Henini M. Handbook of Self Assembled Semiconductor Nanostructures for Novel devices in Photonics and Electronics
Подождите немного. Документ загружается.
Quantum Dot Infrared Photodetectors by Metal-Organic Chemical Vapour Deposition 649
the photoresponse of device A peaked at 6.5 μ m (190 meV) at 77 K and at the bias of 0.4 V
and the half width ( Δ λ / λ
peak
) is only 18%. The narrow photoresponse is a strong indication of a
bound-to-bound transition in the QD energy levels [24] . On the other hand, the photoresponses
from device B and C have peaks at 5.5 μ m (224 meV) and 4.7 μ m (266 meV), respectively. The
half width of device B and C is 46% and 57%, respectively. These broad responses are very differ-
ent from those of device A. They are due to bound-to-continuum transitions [25] . In particular,
looking at device C, we see that the multipeaks appear on the high-energy side above 0.23 eV.
In order to explain this change of the optical transition scheme from device A to device C, we
have used the effective mass embedding method to calculate QD energy levels [26] . The single-
electron Hamiltonian for QDs can be written as:
Hr
mr
Vr()
()
()
2
1
∇∇
*
where the effective mass
mr*( )
is locally varying and
Vr()
is the potential of the electrons deter-
mined by the conduction band offset. The strain effect of quantum dots has been taken into
account in the effective masses and the confi ning potential [27] . We ignored the effect of the
electric fi eld on the energy levels because the electric fi eld is of order 10
4
V/cm and this gives an
energy change of order 5 meV across the height of the QD. The effective masses of InAs and InP
are 0.04 m
0
and 0.077 m
0
respectively, and the electron confi ning potential is 400 meV [28, 29] .
The lens shape and the geometry parameters of the QDs as measured with AFM were inserted
into the calculation. The envelope functions
ψ
n
r()
are the solutions of the Hamiltonian. They
can be obtained by taking a linear superposition of the Bessel and sine functions generated by the
near-cylindrical symmetry of the lens. The number of the basis functions used is 700 [30] .
In order to match the observed energy and satisfy the selection rule we identify the main
photo-transition of device A as the one which links the fi rst excited state to the bound state
which is below the top of the barrier by 24 meV and is expected to lead to a fairly sharp response.
When the height of quantum dots decreases, from the QDs of device A to those of device B,
the confi nement becomes stronger and this pushes the ground state more toward the band edge
of the barrier. According to our calculations, the actual energy level confi guration changes, so
that the most probable photo-transition can change from the bound-to-bound transition to the
bound-to-continuum transition. In device B, the main transition is from the ground state to
the continuum state of the barrier. For device C, the Al
0.48
In
0.52
As layer below the QDs increases
1.0
0.8
0.6
Relative photoresponse (a.u.)
0.4
0.2
0.0
3
45
A
A
B
C
Energy (eV)
0.1 0.2
0.3 0.4 0.5
0.0
0.2
0.4
0.6
0.8
1.0
Relative photoresponse (a.u.)
A: 0.4 V
B: 0.2 V
C: 0.2 V
B
C
T 77 K
678910
11
12
Wavelength (m)
Figure 21.33 Normalized photoresponses of device A, B and C measured by a Fourier transform infrared spectrom-
eter at 77 K under the applied biases. The inset shows the photoresponses as a function of photon energy.
CH021-I046325.indd 649CH021-I046325.indd 649 6/27/2008 5:48:25 PM6/27/2008 5:48:25 PM
650 Handbook of Self Assembled Semiconductor Nanostructures for Novel Devices in Photonics and Electronics
the electron confi ning potential by 300 meV in our calculation [31] , therefore the ground state
moves down slightly compared to device B. In device C, the multiple peaks above 230 meV
are observed at 270, 300, and 340 meV. We believe that those peaks are transitions which link the
ground state of the QDs to the minibands formed by the thin Al
0.48
In
0.52
As layers and InP barriers.
