Single photons from
single quantum dots
Andrea Fiore
Institute of Quantum Electronics and Photonics, EPFL
NCCR Quantum Photonics
Project 2A: Single-photon sources from
self-organized quantum dots
Photons: The long search
1900: Photoelectric effect
1905: Einstein: Light quanta
Statistical properties of light (1955-...):
<I(t)I(t+ )>
PMT
classical
delay
PMT
Coincidence
counts
Hanbury Brown and Twiss experiment (1956)
quantum
Andrea Fiore - NCCR2A
How many particles?
Electronics:
# electrons to process one bit
1970: 10 µ m gate, 10 V
6x107 electrons/gate
2001: 65 nm gate, 1V
2x104 electrons/gate
Photonics:
# photons to transmit one bit
1980: 45 Mbit/s, 1 mW
108 photons/pulse
2001: 10 Gbit/s, 1 mW
8x105 photons/pulse
Where is this all going?
Andrea Fiore - NCCR2A
The day after tomorrow...
Electronics:
Single-electron transistors
Photonics:
Single-photon emitters
and detectors
500 nm
island
gate
Courtesy
CNR-IFN
A very-low-power source ???
What is a "single photon" ???
And what for ???
Ford et al., 1993
Andrea Fiore - NCCR2A
Why single photons
"Quantum-secure" communication uses single photons:
Bob
Alice
01011...
Singlesource
01011...
Eve
• Two polarization coding sets:
=1
=0
=1
=0
1 1 0 0 0 1 0 1 1
• Alice chooses a polarization set randomly:
• Bob selects an analyzer set randomly:
• Only the bits where the same polarization
sets were used are kept
1 1
0 0
1 1
Andrea Fiore - NCCR2A
Eavesdropping on single photons
Bob
Alice
01011...
Singlesource
01011...
A single photon
cannot be split!
Eve
To spy, Eve must intercept the photon and retransmit it:
photon
"wrong"
analyzer
error
By monitoring the error rate Alice and Bob can check
the channel security!
Andrea Fiore - NCCR2A
Quantum cryptography =
Classical money ?
•
2 start-ups
• Quantum cryptography systems being commercialized
• Practical systems much more complicated !
Andrea Fiore - NCCR2A
A close look at single photons
"Classical" light sources (e.g. a laser) are Poissonian:
<n> =1
0 ,4
P (n ) = prob. of n photons
n = average photon number
2
n
0,3 5
n
0 ,3
= n
0,1
0,2 5
0 ,2
0,1 5
0,08
0,06
0 ,1
0,04
0,0 5
0,02
0
<n > =1 0
0,12
P(n )
n
e
P (n ) =
n!
P(n)
n
0,14
0 2
0
4 6 8 1 0 1 2 14 16 1 8
n
0 2 4 6 8 10 12 14 16 18
n
Fluctuations are tied to the statistics
Let's attenuate a laser (=change <n>):
filters
laser
n = 10 6
3
=
10
n
n =1
n =1
Not a singlephoton
source!
Andrea Fiore - NCCR2A
A true single photon source
Typical source used in quantum cryptography:
Attenuated laser, <n> 0.1
P(1)=9%, P(2)=0.45%
• Low bit rate
(Because we are not
changing the statistics...)
• Probability of two-photon pulses not negligible
Security not completely guaranteed
(Brassard et al., PRL 85, 1330 (2000))
For a true single-photon source we must trick Poisson:
Single quantum system:
photon emission
"dead time"
time
"photon
dropper"
Andrea Fiore - NCCR2A
The quest for single-
sources
The simplest single- sources:
• Single atoms
• Single ions
• Single molecules
• ...
Wish list for single- sources:
• Compact
• Electrically pumped
• Room temperature
• Efficient
(Diedrich and Walther, PRL 1987)
electron
photon
A semiconductor LED!
Andrea Fiore - NCCR2A
The challenge
Isolate a single quantum state in a semiconductor
and pump it efficiently
E
Density of states in a semiconductor of volume V:
(E ) =
V
2
2m *
2
h
CB
kT
3/ 2
2
E
Ec
k
VB
In E=20 meV ( kT):
2x1017 states per cm3 in CB
Need to isolate one state !
