Transcript Anthony Krier: Dilute Nitrides for the Mid-infrared

Dilute Nitrides for the Mid-infrared
A. Krier, M. Kesaria, Q. Zhuang
Physics Department Lancaster University, UK
A .V. Velichko, O.Marakovsky, D. Paola, A. Patane
School of Physics and Astronomy, University of Nottingham, UK
A. Polimeni, S. Birindelli, M. Capizzi
Physics Department, CNISM, Sapienza Università di Roma, Italy
PROMIS Summer School
Montpellier 2016
Outline
Motivation
MBE growth on InAs and GaAs
Structural and optical properties
Annealing and Hydrogenation
Devices
N
Motivation
-1
Wavenumber ( cm )
2000
100






2800
3200
Gas sensors - optical absorption;
CH4, CO2, CO
Industrial process control
Spectroscopy
Thermal imaging
Bio-medical diagnostics
Free space optical communications
(3-5 μm)
Military - infrared countermeasures
Transimission ( % )

2400
50
CO
NO2
0
5
HCN
CO2
4.5
HCl
HCN
NO2
CH4
CO
CO2
HCl
4
3.5
Wavelength ( m )
CH4
3
Principal gas absorptions in the mid-infrared
For these applications we need LEDs, lasers and detectors operating
at Room Temperature
Mid-infrared - main challenges
imbalance in the DOS of InAs
Auger recombination (CHSH)
CB
Inter-valence band absorption (IVBA)
Inadequate electrical confinement
-small band offsets
- No SI substrates
InAsN dilute nitride alloys offer some
possibilities for improvement??
1
1’
2’
Eg
HH
2
LH
Addition of N :
- flexible wavelength tailoring
without complex growth
- de-tuning of CHSH resonance
Auger suppression
Δ0
Band anti-crossing
Extended-localized state interaction
E (k ) 
An empirical model
E (k )  ECB (k )
V
V
E ( k )  EN
Anticrossing/repulsion between conduction-band edge and localized states
2
the
band gap
EN  ECB (k )
(
k
)
 EN  Edecreases

2
CB
E ( k ) 
introduces

  Vat low k-value in the CB
minigap(s)
2
2


2.00
GaAsN
E (k) (eV)
E+
1.80
EN
ECB
1.60
E1.40
-5.00
0.00
8
5.00
-1
k (10 m )
W. Shan et al., Phys. Rev. Lett. 82, 1221 (1999)
Dilute nitrides
D. Sentosa, X. Tang,a, and S.J.
Chua, Eur. Phys. J. Appl. Phys.
40, 247–251 (2007)
InAs
N introduces tensile strain (on InAs or GaAs)
disorder and strong bowing
Harris, J. S. Semiconductor Science and Technology 17, 880 (2002)
InN
N
MBE Growth on InAs and on GaAs
V80 Molecular Beam Epitaxy (VG)
with RF Plasma Nitrogen source, As and Sb valved cracker cells (EPI)
Ga, In, Al and dopants GaTe and Be
Sample TG
A0276
A0282
A0285
A0299
A0300
485
420
442
376
450
Flux - As
Flux - N2
Plasma
Power
N Content
6.6x10-6
2.2x10-6
2.2x10-6
2.2x10-6
2.8x10-6
n/a
6.12x10-7
6.12x10-7
6.3x10-7
5.0x10-7
n/a
160
160
160
160
n/a
0.6
0.2
1.0
0.4
%
Large parameter space for InAsN
InAsN successfully grown on InAs with N < 2% and PL observed out to 4.5 µm
For growth on GaAs
Optimum growth at substrate temperatures between 4000C- 4400C
Nitrogen plasma setting fixed at 160 W with flux of 5x10-7 mbar
Growth rate of ~1µm per hour
InAs control sample was grown under the same conditions
X-ray diffraction
2 different layer peaks
obtained - 2 dominant N
compositions
Plastic relaxation
-Vertical and horizontal lattice
deformations obtained
-Gives relaxed lattice const.