Although the spacing is wider compared to a conventional superlattice [32] , the simple calcu-
lation of the energy levels of ten-period 3 nm Al
0.48
In
0.52
As/40 nm InP produced narrow mini-
bands located above the InP band edge by 45, 70 and 100 meV. The calculations reproduce the
transition energy from the ground state to the band edge which is 232 meV; furthermore, the
multiple peaks in the data agree with the calculated energies linking the ground state to the min-
ibands. But all the transitions are broadened by the inhomogeneous size distribution of the quan-
tum dot. We were able to shift the peak detection wavelengths of QDIPs from 6.5 µm to 4.7 µm
using a combination of the change of the quantum dot growth conditions and the addition of
the heterostructure.
400
196
Device A
251
400
400
219
300
232
Device B
Device C
Figure 21.34 The schematic diagram of calculated energy levels of the quantum dots for device A, B and C. The
arrows show the relevant intersubband transitions corresponding to main peaks of each device.
1000
100
10
1
3 2 101
A
B
C
23
Bias (V)
0.1
Responsivity (mA/W)
Figure 21.35 Peak responsivities as a function of bias at 77 K.
The blackbody responsivities ( R
bb
) for device A, B, and C were calculated by measuring the
photocurrent ( I
p
) with a calibrated blackbody source at 800°C. At T 77 K and a bias of 1.5 V,
we obtained peak responsivities ( R
p
) of 77 and 65 mA/W, respectively, in device B and device C as
shown in Fig. 21.35 . The reason for the decrease in the responsivity of B and C when compared
to device A is, we believe, because the transition is changing from what is a relatively sharp high
oscillator strength bound to bound, to a much more diffuse bound-to-continuum transition.
After the noise currents ( i
n
) were measured at T 77 K using a fast Fourier transform spec-
trum analyser, the peak detectivities ( D *) were calculated from D * R
p
( A · Δ f )
1/2
/ i
n
, where
A 1.375 1 0
3
c m
2
is the illuminated detector area and Δ f 1 Hz is the bandwidth. The
detectivities of QDIPs as a function of bias at T 77 K are shown in Fig. 21.36 . The detectivities
CH021-I046325.indd 650CH021-I046325.indd 650 6/27/2008 5:48:26 PM6/27/2008 5:48:26 PM
Quantum Dot Infrared Photodetectors by Metal-Organic Chemical Vapour Deposition 651
of device B and C, which are MWIR QDIPs, were 1.0 1 0
9
cmHz
1/2
/W at a bias of 0.2 V. The
lower detectivities of both device B and C are due to lower peak responsivity when compared to
the responsivity of device A.
21.4.4 InAs QD/InGaAs QW/AlInAs barrier MWIR QDIP
The device structure is shown in Fig. 21.37 and grown by LP-MOCVD [33] . Different from previ-
ous devices, the growth temperature of the whole device structure is 590°C. First, a 0.5 μ m thick
undoped InP buffer layer followed by a 1.0 μ m thick bottom InP contact layer n -type doped to
n 1 . 5 1 0
18
c m
3
was grown. Then the active region was grown, consisting of 25 stacks of
InAs QD/GaInAs QW layers with 29 nm AlInAs barrier layers. The 3.5 nm GaInAs QW layer on
top of each QD layer had a doping level of n 1 1 0
18
c m
3
. Finally, we grew a 0.5 μ m thick
top InP contact layer doped to n 1 . 5 1 0
18
c m
3
. The InAs QDs were grown on the AlInAs
barrier layers via self-assembly with the nominal QD growth rate of 0.5 ML/s and the growth
time of 3.6 s. An array of 400 4 0 0 μ m
2
detector mesas was fabricated.
10
10
10
9
10
8
10
7
Bias (V)
device A
device B
device C
3 2 10
Peak detectivity (cmHz1/2/W)
123
T 77 K
Figure 21.36 Peak detectivities as a function of bias at 77 K.
