Transcript Корреляционные эффекты в Pb1

Mechanisms of the Persistent
Photoconductivity Quenching in
Pb1-xSnxTe(In)
V.I. Chernichkin, D.E. Dolzhenko,
L.I. Ryabova, D.R. Khokhlov
M.V. Lomonosov Moscow State University
Unusual Impurity States
in Pb1-xSnxTe(In)
and on a Way to the Passive
Terahertz Imager
V.I. Chernichkin, D.E. Dolzhenko,
L.I. Ryabova, D.R. Khokhlov
M.V. Lomonosov Moscow State University
Cooperation

M.V. Lomonosov
Moscow State University




Institute of Applied
Physics, Kishinev,
Moldova


Ludmila Ryabova
Dmitry Dolzhenko
Vladimir Chernichkin
Andrey Nicorici
University of Beer
Sheva, Israel


Vladimir Kasiyan
Zinovy Dashevsky

University of
Regensburg



Sergey Ganichev
Sergey Danilov
A.F. Ioffe PhysicalTechnical Institute, StPetersburg

Vassily Bel’kov
Outline



1.
2.
3.






4.
5.
6.
Introduction
Undoped lead telluride-based alloys.
Effects appearing upon doping.
a)
b)
c)
Fermi level pinning effect.
Persistent photoconductivity.
Theoretical model
Terahertz photoconductivity and local metastable states
Pb1-xSnxTe(In)-based terahertz photodetectors.
Summary.
Spectrum of the electromagnetic
radiation
«Terahertz gap»
Terahertz radiation


In this spectral region both radiophysics
methods (at the long-wavelength side)
and optical methods (at the shortwavelength side) work not well
Consequence: absence of good sources
and sensitive detectors of radiation
Areas of application of the Terahertz
radiation





Monitoring of concentration of heavy
organic molecules
Medical applications (oncology,
stomatology)
Meteorology
Security systems (search and detection
of explosives)
Infrared astronomy
Medical applications
Cancer tissue in theTerahertz and in the visible spectral range
Security systems
A boot with a ceramic knife and a plastic explosive
“Semtex” in its sole
Security systems
A polyethylene box under a 10 cm layer of sand.
Pictures are taken in the Terahertz range
Asteroid danger
Maximum of the blackbody radiation
spectral density
l(mm)=3000/T(K)
Sun: T=6000 K, l=500 nm
Earth: T=300 K, l=10 mm
Asteroids: T=10 K, l=300 mm
u=1 THz – Terahertz range!
1/2
Planck HFI
-17
106
10
105
104
10-18
103
10-19
102
101
10-20
100
1
10
Frequency (THz)
Background photon arrival rate (1/s)
Background Photon Noise NEP (W/Hz )
Terahertz astronomy
Russian Space Missions in
Terahertz and Millimeter Ranges

RADIOASTRON



Test launch – 21 January 2011
Launch scheduled for July 2011
MILLIMETRON




Launch scheduled for 2017-2018
The project is accepted by the Russian Space
Agency
Supported by the German Space Agency
Pending support from the European Space Agency
Proton-M
launcher,
L2 orbit,
4500+2100 kg.
SB – space buster DM, SM – service module, WC – warm cabin, TS&EM – thermal screens
& expanding mast, CC – cold cabin, T – telescope.
The Space Observatory in the
single-dish mode
Telescope:
Primary mirror diameter 12 m, surface RMS accuracy 10 mm, diffraction beam 4’’
and field of view 4.5’ at 1.5 THz.
Bolometer arrays:
wavelength ranges
0.2-0.4 mm, and 0.7-1.4 mm
HPBW beam (at 1.5 THz) 4''
Low resolution spectropolarimeter:
wavelength range
0.02-0.8 mm
spectral resolution
R=3
Medium resolution spectrometers:
wavelength ranges
0.03-0.1 mm, and 0.1-0.8 mm
spectral resolution
R = 1000
High resolution spectrometer:
wavelength ranges
0.05 – 0.3 mm
spectral resolution
R = 106
Bolometric sensitivity: at 1 THz, NEP = 10-19 W(s)0.5, A = 100 m2, R=3 and 1 h
integration
5∙10-9 Jy (1 s)
State of the art sensitive
terahertz detectors