and plastic deformation R
Layers with N< 1.2% are
pseudomorphic
Bragg maps narrow in qII
N > 1.2% more diffuse
scattering from misfit
dislocations & defects
Onset of plastic relaxation at
N~ 1.4%
asymmetrical (224) reflections obtained for all samples
N=0.83% - tail indicates vertical N
composition gradient
N=0.34% - thickness fringes – good interface
quality
Growth rate decreases with increasing N
SIMS and TEM analysis
Sample : A0299 InAsN 1% N
1.0E+07
Intensity (cs-1)
1.0E+06
1.0E+05
Ga
As
InAs/GaAs
In
200 nm
1.0E+04
1.0E+03
N
1.0E+02
1.0E+01
InAsN(1%) /GaAs
1.0E+00
0.00
0.50
1.00
1.50
2.00
200 nm
Depth (microns)
N is uniform
No evidence of unintentional impurities (C, O etc.) as-grown InAsN is of high purity
Analysis of secondary ion peaks from CsAsN+ enables accurate N determination
-comparison with XRD data – N content is ~5% larger than determined from XRD
Significant incorporation of non-substitutional N
Higher dislocation density in InAsN – but obtain increase in PL
Localisation, non-uniform PL emission from regions around dislocations?
Raman spectroscopy
Weak InAs modes at 405 and 425 cm−1 and
2nd order InAs optical modes at 435, 450, 460 and 480cm−1
N related
features
Additional N related features at 402, 415, 428 and 443 cm−1
(Wagner et al. N ~ 1.2 %)
2nd order InAs modes
NAS
As -N
N-N
difference spectrum of highest N – lowest N content
443 cm−1 feature - also detected in FTIR
NAs LVM from substitutional 14NAs
402 cm−1 and 415 cm−1 peaks from non-substitutional N-N
or As-N split interstitials, (N antisites or interstitial N) rather
than N-In-N complexes
and As -N produce deviations from Vegard’s law
(Calculations predict N-N split interstitial at 419 cm−1
but also predict that the As-N split interstitial lies well
above the LVM in GaAsN)
Ibanez et al, JAP (2010)
Electrical properties InAsN on GaAs
1
10
T = 293 K
5
3
-1
Ga(AsN)
2
impurity scattering
-2
10
0.0
0.4
0.8
1.2
1.6
N (%)
2
-1 -1
Mobility (m V s )
77K
InAsN
1
0.1
GaAsN
0.01
0.0
0.1
0.2
0.3
0.4
N-content (%)
0.5
0.6
A. Patanè et al Appl. Phys Lett. 93, 25106 (2008)
-3
0.5
x (%)
16
10
1.0
1000 nm n-type InAs(N)
10
T (K)
10
0.0
1
0.4%
0.6%
1.0%
1
T = 293K
2
0.2%
10
nH (cm )
2 -1 -1
 (m V s )
x=0%
4
-1 -1
Phonon scattering
0
10
H (m V s )
In(AsN)
17
10
100
0
Semi-insulating GaAs substrate
N reduces electron mobility
µ is limited by electron scattering by N-atoms, -pairs and clusters
Model for GaAsN predicts a strong reduction of the mobility and
electron mean free path due to the N-levels
Weak dependence of µ on N-content compared to GaAsN due to
the proximity of the N-related states to the CBE
Impurity scattering dominant at high N
Residual carrier conc. increases for N >0.4%
N incorporation introduces native donor states
InAsN - Cyclotron Resonance
0.5
EF= 10 meV
18
2.0%
0.4
EF
-3
n (cm )
10
1.0%
 (eV)
80 meV
40 meV
20 meV
N = 0%
0.3
17
10
-4x10
6
0
-1
k (cm )
4x10
6
Pinning of the Fermi level
16
10 0.0
0.5
1.0
N (%)
1.5
2.0
The increase of electron density with
increasing N indicates a pinning of the Fermi
level and implies a substantial density of
native donor states
O. Drachenko et al. APL 98, 162109 (2011)
Photoluminescence InAsN on InAs
Incorporation of small amounts of N into III-V’s causes
conduction band anti-crossing leading to reduction in
band gap
Good agreement with band anti-crossing model
(60 meV per 1%N)
Long low energy tail appears - localisation
CMN = 2.5 eV at 4 K
caused by uneven nitrogen distribution- composition
fluctuations or point defects
0.355
Band Gap Energy (eV)
0.350
0.345
0.340
0.335
0.330
-4 2
Eg=0.353-[1.1x10 T /(T+100)]
0.325
0
50
100
150
200
Temperature (K)
250
300
6.5
Photoluminescence Lineshape
6.0
5.5
4K
20K
40K
60K
80K
100K
120K
150K
180K
210K
240K
270K
300K
5.0
4.5
Intensity
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.26 0.28 0.30 0.32 0.34 0.36 0.38 0.40
Photon Energy (eV)
PL is Gaussian at low T
As T increases becomes asymmetric with high energy tail
extends well above Eg
Conduction Band
Lineshape - 2 effects
Localization at low T
Free carrier emission at high T
J. Appl. Phys. 108, 103504 (2010)
Valence Band
PL analysis temperature dependence
8
4K
20K
40K
60K
80K
100K
120K
140K
160K
190K
220K
250K
280K
300K
A0299
7
Intensity (a.u.)