SI–InP substrate
0.5 m–InP buffer
1 m n-InP contact (n = 1.5 10
18
cm
3
)
29 nm AlInAs barrier
3.5 nm GaInAs QW
InAs QD
0.5 m n-InP contact (n 1.5 10
18
cm
3
)
25
Figure 21.37 Schematic illustration of the device structure grown with low-pressure metal-organic chemical
vapour deposition.
Spectral responses at several temperatures and applied biases were observed by FTIR in the nor-
mal incidence confi guration without any optical coupling structures ( Fig. 21.38 ). In this device
structure, both the InAs QD layers and GaInAs QW layers are involved in the infrared absorption
CH021-I046325.indd 651CH021-I046325.indd 651 6/27/2008 5:48:27 PM6/27/2008 5:48:27 PM
652 Handbook of Self Assembled Semiconductor Nanostructures for Novel Devices in Photonics and Electronics
process. The coupling of QDs and QWs has been used in other QDIP device structures, such as
dot-in-a-well (DWELL) [34] , where the intersubband transition occurs between the hybrid states
of the quantum dot and the quantum well. In our device structure, we believe the initial state is
not necessarily from a localized “ pure ” quantum dot state but from a delocalized “ mixed ” state of
the quantum well and the quantum dot as shown in the inset of Fig. 21.38a . At an applied bias of
1 V, there are two peaks, around 3.2 μ m and 4.1 μ m, as shown in Fig. 21.38a . The intensity of
the peak around 3.2 μ m does not increase signifi cantly as the temperature increases. The peak
around 3.2 μ m comes from a bound-to-continuum transition where the electrons are photo-
excited from the ground state to a continuum state as depicted in the inset of Fig. 21.38a . This
is the reason why the increase of the temperature does not improve the photoresponse around
3.2 μ m. On the other hand, the photoresponse around 4.1 μ m increases signifi cantly with the
temperature because it comes from a bound-to-bound transition in the InAs QD/GaInAs QW
hybrid states and thus the temperature can help the photo-excited electrons escape to the contin-
uum as depicted in the inset of Fig. 21.38a . For all temperatures except room temperature, at an
applied bias of 5 V ( 2 V for room temperature), the peak around 4.1 μ m was dominant in the
spectral response, as shown in Fig. 21.38b. The strong sensitivity to the applied bias is another
indicator that the transition of the photo-excited electrons takes place between bound states of
the QD/QW hybrid.
The peak responsivity ( R
p
) was measured as a function of bias and temperature as shown in
Fig. 21.39(a) . The responsivity increased with temperature from 120 K to 200 K and started
decreasing above 200 K. The peak responsivity was measured to be 822 mA/W at 150 K and
2.5 3.5 4.0 4.5 5.0 5.5
0.00
0.05
0.10
0.15
0.20
0.25
0.30
Photoresponse (a.u.)
Wavelength (m)
Bias 1V
77 K
120 K
150 K
200 K
240 K
~0.385
eV
InAsQD
GaInAs
AllnAs
~0.305 eV
InAsQD
GaInAs
AllnAs
3.0
(a)
3.0 3.5 4.0 4.5 5.0 5.5 6.0
0
5
10
15
20
25
30
35
40
Photoresponse (a.u.)
Wavelen
g
th (m)
77 K (5V)
120 K (5V)
180 K (5 V)
240 K (5V)
RT (2V)
2.5 3.0 3.5 4.0 4.5 5.0 5.5
0
1
2
3
Room
Temperature
Photoresponse (a.u.)
Wavelength (m)
2V
1V
(b)
Figure 21.38 (a) Photoresponses at different temperatures for 1 V bias; (b) photoresponses at different tempera-
tures for 5 V applied bias and 2 V for room temperature (RT). The inset shows the photoresponses measured at
RT for various biases.
