Quantum Well and Quantum Dot Intermixing for Optoelectronic Device - - PowerPoint PPT Presentation
Quantum Well and Quantum Dot Intermixing for Optoelectronic Device - - PowerPoint PPT Presentation
Quantum Well and Quantum Dot Intermixing for Optoelectronic Device Integration Chennupati Jagadish Australian National University Research School of Physical Sciences and Engineering, Canberra, ACT 0200 AUSTRALIA c.jagadish@ieee.org
Overview
- Introduction
- Methods of Intermixing
- Quantum Wells
- Ion Implantation Induced Interdiffusion
GaAs/AlGaAs, InGaAs/AlGaAs, InP/InGaAs QWs
- Lasers, Photodetectors
- Impurity Free Interdiffusion
GaAs/AlGaAs, InGaAs/AlGaAs, InGaAsN/GaAs QWs
- Integrated Waveguide-Laser
- Quantum Dots
- Suppression of Interdiffusion
- Implantation Induced Interdiffusion
- Summary
Photonic I ntegrated Circuits / Optoelectronic I ntegrated Circuits
- I ntegrated Circuits Show Superior
Performance Over Discrete Devices
- Multi-functional circuits, e.g. WDM
sources
- I ntegrated Transceivers
- Low Cost, Packaging
Photonic I ntegrated Circuits
Different Bandgaps
- n the same chip
WDM Source
- ptical
amplifier passive
- ptical
waveguides
multi
- wavelength
laser diodes
Optical output (to optical fibre)
dielectric passivation implant isolation
Quantum Well Intermixing
- Diffusion of In and Ga across interface creates graded region in the
case of GaAs/InGaAs Quantum Wells
- Changes Bandgap, refractive index, absorption Coefficient
before after
Methods Widely Used for Quantum Well (Dot?) I ntermixing
I mpurity I nduced Disordering, e.g. Zn, Si I mpurity Free I nterdiffusion, e.g. SiO2, SOG I on I mplantation I nduced I nterdiffusion Defects/ I mpurities introduced by these methods enhance atomic interdiffusion Goals: High Selectivity and Low Concentration of Residual Defects while achieving large band gap differences
Widely used in Microelectronics I ndustry Defect Concentration
- I on Dose, I on Mass, I mplant Temperature,
Dose Rate Defect Depth - I on Energy Selective I on I mplantation using Masks
Why I on I mplantation?
Ion implantation induced quantum well intermixing
Vacancy Interstitial
Point defects:
Ion implantation
Intermixing
Schematic of 4 QW structure (40 keV Proton Defect Profile)
QW1 =1.4 nm QW2=2.3 nm QW3=4.0 nm QW4=8.5 nm
10K Photoluminescence Spectra
H.H. Tan et.al.,
- Appl. Phys. Lett.
68, 2401 (1996).
Energy Shifts vs. Proton Dose
900oC, 30 sec QW1 =1.4 nm QW2=2.3 nm QW3=4.0 nm QW4=8.5 nm H.H. Tan et.al.,
- Appl. Phys. Lett.
68, 2401 (1996).
600 650 700 750 800 850 0.86 0.87 0.88 0.89 0.90
InGaAs/InP QW, InP cap InGaAs/InP QW, InGaAs cap
Peak Energy (eV) Annealing Temperature (
- C)
InP 250nm InP 200nm InGaAs 50nm InGaAs QW InP buffer InP buffer
Thermal Stability of InP/InGaAs QWs with InP and InGaAs
10
12
10
13
10
14
10 20 30 40 50 60 70 80 90 InP capped 25
- C
200
- C
Energy Shift (meV) Dose (cm
- 2)
10
12
10
13
10
14
5 10 15 20 25 30 InGaAs capped 25
- C
200
- C
Energy Shift (meV) Dose (cm
- 2)
Implantation Dose (20 keV P) and Temperature Dependence of Energy Shifts in InP/InGaAs QWs
700oC, 60 sec
- C. Carmody, J. Appl. Phys. 93, 4468 (2003)
200 250 300 350 400 450
1x10
3
2x10
3
3x10
3
1 x 10
14 cm
- 2 implanted at 200
- C
1 x 10
14 cm
- 2 implanted
at room temperature random InP cap unimplanted Normalised Yield
Channel
200 250 300 350 400 450 1x10
3
2x10
3
3x10
3
1 x 10
14 cm
- 2 implanted at
room temperature and 200
- C
random InGaAs cap unimplanted Normalised Yield Channel
Damage Accumulation in InP and InGaAs
InP Cap InGaAs Cap
Tuning the Emission Wavelength of GRINSCH Quantum Well Lasers GaAs/AlGaAs QW Lasers
un-implanted dose A dose B
λ1 λ2 λ3
Tuning the wavelength of QW lasers
p-type contact Oxide isolation QW Substrate n-type contact
dose A < dose B λ1 > λ2 > λ3
GRI NSCH QW Laser and 220 keV Proton Defect Profile
Lasing Spectra and L-I Characteristics of GaAs/AlGaAs QW Lasers
(900oC, 60 sec) H.H. Tan and C. Jagadish, Appl. Phys. Lett. 71, 2680 (1997).