6
5
4
3
2
CO2
1
0
-1
2.6
2.8
3.0
3.2
3.4
3.6
3.8
4.0
4.2
4.4
4.6
Wavelength (m)
1.0
300K
Intensity (a.u.)
0.8
0.6
CO2
0.4
0.2
0.0
2.6
2.8
3.0
3.2
3.4
3.6
3.8
4.0
Wavelength (m)
InAsN(1%) exhibits very weak temperature quenching ~ 8x
PL emission obtained up to room temperature without annealing
Peak wavelength near 4 µm – appropriate for CO2 detection
4.2
4.4
4.6
InAsN Photoreflectance
Solid lines are fits to
Where, x is the N content
N does not change Δso
InAsN on GaAs
0.6%N
10
Good agreement with band anti-crossing
model
Inclusion of nitrogen improves the peak
intensity InAsN > InAs on GaAs
Photoreflectance shows Δ0 is constant with
increasing N
Activation energy increases with increasing N
content – CHSH Auger detuning
4K PL
0.4%N
8
Intensity (a.u.)
PL obtained from InAsN on GaAs across the
mid-IR spectral range with addition of small
quantities (~ 1%) of nitrogen
1%N
0.2%N
6
improved PL
InAs/GaAs
4
CO2
2
0
2.6
2.8
3.0
3.2
3.4
3.6
3.8
4.0
4.2
4.4
Wavelength (m)
0
0
0.40
Energy (eV)
Normalised PL Intensity (a.u.)
10
0.35
E0
InAsN/InAs
InAsN/GaAs
InAsN/InAs E0
0.30
InAs/InAs
InAs/GaAs
InAsN(0.6%)/GaAs
InAsN(1%)/GaAs
-1
10
-2
10
InAsN/InAs o
0
BAC model
0.0
0.5
1.0
1.5
2.0
Nitrogen content (%)
2.5
3.0
50
100
150
200
Temperature (K)
250
300
4.6
Adding Sb - MBE growth of InAsSbN
InAs
Adding N to
InAs
Conduction band
Adding Sb to
InAs
Eg
Valence band
Increasing N
Increasing Sb
Tensile strain
Compressive strain
N is hard to incorporate
Use Sb to reduce lattice mismatch increase N
incorporation improve quality
Sb acts as surfactant to maintain 2D growth and
reduces point defects - improves PL
Red-shift of emission wavelength
– need less N to reach longer wavelengths
Sb reduces N surface diffusion length - increases N
incorporation ~ 2.5x
Reduction of Sb segregation induced by N increases Sb incorporation ~1.5x
Photoreflectance
Δso > E0 Auger suppression
Advantage of InAsNSb over InAsN
In-plane strain for layers grown on InAs
can be tuned from tensile to compressive
- Tailor polarization in QW to be either TE or TM
Sb increases confinement in valence band
- dominant polarisation is TE (e1-hh1)
0.8
0.8
Spin orbit splitting In InNAs & InAsNSb
E0
E0+SO
SO
0.5
E0+SO
SO
Fit
Fits for InNAs
0.4
0.4
0.3
0.5
0.3
(c) InNAs
0.0
0.5
1.0
1.5
2.0
Nitrogen concentration (%)
Kudrawiec et al. APL 99, 011904 (2011)
Incorporation of Sb increases Δso and
decreases E0
N does not change Δso
SO
0.6
2.6% Sb
4.1% Sb
7.0% Sb
7.3% Sb
Ref.[31]:
E0
4.8% Sb
This work:
E0
Energy (eV)
Energy (eV)
0.6
E0+SO
0.7
0.7
Both Sb and N reduce E0
(d) In(N)AsSb
0.0
0.5
1.0
1.5
2.0
Nitrogen concentration (%)
~ 5 meV per 1% of Sb
~ 60 meV per 1% N
InAsSbN Photoluminescence
Strong PL at room temperature
- good optical quality
Asymmetric shape
Narrow energy gap – free carrier
emission is important