CH021-I046325.indd 652CH021-I046325.indd 652 6/27/2008 5:48:27 PM6/27/2008 5:48:27 PM
Quantum Dot Infrared Photodetectors by Metal-Organic Chemical Vapour Deposition 653
5 V. In QDIPs or QWIPs, the photocurrent can increase or decrease with the temperature
depending on whether the relaxation to the lower state or the escape to the continuum state is
favourable. Above a certain temperature, the adverse thermal increase of the relaxation of the
photo-excited electrons back to the lower state dominates any improvement in escape [35] . In
our system, that turnover is believed to take place at around 200 K, above which the responsivity
starts decreasing with increasing temperature.
The dark current density of this device was measured as functions of bias and temperature ( Fig.
21.39b ). A remarkably low dark current density was obtained in this device. At 200 K and 5 V,
the dark current density was measured at 163 mA/cm
2
. High dark current usually limits the capa-
bility for high temperature operation in photoconductors. Therefore, it is crucial to achieve a low
dark current with a reasonable photocurrent at high temperature. In QDIPs, low dark currents can
be engineered by introducing a current blocking layer [36] . But this current blocking layer will also
decrease the photocurrent because the dark current and photocurrent follow the same transport
path, as shown previously in section 21.4.2. In our device, the QD layers decrease the dark current
without signifi cantly compromising the photocurrent. We think the InAs QD layers act as mobility
traps for the dark carriers, but do not seriously affect the escape of the photo-excited carriers.
The specifi c detectivity ( D *) was calculated and is shown in Fig. 21.40 . The maximum D *
of 2.8 1 0
11
c m H z
1/2
/W was measured at 120 K. The room temperature detectivity was
6 1 0
7
c m H z
1/2
/W.
The quantum effi ciency η can be obtained from the relation η R
p
h ν / qg where hv is the incom-
ing photon energy, q is the charge of the carrier, and g is the photoconductive gain. As a good
approximation, the noise gain can be used instead of the photoconductive gain [37] . A very high
quantum effi ciency of 35% was obtained in this device for normal incidence. This high quantum
5 4 3 2 101234
5
1
10
100
1000
Peak responsivity (mA/W)
Bias (V)
77 K
120 K
150 K
200 K
225 K
300 K
65432101234
65
1 10
11
1 10
10
1 10
9
1 10
8
1 10
7
1 10
6
1 10
5
1 10
4
1 10
3
1 10
2
1 10
1
1 10
0
1 10
1
225 K
300 K
200 K
180 K
150 K
120 K
77 K
Dark current density (A/cm
2
)
Bias (V)
(b)
(b)
(a)
Figure 21.39 (a) Peak responsivity at different temperatures as a function of applied bias; (b) Dark current density
at different temperatures as a function of applied bias.
CH021-I046325.indd 653CH021-I046325.indd 653 6/27/2008 5:48:28 PM6/27/2008 5:48:28 PM
654 Handbook of Self Assembled Semiconductor Nanostructures for Novel Devices in Photonics and Electronics
effi ciency might be due to the high oscillator strength for the normal incident light and a higher
number of carriers available for the absorption compared to conventional QDIPs where the
number of photoactive carriers is limited by the number of QDs.
21.4.5 Summary
In this section, we presented several QDIP devices based on our InAs QD grown by MOCVD. Our
fi rst device shows a detection wavelength of 6.4 μ m. Its detectivity is 1.0 10
10
cmHz
1/2
/W at
a bias of 0.4 V at T 77 K. This was the fi rst report of D * of an InP-based QDIP by any growth
method. Next, a device structure with an AlInAs current blocking layer is shown. The dark cur-
rent and noise current is substantially reduced although D * is almost the same. The detection
wavelength of our QDIP is tuned by dot engineering. A series of devices with detection wave-
lengths of 6.5, 5.5 and 4.7 μ m are demonstrated. The phototransition scheme is explained by our
theoretical model. Finally, we show a high-performance InAs mid-infrared QDIP, which operates
up to room temperature. The peak detection wavelength is 4.1 μ m. The peak responsivity and the
specifi c detectivity at 120 K were 667 mA/W and 2.8 1 0
11
c m H z
1/2
/W, respectively. Low dark
current density and a high quantum effi ciency of 35% are obtained in this device.