V
(10 nm)3
"Nano"semiconductors
Andrea Fiore - NCCR2A
Semiconductor nanostructures
3D carrier confinement
Bulk semiconductor:
Quantization of energy states
Energy
emission
cond.
band
+
"Quantum Dot"
valence
band
energy
emission
Energy
EC
Ev
energy
Quantum dots behave as "solid-state atoms"
Andrea Fiore - NCCR2A
Nanostructure fabrication
"Straightforward" approach: Fabrication on the nanoscale
• Need high-resolution lithography
• Etching
Nonradiative defects
"Self-organized" methods: Making Nature work for us !
• Strain-driven growth on planar
substrates (project 2A)
substrate
• Growth on patterned substrates
(project 2B)
substrate
Andrea Fiore - NCCR2A
Self-organized growth
Epitaxial growth: Interplay of surface energy and strain
"Standard" (Frank-van der Merve) 2D epitaxial growth:
GaAs
AlAs
substrate
The epitaxial layers "wet"
the surface
2D growth
Stranski-Krastanov 3D growth mode:
InAs
substrate
Strain too large
island formation
Andrea Fiore - NCCR2A
Self-organized quantum dots
MBE growth of InAs on GaAs:
15 nm
200 nm
Atomic Force Microscopy
Transmission Electron Microscopy
Simple growth technique, no
pre-growth patterning required
Random nucleation sites
control of island position
High crystalline quality
High radiative efficiency at RT
Size dispersion
dispersion
no
Emission energy
1.3 µm operation possible
Andrea Fiore - NCCR2A
The objectives
100 nm
QDs
• Fabricate LEDs with single QDs
• Characterize single photon emission
• Application to quantum communication
This talk:
• Growth and characterization of QDs
• QD devices
• Towards a single QD LED
Andrea Fiore - NCCR2A
Quantum Dot optimization
Specifications:
• High radiative efficiency
• Wavelength 1300 nm
• Low areal density ( 109 cm-2)
•T
diffusion
and fewer QDs
• In flow
surface
larger
Growth parameters:
• Temperature
• Growth rate
• Capping procedure
In atoms
InAs islands
more atoms on
more QDs
Andrea Fiore - NCCR2A
Growth optimization
Areal density:
0.075 ML/s
1.9x1010 cm-2
0.037 ML/s
1.4x1010 cm-2
Emission wavelength:
0.15 ML /s
PL p ea k (nm )
0.075ML /s
8
0.037 ML /s
0.019 ML /s
0.009 ML /s
6
4
2
0
1000
1 300
44
1 280
42
1 260
40
1 240
38
1 220
36
1 200
34
1 180
32
1 160
1100
1200
1300
1400
W avelength (nm )
Chen et al., JAP 91, 6710 (2002)
6x109 cm-2
0
0.0 4
0.0 8
0.12
F WHM
Intensity (arb. un.)
10
0.019 ML/s
30
0.1 6
G row th rate (M L/s )
Andrea Fiore - NCCR2A
Growth: Capping
GaAs capping
In segregation
InGaAs capping:
InGaAs
In-rich
W avelength (n m )
14 50
14 00
13 50
• Suppression of In segregation
13 00
• Reduction of strain
12 50
0 .1 5 M L/s
0.0 09 ML/s
12 00
11 50
0
• Bigger QDs
0.0 5 0 .1 0.1 5 0 .2 0.2 5 0.3 0 .35
In co m po sition in cap
Andrea Fiore - NCCR2A
Radiative properties of
optimized QDs
110 0
10 W /cm
2
5K
10 0K
15 0K
20 0K
25 0K
30 0K
20
15
10
5
0
1100
12 00
1300
1 400
Dec ay tim e (ps )
Inten sity (arb. un.)
25
100 0
900
800
4 00
700
3 00
600
2 00
1 00
500
0
0
400
300
100K
0
10 0 0
200 0
T im e [ ps ]
50
100
15 0
200
250
30 0
Te mpera ture ( K)
W avelength (nm )
Markus et al., APL 80, 911 (2002)
Conclusions:
• High radiative efficiency 20% at RT
• Long carrier lifetime 1ns
• 1300 nm emission at RT
• Density
6x109 cm-2
to be decreased
Andrea Fiore - NCCR2A
QDs in "standard" devices:
LEDs
Au
m irro r
- c a v it y
-8
6 10
-8
293 K
5 10
-8
21 A/cm
4 10
-8
3 10
-8
2 10
-8
1 10
-8
0
1100
2
0
1.5
293 K
2
1200
1300
W avelen gth (nm )
Cu rren t density (A/cm )
50
100
150
200
1400
1.5
1.0
1
0.50
0.5
0
0
0.0
50 100 150 200 250 300 350
Cu rren t (mA)
Extern al q uantu m efficiency (%)
2
7 10
Lig ht (mW )
In ten sity (arb. un.)