Especially > 100 K
High energy tail extends well above Eg
Latwoska et al, Appl. Phys. Lett 102, 122109 (2013)
Gaussian at low T
PL peak lower than Eg determined from PR
Characteristic S-shape but with weak carrier
localisation
- Stokes shift <10 meV
smaller than for InAsN
Composition fluctuations or point defects reduced
due to surfactant effect of Sb
Photoluminescence curve fitting
Fit using
Includes localized and band-band transitions
A = scaling factor
Ecr = energy of crossover between equations
K = smoothing parameter
σ relates to slope of DOS
Set K = kBT/σ and Ecr = Eg + kBT/σ
n= 0.5 to 2 for momentum conserving non-conserving
transitions
Black arrows – Eg determined from PL fitting
Red arrows – PL peak
Best fit when n=1
Note the difference which increases with T
Latwoska et al, Appl. Phys. Lett 102, 122109 (2013)
Temperature dependence of bandgap
Comparison of change in energy gap with T
InNAsSb 65 meV
whereas 1% N in GaAs reduces T dependence of Eg by 40%
InAs
66 meV
InSb
62 meV
BAC model gives good agreement
T dependence of Eg in InNAsSb is not
sensitive to N due to large separation
between EN and EM (~ 1 eV)
Rapid thermal annealing of InAsN/GaAs
1.0
349
1.0
0.6
Normalised PL Intensity
0.8
PL Intensity (a.u)
PL Intensity (arb. units)
0.8
0.6
0.4
0.2
0.0
400
0.4
500
550
0
As grown
0
RTA 450 C
0
RTA 500 C
0
RTA 550 C
0.2
0.0
0.30
450
Annealing Temperature ( C)
0.35
0.40
0.45
0.50
0.8
As grown
0
RTA 450 C
0
RTA 500 C
0
RTA 550 C
Peak Energy (meV)
1.0
348
347
346
345
344
343
342
341
0.6
340
400
450
500
550
0
Annealing Temperature ( C)
0.4
0.2
0.0
0.25
Energy (eV)
0.30
0.35
0.40
0.45
0.50
Energy (eV)
Rapid thermal annealing (RTA): 30 sec in N2
4K PL enhanced by 25-30 times due to a reduction in SRH recombination
RT PL intensity increased only moderately
RTA causes reduction in point defects but not in Auger recombination
Slight blue shift (10 meV) in PL emission peak energy: due to small decrease in N ~0.1%
Increase in PL efficiency of RTA InAsN/InAs
PL intensity (a.u.)
1
1.6
0.1
0.01
1.2
400
450
500
550
o
Annealing temperature ( C)
0.8
As grown
0
RTA 450 C
0
RTA 500 C
0
RTA 550 C
0.4
0.0
0.35
0.40
0.45
0.50
Energy (eV)
Peak Energy (meV)
Normalised PL Intensity
2.0
PL Intensity (a.u.)
378
1.0
0.8
376
374
372
370
368
366
364
362
0.6
360
400
450
500
0.4
As grown
0
RTA 450 C
0
RTA 500 C
0
RTA 550 C
0.2
0.0
0.35
0.40
0.45
0.50
Energy (eV)
4K PL enhanced by 25-30 times due to reduction in SRH recombination
Slight blue shift (17 meV) in PL emission peak energy: due to decrease in N ~0.15%
Kesaria, M. et. al . Infrared Physics & Technology 2015, 68, 138-142
550
o
Anneling Temperature ( C)
0.55
Passivation of localized states & Temperature Quenching
0
10
0
RTA 500 C
As grown
0.34
InAsN(1%)/GaAs
0.33
Eg(T)=Eg(0) –
InAsN(1%)/GaAs
Normalized PL Intensity (a. u.)