21.5 InP-based QDIP focal plane arrays
Infrared imaging systems with high operating temperature will reduce imaging system cost and
complexity by reducing the cooling requirements normally associated with running at cryogenic
temperatures. The ability to maintain a photocurrent up to high temperatures with a low dark
current makes QDIPs especially suitable for FPA applications where large dark currents, like
those associated with QWIPs operating at high temperatures, would prevent high temperature
operation due to saturation of the signal collecting capacitor on the readout integrated circuit
(ROIC). Other detector technologies with potential for high operating temperature in the MWIR
window include HgCdTe [4] and type II superlattices [38] .
Here we present an infrared focal plane array with a very low dark current operating at temper-
atures up to 200 K. The FPA is based on our high temperature high performance QDIP shown in
21.4.4. The responsivity of this detector increases with increasing temperature up to 180 K, then
remains fairly constant up to 200 K and begins to drop above 200 K. The photocurrent spectrum
and signal are both measurable up to room temperature for this device. The persistence of the
photocurrent up to room temperature gives this device the potential for high temperature imaging
provided that the noise and dark currents are suitably low. The dark current density of the detec-
tor was very low even at high temperatures. This makes the detector a candidate for high-temper-
ature FPA applications because the ROIC readout capacitor will not be saturated too quickly.
A 320 256 focal plane array with 30 μ m pitch and 25 μ m 2 5 μ m detectors was fabricated
[39] . The FPA pixel defi nition and metallization were essentially the same as for the test mesa
90 120 150 180 210 240 270 300
10
8
10
9
10
10
10
11
Detectivity (cmHz
1/2
/W)
Operation temperature (K)
Figure 21.40 Maximum detectivity at different temperatures.
CH021-I046325.indd 654CH021-I046325.indd 654 6/27/2008 5:48:29 PM6/27/2008 5:48:29 PM
Quantum Dot Infrared Photodetectors by Metal-Organic Chemical Vapour Deposition 655
fabrication [40] . After array fabrication, the Indigo ISC9705 ROIC is hybridized to the FPA by
indium bump bonding. Then, the FPA substrate was thinned using mechanical lapping and pol-
ishing. Finally, the hybridized die was then mounted in and wire bonded to a leadless ceramic
chip carrier.
The FPA was tested using a CamIRa system from SE-IR Corp. The imaging system cryostat was
equipped with a Ge window with 94% transmission and a midwave, f/2 ASIO series lens from
Janos Technology with 90% transmission. All imaging and measurements were taken with a
300 K background.
T 130 K T 200 K
Figure 21.41 Focal plane array imaging taken at 130 K and 200 K, which was also the maximum operating tem-
perature of the array.
Images obtained from the focal plane array are shown in Fig. 21.41 . During testing, imag-
ing was achieved at operating temperatures from 77 K to 200 K. The imaging tests were carried
out at a fi xed frame rate of 32.64 Hz and, depending on the operating temperature, with applied
biases from 1 V to around 3 V and integration times from 0.34 ms up to 30.41 ms. Two-point
non-uniformity correction was used for the imaging. Imaging of human targets was possible up
to around 150 K, and a hot soldering iron could be imaged up to 200 K. Above 200 K the dark
current of the detector became too high for imaging.