Multi-Step Implantation Scheme for Improved GaAs/AlGaAs QW Laser Performance
Tuning the Detection Wavelength of Quantum Well Infrared Photodetectors (QWIPs)
Quantum Well I nfrared Photodetectors
after λ´
1
λ´2 before λ1
Quantum well intermixing
SI-Substrate 1.3 µm GaAs (Si: 2×1018cm-3)
50 nm Al0.3 Ga0.7As barrier 50 nm Al0.3 Ga0.7As barrier 50 nm Al0.3 Ga0.7As barrier 50 nm Al0.3 Ga0.7As barrier
2 µm GaAs cap (Si: 2×1018cm-3) ×48
4.5 nm GaAs (Si: 2×1018cm-3)
0.5 µm AlAs buffer
QWIP structure Grown by MBE
2 4 6 8 10 12 14 0.0 0.5 1.0 1.5 2.0
Bottom contact+substrate Top contact QWs
Displacement Density (a. u.) Implantation Depth (µm)
Defect distribution profile of 0.9 MeV Protons
Metal contact Multi-QWs Metal contact Bottom contact Substrate
un-implanted dose A dose B
λ1 λ2 λ3 λ1 < λ2 < λ3 dose A < dose B
Top contact
Tuning the wavelength of QWIP
4 6 8 10 12 14 0.0 0.2 0.4 0.6 0.8 1.0 1.2 un-implanted 1×1016 cm-2 2×1016 cm-2 3×1016 cm-2 4×1016 cm-2
Photoresponse (a. u.) Wavelength (µm)
QWIP spectral response
950oC, 30 sec M.B. Johnston et.al,
- Appl. Phys. Lett.
75, 923 (1999).
- L. Fu et al,
- Appl. Phys. Lett.
78, 10 (2001).
- 3
- 2
- 1
1 2 3 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.10 0.11 0.12 0.13
3×1016 cm-2 1×1016 cm-2 2×1016 cm-2 reference
Responsivity (mA/W) Bias (V)
Responsivity
- L. Fu et al, Appl. Phys.
- Lett. 78, 10 (2001).
- 6
- 4
- 2
2 4 6 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14
1e16/RTA/1e16/RTA 2e16/RTA Response (mA/W) Bias (V) reference
Responsivity
- L. Fu et al, Infrared Phys. & Technol. 42, 171 (2001).
One-step implant-anneal sequence: 0.9 MeV 2×1016 cm-2 / 950ºC 30 s Two-step implant-anneal sequence: 0.9 MeV 1×1016 cm-2 / 950ºC 30 s /0.9 MeV 1×1016 cm-2 / 950ºC 30 s
4 5 6 7 8 9 10 11 12 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
∆λ/λ = 15.95% ∆λ/λ = 17.43% un-implanted reference
- ne-step implant-annealed sample
two-step implant-annealed sample
Photoresponse (a. u.) Wavelength (µm)
Relative spectral response
- L. Fu et al, Infrared
- Phys. & Technol.
42, 171 (2001).
I mpurity Free Vacancy Disordering
Silicon dioxide acts as a sink for Ga out-diffusion
(i) Creation of Ga vacancies, (ii) Diffusion of Ga Vacancies
QW Dielectric Film (Silicon Dioxide) Ga Vacancy Ga Atom
Maintains Good Crystal Quality Low Concentration of Residual Defects Low Concentration of Electrically Active Defects Relatively Simple Technique and No Residual I mpurities in the Active Regions
Why I mpurity Free Vacancy Disordering?
Experimental conditions
- Spin-on glass (un-doped and Ga-doped): 3000
rpm for 30 s, baking at 400°C for 15 min
- SiO2: Plasma enhanced chemical vapour
deposition (PECVD)
- TiO2: E-beam evaporation
- RTA: 700 ºC to 900ºC for 30 s
- Low temperature photoluminescence
IFVD using doped spin-on layers GaAs/AlGaAs 2 QW structure
un-doped
Wavelength (nm)
P-doped
PL intensity (a. u)
650 675 700 725 750 775
Ga-doped
- L. Fu et al., Appl. Phys. Lett.