Photoluminescence peak energy (eV)
0.35
𝛼 T2
as-grown
-1
10
InAs/GaAs
-2
10
InAs/GaAs
o
RTA at 550 C
o
RTA at 500 C
o
RTA at 450 C
InAsN/GaAs
(T+β)
0.32
0
50
100
150
200
250
300
0
50
100
150
200
250
300
Temperature (K)
Temperature (K)
Conduction Band
Varshni fit
β =100-110 K
α =5×10-4 (eV/K)
Eg(0)=0.343-0.349 eV
Introduction of N reduces
thermal quenching
Valence Band
RTA reduces localisation and improves N compositional
uniformity
But, RTA increases thermal
quenching rate
Localisation is reduced
H irradiation - Increase in PL efficiency
TH = 250 C
H-dose H0 = 1×1016 cm-2
[N]=0, InAs
PL efficiency first increases with
H dose, reaches a maximum
value
then it saturates and decreases
at higher H doses
The same behavior is observed
for [N]≠0, In(AsN)
Notice that hydrogenation conditions and H doses resulting in maximum increment
in PL efficiency depend on [N]
Hydrogen diffusion and trapping: SIMS
TH = 250 C
[N]=0
[H]=10×1016 cm-2
D diffuses see inset to the InAs/GaAs interface,
where it is trapped and accumulates
[N]=1.1%
[H]=1×1016 cm-2
X-TEM
X-TEM images highlight a non-homogeneous
concentration of threading dislocations at the
InAs/GaAs highly mismatched interface
D diffuses and traps (at N sites)
D does not reach the InAs/GaAs interface
Passivation of localized states
[N]=1.1%
as-grown
0.30
0.28
PL Intensity (arb. un.)
0.32
0
160 K
0.30
0.28
0
- S-shaped of Epeak vs T
100 K
60 K
12 K
0.30
0.32
0.34
0.36
Energy (eV)
100
150
0.38
200
Temperature
(K)
0.34
0.32
As-grown samples:
- High density of localized states
220 K
50
PL Intensity (arb. un.)
Photoluminescence peak energy (eV)
0.34
(a)
250
300
[N]=1.1%
dH=H0
240 K
- PL efficiency increases
160 K
80 K
- localized states are passivated, Epeak
12 K
0.32
0.34
0.36
(b)
0.38
Energy (eV)
50
100
150
200
Temperature (K)
[H] = 1×1016 cm-2
Hydrogenated samples:
280 K
0.30
- RT emission is not achieved
250
300
increases
- a sizable RT emission is achieved
[N] =1.1%
dH = 1H0
273 K
PL intensity (arb. units)
Integrated PL intensity (arb. units)
Thermal recovery of the as-grown PL efficiency
[H] = 1×1016 cm-2
I (T )   I 0i e
573 K
613 K
 vait  e(  Eai / KBT )
i
vai  3400 cm 1
Ta=673 K
733 K
a.g.
0.32
0.34
attempt frequency value
0.36
Energy (eV)
300
[N]=1.1%
400
500
600
700
Annealing temperature (K)
Two activation energies
typical of N-H local modes
a.g.
800
Eai  2.14 eV ; 2.40 eV
two different nonradiative defects are passivated by H
c
InAsN QW lasers on InP
InAsN ridge lasers operating up to 2.6 µm have been demonstrated – grown by gas source MBE
limited by N incorporation and critical thickness
4 QW InAsN/InGaAs on InP (5μs pulse width, 500 Hz repetition rate)
Max. operating temperature 260 K with T0 = 110 K
Decreasing growth temp incorporates more N
….but reduces QW quality
D. K. Shih, H, H. Lin, and Y. H. Lin, IEE Proc. Optoelectronics 150, 253 (2003)
InAsSbN / InAs MQWs
100 nm InAs Capping
Layer
10x InAsNSb /InAs QW
(12x24 nm)
200 nm InAs Buffer
Layer
Growth of the
MQWs calibrated
using the same
growth method of
previously grown
InAsNSb bulk layers
InAs substrate
200 nm InAs Buffer layer grown at 480°C
10x InAsSbN/InAs QW grown at 420°C
• Growth rate of 0.5µm per hour
-6
• Nitrogen plasma setting fixed at 160 W with flux of 6×10 mbar
100 nm InAs Capping Layer grown at 480°C
As flux kept at minimum for growth of InAs layers
∆EV =
102meV
InAs
hh1 = 9meV
hh2 = 36meV
InAs0.92Sb0.08 InAs
InAsSbN/InAs MQW 4K photoluminescence
7
5
4
3
2
Peak Wavelength
tot=3.62m
e - hh1
4K
0.009
=3.68m
0.008
e1-hh1
e - lh1
=3.48m
2.6
2.8
3.0
3.2
3.4
4K
N =1%, Sb 6%
4.38 m Bulk
0.007
e1-lh1
3.6
3.8
4.0
4.2
4.4
4.6
Wavelength (m)
Intensity (a.u.)