Transition edge sensors
Hot electron bolometers
Ge(Ga) blocked impurity band detectors
Kinetic inductance detectors
Problems (as I see them)




Very low operating temperature
< 150 mK
NEP not better than 4*10-19 W/Hz1/2 in
the lab and not better than
10-17 W/Hz1/2 in real space missions
Quite poor dynamic range
Problems with arrays
Alternative possibility
Doped lead telluride-based alloys
Undoped Lead Telluride-Based
Alloys
PbTe: narrow-gap semiconductor:
 1. Cubic face-centered lattice of the
NaCl type
 2. Direct gap Eg = 190 meV at T = 0 K
at the L-point of the Brillouin zone
 3. High dielectric constant   103.
-2 m .
 4. Small effective masses m  10
e
Pb1-xSnxTe Solid Solutions:
200
100
0
-100
E, meV
Ec
0
0.1
0.2
0.3
0.4
x
Ev
Origin of free carriers:
deviation from stoikhiometry  10-3.
As-grown alloys: n,p  1018-1019 cm-3
Long-term annealing: n,p > 1016 cm-3
Effects Appearing
upon Doping
Fermi Level Pinning Effect.
n
18
-3
7 10 cm
NIn
p
PbTe(In), NIn > Ni
Consequences
1.
Absolute reproducibility of the sample
parameters independently of the growth
technique. Therefore low
production
costs.
2.
Extremely high spatial homogeneity.
3.
High radiation hardness (stable to
hard radiation fluxes up to 1017 cm-2)



Fermi Level Pinning in the
Pb1-xSnxTe(In) Alloys.
200
E, meV
150
100
Ec
EF
n-type metal
50
0
0,0
0,1
0,2
0,3
-50
p-type metal
-100
Ev
semiinsulating state
0,4 x
Persistent Photoconductivity
10
5
10
4
10
3
10
2
10
1
10
0
2
R, Ohm
3
2'
3'
1
-1
4
1'
-2
4'
10
10
0
5
10
15
20
25
100/T, K
-1
Temperature dependence of the sample resistance R
measured in darkness (1-4) and under infrared
illumination (1'-4') in alloys with x = 0.22 (1, 1'), 0.26 (2, 2'),
0.27 (3, 3') and 0.29 (4, 4')
Photoconductivity Kinetics
s
4
 > 10 s
Persistent
photoconductor
Ordinary
photoconductor
radiation
on
radiation
off
t
Long lifetime
of the photoexcited
electrons
is due to a barrier
between local and
extended electron
states –
DX-like
impurity centers.
Shubnikov – de Haas oscillations
induced by illumination
Mixed valence model: picture
Te
Te

-e
3+
In
Te
Te
Te
Te
-e
+
2+
In
Te
Te
Te
Te
-e
Te
In
Te

Model for long-term relaxation
processes
Te
Te
hu
e
In
Te
Te
Free electron
In the conduction band
Bound state
Of one electron
Configurationcoordinate diagram
Etot = Eel + Elat =
= (Ei-)n +2/20
(n = 0,1,2) –
number of localized
electrons
Bound electron,
The lattice is locally deformed


E2 – ground local
state;
E1 – metastable
local state
Photoconductivity kinetics
s
4
 > 10 s
Persistent
photoconductor
Ordinary
photoconductor
radiation
on
radiation
off
t
Fast relaxation
is due to transitions
to the metastable state,
slow relaxation
corresponds to
transitions to the
ground local state
Local metastable states
The metastable states are responsible
for appearance of a range of strong
effects:




Enhanced diamagnetic response up to 1% of ideal
Enhancement of effectic dielectric permittivity up to
105 at TeraHertz illumination
Giant negative magnetoresistance up to 106
Persistent photoconductivity in the terahertz spectral
range
Spectral response

Two approaches


Low-background: sample screened from
the background radiation, low-intensity
sources
High-background: sample is not screened
from the background radiation, highintensity sources
High-background approach