The FPA photocurrent was measured using an extended area blackbody source from CI
Systems. The photocurrent of the FPA was measured with an extended area blackbody set at
35°C. The responsivity and conversion effi ciency could then be calculated by dividing the meas-
ured photocurrent by the appropriate optical input. The mean responsivity and mean conversion
effi ciency of the FPA were 34 mA/W and 1.1%, respectively. The temporal noise of the focal plane
array was measured by taking the standard deviation of the FPA signal. A common metric for
focal plane array performance is the noise equivalent temperature difference (NEDT). In order to
measure the NEDT, the FPA signal is measured for two different temperature targets provided by
the extended area blackbody (25°C and 35°C). The differential signal for the two temperatures
divided by the temporal noise gives the signal to noise ratio (SNR), and the target temperature
difference divided by the SNR gives the NEDT. At 120 K the average NEDT was 344 mK and the
percentage of functional pixels was 99%. The histogram of the NEDT is shown in Fig. 21.42 . It
should be noted that the NEDT measurements were taken without two-point non-uniformity
correction. The current injection effi ciency, which is the ratio of the integrated current to the
photocurrent, of the array was calculated using:
η
ω
inj
ph
m
m
m
IRg
Rg
jC R
g
int det
det
det det
det
I
R
1
1
1
1
(21.33)
where g
m
is the ROIC transconductance, R
det
is the detector differential resistance, and C
det
is the
detector capacitance of 8.3 1 0
1 4
F. The injection effi ciency was 99% at 120 K. At 200 K the
injection effi ciency was still 98% due to the high differential resistance even at high temperature.
CH021-I046325.indd 655CH021-I046325.indd 655 6/27/2008 5:48:29 PM6/27/2008 5:48:29 PM
656 Handbook of Self Assembled Semiconductor Nanostructures for Novel Devices in Photonics and Electronics
In conclusion, focal plane array imaging based on an InP-based, InAs QD/InGaAs QW/InAlAs
barrier detector with imaging at temperatures up to 200 K has been demonstrated. The very low
dark current of the device enabled the high temperature imaging capability. The improvement of
performance for this FPA will need to come primarily in two areas. The fi rst is achieving a further
reduction of the dark current at high temperatures, and the second is achieving a higher respon-
sivity at low bias.
21.6 Summary
In this chapter, InAs QDIPs based on the GaInAs/AlInAs/InP system were demonstrated by
MOCVD. The dependence of the dot formation on the MOCVD growth conditions were systemati-
cally studied and characterized. High density, uniform InAs QDs were obtained at the optimum
conditions. Based on such QDs, several QDIP device structure were fabricated and characterized.
The wavelength tunability of the QDIPs was realized within a certain range. With the improve-
ment of structure design and material quality, some of the devices ’ have outperformed the
corresponding QWIP. A 320 256 MWIR FPA based on our high performance QDIP was dem-
onstrated. Imaging at temperatures up to 200 K was demonstrated.
Despite the exciting progress that has been made in this fi eld, the performance of QDIP devices
is still from desired. In order to further improve the device performance, QDIPs can be improved
at several levels:
● QDs are the key element of the device. The dot size directly affects the detection wave-
length. The shape of the dot may affect the oscillator strength and thus quantum effi -
ciency. The density and uniformity affect device performance such as responsivity and
detectivity. Thus it is crucial to have high density and uniformity QDs and still maintain
the right size.
● Proper doping of quantum dots is necessary to maintain the balance between high respon-
sivity and low dark current, which is the key to high-temperature operation.
● The choice and quality of barrier material are important. A barrier provides confi nement
of the QDs and so affects the detections ’ wavelength as well as transport properties.
● The stack number of QD layers in the QDIP active region can be increased to improve the
absorption. Ideally, each layer of QD stacks would be the same, thus QDIP would be like
QWIP, in which detectvity D
*
scales with the square root of the number of stacks
N
0 0.2 0.4 0.6 0.8 1
0
500
1000
1500
2000
2500
3000
NEDT histogram
NEDT (K)
Number of pixels
Figure 21.42 Histogram of the focal plane array noise equivalent temperature difference (NEDT) at 120 K. The
mean NEDT was 344 mK.
CH021-I046325.indd 656CH021-I046325.indd 656 6/27/2008 5:48:30 PM6/27/2008 5:48:30 PM
Quantum Dot Infrared Photodetectors by Metal-Organic Chemical Vapour Deposition 657
within a certain range [41] . In reality, however, due to the strain, the quality of stacks
might degrade with the increase in stacks, which would limit the success of this approach.