7, 1171 (2002).
- L. Fu et. Al., Appl. Phys. Lett.
76, 837-839 (2000).
Substrate
SiO2
QWs
IFVD using doped spin-on layers InGaAs/AlGaAs structure
wavelength (nm)
un-doped
860 880 900 920 940 960
Ga-doped
PL Intensity (a. u.)
P-doped
- L. Fu et. al.,
- J. Appl. Phys.
92, 3579 (2002)
I ntegration of a Waveguide and a Laser Diode Using I FVD
Lateral waveguiding
P++ GaAs contact layer DQW active layer 1.9 µm 0.1 µm Ridge in the waveguide part Ridge in the gain part waveguide active (gain)
La =0.87 mm (laser diode) Lp = 2.5 mm (waveguide)
- M. Buda et.al.,
- J. Electrochem. Soc.
150, G481 (2003).
Quantum Dot Photonic I ntegrated Circuits
Why Quantum Dots?
Three Dimensional carrier confinement Leads to Atom Like Density of States Low Threshold Current Lasers High Quantum Efficiency High Thermal Stability (To) Lasers Lasers operating at 1.3 & 1.55 um on GaAs (VCSELs) Normal I ncidence Operation of QDI Ps
Self Assembled Growth of Quantum Dots
Layer by Layer Growth Lattice matched Systems, e.g. AlGaAs on GaAs Direct I sland Growth Large lattice mismatch Very High I nterfacial Energy, e.g GaN on Saphire Layer by Layer followed by I sland Nucleation Dissimilar Lattice Spacing, Low I nterfacial Energy e.g. I nAs on GaAs
Volmer-Weber Growth Mode Stranski-Krastanow Growth Mode Frank- van der Merwe Growth Mode
Quantum Dot Intermixing
- Large surface area to volume ratio
- Non-uniform composition profile
- Large strain field around the dots
Growth details
- Aixtron 200/4 MOCVD
reactor
– rotation – IR lamps
- TMGa, TMIn, AsH3 &
PH3
- Single layer
In0.5Ga0.5As dots
- Dot growth ~500-550°C
- GaAs cap at 650°C
Characterization
- AFM and PL
GaAs GaAs
300nm 300nm 5- 6ML 50% InGaAs
S-I GaAs Substrate
Amount of Material
- Increase material:
– Density increases until saturation – Size decreases
6.5ML 5.8ML 5ML 4ML
Temperature - AFM
500°C 520°C 550°C
Height increases with increasing temperature Less incoherent dots with increasing temperature
Experimental Details
- Samples annealed for 30sec
– Proximity capped RTA – Anneals at 700, 800, 850 and 900°C
Spin on Glass, PECVD SiO2, E-Beam TiO2
- Photoluminescence
– 10K – Cooled Ge detector – Argon-ion laser at 514.5nm
Thermal stability of single layer QDs
880 920 960 1000 1040 1080 1120 1160 0.0 0.2 0.4 0.6 0.8 1.0 1.2
Wavelength (nm) Intensity (a.u.)
as-grown 750
OC
800
OC
850
OC
900
OC
12K Photoluminescence Spectra
I VFD of single layer QDs Spin on Glass (SOG)
920 960 1000 1040 1080 1120 1160 0.0 0.2 0.4 0.6 0.8 1.0
Wavelength (nm) Intensity (a.u.)
as-grown annealed ref undoped SOG ga SOG ti SOG
Suppression of thermal interdiffusion by TiO2 as function of RTA temperature
20 40 60 80 100 120 140 160 (a)
RTA TiO 2+RTA
PL energy shift (meV)
680 700 720 740 760 780 800 820 840 860 0.4 0.5 0.6 0.7 0.8 0.9 1.0 (b)
FWHM of annealed FWHM of as-grown Annealing temperature (ºC)
Substrate
TiO2
- L. Fu et.al.,
- Appl. Phys. Lett.
82, 2613 (2003)
Suppression of interdiffusion using bi- layer SiO2 + TiO2
900 950 1000 1050 1100 1150 1200 0.0 0.2 0.4 0.6 0.8 1.0 1.2
as-grown RTA only SiO2 + RTA SiO2 + TiO2 + RTA
PL intensity (Normalized) Wavelength (nm)
E1 E2
Substrate
SiO2 TiO2 RTA: 850ºC 30 s
Suppression of interdiffusion using bi- layer SiO2 + TiO2 as a function of SiO2 thickness
20 40 60 80 100 120 140
- 60
- 40
- 20
20 40 60 80 100 120
E1 E2
PL energy shifts (meV) Thickness of SiO2 (nm)
- L. Fu et al, Appl. Phys. Lett. 82, 2613 (2003).