Laser wavelengths:
90, 148, 280, 496 mm
Pulse length: 100 ns
Power in a pulse: up to 30 kW
Sample temperature: 4.2 – 300 K
Samples: single crystalline
Pb0.75Sn0.25Te(In), polycrystalline
PbTe(In) films
Fermi Level Pinning in the
Pb1-xSnxTe(In) Alloys.
200
E, meV
150
100
Ec
EF
n-type metal
50
0
0,0
0,1
0,2
0,3
-50
p-type metal
-100
Ev
semiinsulating state
0,4 x
Photoconductivity kinetics
1.4
l = 280 mm
1.2
1.0
2
hu
4.2 K
RL >> RS
0.8
0.6
0.4
shape of the laser pulse (arb.un.)

s /s*10
RS
RL
0.2
0.0
-0.2
25 K
-0.4
0
1
2
t (ms)
3
4
5
6
Time profile of a laser pulse and photoconductivity kinetics
at different temperatures
Photoconductivity mechanisms


Negative photoconductivity: electron
gas heating, change in electron mobility
Positive photoconductivity: generation
of non-equilibrium electrons from
metastable impurity states, change in
free electron concentration
Dependence of the photoresponse
amplitude on the radiation
wavelength for Pb0.75Sn0.25Te(In)
90mm
6
5
s/s0*10
2
4
3
280mm
2
148mm
1
496mm
0
0
20
40
80
60
100
120
Considerable photoresponse
is observed up the wavelength
of 496 mm
which is
more than two times higher
than the previous record
value of 220 mm
observed for uniaxially stressed
Ge(Ga)
-1
 (cm )
Linear extrapolation of the quantum efficiency to the zero photoresponse
gives the cut-off energy Еred=0!
Kinetics of the terahertz
photoresponse in PbTe(In)
-4
s0=3.3*10 Ohm
0.4
-4
s0=4*10 Ohm
T=4.2 K
-1
-1
l = 90 mm
1,0
ss0
100*ss0
0,5
0.2
0,0
l = 280 m m
-0,5
0
0.0
1
2
t, m s
3
4
stationary state
l = 148 mm after 200 laser pulses pass
-1
0
1
2
t, ms
3
4
Equ
E, meV
100
Equ
EqF
EqF
80
E, meV
60
40
EF
60
20
40
0
20
-20
0
Ec
PbTe(In)
Equ
EqF
Ec
EF
-40
Pb0.75Sn0.25Te(In)
New type of local states in
semiconductors
A new type of semiconductor local
states which are linked not to a definite
position in the energy spectrum, but to
the quasiFermi level position which may
be tuned by photoexcitation.
Low-background approach
s
4
 > 10 s
Persistent
photoconductor
Ordinary
photoconductor
radiation
on
radiation
off
t
Integration
increases
the signal-to noise
ratio
but
It is important
to be able to
quench fast
the persistent
photoconductivity
Quenching of the Persistent
Photoconductivity


1. Thermal quenching: heating to 25
K and cooling down: too slow process.
2. Microwave quenching: application
of microwave pulses to the samples
f = 250 MHz, P = 0.9 W, t = 10 ms
Mechanism of the radiofrequency
quenching: experimental
Illumination at the wavelength 200 mm
We have measured conductivity at the
point 1 (100 ms after the pulse) и 2
(900 ms after the pulse)
50
P, mW
40
30
point 2
t=900 ms
point 1
t=100 ms
20
10
0
0
200
400
600
t, ms
800
1000
Measured values:
s1
s2 (s2-s1)/s1
as a function of
- radiofrequency in a pulse f (70 MHz-3
GHz)
- pulse length t (1-64 ms)
- power in a pulse P (up to 70 mW)
Dependence of the “quenching level”
of the radiofrequency
2,5
T=32 ms
2,0
Quenching is more effective
at low frequencies.
s1, mS
1,5
1,0
P
16mW
32 mW
79 mW
0,5
0,0
0
500
1000 1500 2000 2500 3000 3500
f, MHz
The quenching efficiency rises
with increasing power in a pulse
Dependence of s on the
radiofrequency f
12
15
19
25
20
smS
Too effective quenching
at low frequencies leads
to the photoresponse decrease!
T=32 ms
15
P
16mW
32 mW
79 mW
10
The photoresponse decreases
at high frequencies, too.
5
0
0
500
1000
1500
f, MHz
2000
2500
3000
There exists an optimal
in the radiofrequency
region of quenching
Dependence of the radiofrequency
corresponding to the maximal signal on
the radiofrequency pulse length
1000
As the quenching pulse length
increases, the radiofrequency
corresponding to
the signal maximum
saturates.
f (smax), Mhz
800
600
The saturation level increases
with increasing power in a pulse.
400
P
16mW
32 mW
79 mW
200
0
10
20
T, ms
30
40
50
Dependence of the relative signal
amplitude on the pulse length
smax, /s1, %
40
30
P
16mW
32 mW
79 mW
The relative signal amplitude
may reach 40%!
20
10
0
0
10
20
T, ms
30
40
50
Conclusions of the quenching
features