● An external scheme may also used to improve the device performance. Resonant cavity
enhanced (RCE) detectors have been successfully used in near-infrared telecommunica-
tions. The internal quantum effi ciency of QDIPs may also be enhanced by RCE structure.
This may be realized by placing a pair of distributed Bragg refl ectors (DBR). One limit to
this approach is that it requires the growth of many quarter wavelength layers in order
to achieve the desired refl ectivity. Such growth will be tedious and more and more diffi -
cult with the increase in the wavelength. An alternative is to make the DBR with higher
index contrast material, for example with dielectric materials. This will have to be done
after the growth and may require the removal of the substrate. The normal trade-off of
the DBR scheme is a reduction in the bandwidth of the photoresponse. However, a QDIP
has a natural narrow bandpass detection scheme that is compatible with resonant cavity
technology.
● Besides the improvement of QDIP itself, the QDIP FPA requires extra consideration such as
the matching of the device operation parameters with characteristics of the ROIC.
References
1. K. Mukai and M. Sugawara, Semiconductor and Semimentals. Volume 60, Chapter 5 (Acadamic
Press, San Diego, 1999).
2. Y. Arakawa and H. Sakaki , Appl. Phys. Lett. 40 , 939 ( 1982 ) .
3. M. Asada , Y. Miyamoto , and Y. Suematsu , IEEE. J. Quantum Electron . QE-22 , 1887 ( 1986 ) .
4. A. Piotrowski , P. Madejczyk , W. Gawron , K. Klos , J. Pawluczyk , M. Grudzien , J. Piotrowski , and
A. Rogalski , Proc. of SPIE 5732 , 273 ( 2005 ) .
5. B.F. Levine , J. Appl. Phys. 74 , R1 – R81 ( 1993 ) .
6. V. Ryzhii , Semicond. Sci. Technol. 11 , 759 – 765 ( 1996 ) .
7. L. Goldstein , F. Glas , J. Marzin , M. Charasse , and G. LeRoux , Appl. Phys. Lett. 47 , 1099 ( 1985 ) .
8. D. Gershoni , H. Temkin , G.J. Dolan , J. Dunsmuir , S.N.G. Chu , and M.B. Panish , Appl. Phys. Lett.
53 , 995 ( 1998 ) .
9. S. Gangopadhyay and B.R. Nag , Nanotechnology 8 , 14 ( 1997 ) .
10. M. Califano and P. Harrison , Phys. Rev. B 6 1 , 10959 ( 2000 ) .
11. E.P. Pokatilov , V.A. Fonoberov , V.M. Fomin , and J.T. Devreese , Phys. Rev. B 64 , 245328 ( 2001 ) .
12. J. Williamson , L. Wang , and A. Zunger , Phys. Rev. B 62 , 12963 ( 2000 ) .
13. E. Rosencher and B. Vinter . Optoelectronics , ( Cambridge University Press , Cambridge , 2002 ) .
14. K.K. Choi , Appl. Phys. Lett. 65 , 1268 ( 1994 ) .
15. Z. Ye , J.C. Campbell , Z. Chen , E.-T. Kim , and A. Madhukar , Appl. Phys. Lett. 83 , 1234 ( 2003 ) .
16. W.A. Beck , Appl. Phys. Lett. 63 , 3589 ( 1993 ) .
17. B.F. Levine , A. Zusman , S.D. Gunapala , M.T. Asom , J.M. Kuo , and W.S.J. Hobson , Appl Phys. 7 2 ,
4429 ( 1992 ) .
18. V. Ryzhii , Jap J. Appl. Phys. 4 0 , L148 ( 2001 ) .
19. V. Ryzhii , I. Khmyrova , and V. Mitrin , Semic. Science and Technology 19 , 8 ( 2004 ) .
20. Z.M. Ye , J.C. Campbell , Z.H. Chen , E.T. Kim , and A. Madhukar Noise and photoconductive gain in
InAs quantum-dot infrared photodetectors , Appl. Phys. Lett. 83 , 1234 ( 2003 ) .