n+ GaAs
Andrea Fiore - NCCR2A
QD lasers
Laser performance:
• Jth<100 A/cm2
•
ext=30-40%
7 10
-4
6 10
-4
5 10
-4
4 10
-4
3 10
-4
2 10
-4
1 10
1 4 7 A /cm2
1 3 6 A /cm2
-4
0
3 QD layers, 18 µ m strip e
2 mm, 3 QD lay e rs, RT
0 10
1 10 0 11 50 12 0 0 12 50 1 30 0 1 35 0 1 40 0
W a ve le ng th (nm )
40
P ow er (m W)
Intens ity (a.u.)
• CW up to 80 C
R T , pu lsed
35
30
25
20
15
10
2m m
5
5m m
0
3m m
0
50 100 150 200 250 300 35 0 40 0
Current (m A)
Andrea Fiore - NCCR2A
Simple approaches to
single-QD devices
Nanomesas for optical pumping:
1 QD
100 nm nanomesa by Ebeam
lithography and reactive-ion etching
Shadow-mask LED:
300 nm opening in metal contact
by Ebeam lithography and lift-off
It works! (Yuan at al., Science 2002)
But: Very low efficiency...
(Collaboration with CNR-IFN Rome)
Andrea Fiore - NCCR2A
Towards single-QD LEDs
Nano LED: (0.1x0.1 µ m2 = 1 QD)
Macroscopic LED: (100x100 µ m2 = 106 QDs)
QDs
?
Critical issues:
• High spatial resolution needed
• Alignment critical !
• Lateral carrier loss (current spreading,
nonradiative recombination)
Andrea Fiore - NCCR2A
Scaling of LEDs: Issues
"Standard" LED structures:
Current spreading &
NR recombination
Current spreading
NR rec
10
3
10
1
10
-1
10
-3
10
-5
10
-7
10
-9
Qua ntum efficiency (%)
2
Curre nt de nsity (A/c m )
• Large effective device area
• Decrease in efficiency
40 0 um
20 0 um
10 0 um
50 um
0
1
2
3
Volt age (V)
11
10
9
8
7
6
5
wit h re cess etch
4
wit h ox ide a perture
no current confinement
3
0
100
200
300
400
500
D evice siz e (µ m )
Current density and efficiency do not scale with device size
Difficult to fabricate < 10 µm devices
Andrea Fiore - NCCR2A
A new approach to nano-LEDs
Al2O3
Au
GaAs
AlGaAs
QDs
• Epitaxial growth
• Optical lithography & etching
• Selective oxidation
• Metallization
300 nm current aperture
oxid. edge
oxid. edge
•
•
•
Fiore et al., to be published in APL
Down to 100 nm current apertures
with simple optical lithography
Can tailor both current and
optical confinement
Bandgap engineering to reduce
current spreading
Andrea Fiore - NCCR2A
10
4
10
2
10
0
10
-2
10
-4
10
-6
10
-8
1.5 10
293 K
3 0 .m
9.1 .m
1.8 .m
1.0 .m
8 30 nm
6 00 nm
0
0.2
30 um
9.1 u m
1.8 u m
1.0 u m
830 n m
600 n m
-8
29 3 K
L ig ht (W )
2
C ur rent density (A/cm )
NanoLED performance
0.4
0.6
0.8
V oltage (V)
1
1.2
1 10
-8
5 10
-9
0 10
0
0
1 10
-5
C urrent (A)
2 10
-5
Good scaling of current-voltage and efficiency characteristics
Suppressed current spreading
Suppressed carrier diffusion in the QDs
Negligible nonradiative recombination
Andrea Fiore - NCCR2A
Conclusions & Perspectives
Where we are:
QD growth optimized
710
Efficient QD devices
Intensity (a.u.)
610
510
410
310
210
110
-8
-8
-8
293K
21A /cm
2
-8
-8
-8
-8
0
1 05 0
1 15 0
1 25 0
1 35 0
1 45 0
Wavelength(nm)
Technology for single-QD LEDs
The future:
Even smaller: Getting the single-QD LED
Playing with single QDs and single photons…
Andrea Fiore - NCCR2A