The thermal mechanism of quenching is
excluded
The mechanism related to the electron gas
heating is likely
As the radiofrequency decreases, the power
in a pulse or the pulse length increase, the
quenching efficiancy rises
At the same time it is easy to destroy the
“photosensitive state” of a sample if the
quenching pulse is “too effective”
Operation of an “integrating”
photodetector
I3
s
Options:
1. Internal modulation
I1 < I2 < I3
Radiation intensity is constant,
registration of the signal using
a lock-in amplifier
I2
at the frequency of quenching
I1
2. External modulation
Modulation of the radiation
intensity, registration of the
signal using a lock-in amplifier
at the frequency of modulation
t
Low-temperature insert
3
4
6
2
1
1
2
6
5
1
2
3
4
5
6
Blackbody
Thermal shield1
Thermal shield2
Thermal shield3
Stop aperture
Sample holder
Internal modulation






Single photodetector operating in the regime of the
periodical accumulation and successive fast
quenching of the photosignal.
operating temperature 4.2 K;
wavelenghth below 1100 mm (defined by the stop
aperture diameter);
area 300*200 mm;
quenching rate 1000 Hz;
lock-in amplifier integration time 1 s (bandwidth
1Hz);
NEP = 8*10-17 W/Hz1/2
Problems

s


t
Light off
Possible thermal
leaks
Measurements with
a filter
Question with
transients
Usual set up
Input 300 K window
Sample
50 mK
cold
finger
Input 1.5 K filter
Blackbody
Chopper
T=300K
Background power 1.4 * 10-13 W
Fluctuations 7.3 * 10-18 W/Hz1/2
Electrical connections
Driving
pulse
generator
U=0.1V
Lock-in
amplifier
RL=1k
input
DC-RF
filter
RF
generator
K=1000
output
Rs
Heater wires
Blackbody
Thermocouple
Oscilloscope
Blackbody temperature
modulation
3,0
T, K
2,5
2,0
1,5
1,0
0
5000
10000
t, ms
15000
Performance at 1.57 K









Single photodetector operating in the regime of the periodical
accumulation and successive fast quenching of the photosignal.
operating temperature 1.57 K;
wavelenghth 350 mm (defined by the filter, Q=4);
area 300*200 mm;
quenching rate 1000 Hz;
blackbody modulation rate 0.3 Hz;
lock-in amplifier integration time 100 s;
Blackbody temperature providing S/N=1 Tbb=2.7 K
NEP ~ 6×10-20 W/Hz1/2 !!!
WOW!!!
BUT
Problems





No control on the signal form
Possible thermal leaks
Possible radiation leaks
Possible influence of the off-band
transmission of the filter
Possible cross-talks of the blackbody
heater and the measurement circuit
Therefore
No firm conclusion yet
Summary


We have observed a new type of semiconductor local
states which are linked not to a definite position in
the energy spectrum, but to the quasiFermi level
position
We have demonstrated
NEP = 6×10-20 W/Hz1/2
for a single photodetector operating in the regime of
the periodical accumulation and successive fast
quenching of the photosignal, with the operating
temperature 1.57 K at the wavelength of 350 μm
HOWEVER
further tests are needed to confirm this
Directions of the future
activities







Measurements of the photon noise
Single photon counting? Why not
Development of the portable readout electronics
Development of linear arrays and full-scale arrays
Development of tunable terahertz filters
Development of a system for passive terahertz vision
in medical applications
Investigation for possibilities of application in space
missions



2-nd International Conference
"Terahertz and Microwave radiation:
Generation, Detection and Applications“
Moscow, June 20-22
Tera2012.phys.msu.ru