21. W. Zhang , H. Lim , M. Taguchi , S. Tsao , B. Movaghar , and M. Razegh i , Appl. Phys. Lett. 8 6 ,
191103 ( 2005 ) .
22. J. Phillips ,
J. Appl. Phys. 9 1 , 4590 ( 2002 ) .
23. Y.M. Qiu and D. Uhl , J. Cryst. Growth 257 , 225 ( 2003 ) .
24. E.T. Kim , A. Madhukar , Z. Ye , and J.C. Campbell , Appl. Phys. Lett. 84 , 3277 ( 2004 ) .
25. D. Pan , E. Towe , and S. Kennerly , Appl. Phys. Lett. 73 , 1937 ( 1998 ) .
26. D. Gershoni , H. Temkin , G.J. Dolan , J. Duinsmuir , S.N.G. Chu , and M.B. Panish , Appl. Phys. Lett.
53 , 995 ( 1988 ) .
27. M.A. Cusack , P.R. Briddon , and M. Jaros , Phys. Rev. B 54 , R2300 ( 1996 ) .
28. R. People , J. Appl. Phys. 62 , 2551 ( 1987 ) .
CH021-I046325.indd 657CH021-I046325.indd 657 6/27/2008 5:48:30 PM6/27/2008 5:48:30 PM
658 Handbook of Self Assembled Semiconductor Nanostructures for Novel Devices in Photonics and Electronics
29. M. Holm , M.-E. Pistol , and C. Pryor , J. Appl. Phys. 92 , 932 ( 2002 ) .
30. H. Lim , W. Zhang , S. Tsao , T. Sills , J. Szafraniec , K. Mi , B. Movaghar , and M. Razeghi , Phys. Rev.
B 72 , 085332 ( 2005 ) .
31. M. Allovon and M. Quillec , IEE Pro. 139 , 148 ( 1992 ) .
32. M. Razeghi, The MOCVD Challenge, Vol. 1 (Adam Hilger, Bristol, 1989).
33. H. Lim , S. Tsao , W. Zhang , and M. Razeghi , Appl. Phys. Lett. 90 , 131112 ( 2007 ) .
34. S. Krishna , J. Phys. D: Appl. Phys. 38 , 2142 – 2150 ( 2005 ) .
35. H. Lim , W. Zhang , S. Tsao , T. Sills , J. Szafraniec , K. Mi , B. Movaghar , and M. Razeghi , Phys. Rev.
B 72 , 085332 ( 2005 ) .
36. S.Y. Wang , S.D. Lin , H.W. Wu , and C.P. Lee , Appl. Phys. Lett. 78 , 1023 ( 2000 ) .
37. Z. Ye , J.C. Campbell , Z. Chen , E.-T. Kim , and A. Madhukar , Appl. Phys. Lett. 83 , 1234 ( 2003 ) .
38. M. Razeghi , Y. Wei , A. Hood , D. Hoffman , B.M. Nguyen , P.Y. Delaunay , E. Michel , and
R. McClintock , Proc of SPIE 6206 , 62060N – 1 ( 2006 ) .
39. S. Tsao , H. Lim , W. Zhang , and M. Razeghi , Appl. Phys. Lett. 90 , 201109 ( 2007 ) .
40. J. Jiang , K. Mi , S. Tsao , W. Zhang , H. Lim , T. O’Sullivan , T. Sills , M. Razeghi , G.J. Brown , and
M.Z. Tidrow , Appl. Phys. Lett. 84 , 2232 ( 2004 ) .
41. K.K. Choi . The Physics of Quantum Well Infrared Photodetectors , ( World Scientifi c Publishing Co. ,
Singapore , 1997 ) .
CH021-I046325.indd 658CH021-I046325.indd 658 6/27/2008 5:48:30 PM6/27/2008 5:48:30 PM