Engineering of Disorder in MBE grown Ultra-High Mobility 2D Electron System

1
Engineering of Disorder in MBE grown Ultra-High
Mobility 2D Electron System
Vladimir Umansky Braun Center for Submicron
Research Weizmann Institute of Science, Rehovot,
Israel
Collaborators Moty Heiblum group (Braun Center
for Submicron Research) Jurgen Smet group
(Max-Planck-Institut für Festkörperforschung,
Stuttgart)
2
Preface 2DEG and Mesoscopic Physics
Mobility 25,000 cm2/Vs
3
Electron mobility progress
4
Outlook

• 2D Electron Gas - basics
• DX centers why we are lucky to have them?
• How to observe 5/2 quasiparticles ?
• New ideas for band gap engineering
• Ultra High Mobility. Is it enough ?
• How to control disorder ?
• Conclusions

5
2DEG in AlGaAs/GaAs
2DEG in AlGaAs/GaAs -scattering
AlGaAs(x0.3)
GaAs
Tlt1K
Doping
Spacer (d)
BG
RI
2DEG
?Ec
E0
EF
2DEG Total Depth (D)
6
DX centers
The standard 2DEG structure
In the dark Pros Frozen charge (in the dark)
allows gating Cons Low doping efficiency (in the
dark) ? high RI scattering After Illumination in
the dark Pros Almost double density after
illumination ? high mobility. Cons Parallel
conduction/gate instability.
7
Applications
Gateable 2DEG QDs, QPC, Spin-pump, Quantum shot
noise, etc
Deep structures Measurements after illumination
Shallow structures Measurements in the dark
8
5/2 in the standard 2DEG
Data from 1998
Standard Al0.36Ga0.64As/GaAs 2DEG Mobility
14 106 cm2/Vs Density 2.2 1011 cm-3
Measurements After illumination
Ungateable
5/2
9
How to Achieve Ultra-High Mobility ?
Background Impurity Scattering

• MBE system design
• Raw materials (i.e. Gallium (7N) ? 251015
  cm-3 ) ()
• Optimal growth conditions (rate, temperature,
  III/V ratio, etc)
• Optimal 2DEG structure design
• Optimal growth sequence design

() background impurity density 11014 cm-3
limits mobility by 12 106 cm2/Vsec
10
Double Side Doping
For the same spacer width
Concern Interface scattering in QW ? Inverse
interface
Used first by L. Pfeiffer to produce samples with
gt 30 106 cm2/Vsec
11
Doping in Short Period Super-Lattice

• Higher transfer efficiency
• Higher mobility due to better screening by X
  electrons
• No parallel conductance due to 3 times shorter
  Bohr radius

Short Period Super-Lattice - SPSL
12
Results on Electron Mobility
RIBER MBE32 machine
Uniform Doping in Al0.35Ga0.65As
EF
2DEG
SPSL d-doping
36x106cm2/Vs
2DEG in QW
EF
SPSL d-doping
13
Is Mobility a Relevant Parameter for FQHE ?
14
BG scattering vs RI scattering
Spacer 80 nm
BG limited mobility 16 106 cm2/Vs
For spacer gt 80 nm contribution of RI scattering
lt 1315
15
Mobility, Disorder FQHE

• In high mobility 2DEG the main scattering
  mechanism BG scattering
• BG impurities 1013 cm-3 in 30 nm QW? average
  distance 2 mm
• RI disorder potential characteristic length ?
  spacer ? 80100 nm

FQHE is governed by RI induced disorder
16
How to control the RI disorder?
Introduce Spatial Correlations between Ionized
Donors !!!
g 20
17
Over-doping FQHE
SPSL d-doping
EF
2DEG
Uniform Doping in Al0.35Ga0.65As

• Concern Over-doping leads to Parallel
  conductance
• Minimal Doping 21011 cm-2
• Average distance between donors 200 ?
• Bohr Radius for X-electron 2030 ? ?
  over-doping of 25 times looks feasible

18
Application for 5/2
EF
SPSL d-doping
19
Measurements of ¼ electrons charge
20
Theres no such thing as a free lunch
Double side doped 2DEG n(3.03.3)1011 cm-2,
m(2933)106 cm2/Vs
g 2
g 2.3
g 2.5
5/2
21
Phase transition in Donor layer (s)
g 2
0
2
1
0
g 2.3
g2.3 g1.1
22
Phase Transition in Disordered 2DES
23
Ideal 2D system for mesoscopic device
24
Engineering of Disorder Doping Schemes

• Using another AlAs-GaAs SPSL for doping
• Using multiple doping layers in SPSL
• Using shallow DX centers in AlGaAs

25
Conclusions

• High mobility (low total scattering rate) is just
  a precondition to obtain very low disordered 2D
  systems.
• FQHE is governed by RI induced disorder
• Spatial Correlations of Remote Ionized Donors are
  necessary to obtain perfect 5/2 FQHE