Thermal expansion coefficient
Material α (ºC-1) GaAs 6.86 × 10-6 SiO2 0.52 × 10-6 TiO2 8.19 × 10-6
Thermal stress effect
- Compressive stress on GaAs
surface ⇒ favourable for VGa diffusion ⇒ enhanced interdiffusion
QDs
Substrate
tensile compressive SiO2
Substrate
compressive tensile TiO2
Tensile stress on GaAs surface ⇒ unfavourable for VGa diffusion ⇒ inhibited interdiffusion Compressive stress on GaAs surface ⇒ favourable for VGa diffusion ⇒ enhanced interdiffusion
Stacked Dots
10-150A 300A
GaP la ye r InGaAs dots GaAs substra te 2000A GaAs c a p
Growth of stacked layers
Top layer of 3- layer stack. With smoothing Density ~4x1010/cm2 Single Layer Density ~4x1010/cm2 Top layer of 3- layer stack. Original conditions Dots smaller than for single layer. Density ~2x1010/cm2 Top layer of 3-layer stack. Low V/III, with GaP layers. Dots slightly flatter. Density ~3x1010/cm2
I FVD of stacked QDs
920 960 1000 1040 1080 1120 1160 0.0 0.2 0.4 0.6 0.8 1.0
Intensity (a.u.) Intensity (a.u.) Wavelength (nm) Stacked layer
0.0 0.2 0.4 0.6 0.8 1.0
Single Layer as-grown
- ann. ref
SiO2 SiO2/TiO2 TiO2
Implantation
(Single QD Layers)
900 950 1000 1050 1100 1150 1200 0.0 0.2 0.4 0.6 0.8 1.0
Intensity (a.u.) Wavelength (nm)
as-grown annealed 750
OC, 30s
5x10
13H/cm 2
1x10
14H/cm 2
1x10
15H/cm 2
900 950 1000 1050 1100 1150 1200 0.0 0.2 0.4 0.6 0.8 1.0
Wavelength (nm) Intensity (a.u.)
as-grown annealed 1x10
11 As/cm 2
5x10
11 As/cm 2
1x10
12 As/cm 2
5x10
12 As/cm 2
Protons Arsenic Ions
- P. Lever, Appl. Phys. Lett. 82, 2053 (2003)
Stacked QDs -H implant
1000 1020 1040 1060 1080 1100 1120 1140 1160 1180 1200 0.0 0.2 0.4 0.6 0.8 1.0 1.2
Wavelength (nm) Intensity (a.u.)
as-grown annealed 750
OC, 30s
5x10
13H/cm 2
1x10
14H/cm 2
Energy Shifts – As Ion Dose, Annealing Temperature (Single Layer of QDs)
10
11
10
12
20 40 60 80 100 120 140
Dose (As/cm
2)
Energy Shift (meV)
annealed at 700
OC, 30s
annealed at 750
OC, 30s
annealed at 800
OC, 30s
Energy Shifts vs Implant Temperature (Protons and Arsenic Ions)
20 40 60 80 100 120 140 160 180 200 220 20 40 60 80 100 120 140
Implant Temperature (
OC)
Energy Shift (meV)
5x10
11 As/cm 2
1x10
12 As/cm 2
5x10
14 H/cm 2
1x10
15 H/cm 2
Energy Shift and PL Intensity Vs. Displacement Density
10
19
10
20
10
21
20 40 60 80 100 120 140
Displacements (cm
- 3)
Energy Shift (meV)
H As
10
19
10
20
10
21
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Displacements (cm
- 3)
Intensity (norm)
H As
Summary
Quantum Well and Quantum Dot I ntermixing Techniques are promising for Optoelectronic Device I ntegration Understanding defect generation, diffusion and annihilation processes are important for achieving QWI and QDI Dopant diffusion issues need to be taken into consideration for Device Structures such as Lasers and Photodetectors
Acknowledgements
Australian Research Council – Funding
Past and Present Members of Semiconductor Optoelectronics And Nanotechnology Group, ANU (H.H.Tan, L. Fu, P. Lever, M. Buda, P.N.K. Deenapanray, J. Wong-Leung, Q. Gao, C. Carmody, R.M. Cohen,
- A. Clark, G. Li, M.I . Cohen, S. Yuan, Y. Kim, K. Stewart,
- P. Gareso, A. Allerman…)
Many Collaborators in Australia (Mike Gal, M.B. Johnston. P.Burke, L.V. Dao, P.Reece, B.Q. Sun, David Cockayne. Zou Jin) and Overseas