Relative Intensity (a.u.)
6
1.8W
1.6W
1.2W
1W
0.8W
0.6W
0.5W
0.4W
0.2W
0.1W
0.06W
0.03W
0.006
3.62 m MQW
0.005
0.004
0.003
0.002
0.001
1
0.000
0
-1
3.0
-0.001
2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 4.2 4.4 4.6 4.8 5.0
3.2
3.4
3.6
3.8
4.0
Wavelength (m)
No blue-shift with excitation power
- Type I QW
3.48 µm
3.62 µm (expt.)
4.2
4.4
Wavelength (m)
Band alignment determined by modification of
InAsSb - Type II alignment with conduction and
valence band offsets of 39 & 82 meV
ADDITION OF N :
• Reduction in overall strain
band gap
Reduction of
• Conduction band further reduced by BAC model
Reduction of 63 meV
50
InAsSbN MQW LED
25mA
50mA
75mA
100mA
150mA
200mA
300 K EL
40
EL Intensity (a.u.)
p-i-n diode containing 10x
InAsSbN QW in active region N =1%, Sb 6%
p InAs
InAsNSb MQW
n InAs
30
C-H
absorption
20
InAs (100) substrate
10
p+-InAs
n+-InAs
0
2500
3000
3500
Wavelength (nm)
Longest wavelength dilute nitride
light emitting device to date
0.07
12
4 K EL
Intensity (a.u.)
0.05
10
Outout Power (W)
30mA
50mA
75mA
100mA
150mA
4K
0.06
8
6
4
0.04
0.03
0.02
InAsSbN e-hh1
InAsSb e-hh1
InAsSb e-hh2
0.01
2
0.00
2.5
0
50
100
150
200
Current (mA)
250
300
3.0
3.5
4.0
4.5
Wavelength (m)
LED output power : 6 µW at 100 mA drive current and internal RT efficiency ~ 1%
4000
4500
Comparison with InAsSb
0.012
Intensity (a.u.)
0.010
4K
20K
40K
60K
80K
100K
120K
140K
160K
190K
220K
250K
280K
300K
InAsSbN MQW LED
N =1%, Sb 6%
0.008
0.006
0.004
0.002
2400
2600
2800
3000
3200
3400
3600
3800
4000
InAsSbN e-hh1
InAsSb e-hh1
InAsSb e-hh2
4200
Wavelength (nm)
Comparison of the temperature dependence
of the EL with that of type II InAsSb/InAs
reveals more intense emission at low
temperature
Improved temperature quenching up to
T~200 K where thermally activated carrier
leakage becomes important and further
increase in the QW band offsets is needed
Increasing the nitrogen content above 0.5%
reduces the band gap sufficiently such that
the energy gap Eo becomes less than Δso
effectively detuning the CHSH Auger
recombination mechanism
InAsSbN MQW p-i-n photodetector
0.1
4K
20K
40K
60K
80K
100K
120K
140K
160K
190K
220K
250K
280K
300K
Current (A)
0.01
1E-3
1E-4
1E-5
EL emission (a.u)
-0.5
1.5
2.0
2.5
3.0
3.5
4.0
0.5
1.0
1.5
10
9
8
7
6
2
0.1
1.0
0.0
Voltage (V)
R0A (cm )
1
1E-6
R0A ~1/n
5
4
2
(R
 1/n
R 0AA)~1/n
3
2
0
4.5
Wavelength (m)
3
4
5
6
7
8
9
10
11
12
13
-1
1000/T (K )
A0363
4.00E+017
3.50E+017
3.00E+017
2.50E+017
17
-2
NA = 8.3x10 cm
-3
2.00E+017
-2
Cut-off λ ~ 4 μm
Ideality factor = 1.6
R0A
T<120 K generation-recombination dominates
T>220K diffusion limited recombination is
dominant
Capacitance at 0V =2.54 nF
Built in potential = 0.19 V
Carrier concentration = 8.3x1017 cm-3
2
C (F )
Photoresponse (a.u.)
10
1.50E+017
1.00E+017
5.00E+016
2
2
1/C =2(Vbi-V)/A qNA
Slope = NA
Vbi=0.19V
x-intercept = Vbi
0.00E+000
-0.20 -0.15 -0.10 -0.05 0.00
0.05
0.10
Voltage (V)
0.15
0.20
0.25
0.30
0.35
In(AsN) Resonant Tunnelling Diodes (RTDs)
A multi-layered p-i-n RTD structure grown by MBE with an InAsN QW layer
embedded between two InAlAs barriers.
