and many-body effect.
Our recent results and motivations
carrier temperature : 10K
High electron density : 6x105 cm-1
I    

0
A   

0
1
'
1
'
L DOS    'd ' f mee ,mh ,  ,T   f mhe ,mh ,T  


L DOS    'd ' 1  f mee ,mh ,  ,T  
Higher temperature
Band-edge absorption ?
’04 T. Ihara et. al.
Band edge & Fermi edge
induced by 1D DOS and
Fermi’s golden rule
Lower temperature
Fermi edge singularity ?
Low density
Band-edge absorption ?
Band gap renormalization ?
Metal-Insulator crossover ?
Bound states ?
1D screening ?
Sample structure of n-type doped T-wire
＜fabrication＞
MBE with cleaved edge
overgrowth method
＜size of wire＞
14 x 6nm x 4mm(single)
＜doping＞
①Si modulation doping
②FET gate structure
→tunable electron density
＜measurement＞
Temperature-elevated
micro-PL spectra &
resonant PLE spectra
Results : PL and PLE spectra at various temperature
T=50K (k T/Ef ～1)
B
non-degenerate 1DEG
high
Temperature
A sharp absorption
peak (BE) appears.
T=10K
(kBT/Ef ～0.2)
degenerate 1DEG
Fermi-edge absorption
onset (FE) appears.
low
Sharp absorption peak at 50K
- Same energy with PL peak. - Good agreements with calculations.
- Characteristic of 1D electron systems
Band-edge absorption peak induced by 1D DOS divergence !!
Results : PL and PLE spectra at various electron density
ne=6x105cm-1 ～ 3x105cm-1
degenerate 1DEG (kBT/Ef < 0.5)
Electron density
-Fermi edge absorption onset
-Band edge emission
-Band gap renormalization
high
ne=3x105cm-1 ～ 1.5x105cm-1
non-degenerate 1DEG (kBT/Ef > 0.5)
Sharp band-edge absorption
low
ne
< 1.5x105cm-1
Discrete peaks of bound states
Sharp band-edge absorption appears at ne = 1.5x105cm-1 ～ 3x105cm-1
Discrete symmetrical peaks appear only at low densities (ne < 1.5x105cm-1 )
Results : PL and PLE peak energy plot with Ef and ne plot
Band gap renormalization
※ Eb～13meV (for neutral exciton)
0cm-1
0meV
(1.5685+0.013
)
2x105cm-1 6x105cm-1
-15meV
(1.5665)
-17meV
(1.5645)
Discrete peak of bound states
-Exciton peak disappears ne < 1.5x105cm-1
Problem :
Sharp asymmetrical absorption peak (▲) at the lowest energy (1.5665eV)
※ almost 2 meV below the neutral excitons
Band-edge absorption ? Bound states ? Many-body objects ? or Mixture of them ?
Our assignment
Discussion : What is the asymmetrical absorption peak (▲)?
Asymmetrical absorption peak
(▲) at the lowest energy
※ almost 2 meV below the neutral excitons
The asymmetrical peak
structure is observed even at
high temperature (50K).
 Band edge divergence
?
Bound states with continuum
Many-body object (FES)
Sharp asymmetrical peak at Vg=0.2V seems to originate from Band-edge divergence.
⇒ 1D bound states disappear at such low density (～1.5x105 cm-1) !?
Discussion : Difference between 1D and 2D systems.
Asymmetrical absorption peak (▲) at the
lowest energy
※ almost 1.5 meV below the neutral excitons
In the case of 2D electron systems
The asymmetrical peak
structure is not observed at
high temperature (50K).
Band edge divergence
Bound states with continuum
 Many-body object (FES)
Sharp asymmetrical peak at Vg=0.4V seems to originate from many-body effect.
⇒ The origin of the lowest energy peak (▲) is different in 1D and 2D systems !?
Summary
Low-temperature PL and PLE spectra are studied in an n-type modulation-doped
T-shaped single quantum wire with a gate to tune electron densities.
With non-degenerate 1D electron gas, band-edge absorption exhibits a sharp bandedge-divergence of 1D density of states.
When the dense 1D electron gas is degenerate at a low temperature, we observe a
band-edge emission peak and a Fermi-edge absorption onset.
Problems (Further investigation …)
- Bound state and/or Band edge
- PL and PLE measurement at lower temperature (～1K)
- 1D / 2D electron system with large hole effective mass
※ In the case of 2D electron systems
※ Estimation of Carrier Temperature
1
fe 
1  exp  e   e 
fh 
1
I    fe f h
1  exp  h   h 
A I 
A    1  f e 
1  f e h 

fe fh
temperature
A  I  exp  
PL (I)
PLE
PL( I )  exp  
with T=10K
with T=5K
with T=15K
Good agreements between PLE and PL×exp[β(ћω)] at T=10K