0.1µm p+ InAs X1018
0.3µm p+ InAs X1017
10nm InAlAs
10nm InAsN
10nm InAlAs
20 nm InAs
Energy (eV)
20 nm InAs
InAs: n-type
E1
0.2
In(AsN)
0.3µm n+ InAs X1017
0.1µm n+ InAs X1018
In(AlAs)
0.4
n-InAs
0.0
-50
0
z (nm)
Optical mesa diodes and
TO5 headers for electrical
and optical studies
p-InAs
Fermi level
50
Low-T current-voltage characteristics
• Extended negative differential resistance (NDR);
• Ohmic region at low bias voltages;
• Additional resonance features in the differential conductance.
T = 2K
10
dI/dV (mS)
I (mA)
0.5
0.0
-0.5
E1
D
0
0.0
0.0
0.1
V (V)
0.2
0.2
0.3
0.4
Following the NDR from low to room temperature
The NDR is not observed in InAs and is weakly affected by T
15K
30
45
60
75
90
100
110
120
130
140
160
180
200
220
240
260
270
280
296
300
abs(I) (mA)
In(AsN)
4
D
2
0
abs(I) (mA)
InAs
4
2
Two transport processes
Thermal diffusion
Zener tunnelling
through N-related states
(50 meV below CB)
Diffusion
Ea
Tunnelling
0
0.0
0.1
0.2
V (V)
0.3
Magneto-tunnelling Spectroscopy (MTS)
An e- travelling a distance s under the
effect of an electric field F and a
transverse magnetic field B gains
Δky
Bx
Energy   qF  s
e
Wave vector k  qs  B / 
emitter
0D state
The tunnelling current can be expressed as an overlap integral of emitter and 0D
states in k-space:
I  y em( k )y 0 D ( k  k )dk ~ y 0 D ( k )
Mapping |ψ|2
in k-space
2
2
k ~ B
yemitter
y0D
k
Sakai, J.-W. et al Probing the wave function of quantum confined states by resonant magnetotunneling, Phys. Rev. B 48, 5664 (1993)
Vdovin E.E. et al. Imaging the Electron Wave Function in Self-Assembled Quantum Dots. Science 290, 122-124 (2000)
Summary
Successful MBE growth of InAsN directly onto InAs and GaAs substrates with N up to
~ 2%
Behaviour of N in InAs different to N in GaAs
Fermi level pinning and native donor states
PL covers the mid-infrared (2-5 μm) spectral range in good agreement with the BAC
model
Localisation and free carrier effects are important in interpretation of PL spectra
N reduces band gap but has little effect on T sensitivity
Photoreflectance shows N has no effect on Δo , Auger CHSH de-tuning is possible
Addition of Sb increases N incorporation –structural and optical properties
- improved and bright PL obtained from Type I InAsSbN/InAs MQWs
First long wavelength dilute N LED operating at 300 K
good prospects for device applications if electron concentration can be controlled
RTA reduces native point defect concentration & improves material uniformity
4K PL intensity increased by ~ 20x
Elimination of deep fluctuations in conduction band leads to reduced localization in
annealed samples but causes an increase in thermal quenching
Increasing doses of H irradiation of In(AsN) leads to a passivation of localised states
and an increase in PL efficiency ~20x at low H doses
First InAsN resonant tunnelling diodes – MTS to probe localised defect states
41
Acknowledgements
A. Patane
Nottingham University
Transport measurements
R. Beanland & A. Sanchez
University of Warwick
TEM
J. Ibanez
University of Madrid
Raman spectroscopy
R. Kudrawiec
Institute of Physics, Wroclaw
Photoreflectance
O. Drachenko
M. Helm
Helmholtz-Zentrum
Dresden-Rossendorf
Cyclotron resonance studies
M. Schmidbauer
Leibniz-Institute, Berlin
X-ray diffraction measurements
CNISM and Physics Department
Sapienza Università di Roma, Italy
Hydrogenation studies
S. Birindelli, M. De Luca,
A. Polimeni and M. Capizzi
We are grateful for financial support provided from EPSRC (EP/J015849/1) and the EU Marie Skłodowska-Curie ITN
-PROMIS (H2020-MSCA-ITN-2014-641899)